Photo Multiplier Handbook

Contents

1 . l n t r o d u c t i o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Early development, photoemitter and secondary-emitter development, applications development,photomultiplier and solid-state detectors compared

2. Photomultiplier Design . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10Photoemission, practical photocathode materials, opaque and semitransparent photocathodes, glasstransmission and spectral response, thermionic emission, secondary emission, time tag in photoemission andsecondary emission

3. Electron Optics of Photomultipliers , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Electron-optical design considerations, design methods for photomultiplier electron optics, specificphotomultiplier electron-optical configurations, anode configurations

4. Photomultiplier Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . 36Photocathode-related characteristics, gain-related characteristics, dark current and noise, time effects, pulsecounting, scintillation counting, liquid scintillation counting, environmental effects

5. Photomultiplier Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

Summary of selection criteria, applied voltage considerations, mechanical considerations, optical considera-tions, specific photomultiplier applications

Appendix A. Typical Photomultiplier Applications and Selection Guide . . . . . . . . . . . 119

Appendix B. Glossary of Terms Related to Photomultiplier lubes and Their Applica- I. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

Appendix C. Spectral Response Designation Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132

Appendix D. Photometric Units and Photometric-to-Radiant Conversion. . . . . . . . . . . . . I 37

Appendix E. Spectral Response and Source-Detector Matching . . . . . . . . . . . . . . . . . . . I 43

Appendix F. Radiant Energy and Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148

Appendix G. Statistical Theory of Noise in Photomultiplier Tubes. . . . . . . . . . . . . . . I 60

I n d e x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 7

information furnished by BURLE INDUSTRIES, INC. is believed to be accurate and reliable. However, no responsibility of liabilityis assumed by BURLE for its use, nor for any infringement of patents or other rights of third parties which may result from its use. Nolicense is granted by implication or otherwise under any patent or other rights of BURLE INDUSTRIES, INC.

BURLE ® and BURLE INDUSTRIES, INC.® are registered trademarks of BURLE TECHNOLOGIES, INC. Marca(s) Registrada(s).Copyright © 1980 by BURLE TECHNOLOGIES, INC. All rights reserved. No part of this book may be copied or reproduced in anyform or by any means -- graphic, electronic, or mechanical or incorporated into any information storage or retrieval system, withoutthe written permission of the copyright owner.

Supersedes PMT-62, 8-80 Printed in U.S.A./ 10-89TP-136

2

1. Introduction

The photomultiplier is a very versatile andsensitive detector of radiant energy in theultraviolet, visible, and near infrared regionsof the electromagnetic spectrum. A schemat-ic diagram of a typical photomultiplier tubeis given in Fig. 1. The basic radiation sensoris the photocathode which is located inside avacuum envelope. Photoelectrons are emit-ted and directed by an appropriate electricfield to an electrode or dynode within theenvelope. A number of secondary electronsare emitted at this dynode for each imping-ing primary photoelectron. These secondaryelectrons in turn are directed to a seconddynode and so on until a final gain ofperhaps 106 is achieved. The electrons fromthe last dynode are collected by an anodewhich provides the signal current that is readout.

PHOTOELECTRONS

92cs-32288

Fig. 1 - Schematic representation of a photo-multiplier tube and its operation

For a large number of applications, thephotomultiplier is the most practical or sen-sitive detector available. The basic reasonfor the superiority of the photomultiplier isthe secondary-emission amplification thatmakes it possible for the tube to approach“ideal” device performance limited only bythe statistics of photoemission. Amplifica-tions ranging from 103 to as much as 108

provide output signal levels that are com-patible with auxiliary electronic equipment

without need for additional signal amplifica-tion. Extremely fast time response with risetimes as short as a fraction of a nanosecondprovides a measurement capability in specialapplications that is unmatched by otherradiation detectors.

EARLY DEVELOPMENTThe development (history) of the photo-

multiplier is rooted in early studies of secon-dary emission. In 1902, Austin and Starke1

reported that the metal surfaces impacted bycathode rays emitted a larger number of elec-trons than were incident. The use of secon-dary emission as a means for signalamplification was proposed as early as1919.2 In 1935, Iams and Salzberg 3 of RCAreported on a single-stage photomultiplier.The device consisted of a semicylindricalphotocathode, a secondary emitter mountedon the axis, and a collector grid surroundingthe secondary emitter. The tube had a gainof about eight. Because of its better frequen-cy response the single-stage photomultiplierwas intended for replacement of the gas-filled phototube as a sound pickup formovies. But despite its advantages, it sawonly a brief developmental sales activitybefore it became obsolete.

Multistage DevicesIn 1936, Zworykin, Morton, and Malter,

all of RCA4 reported on a multistagephotomultiplier. Again, the principal con-templated application was sound-on-filmpickup. Their tube used a combination ofelectrostatic and magnetic fields to directelectrons from stage to stage. A photographof a developmental sample is given in Fig. 2.Although the magnetic-type photomultiplierprovided high gain, it had several dif-ficulties. The adjustment of the magneticfield was very critical, and to change the gainby reducing the applied voltage, themagnetic field also had to be adjusted.

3

Photomultiplier Handbook

Another problem was that its rather wideopen structure resulted in high dark currentbecause of feedback from ions and lightdeveloped near the output end of the device.For these reasons, and because of thedevelopment of electrostatically focusedphotomultipliers, commercialization did notfollow.

Fig. 2 - Magnetic-type multistage photomul-tiplier reported by Zworykin, Morton, andMalter in 1936.

The design of multistage electrostaticallyfocused photomultipliers required ananalysis of the equipotential surfaces be-tween electrodes and of the electron trajec-tories. Before the days of high-speed com-puters, this problem was solved by amechanical analogue: a stretched rubbermembrane. By placing mechanical models ofthe electrodes under the membrane, theheight of the membrane was controlled andcorresponded to the electrical potential ofthe electrode. Small balls were then allowedto roll from one electrode to the next. Thetrajectories of the balls were shown to cor-respond to those of the electrons in the cor-responding electrostatic fields. Workingwith the rubber-dam analogue, both J.R.Pierce5 of Bell Laboratories and J.A.Rajchman6 of RCA devices linear arrays ofelectrodes that provided good focusing prop-erties. Although commerical designs did notresult immediately from the linear dynodearray, The Rajchmann design with somemodifications eventually was, and still is,used in photomultipliers-particularly forhigh-gain wide-bandwidth requirements.First Commercial Devices

The first commercially successfulphotomultiplier was the type 931. This tubehad a compact circular array of ninedynodes using electrostatic focusing. Thefirst such arrangement was described byZworykin and Rajchman.7 Modificationswere later reported by Rajchmann and

4

Snyder 8 and by Janes and Gloverg, all ofRCA. The basic electron-optics of the cir-cular cage was thus well determined by 1941and has not changed to the present timealthough improvements have been made inprocessing, construction, and performanceof the 931A product.

The success of the 931 type also resultedfrom the development of a much improvedphotocathode, Cs3Sb, reported by Gorlich 10

in 1936. The first experimental photo-multipliers had used a Ag-O-Cs photocath-ode having a typical peak quantum efficien-cy of 0.4% at 800 nm. (The Ag-O-Cs layerwas also used for the dynodes.) The newCs3Sb photocathode had a quantum effi-ciency of 12% (higher today) at 400 nm. Itwas used in the first 931’s, both as aphotocathode and as a secondary-emittingmaterial for the dynodes.

PHOTOEMITTER AND SECONDARY-EMITTER DEVELOPMENT

Photocathode MaterialsMuch of the development work on

photomultiplier tubes has been concernedwith their physical configuration and therelated electron optics. But a very importantpart of the development of photomultipliertubes was related to the photocathode andsecondary-emission surfaces and their pro-cessing. RCA was very fortunate during the1950’s and 60’s in having on its staff, prob-ably the world’s foremost photocathode ex-pert, Dr. A.H. Sommer. His treatise onPhotoemissive Materials1 1 continues to pro-vide a wealth of information to allphotocathode process engineers.

Sommer explored the properties ofnumerous photocathode materials-par-ticularly alkali-antimonides. Perhaps hismost noteworthy contribution was themultialkali photocathode (S-20 spectralresponse). This photocathode, Na2KSb:Cs,is important because of its high sensitivity inthe red and near infrared; the earlier Cs3Sbphotocathode spectral response barely ex-tends through the visible, although it is verysensitive in the blue where most scintillatorsemit.

Bialkali photocathodes were also de-veloped by Sommer and have proven to be

better in some applications than the Cs3Sbphotocathode. Thus, the Na2KSb photo-cathode has been found to be stable at highertemperatures than Cs3Sb and, in addition,has a very low dark (thermal) emission. Ithas been particularly useful in oil-well-logging applications. Another bialkali pho-tocathode, K2CsSb, is more sensitive thanCs3Sb in the blue and is, therefore, used byRCA to provide a better match to theNaI:Tl crystals used in scintillation count-ing.

Dynode MaterialsThe first secondary-emission material

used practically by RCA was the Ag-O-Cssurface. But with the development of theCs3Sb material for photocathodes, it wasfound that this material was also an excellentsecondary emitter. Other practical secondaryemitters developed during the early years ofphotomultiplier development were MgO:Cs(often referred to as “silver-magnesium”)and BeO:Cs (“copper-beryllium”).

In the early 1960’s, R.E. Simon12 whileworking at the RCA Laboratories developedhis revolutionary concept of Negative Elec-tron Affinity (NEA). Electron affinity is theenergy required for an electron at theconduction-band level to escape to thevacuum level. By suitably treating the sur-face of a p-type semiconductor material, theband levels at the surface can be bentdownward so that the effective electron af-finity is actually negative. Thermalized elec-trons in the conduction band are normallyrepelled by the electron-affinity barrier; theadvantage of the NEA materials is that theseelectrons can now escape into the vacuum asthey approach the surface. In the case ofsecondary emission, secondary electrons canbe created at greater depths in the materialand still escape, thus providing a muchgreater secondary-emission yield. In the caseof photoemission, it has been possible toachieve extended-red and infrared sen-sitivities greater than those obtainable withany other known materials. The first prac-tical application of the NEA concept was tosecondary emission.Simon and Williams l3

An early paper bydescribed the theory

and early experimental results of secondary-emission yields as high as 130 at 2.5 kV forGaP:Cs.

Introduction

APPLICATIONS DEVELOPMENTAstronomy and Spectroscopy

Early applications of the photomultiplierwere in astronomy and spectroscopy.Because the effective quantum efficiency ofthe photomultiplier was at least ten timesthat of photographic film, astronomers werequick to realize the photomultiplier tube’sadvantage. Furthermore, because the outputcurrent of the photomultiplier is linear withincident radiation power, the tube could beused directly in photometric and spec-trophotometric astronomy. The type 1P28, atube similar to the 931 but having anultraviolet-transmitting envelope was par-ticularly useful in spectroscopy. The size andshape of the photocathode were suitable forthe detection and measurement of line spec-tra and the very wide range of available gainproved very useful. 14

Radar JammerA totally unexpected application for the

new photomultiplier tube occurred duringWorld War II. The development of radar fordetecting and tracking aircraft led to thesimultaneous need for wideband electronic-noise sources as radar jammers. Althoughother sources of noise were tried, thephotomultiplier proved to be most suc-cessful. The advantage of the tube was itshigh gain (107) and wide band width (severalhundred MHz). As a noise source the tubewas operated with a non-modulated inputlight source and with high gain. The outputamplifier photoelectric shot noise was“white” and thus indistinguishable fromnatural noise sources. This application ofphotomultiplier tubes resulted in productionof thousands per month compared withprevious production measured in only hun-dreds per year.Scintillation Counting

A proliferation of photomultiplier designsfollowed the invention of the scintillationcounter shortly after World War II.15,16 Thephotomultiplier tubes were designed withsemitransparent photocathodes deposited onan end window which could be coupleddirectly to the scintillator. The principalscintillator used, NaI doped with thallium,was discovered by Hofstadter17. Much ofthe development work on photomultiplier

5


tubes during this period was reported byRCA and its competitors in the biannualmeetings of the Scintillation Counter Sym-posium. These symposia were reported fullyin the IRE (and later the IEEE) Transactionson Nuclear Science beginning with themeeting in Washington, January 1948. Thescintillation counter became the most impor-tant measurement instrument in nuclearphysics, nuclear medicine, and radioactivetracer applications of a wide variety.Headlight Dimmer

During the 1950’s, RCA collaborated withthe General Motors Company (Guide-LampDivision) on a successful headlight dimmer.The photoelectric headlight dimmer-firstmade available only on Cadillacs andOldsmobiles-basically used a tube similarto the 931A, but redesigned and tested to theauto manufacturer’s particular require-ments. The optical engineering problem wasto sense the oncoming headlights or tail-lights being followed without responding tostreet and house lights. Vertical and horizon-tal angular sensitivity was designed to matchthe spread of the high beams of the automo-bile. A red filter was installed in the opticalpath to provide a better balance between sen-sitivity to oncoming headlights and to tail-lamps being followed. The device achieved aremarkable success, probably because of thenovelty, and thousands of photomultipliertubes were used. But today, one rarely sees aheadlight dimmer.

Medical Diagnostic EquipmentIn recent years two medical applications

have used large numbers of photomultipliertubes and have spurred further develop-ments and improvements. The gamma cam-era18 is a sophisticated version of the scin-tillation counter used medically for locatingtumors or other biological abnormalities. Aradioactive isotope combined in a suitablecompound is injected into the blood streamor ingested orally by the patient. Theradioactive material disintegrates and gam-ma rays are ejected from preferential loca-tions such as tumors or specific organs. Alarge crystal intercepts the gamma rays andscintillates. Behind the crystal are photo-multiplier tubes, perhaps 19, in hexagonalarray. The location of the point of scintilla-tion origin is obtained by an algorithm

6

depending upon the individual signals fromeach of the photomultipliers. Counting iscontinued until several hundred thousandcounts are obtained and the organ in ques-tion is satisfactorily delineated. The locationof each scintillation is represented by a pointon a cathode-ray-tube presentation.

The Computerized Axial Tomographic(CAT) scanner was introduced to this coun-try in 1973. The device uses a pencil or fan-beam of X-rays which rotates around the pa-tient providing X-ray transmission datafrom many directions. A scintillator coupledto a photomultiplier detects the transmittedbeam-as an average photomultiplier cur-rent-and a computer stores and computesthe cross-section density variation of the pa-tient’s torso or skull. The photomultipliersare1/ or 3/4-inch end-on tubes whichcouple to the scintillator, commonly BGO(bismuth germanate). Each unit is equippedwith as many as 600 photomultipliers.

PHOTOMULTIPLIERS AND SOLID-STATE DETECTORS COMPARED

In some applications either a photomulti-plier or solid-state detector could be used.The user may make his choice on the basis offactors such as cost, size, or previous ex-perience. In other applications, the choicemay be dictated by fundamental propertiesof the photomultiplier or the solid-statedetector. A discussion follows of some of thecommon applications favoring one or theother detector with reasons for the choice. Asummary presents the principal considera-tions the user must apply in making a choicein an application for which he requires aphotodetector. This information should beparticularly useful to the designer who is notwell acquainted in this field.Photomultiplier Features

The photomultiplier is unique in its abilityto interface with a scintillation crystal andnot only count the scintillations but measuretheir magnitude and time their arrival. Mostscintillators emit in the blue and near ultra-violet. This spectral output obviously favorsthe photomultiplier having a photocathodewith high quantum efficiency in the shortwavelength range. On the other hand a sili-con p-i-n diode is relatively poor in this partof the spectrum but does best in the red andnear infrared. The most important factor,

probably, is the gain of the photomultiplierwhich permits the measurement of the verysmall signals from individual scintillationswith a good signal-to-noise ratio, limitedprimarily by the statistics of the number ofphotoelectrons per pulse. Finally, the shortrise time of the photomultiplier using fastscintillators permits time-of-flight measure-ments to be made in nuclear physics.

Although the CAT scanner equipmentalso uses photomultipier tubes to detect thescintillations in bismuth germanate (BGO)crystals, the situation is somewhat differentfrom the scintillation counting applicationsdiscussed above. In the CAT scanner theX-rays produce a broad band of pulseheights and no attempt is made to single outand detect single scintillation events. Thephotomultiplier is used in an analog mode todetect the level of radiation incident on thecrystal. In the CAT scan operation thetypical machine scans the patient in a fewseconds and the level of irradiance from thecrystal onto the photomultiplier is relativelyhigh so that only a relatively low gain photo-multiplier is required. Furthermore, thespeed of response requirement for thephotomultiplier is relatively modest-per-haps a few hundred microseconds. Still, theprincipal advantage of using a photomulti-plier in this application for the detection ofthe radiant signal is its good signal-to-noiseratio. This ratio is very important to the pa-tient because a reduction in its signal-to-noise ratio would have to be made up forwith an increased X-ray dose. Nevertheless,there is interest and development activityaimed at replacing the photomultiplier withsilicon p-i-n detectors. Two factors couldfavor the alternate use of a silicon cell: (1) abetter scintillator (BGO is almost an order ofmagnitude less sensitive than NaI:Tl; (2) afaster scanning machine (a very desirabletechnological advance because is wouldminimize effects of body motions). Both ofthese factors would result in a largerphotocurrent and could bring the signal levelfor the silicon detector to the point where thefundamental signal-to-noise ratio from theX-ray source would not be degraded. Suchdevelopments may be anticipated becausethe silicon detector would also have the ad-vantage of smaller size and perhaps lowercost.

Introduction

As a result of increasing concern about en-vironment, pollution monitoring is becom-ing another important application for photo-multiplier tubes. For example, in the moni-toring of NOx the gas sample is mixed withO 3 in a reaction chamber. A chemilum-inescence results which is measured using anear-infrared-pass filter and a photomulti-plier having an S-20 spectral response. Al-though the radiation level is very low, NOcan be detected down to a level of 0.1 ppm.The advantage of the photomultiplier in thisapplication is again the high gain and goodsignal-to-noise ratio (the photomultiplier iscooled to 0°C to reduce dark-current noise)even though the radiation spectrum is ob-served near the threshold of the S-20 spectralrange.

In another pollution-monitoring applica-tion, SO2 is detected down to a level of 0.002ppm. Here, the sample containing SO2 is ir-radiated with ultraviolet and the excited SO2molecules fluoresce with blue radiation thatis detected with a combination of a narrow-band filter and photomultiplier. Very weaksignals are detected and again it is the highgain, good signal-to-noise ratio and, in addi-tion, good blue sensitivity which makes thedetection and measurement of small contam-inations of SO2 possible.

Spectroscopy is one of the very early ap-plications for photomultipliers. The widerange of radiation levels encountered isreadily handled by the approximately loga-rithmic gain variation of the photomultiplierwith voltage. At very low signal levels, thesignal-to-noise capability of the photomulti-plier is essential. Because photomultiplierspectral response (with quartz or ultraviolet-transmitting-glass windows) covers the rangefrom ultraviolet to near infrared, thephotomultiplier is the logical choice for spec-troscopic applications, except in the infraredregion of the spectrum.Photocell* Features

Because of their small size and low cost,CdSe and CdS type photocells are the logicalselection for applications such as automaticexposure control in photographic cameras orvarious inspection and counting require-ments.*“Photocell” is used here to indicate a photosensitivedevice in which the charge transport takes place througha solid as compared with “phototube” in which thecharge transport is through a vacuum.

7


Many p-i-n silicon cells are used in com-bination with lasers or LED’s (light emittingdiodes). Here, one of the principal advantagesof the silicon cell is its good response in thenear infrared out to 1100 nm. In combina-tion with the Nd:YAG laser emitting at 1060nm, the silicon cell is used widely in laserranging and laser tracking. A similar ap-plication utilizes an LED emitting near 900nm with a silicon cell for automatic rangingfor special camera equipment. Size and in-frared sensitivity are again the importantqualifications.

A rapidly growing application for photo-cells is for fiber-optic communicationsystems. LED’s are coupled to the fibers andthe detector may be a p-i-n diode or, for abetter signal-to-noise ratio, a silicon ava-lanche diode. The qualifying attributes forthe choice of detector are size, near infraredsensitivity, adequate speed of response, andgood signal-to-noise ratio.

Smoke detectors now use large numbers ofLED’s and p-i-n silicon cells. Again size,cost, and infrared sensitivity are the impor-tant qualifications.Characteristics Comparison Summary

Spectral Response. Photomultipliers canbe obtained with good spectral sensitivity inthe range 200 to 900 nm. Silicon cells haverather poor blue sensitivity, but are excellentout to 1100 nm. In general, then, the photo-multiplier is to be preferred for applicationsinvolving the shorter wavelengths, althoughother factors may override this considera-tion.

Speed of Response. If very fast response isrequired, the photomultiplier is usually thebest choice of a detector. Photomultipliersare available with rise times (10 to 90%) of 1or 2 nanoseconds using a 50-ohm load. Theinherent rise time of silicon cells may be inthe range 10 to 20 nanoseconds, dependingupon the area of the cell. However, becauseof the cell’s capacitance, the effective risetime is much longer depending upon thechoice of load resistance. For example, witha 1-megohm load resistance, the rise timemay be of the order of 20 microseconds. Afairly large load resistance must be chosen tomaintain good signal-to-noise characteristicsfor the silicon cell. Silicon avalanche photo-diodes can have rise times as short as 2nanoseconds. Gain for an avalanche photo-

8

diode can be of the order of 100, but the sen-sitive area is small-about 0.5 squaremillimeter.

Sensitive Area. Photomultiplier tubes aremade in a variety of sizes so that many dif-ferent optical configurations can be accom-modated. The largest photocathode areaavailable in commercial RCA photomulti-plier tubes has a nominal diameter of 5 in-ches and a minimum useful area of 97 squarecentimeters. By way of contrast, the 1/2-inchside-on photomultiplier has a projected pho-tocathode area of 0.14 square centimeter.Silicon p-i-n diodes are available with sen-sitive areas generally not larger than 1 squarecentimeter; and avalanche silicon cells, 0.005square centimeter. In many applications, afairly large area is required, e.g., coupling toa cathode-ray tube or a large scintillator.This requirement generally indicates the useof a photomultiplier tube. Silicon cells are atan advantage when the source is small fordirect coupling or for lens imaging.

Temperature. Photomultipliers are gener-ally not rated for operation at temperatureshigher than 75° C. Exceptions are photomul-tipliers having a Na2KSb photocathode. Thisbi-alkali photocathode can tolerate temper-atures up to 150° C or even higher for shortcycles. In oil-well logging measurements thisconsideration is important. Photocathodesensitivities and gain change very little withtemperature, but dark current does increaserapidly. Dark currents at room temperatureare of the order of 10 - l5 ampere at the pho-tocathode and double about every 10° C.Silicon cells are rated from - 50 to 80° C.Sensitivities are also relatively independentof temperature. But dark current which maybe 10-7 ampere at room temperature, alsotends to double about every 10° C.

Signal-to-Noise Ratio. At very low lightlevels, the limitation to detection andmeasurement is generally the signal-to-noiseratio. One way of describing the limit todetection is to state the Equivalent Noise In-put (ENI) or the Noise Equivalent Power(NEP). The NEP is the power level into thedevice which provides a signal just equal tothe noise. Most often the bandwidth isspecified as 1 hertz and the wavelength of themeasurement is at the peak of the spectralresponsivity. ENI is the same type of specifi-cation except the unit instead of power may

be luminous flux.For a photomultiplier such as one used for

spectroscopy, the NEP at room temperatureat 400 nm is about 7 x 10-16 watts, or theEN1 is about 7 x 10 - l3 lumens. Bothspecifications are for a l-hertz ban

dwidth.

For a p-i-n silicon photocell, the NEP at 900nanometers may be of the order of 2 x10- watts, or the EN1 of 1.5 x 10- l1

lumens. Both values are for a l-hertz band-width. Thus, the photomultiplier is clearlysuperior in this category. Also it should bepointed out that the silicon diode must becoupled into a load resistance of about 5megohms in order to avoid noise dominationfrom the coupling resistor. Unfortunately,this large resistance then increases the effec-tive rise time of the silicon device to about100 microseconds. The NEP of a silicon ava-lanche photodiode is about 10-14 watt at900 nanometers or the ENI is 8 x 10 - l3lumens, both for a l-Hz bandwidth. Thelumen in these descriptions is that from atungsten source operating at 2856 K colortemperature. Peak emission for such asource is near 1000 nm and thus closelymatches the spectral peak of the silicondevices.

Gain. A photomultiplier can have a gainfactor, by which the fundamental photo-cathode signal is multiplied, of from 103 to108. Silicon avalanche photodiodes have again of about 100. Silicon p-i-n diodes haveno gain. The high gain of the photomulti-plier frequently eliminates the need ofspecial amplifiers, and its range of gain con-trolled by the applied voltage provides flex-ibility in operation.

Stability. Photomultiplier tubes are notnoted for great stability although for lowanode currents and careful operation theyare satisfactory. When the light level isreasonably high, however, the very goodstability of the silicon p-i-n cell is a con-siderable advantage. The silicon cell makes aparticularly good reference device for thisreason. In fact, the National Bureau of Stan-dards has been conducting special calibra-tion transfer studies using p-i-n silicondiodes.

REFERENCES1. H. Bruining, Physics and applications

of secondary electron emission, (McGraw-Hill Book Co., Inc.; 1954).

Introduction

2. J. Slepian, U.S. Patent 1, 450, 265,April 3, 1923 (Filed 1919).3. H. E. Iams and B. Salzberg, “The

secondary emission phototube,” Proc. IRE,Vol. 23, pp. 55-64 (1935).4. V.K. Zworykin, G.A. Morton, and L.

Malter, "The secondary-emission multipli -er-a new electronic device,” Proc. IRE,Vol. 24, pp. 351-375 (1936).5 . J .R. Pierce , “Electron-multiplier

design,” Bell Lab. Record, Vol. 16, pp.305-309 (1938).6. J.A. Rajchman, “Le courant residue1

dans les multiplicateurs d’electrons elec-trostatiques,” These L’Ecole PolytechniqueFederale (Zurich, 1938).7. V.K. Zworykin and J.A. Rajchman,

“The electrostatic electron multiplier, Proc.IRE, Vol. 27, pp. 558-566 (1939).

8. J.A. Rajchman and R.L. Snyder, “Anelectrostatically focused multiplierphototube,” Electronics, Vol. 13, p. 20(1940).

9. R.B. Janes, and A.M. Glover, “Recentdevelopments in phototubes,” RCA Review,Vol. 6, pp. 43-54 (1941). Also, A.M. Glover,“A review of the development of sensitivephototubes,” Proc. IRE, Vol. 29, pp.413-423 (1941).10. P. Gorlich, “Uber zusammengesetzte,durchsichtige Photokathoden,” 2. Physik,Vol. 101, p. 335 (1936).11. A.H. Sommer, Photoemiss ivematerials, John Wiley and Sons; 1968.12. R.E. Simon, Research in electronemission from semiconductors, QuarterlyReport, Contract DA 36-039-AMC-02221(E) (1963).13. R.E. Simon and B.F. Williams,“Secondary-electron emission,” IEEETrans. Nucl. Sci., Vol. NS-15, pp. 166-170(1968).14. M.H. Sweet, “Logarithmic photomul-tiplier tube photometer,” JOSA, Vol. 37, p.432 (1947).15. H. Kallmann, Natur u Technik (July1947).16. J.W. Coltman and F.H. Marshall, “Aphotomultiplier radiation detector,” Phys.Rev., Vol. 72, p. 582 (1947).17. R. Hofstadter, “Alkali halide scintilla-tion counters,” Phys. Rev., Vol. 74, p. 100(1948).18. H.O. Anger, “Scintillation camera”,Rev. Sci. Instr., Vol. 29, pp. 27-33 (1958).

9

2. Photomultiplier Design

PHOTOEMISSIONThe earliest observation of a photoelectric

effect was made by Becquerel in 1839. Hefound that when one of a pair of electrodesin an electrolyte was illuminated, a voltageor current resulted. During the latter part ofthe 19th century, the observation of aphotovoltaic effect in selenium led to thedevelopment of selenium and cuprous oxidephotovoltaic cells.

The emission of electrons resulting fromthe action of light on a photoemissive sur-face was a later development. Hertz dis-covered the photoemission phenomenon in1887, and in 1888 Hallwachs measured thephotocurrent from a zinc plate subjected toultraviolet radiation. In 1890, Elster andGeitel produced a forerunner of the vacuumphototube which consisted of an evacuatedglass bulb containing an alkali metal and anauxiliary electrode used to collect thenegative electrical carriers (photoelectrons)emitted by the action of light on the alkalimetal.Basic Photoelectric Theory

The modern concept of photoelectricitystems from Einstein’s pioneer work forwhich he received the Nobel Prize. Theessence of Einstein’s work is the followingequation for determining the maximumkinetic energy E of an emitted photoelec-tron:

(1)Eq. (1) shows that the maximum energy ofthe emitted photoelectron mv2/2 is propor-tional to the energy of the light quanta hv

must be given to an electron to allow it toescape the surface of a metal. For eachmetal, the photoelectric effect is character-

ed in electron-volts.

10

In the energy diagram for a metal shownin Fig. 3, the work function represents theenergy which must be given to an electron atthe top of the energy distribution to raise itto the level of the potential barrier at themetal-vacuum interface.

METAL

FERMIENERGY

92CS - 32289

Fig. 3 - Energy mode/ for a metal showingthe relationship of the work function and theFermi level.

According to the quantum theory, onlyone electron can occupy a particular quan-tum state of an atom. In a single atom, thesestates are separated in distinct “shells”; nor-mally only the lower energy states are filled.In an agglomeration of atoms, these statesare modified by interaction with neighboringatoms, particularly for the outermost elec-trons of the atom. As a result, the outerenergy levels tend to overlap and produce acontinuous band of possible energy levels, asshown in Fig. 3.

The diagram shown in Fig. 3 is for atemperature of absolute zero; all lowerenergy levels are filled. As the temperature isincreased, some of the electrons absorb ther-mal energy which permits them to occupy

scattered states above the maximum level forabsolute zero. The energy distribution ofelectrons in a particular metal is shown inFig. 4 for several different temperatures. Atabsolute zero, all the lower states are oc-cupied up to the Fermi level. At highertemperatures, there is some excitation to up-per levels. The electron density at a par-ticular temperature is described by theFermi-Dirac energy-distribution function,which indicates the probability of occupa-tion f for a quantum state having energy E:

f = 1 (2)

When E is equal to Ef, the value of f is 1/2.It is customary to refer to the energy of levelEf, for which there is a 50-per-cent prob-ability of occupancy, as the Fermi level. Atabsolute zero, the Fermi level corresponds tothe top of the filled energy distribution.

If the energy derived from the radiantenergy is just sufficient to eject an electron atthe Fermi level, the following relation exists:

(3)

radiation, is related to the long-wavelength

(4)

The relationship may be rewritten to relatethe long-wavelength limit to the work func-tion, as follows:

(5)

Because some of the electrons occupystates slightly higher than the Fermi level, asshown in Fig. 4, excitation of these electronsproduces an extended response at the redthreshold of the spectral-response character-istic. As a result, there is no abrupt redthreshold at normal temperatures, and thetrue work function cannot be obtained in asimple manner from the spectral-responsemeasurement. However, a universal func-tion devised by Fowler can be used to predictthe shape of the spectral-response curve nearthe threshold; the work function can then becalculated from these data.

Photomultiplier Design

fig. 4 - Energy distribution of conductionelectrons in potassium at temperatures of 0,200, and 1033 degrees Kelvin based onelementary Sommerfeld theory. (ref. 19)

Work Function and Spectral ResponseMeasurement of the work function and

spectral response for clean metal surfaceshas been of considerable importance in thedevelopment of photoelectric theory.

Work functions for pure metals are in therange 2 to 5 electron-volts. (See A.H. Som-mer, Ref. 11, Table 3.)

Fig. 5 shows spectral-response curves forthe alkali metals. The curves indicate aregular progression of the wavelength formaximum response with atomic number.The most red-sensitive of these metals iscesium, which is widely used in the activa-tion of most commercial phototubes.

The energy distribution of emitted photo-electrons has been measured for a number ofmetals and photosurfaces. Typical resultsare shown in Fig. 6 for a potassium film of20 molecular layers on a base of silver.20 Themaximum emission energy corresponds tothat predicted by the Einstein photoelectricequation.

Quantum EfficiencyBecause a quantum of radiation is neces-

sary to release an electron, the photoelectriccurrent is proportional to the intensity of theradiation. This first law of photoelectricityhas been verified experimentally over a widerange of light intensities. For most materials,the quantum efficiency is very low; on thebest sensitized commercial photosurfaces,the maximum yield reported is as high as oneelectron for three light quanta.

11


WAVELENGTH-NANOMETERS

92CS-32291

Fig. 5 - Spectral-response characteristics for the alkali metals showinggression in the order of the periodic table. (ref. 19)

An ideal photocathode has a quantum ef-ficiency of 100 per cent; i.e., every incidentphoton releases one photoelectron from thematerial into the vacuum. All practicalphotoemitters have quantum efficienciesbelow 100 per cent. To obtain a qualitativeunderstanding of the variations in quantumefficiency for different materials and for dif-ferent wavelengths or photon energies, it isuseful to consider photoemission as a pro-cess involving three steps: (1) absorption of aphoton resulting in the transfer of energy

Fig. 6 - Energy distribution of photoelec-trons from a potassium film. (ref. 20)

12

regular pro-

from photon to electron, (2) motion of theelectron toward the material-vacuum inter-face, and (3) escape of the electron over thepotential barrier at the surface into thevacuum.

Energy losses occur in each of these steps.In the first step, only the absorbed portionof the incident light is effective and thuslosses by transmission and reflection reducethe quantum efficiency. In the second step,the photoelectrons may lose energy by col-lision with other electrons (electron scat-tering) or with the lattice (phononscattering). Finally, the potential barrier atthe surface prevents the escape of some elec-trons.

Metallic and Semiconductor MaterialsThe energy losses described vary from

material to material, but a major differencebetween metallic and semiconductingmaterials makes separate consideration ofeach of these two groups useful. In metals, alarge fraction of the incident visible light isreflected and thus lost to the photoemissionprocess. Further losses occur as the photo-electrons rapidly lose energy in collisions withthe large number of free electrons in themetal through electron-electron scattering.As a result, the escape depth, the distancefrom the surface from which electrons canreach the surface with sufficient energy toovercome the surface barrier, is small, and istypically a few nanometers. Finally, the

work function of most metals is greater thanthree electron-volts, so that visible photonswhich have energies less than three electron-volts are prevented from producing electronemission. Only a few metals, particularly thealkali ones, have work-function values lowenough to make them sensitive to visiblelight. Because of the large energy losses inabsorption of the photon and in the motionof the photoelectron toward the vacuum (thefirst and second steps described above), eventhe alkali metals exhibit very low quantumefficiency in the visible region, usually below0.1 per cent (one electron per 1000 incidentphotons)l As expected, higher quantum effi-ciencies are obtained with higher photonenergies. For 12-electron-volt photons,quantum yields as high as 10 per cent havebeen reported.

Fig. 7 - Simplified semiconductor energyband model.

The concept of the energy-band modelsthat describe semiconductor photoemitters isillustrated in its simplest form in Fig. 7. Elec-trons can have energy values only within welldefined energy bands which are separated byforbidden-band gaps. At 0 K, the electronsof highest energy are in the so-called valenceband and are separated from the empty con-duction band by the bandgap energy EG.The probability that a given energy level maybe occupied by an electron is described byFermi-Dirac statistics and depends primarilyon the difference in energy between the levelunder consideration and Fermi level. As afirst approximation, it may be said that anyenergy levels which are below the Fermi levelwill be filled with electrons, and any levelswhich are above the Fermi level will be emp-


ty. At temperatures higher than 0 K, someelectrons in the valence band have sufficientenergy to be raised to the conduction band,and these electrons, as well as the holes in thevalence band created by the loss of electrons,produce electrical conductivity. Because thenumber of electrons raised to the conductionband increases with temperature, the con-ductivity of semiconductors also increaseswith temperature. Light can be absorbed byvalence-band electrons only if the energy ofthe photon is at least equal to the band-gapenergy EG. If, as a result of light absorption,electrons are raised from the valence bandinto the conduction band, photoconductivityis achieved. For photoemission, an electronin the conduction band must have energygreater than the electron affinity EA. Theadditional energy EA is needed to overcomethe forces that bind the electron to the solid,or, in other words, to convert a “free” elec-tron within the material into a free electronin the vacuum. Thus, in terms of the modelof Fig. 7, radiant energy can convert an elec-tron into an internal photoelectron (photo-conductivity) if the photon energy exceedsEG and into an external photoelectron(photoemission) if the photon energy ex-ceeds (EG + EA). As a result, photons withtotal energies Ep less than (EG + EA) cannotproduce photoemission.

The following statements can therefore bemade concerning photoemission in semicon-ductors. First, light absorption is efficient ifthe photon energy exceeds EG. Second,energy loss by electron-electron scattering islow because very few free electrons are pres-ent; thus, energy loss by phonon scattering isthe predominant loss mechanism. Theescape depth in semiconductors is thereforemuch greater than in metals, typically of theorder of tens of nanometers. Third, thethreshold wavelength, which is determinedby the work function in metals, is given bythe value of (EG + EA) in semiconductors.Synthesis of materials with values ofEG + EA) below 2 electron-volts has

demonstrated that threshold wavelengthslonger than those of any metal can be ob-tained in a semiconductor. Semiconductors,therefore, are superior to metals in all threesteps of the photoemissive process: they ab-sorb a much higher fraction of the incidentlight, photoelectrons can escape from a

13


greater distance from the vacuum interface,and the threshold wavelengths can be madelonger than those of a metal. Thus, it is notsurprising that all photoemitters of practicalimportance are semiconducting materials.Negative-Electron-Affinity Materials 20a

In recent years, remarkable improvementsin the photoemission from semiconductorshave been obtained through deliberatemodification of the energy-band structure.The approach has been to reduce the elec-tron affinity, EA, and thus to permit theescape of electrons which have been excitedinto the conduction band at greater depthswithin the material. Indeed, if the electronaffinity is made less than zero (the vacuumlevel lower than the bottom of the conduc-tion band, a condition described as“negative electron affinity” and illustratedin Fig. 8), the escape depth may be as much

92CS-32294

Fig. 8 - Semiconductor energy-band modelshowing negative electron affinity.

as 100 times greater than for the normalmaterial. The escape depth of a photoelec-tron is limited by the energy loss suffered inphonon scattering. Within a certain periodof time, of the order of 10-12 second, theelectron energy drops from a level above thevacuum level to the bottom of the conduc-tion band from which it is not able to escapeinto the vacuum. On the other hand, theelectron can stay in the conduction band inthe order of 10-10 second without furtherloss of energy, i.e., without dropping intothe valence band. If the vacuum level isbelow the bottom of the conduction band,the electron will be in an energy state fromwhich it can escape into the vacuum for aperiod of time that is approximately 100times longer than if an energy above the bot-tom of the conduction band is required forescape, as in the materials represented byFig. 7. Therefore, a material conforming to

the conditions of Fig. 8 has greatly increasedescape depth. Under such circumstances, thephotosensitivity is significantly enhanced.Substantial response is observed even forphotons with energies close to that of theband gap where the absorption is weak. Effi-cient photoemission in this case results onlybecause of the greater escape depth.

The reduction of the electron affinity isaccomplished through two steps. First, thesemiconductor is made strongly p-type bythe addition of the proper “doping” agent.For example, if gallium arsenide is the hostmaterial, zinc may be incorporated into thecrystal lattice to a concentration of perhaps1,000 parts per million. The zinc producesisolated energy states within the forbiddengap, near the top of the valence band, whichare normally empty, but which will acceptelectrons under the proper circumstances.The p-doped material has its Fermi level justabove the top of the valence band. The se-cond step is to apply to a semiconductor asurface film of an electropositive materialsuch as cesium. Each cesium atom becomesionized through loss of an electron to ap-type energy level near the surface of thesemiconductor, and is held to the surface byelectrostatic attraction.

The changes which result in the energy-band structure are two-fold. In the firstplace, the acceptance of electrons by thep-type impurity levels is accompanied by adownward bending of the energy bands.This bending can be understood by observ-ing that a filled state must be, in general,below the Fermi level; the whole structurenear the surface is bent downward to ac-complish this result. In the second place, thepotential difference between the chargedelectropositive layer (cesium) and the bodycharge (filled zinc levels) results in a furtherdepression of the vacuum level as a result ofa dipole moment right at the surface.

Another way to describe the reduction ofthe electron affinity is to consider the surfaceof the semiconductor as a capacitor. Thecharge on one side of the capacitor isrepresented by the surface layer of cesiumions; the other charge is represented by theregion of filled acceptor levels. The reduc-tion in the electron affinity is exactly equalto the potential difference developed acrossthe capacitor.

14

In a more rigorous analysis, the amountby which the energy bands are bent is foundto be approximately equal to the band-gap,and the vacuum level is lowered until the ab-sorption level of the electropositive materialis essentially at the top of the valence band.

PRACTICAL PHOTOCATHODEMATERIALS

Research on commercially useful photo-emitters has been directed primarily towarddeveloping devices sensitive to visible radia-tion. The first important commercial photo-surface was silver-oxygen-cesium. This sur-face, which provides a spectral responsedesignated S-l, is sensitive throughout theentire visible spectrum and into the infrared.Although it has rather low sensitivity andhigh dark emission, the good response in thered and near-infrared still recommends itsuse in special applications although otherphotocathodes are more generally used inphotomultipliers today.

200 4 0 0 600 800 1000

WAVELENGTH - NANOMETERS

92CM-32295

Fig. 9 - Typical spectra/-response curves,with 0080 lime-glass window for (a) silver-oxygen-cesium (Ag-O-Cs), (b) cesium-antimony (Cs3Sb), (c) multialkali or trialkali(Na2KSb:Cs*The terminology “:Cs” indicates trace quantities ofthe element.


The photocathodes most commonly usedin photomultipliers are cesium-antimony(Cs3Sb), multialkali or trialkali (Na2KSb:Cs),* and bialkali (K2CsSb). Anotherbialkali photocathode (Na2KSb) is par-ticularly useful at higher operating tempera-tures because of its stability. Recently, therubidium-cesium-antimony (probably Rb2Cs Sb) photocathode has been introducedbecause of its favorable blue sensitivity.Typical spectral response curves for thesematerials are shown in Figs. 9 and 10. Addi-tional information about these and otherphotocathodes of practical importance isshown in Table I.

3 0 0 4 0 0 5 0 0 6 0 0 7 0 0

WAVELENGTH -NANOMETERS

92CM-32296

Fig. 10 - Typical spectral-response curvesfor various photocathodes useful in scintilla-tion counting applications. The variation inthe cutoff at the low end is due to the use ofdifferent envelope materials.

The long-wavelength response of the mul-tialkali photocathode has been extended byprocessing changes including the use of anincreased photocathode-film thickness at theexpense of the short-wavelength response.Fig. 11 shows two typical spectral-responsecurves of the ERMA types (Extended RedMulti-Alkali) II and III compared with theS-20 response of the conventionally pro-cessed multialkali photocathode.

15


Table INominal Composition and Characteristics of

Various Photocathodes

NominalComposition

Cs3Sb 0

Cs3Sb 0

Cs3Sb S

Na2KSb S

Na2KSb S

K2CsSb 0

K2CsSb 0

K2CsSb S

K2CsSb S

K2CsSb S

K2CsSb S

Rb2CsSb S

Na2KSb:Cs 0

Na2KSb:Cs S

Na2KSb:Cs S

Na2KSb:Cs S

Na2KSb:Cs Sd

Na2KSb:Cs S

GaAs: Cs-0 O e

Photo-cathode

JEDECEnvelopeMateriala

ResponseDesignation

Conversion LuminousFactorb Respon-

sivitv(lumen/watt) (uA/lumen) Response (mA/watt) (percent)

Wave-length of Respon-Maximum sivitv

QuantumEfficiency

DarkEmissionat 25° C(fA/cm2)

0080 S-4 950 40 380 38 12

9741 S-5 1244 40 340 50 18

0080 S-11 857 70 400 60 19

7056 S-24 1250 40 380 50 16

Sapphire 1400 43 380 60 19

0080 938 60 380 56 18

7740 1083 60 400 65 20

B270 1111 90 380 100 33

0080 1120 80 380 90 29

7740 1140 71 420 82 24

9741 1240 56 380 70 230080 948 100 420 95 28

9741 510 75 380 38 12

7740 ERMA IIIc 160 180 575 29 6

0080 ERMA IIc 250 200 550 50 11

0080 S-20 480 135 390 65 21

0080 230 300 530 70 16

7056 432 117 420 51 15

9741 115 720 800 80 12

a Numbers refer to the following glasses:0080 - Corning Lime Glass9741 - Corning Ultraviolet Transmitting Glass7056 - Coming Borosilicate Glass7740 - Corning Pyrex GlassB270 - Schott BK270

b These conversion factors are the ratio of the radiantresponsivity at the peak of the spectral response char-acteristic in amperes per watt to the luminous respon-sivity in amperes per lumen for a tungsten lamp oper-ated at a color temperature of 2856 K.

0.2

0.3

3

.0003

.0003

.02

.02

.02

.02

.08

.3

0.4

1.2

92.

c A BURLE designation for “Extended-Red Multial-kali.”

d Reflecting substrate.

e Single crystal.

0 = Opaque

S = Semitransparent

16


The negative-electron-affinity materialsdescribed earlier are used in both opaqueand semitransparent photocathodes. Spec-tral response curves for GaAs:Cs-O andInGaAs:Cs-O are shown in Fig. 12. Therehas been considerable interest in detectors

92CS-32298

Fig. 12 - Spectra/ response characteristics Fig. 13 - Spectral absorptance of the Ag-O-Cs and the K2CsSb semitransparent photo-

pared with the S-1 characteristic (Ag-O-Cs). cathodes.

for 1060 nm, the wavelength of the Nd:YAGlaser. For comparison of sensitivities at thiswavelength, the spectral response of the Ag-0-Cs photocathode is also shown. TheInGaAs:Cs-O photocathode has the higherresponsivity at 1060 nm. The NEA photo-cathodes are generally fairly small comparedwith the large semitransparent photocath-odes used for scintillation counting. Stabili-ty, especially for the longer-wave-lengthNEA photocathodes, is a problem unless thephotomultiplier output current are kept low.

OPAQUE AND SEMITRANSPARENTPHOTOCATHODES

Photocathodes may be classified asopaque or semitransparent. In the opaquephotocathode, the light is incident on a thickphotoemissive material and the electrons areemitted from the same side as that struck bythe radiant energy. In the second type, thesemitransparent photocathode, the photo-emissive material is deposited on atransparent medium so that the electrons areemitted from the side of the photocathodeopposite the incident radiation.

Because of the limited escape depth ofphotoelectrons, the thickness of thesemitransparent photocathode film is

8 0

0 200 4 0 0 6 0 0 800 1000 1200

WAVELENGTH -NANOMETERS

92cs -32299

17


critical. If the film is too thick, much of theincident radiant energy is absorbed at adistance from the vacuum interface greaterthan the escape depth; if the film is too thin,much of the incident radiant energy is lost bytransmission.

02 0 0 300 4 0 0 5 0 0 6 0 0 7 0 0 8 0 0 9 0 0

WAVELENGTH- NANOMETERS

9 2 C S - 3 2 3 0 0

Fig. 14 - Spectral absorption coefficient data

tocathodes.

The radiant spectral flux absorption of asemitransparent photocathode varies withwavelength as illustrated in Figs. 13 and 14.The ordinate in Fig. 13 is absorptance whichis defined as the ratio of the radiant flux ab-sorbed by the layer to that incident upon it.(Absorptance plus reflectance plus transmit-tance add to unity.) The data shown in Fig.14 are spectral absorption coefficients. The

of the photocathode layer is given by

(6)

that the data in Fig. 14 are given in units ofmicrometers - l so that in Eq. 6, d must begiven in micrometers. Typical thickness of aNa2KSb:Cs photocathode is about 0.030

Thus, at 400 nm the photocathode ab-sorbs 87% of the flux which is not reflected.The spectral response of semitransparentphotocathodes can be controlled to some ex-tent by thickness variation. With increasingthickness in the case of alkali antimonides,blue response decreases and red response in-creases.

WAVELENGTH - NANOMETERS9 2 C S - 3 2 3 0 1

Fig. 15 - Ultraviolet transmittance cut off of various glasses and crystals used inphotomultiplier photocathode windows. Data are all for 1 mm thickness.

18

GLASS TRANSMISSION ANDSPECTRAL RESPONSE

Although photocathode spectral responseis determined primarily by the nature of thephotocathode surface, especially in the visi-ble and long wavelength cut-off regions, theshort-wave length cut-off characteristic ofall photocathodes is determined by thetransmission of the window to the photo-cathode. Utraviolet cut-off characteristicsof a number of glasses or crystals which havebeen used in photomultiplier fabrication areshown in Fig. 15.

tion. Photomultiplier tubes made with fusedquartz windows are useful in Cerenkovcounting applications where the spectralenergy distribution increases with decreasing

terval. Fused quartz is also a useful materialin liquid scintillation counting because of itsminimum contamination with 40K which cancause unwanted background counts.

The data presented in this figure are all for1 mm thickness. The data also include lossesfrom reflection. For most glasses with an in-dex of refraction of about 1.5, the reflectionloss is about 4% at each surface. The loss ishigher in high index-of-refraction materialsuch as sapphire. Although some photo-multipliers, especially those of small sizemay have window thickness of 1 mm or less,larger face plates are generally thicker inorder to provide adequate strength. Thetransmittance varies with thickness accord-ing to the following relationship

Less expensive than synthetic fused quartzis Corning 9741 glass, which is frequentlyused in photomultipliers designed for thenear ultraviolet. It is a Kovar-sealing glassbut has the disadvantage of possibleweathering over long periods of exposure tothe atmosphere.

Another glass for the near ultraviolet isCorning 9823, which seals to 0120 lead glassand which can be used with Dumet metal

leads. This glass is somewhat inferior to 9741at the shortest wavelengths, but is better forwavelengths longer than 240 nm.

(7)

where k is a factor (approximately 0.92 formost glasses) dependent upon the surface

tion, and t is the thickness.

Glass type 7056 is selected for its good op-tical quality. It is a hard glass which seals to7052 and Kovar. Pyrex type 7740 is selectedprimarily for its low content of 40K and isused in liquid scintillation-counting applica-tions. Lime glass is the least expensive; it is asoft glass which seals to lead glass, type0120.

The window extending the furthest intothe ultraviolet is LiF. A few tubes are madewith this material but, because its fabrica-tion is difficult, LiF face plates are only usedfor special applications of spectroscopy.Transmission extends to wavelengths shorterthan the Lyman-alpha limit of 121.5 nm.

Corning 9025 is a special non-browningglass. It is doped with cerium and resistsdarkening from exposure to ionizing radia-tion. One application is in satellites whichmust pass through space regions of high-intensity ionizing radiation.

THERMIONIC EMISSION

Sapphire windows (ultraviolet grade) aregood down to about 150 nm and are easier touse than LiF. Sapphire can be sealed toKovar by a metalizing and brazing tech-nique. Special photomultiplier tubes havingsapphire windows were used in the HEAOprogram (High Energy Astronomical Obser-vatory).

Current flows in the anode circuit of aphotomultiplier tube even when it isoperated in complete darkness. The dc com-ponent of this current is called the anodedark current, or simply the dark current.This current and its resulting noise compo-nent usually limit the lower level ofphotomultiplier light detection. As a result,the anode dark-current value is nearlyalways given as part of the data for any tube.

Except for the extended short wavelength There are several sources of dark currentcutoff of sapphire, Suprasil, an ultraviolet in a photomultiplier tube: ohmic leakage,grade of synthetic fused quartz, has a better thermionic emission, and regenerative ef-transmission characteristic. Sapphire suffers fects. Ohmic leakage may result from con-some loss in transmission by reflection taminations on the insulators within thebecause of its relatively high index of refrac- tube, on the outside of the tube envelope, or

19



on the base. Thermionic emission generallyoriginates from the photocathode itself andis amplified by the gain of the multiplier sec-tion. Some emission may also come from thesecondary-emission dynode surfaces. Regen-erative effects can occur in the tube par-ticularly if it is operated with high voltageand high gain. (Regenerative effects are dis-cussed in more detail in a later section onDark Current and Noise.) The following dis-cussion relates to the origin of thermionicemission current and gives practical valuesfor this current in photomultiplier photo-cathodes.

In a metal, the electrons which escape asthermionic emission are generally from thetop of the conduction band (see Fig. 3).Thus, the work functions for photoemissionand for thermionic emission are the same.Thermionic emission as a function of work

ture T in degrees Kelvin is given by thefamiliar Richardson equation:

where j is the thermionic current density; e,the electron charge; m, the electron mass; k,Boltzman’s constant; and h, Planck’s con-stant. If the constants before the exponentialexpression are given in mks units, the equa-tion (8) expressed in amperes per meter2

becomes

For semiconductor photocathodes, thework functions of photoemission and ther-mionic emission may be quite different. Thework function for photoemission (see Fig. 7)is the electron energy corresponding to theheight from the top of the valence band tothe vacuum level, or EA (the electron affini-ty) plus EG (the forbidden gap, i.e., the

*An electron volt is the energy acquired by an electronin being accelerated through a drop in potential of onevolt. In equations such as (8), the value of kT in the ex-ponent must also be expressed in electron volts. It is the

in volts must be multiplied by the electron charge, e, incoulombs (1.6 x 10 - 19) to obtain the work function injoules. The value of kT may then also be expressed injoules.

separation of valence and conductionbands). For an intrinsic semiconductor, ther-mionic emission originates from the valenceband, as does photoemission, but the “workfunction” is not the same as for photoemis-sion. In the case of an intrinsic semiconduc-tor, thermionic-emission density can be ex-pressed as

For an impurity semiconductor wherethermionic emission originates from the im-

IOI

Fig. 16 - Variation of thermionic-emissioncurrent density from various photocathodesused in photomultiplier tubes as a function ofreciprocal temperature. Thermionic emissionmultiplied by the gain of the photomultiplieris a principal source of anode dark current.

20

purity centers, the equation for thermionicemission may be written as follows21:

where EF is the Fermi level energy referencedfrom the top of the valence band, and no isthe impurity concentration.

Typical dark-emission current-densitycharacteristics for various photocathodes areshown in Fig. 16 as a function of reciprocaltemperature. The current density is plottedon a logarithmic scale to show theexponential-like character of the emission.Anode dark current of the photomultiplierresults from the cathode emission multipliedby the gain of the tube. Thermionic darkemission varies from tube to tube of thesame type, probably because of the variationin no. Note that some of the curves showsubstantial curvature at the low temperatureend. Some of this curvature is explained bythe variation of T in the T2/ T3/4 = T5/4

term. But an explanation for the greater partof the curvature may be the presence ofpatches of different impurity level or con-centration, or it may be that the impurityconcentration itself is a function of tempera-ture. Exposure to temperatures above thenormal operating range sometimes results inpermanent reduction in dark current. Mostphotocathodes are p-type semiconductorsone result of which is a lower dark emissionthan for n-type semiconductors because ofthe reduced Fermi-level energy.

SECONDARY EMISSIONWhen electrons having sufficient kinetic

energy strike the surface of a material,secondary electrons are emitted. The

fined as follows:

(12)

where NS is the average number of secondaryelectrons emitted for Ne primary electronsincident upon the surface.


Fundamentals of Secondary EmissionThe physical processes involved in secon-

dary emission are in many respects similar tothose already described under Photoemis-sion. The main difference is that the impactof primary electrons rather than incidentphotons causes the emission of electrons.The steps involved in secondary emission canbe stated briefly as follows:

1. The incident electrons interact withelectrons in the material and excite them tohigher energy states.

2. Some of these excited electrons movetoward the vacuum-solid interface,

3. Those electrons which arrive at the sur-face with energy greater than that repre-sented by the surface barrier are emitted intothe vacuum.

When a primary beam of electrons im-pacts a secondary-emitting material, theprimary-beam energy is dissipated within thematerial and a number of excited electronsare produced within the material. Thenumbers of excited electrons produced areindicated in Fig. 17 for primary energiesvarying from 400 to 2200 electron-volts. Thetotal number of excited electrons producedby a primary is indicated by the area of theindividual rectangles in the figure. These ap-proximate data are based on experimentaldata assuming that the range of primaryelectrons varies as the 1.35 power of theprimary energy and that the number of elec-trons excited is uniform throughout theprimary range.

0.5

Fig. 17 - The processes of secondarysion. See text for explanation.

-NANOMETERS92cs-32303

emis-

21


As an excited electron in the bulk of thematerial moves toward the vacuum-solid in-terface, it loses energy as a result of colli-sions with other electrons and opticalphonons. The energy of the electron is veryrapidly dissipated as a result of these col-lisons, and it is estimated that the energy ofsuch an electron will decay to within a fewtimes the mean thermal energy above thebottom of the conduction band within10-12 second. If the electron arrives at thevacuum-solid interface with energy belowthat required to traverse the potential bar-rier, it cannot escape as a secondary elec-tron. Therefore, only those electrons excitednear the surface of the material are likely toescape as secondary electrons. The probabil-ity of escape for an excited electron is as-sumed to vary exponentially with the excita-tion depth, as indicated in Fig. 17. If the pro-duct of the escape function and the numberof excited electrons (which is a function ofprimary energy and depth) is integrated, a

obtained, as indicated in the insert at the topof Fig. 17. The model which has been as-s u m e d t h u s e x p l a i n s t he gene ra lcharacteristics of secondary emission as afunction of primary energy. Secondary-emission yield increases with primaryenergy, provided the excited electrons areproduced near the surface where the escapeprobability is high. As the primary-electronenergy increases, the number of excited elec-trons also increases, but the excitation oc-curs at greater depths in the material whereescape is much less probable. Consequently,the secondary-emission yield eventuallyreaches a maximum and then decreases withprimary energy.Secondary Emitter Materials

Experimental secondary-emission-yieldvalues are shown as a function of primary-electron energy in Fig. 18 for MgO, a tradi-tional secondary-emission material, and forGaP:Cs, a recently developed negative-electron-affinity material. Also shown is acalculated curve from Simon and Williams13

based on their model of the GaP:Cs emitter.Although both MgO and GaP:Cs display thegeneral characteristics of secondary emissionas a function of primary energy, as expectedfrom the model illustrated in Fig. 17, thesecondary-emission yield for GaP:Cs in-

22

creases with voltage to much higher values.Even though electrons are excited ratherdeep in the negative-electron-affinitymaterial and lose most of their excess energyas a result of collisions, many still escape in-to the vacuum because of the nature of thesurface barrier.

PRIMARY ENERGY - keV92cs -32304

Fig. 18 - Typical experimental curve of sec-ondary-emission yield as a function of .primary-electron energy in GaP:Cs and MgO.Also shown in calculated curve for GaP:Csfrom Simon and Williams.13

Because the GaP:Cs material is more dif-ficult to handle than more conventionalsecondary emitters, and, therefore, results inhigher cost, its use as a dynode has beenrestricted to applications where the very highsecondary emission is particularly advan-tageous. It is used as, for example, inphotomultiplier applications benefitting bythe reduction in statistical noise, or in thedesign of photomultipliers having fewerstages for a given amplification. The use offewer stages also reduces the variation ofgain with voltage changes.

In the development of photomultipliertubes it has been found that the photocath-ode material may also be useful as a secon-dary emitter22. Such was the case in some ofthe first photomultipliers developed whichused a Ag-O-Cs photocathode and Ag-O-Csdynodes. Because of its high dark-emissioncurrent and its instability, especially atmoderate current-densitymaterial is no longer used.

levels, this

Other photocathode materials which alsoserve as secondary emitters are Cs

3Sb, Rb-

Cs-Sb, K2CsSb, and Na2KSb:Cs. Because

the processing of these secondary emitters is


not identical in most cases to that of the cor-responding photocathode, the particularchemical formulations specified here maynot be accurate. Secondary emission ratiosfor these and other materials are shown inFig. 19.

Fig. 19 - Secondary emission ratios for anumber of materials which have been used asdynodes in photomultipliers as a function ofaccelerating voltage of the primary electrons.

Very high secondary-emission yields havebeen reported23,24 fo r Na 2KSb:Cs , themulti-alkali photocathode (S-20 response).Some photomultiplier with thismaterial as the secondary emitter althoughits processing is complex. The very highyields, particularly at high primary energies,would suggest that the material has an effec-tive negative electron affinity similar to thatof GaP:Cs. This explanation may also holdtrue for K-Cs-Sb, which is used is some tubeshaving this type of photocathode. A materialthat has been very commonly used is Cs3Sbcorresponding to the photocathodes with S-4or S-l1 spectral responses. It has goodsecondary emission in the practical workingrange near 100 volts. Rb-Cs-Sb is a rathernew material which is just coming into usebecause the corresponding photocathode hasgood properties. All of the alkali an-timonides mentioned here have limitations.They cannot tolerate exposure to air andthey are damaged by temperatures in excessof 75 degrees C. In addition, stability suffers

c m -2.

A very practical secondary emitter can bemade from an oxidized silver-magnesiumalloy25,26 containing approximately 2 percent of magnesium. Although silver-magnesium dynodes do not have as high asecondary-emission ratio as some of thematerials mentioned above (see Fig. 18), thematerial is easily processed and is morestable at relatively high currents. In addi-tion, it can tolerate higher temperatures.This surface has a low thermionicbackground emission which is important inapplications requiring detection of low-levellight. When it is activated with cesium, gainis somewhat higher. Without the cesium ac-tivation, the oxygen-activated silver-magnesium layer has been used effectively indemountable systems for detecting ions andother particles.

A material having characteristics verysimilar to those of silver-magnesium is anoxidized layer of copper-berylliumalloy27,28,29in which the beryllium compo-nent is about 2 per cent of the alloy. Secon-dary emission is usually enhanced by thebake-out in cesium vapor. A secondary-emission characteristic of the cesium-activated copper-beryllium material is shownin Fig, 19. Because of the advantages inhandling and the manufacturing cost, thecopper-beryllium is largely taking the placeof silver-magnesium in applications requir-ing low dark emission and stability atrelatively high current densities. The lowersecondary-emission yield is usually compen-sated for in photomultiplier design by ap-plication of higher voltage or by an increasein the number of dynode stages.

92CS - 32306

Fig. 20 - Typical secondary-electron energydistribution; peak at right is caused by re-flected primary electrons.

23


When secondary electrons are emitted intothe vacuum, the spread of emission energiesmay be quite large, as illustrated in the curveof Fig. 20 for a positive-electron-affinityemitter. The peak at the right of the curvedoes not represent a true secondary, butrather a reflected primary. Data are notavailable for the emission energies from anegative-electron-affinity material, but theyare expected to be considerably less than forpositive-affinity materials.

TIME LAG IN PHOTOEMISSION ANDSECONDARY EMISSION

Because both photoemission and secon-dary emission can be described in terms ofthe excitation of electrons within the volumeof the solid and the subsequent diffusion ofthese electrons to the surface, a finite timeinterval occurs between the instant that aprimary (photon or electron) strikes a sur-face and the emergence of electrons from thesurface. Furthermore, in the case of secon-dary emission, the secondaries can be ex-pected to reach the surface over a period oftime. Within the limitations of a mechanisticapproach to a quantum phenomenon, timeintervals for metals or insulators of the orderof 10-13 to 10-14 second may be estimatedfrom the known energy of the primaries,their approximately known range, and theapproximately known diffusion velocities ofthe internal electrons. In negative-electron-affinity semiconductors, it is known that thelifetime of internal “free” electrons havingquasi-thermal energies (i.e., electrons nearthe bottom of the conduction band) can beof the order of 10-10 second.

Thus far, experiments have provided onlyupper limits for the time lag of emission. Inthe case of secondary emission, a variety ofexperiments have established limits. Severalinvestigators30,31,32

have deduced limitsfrom the measured performance of electrontubes using secondary emitters. Others,making direct measurements of these limits,have determined the time dispersion ofsecondary emission by letting short electronbunches strike a target and comparing theduration of the resulting secondary buncheswith the measured duration of the primarybunch. Bythis means an upper limit of6x10-l 2 second was determined forplatinum33 and an upper limit of 7 x 10-11 1

24

second for an MgO layer34 formed on thesurface of an AgMg alloy.

The upper limit for the time lag in photo-emission, however, is not well established.From careful measurements of the time per-formance of fast photomultipliers it can beinferred that the limit must be less than10 - 10 seconds.

While these limits are a useful guide to thetype of time performance to be expected inpresent photomultipliers, they will probablyhave less significance as photomultipliers us-ing new semiconducting photoemitters andsecondary emitters are developed. Semicon-ductors having minority-carrier lifetimes ofthe order of microseconds are now available.Probably, by combination of this character-istic with negative electron affinity, highergains and quantum efficiencies can beachieved, but at a sacrifice of time responseor band width. However, the first generationof negative-electron-affinity emitters (e.g.,Gap) has actually resulted in photomulti-pliers having better time performancebecause a smaller number of stagesoperating at higher voltage can be used. Atthis time it can only be concluded that in thefuture photomultipliers will probably bedesigned to match in more detail the re-quirements of a particular use.

REFERENCES

19. V.K. Zworykin, and E.G. Ramberg,Photoelectricity and its Application, JohnWiley and Sons, Inc., New York, (1949).20a. H. Rougeot and C. Baud, “NegativeElectron Affinity Photoemitters,” Advancesin Electronics and Electron Physics, Vol. 48,Edited by L. Marton, Academic Press, 1979.20. J.J. Brady, “Energy Distribution ofPhotoelectrons as a Function of theThickness of a Potassium Film,” Phys.Rev., Vol. 46, (1934).21. D.A. Wright, Semi-conductors,Methuen and Co., New York (1955).22. A.H. Sommer, “Relationship betweenphotoelectric and secondary electron emis-sion, with special reference to the Ag-O-Cs(S-l) photocathode,” J. Appl. Phys., Vol.42, pp. 567-569, (1971).23. A.A. Mostovskii, G.B. Vorobeva, andG.B. Struchinskii, Soviet Physics-SolidState, Vol. 5, p. 2436, (1964).

24. C. Ghosh and B.P. Varma, “Secon-dary emission from multialkali photo-cathodes,” J. Appl. Phys., Vol. 49, pp.4554-4555, (1978).25. V.K. Zworykin, J.E. Ruedy, and E.W.Pike, “Silver-magnesium alloy as a secon-dary emitting material,” J. Appl. Phys.;Vol. 12, (1941).26. P. Rappaport, “Methods of processingsilver-magnesium secondary emitters forelectron tubes,” J. Appl. Phys., Vol. 25, pp288-292, (1954).27. J.S. Allen, “An improved electronmultiplier particle counter,” Rev. Sci. Instr.,Vol. 18, (1947).28. A.M. Tyutikov, “Structure and secon-dary emission of emitters of activatedberyllium bronze, " Radio Engineering andElectronic Physics, (Translated from theRussian and published by the IEEE), Vol. 8No. 4, pp. 725-734, (1963).


29. A.H. Sommer, “Activation of silver-magnesium and copper-beryllium dynodes,”J. Appl. Phys., Vol. 29, pp. 598-599, (1958).30. G. Diemer and J.L.H. Jonker, “Onthe time delay of secondary emission,”Philips Research Repts., Vol. 5, p. 161(1950).3 1. M.H. Greenblatt and P.A. Miller, Jr.,“A microwave secondary-electron multi-plier,” Phys. Rev., Vol. 72, p. 160 (1947).32. C.G. Wang, "Reflex Oscillators using.secondary-emission current,” Phys. Rev.,Vol. 68, p. 284 (1945).33. I.A.D. Lewis and F.H. Wells, Milli-Microsecond Pulse Techniques, PergamonPress (1959).34. K.C. Schmidt and C.F. Hendee,“Continuous-channel electron multiplieroperated in the pulse saturated mode,”IEEE Trans. Nucl. Sci., Vol. NS-13, No. 3,p. loo (1966).

25


3. Electron Optics of Photomultipliers

ELECTRON-OPTICAL DESIGNCONSIDERATIONS

One of the primary design considerationsin a photomultiplier tube is the shaping andpositioning of the dynodes (usually in a re-current geometrical pattern) so that all thestages are properly utilized and no electronsare lost to support structures in the tube ordeflected in other ways. Although it is notnecessary that the electrons come to a sharpfocus on each succeeding stage, the shape ofthe fields should be such that electrons tendto return to a center location on the nextdynode, even though the emission point isnot at the optimum location of the precedingdynode. If this requirement is not met, theelectrons increasingly diverge from thecenter of the dynode in each successivedynode stage. This effect in turn can lead tothe skipping of stages and loss of gain.Magnetic fields may be combined with elec-trostatic fields to provide the required elec-tron optics, although today most photomul-tipliers are electrostatically focused. Inaddition to providing good collection ofsecondary electrons from stage-to-stage, it isimportant for some applications to minimizethe time spread of electron trajectories. Forthis purpose is is useful to provide strongelectric fields at the surfaces of the dynodesto assure high initial acceleration of the elec-trons. Also important may be the design ofconfigurations which provide nearly equaltransit times between dynodes regardless ofthe point of emission on the dynode.

In scintillation counting applications, fair-ly large photocathode areas are required forefficient scintillator coupling. Ideally, thephotocathode should be semitransparentand located on the flat window of the tube.This requirement poses a special problem indesign of efficient photocathode-to-first-

26

dynode electron optics. If all of the emittedphotoelectrons are not guided properly tothe first dynode, the signal-to-noise ratio ofthe photomultiplier is degraded and poorerpulse-height-resolution characteristics result.

The region between the last dynode andthe anode is also of special electron-opticalconcern. The withdrawal field at the lastdynode should be large to minimize spacecharge effects which limit the linearity andmagnitude of output current pulses. Anotherconsideration in high-speed photomultipliersis to provide a structure which is matched toappropriate transmission lines.

The complete electron-optical configura-tion of the photomultiplier must also be suchas to avoid regenerative effects. For exam-ple, there should not be an open path in thetube through which occasional ions or lightcould feed back from the output end to thephotocathode.

DESIGN METHODS FOR PHOTO-MULTIPLIER ELECTRON OPTICS

Before the days of high-speed computers,photomultiplier electron-optical problemswere often solved by means of mechanicalanalogues such as a stretched rubber mem-brane. When mechanical models of the elec-trodes were placed under such a membraneand their height adjusted to correspond tothe desired electrical potential, the height ofthe membrane in the spaces between the elec-trodes corresponded to the equivalent elec-trical potential. Small balls were then al-lowed to roll from one electrode to the next.The trajectories of the balls corresponded tothose of the electrons in the photomultiplierstructure. With appropriate model design,friction and depression of the membrane by

Electron Optics of Photomultipliers

the ball were made negligible. This model,however, was only valid for geometries inwhich the electrodes could be assumed to becylindrical surfaces (generated by a lineparallel to a fixed direction and movingalong a fixed curve) sufficiently long to beconsidered infinite in extent. Applicationwas made to numerous dynode configura-tions.

Another useful analogue was the resis-tance network such as a two-dimensionalarray of connectors having equal spacingvertically and horizontally. This array repre-sented a cross-section plane through cylin-drical surfaces again assumed to be infinitein length. Resistors of equal value were con-nected between adjacent connectors bothvertically and horizontally. Points in the ar-ray corresponding to an electrode were allconnected to the same potential. Equipoten-tial lines between electrodes could then bedetermined by observing the potential valueson the connectors between the electrodes.When the equipotential lines had been deter-mined, electron trajectories could be readilycalculated between closely spaced equipoten-tial surfaces.

In another variation of the resistance net-work, the vertical distribution of resistancevalues was made logarithmic instead of uni-form. This array then corresponded to anaxially symmetrical system with the axis ap-proximated across the top of the board. Anexample of a relevant problem is the regionbetween the photocathode and first dynodefor end-on-type photomultipliers where theaxis passes through the center of the photo-cathode.

Only very simple electron-optical systemscan be solved in closed form, which requires

However, by using relaxation techniqueswith a computer, the potential distributionon a set of points confined within definedboundary conditions can be determined.Secondly, using the force equations, electronpaths can be traced. Computer programs ex-ist for solving various cases with or withoutsymmetry and with irregular potential boun-daries.

An electron-optical design for a photo-multiplier may be arrived at from computer-developed equipotential lines and electrontrajectories plotted on a plan showing the

electrode configurations. Collection efficien-cy and time response may be predicted froman analysis of the electron trajectories. Col-lection efficiency at the first dynode is de-fined as the ratio of the number of photo-electrons which land upon a useful area ofthe dynode to the number of emitted photo-electrons. If all the photoelectrons begintheir trajectories at the surface of the photo-cathode with zero velocity, 100% collectionwould be possible. Because of the finite in-itial velocities, however, some electronsbegin their trajectories with unfavorableangles of launch and are not collected on auseful area.

In modern photomultiplier structures,first-dynode collection efficiencies rangefrom 85 to 98 per cent. Ideally, the emittedphotoelectrons should converge to a verysmall area on the first dynode. In practice,this electron-spot diameter is usually lessthan 1/4 of the cathode diameter, dependingupon the tube type and focusing structure.

SPECIFIC PHOTOMULTIPLIERELECTRON-OPTICAL

CONFIGURATIONSCircular-Cage Structure withSide-On Photocathode

The first commercially successful photo-multiplier design was based on a circular ar-ray of photocathode and dynodes-as in the931A. This design is depicted schematicallyin Fig. 21. The photocathode is of the“opaque” type; i.e., electrons are emittedfrom the same side as the photocathode is il-luminated. This particular type of tube isrelatively inexpensive and is useful for ap-plications such as spectroscopy where highsensitivity is required but the photocathodearea need not be large. In this case the bestcollection is from the side of the photocath-ode near the first dynode. In the part of thephotocathode near the apex formed by thegrill and the photocathode, the electrostaticfields are small and collection efficiency ispoor. The time of response of the tube isshort because of its small size and high inter-dynode field strengths, and there is relativelysmall feedback from the anode end of thetube back to the photocathode.

27


Circular-Cage Structure withEnd-On Photocathode

Fig. 22 illustrates the use of a circular cagecoupled to a flat semi-transparent photo-cathode. Such a configuration is useful inscintillation counting where the flat photo-cathode is coupled to the scintillating crystal.In order to improve the collection efficiencyof photoelectrons by the first dynode, theelectrode which was the photocathode inFig. 21 has been modified to provide a larger

l-10: DYNODESI I : ANODE 92CS-32308

fig. 22 - An end-on photomultiplier structureutilizing a circular dynode arrangement. Thistype of tube would be useful as a detector forscintillation counting.

28

useful first-dynode area in the end-on con-struction. In addition, a skewed field is pro-vided by the focusing electrode which resultsin a more favorable pattern of impactingphotoelectrons on the first dynode. Thisend-on construction still has a relativelyshort time response because of the focuseddynode arrangement, although the transit-time spread of electrons traversing thephotocathode-to-first-dynode space in-creases the time of response as comparedwith that of the simple circular side-on struc-ture of Fig. 21.

Box-and-Grid Structure withEnd-On Photocathode

The box and grid dynode structure shownin Fig. 23 is also useful in scintillation count-ing because of the relatively large flat semi-transparent photocathode. This configura-tion has the advantage of providing a ratherlarge entrance area to collect photoelectronsso that collection efficiency is very nearly100%. This feature is important in providinggood pulse-height resolution in scintillationcounting. The individual dynode boxes areopen at the exit end and have a grid at the en-trance. The grid provides an electric fieldwhich penetrates the preceding dynode re-gion and aids in the withdrawal of secondaryelectrons. The grid also eliminates a retard-ing field that would be caused by the poten-tial of the preceding dynode. Because thefield penetration is rather weak, the electron


INCIDENT‘RADIATION

INCIDENTRADIATION

Fig. 23 - The box-and-grid multiplier structure.

FACEPLATE

INCIDENTRADIATION

Fig. 24 - The Venetian-blind multiplier structure.

transit time between dynodes is relativelyslow and has a rather large time spread.

Venetian-Blind StructureAnother structure which provides good

collection of photoelectrons, but again israther slow in response time, is the venetian-blind photomultiplier shown in Fig. 24.Good collection of photoelectrons is aidedby the size of the photoelectron collectingarea which can be even larger than that ofthe box-and-grid construction. The relativelyslow time response is the result of the weakelectric field at the surfaces of the dynodevanes. The structure is very flexible as to thenumber of stages. Some of the secondary

electrons are lost because of the interpositionof the grids between stages, as with the box-and-grid construction.

Tea-Cup StructureA recent innovation in front-end design is

the so-called “tea-cup” photomultiplier,named after its large first dynode. Secondaryelectrons from the first dynode are directedto an opening in the side of the tea-cup andthence to the second dynode. Fields betweenthe photocathode region and the firstdynode region are separated by a very finegrid structure. The particular advantage ofthis design is that the collection efficiency ofthe large first dynode is good not only for

29


92CM-32311

Fig. 25 - Tea-cup photomultiplier showing photoelectron paths directly from thephotocathode and those initiated by light transmitted through the photocathodeand striking the side wall which also has an active photoemissive layer. Also in-dicated are equipotential lines in the region between photocathode and firstdynode.

photoelectrons emitted from the photocath-ode but for photoelectrons emitted from theactivated side walls between the first dynodeand the photocathode as shown in Fig. 25.The side-wall photoemission results fromlight which passes through the front semi-transparent photocathode. The increasedphotoemission and collection efficiency im-proves the pulse-height resolution in scin-tillation counting applications. Indicated onthe diagram are equipotential lines and pho-toelectron paths showing the collection ofelectrons from the front surface and fromthe side walls.

In-Line Dynode Structure withCurved Photocathode

The planar-photocathode design, such asthat shown in Figs. 22-25, provides excellentcoupling to a scintillation crystal, but itstime response is not as good as that of aspherical-section-photocathode design. Thespherical-section photocathode shown inFig. 26 when coupled to a high-speed elec-tron multiplier provides a photomultiplierhaving a very fast time response. Some tubetypes utilize a spherical-section photocath-ode on a plano-concave faceplate to facili-tate scintillator coupling. This design is


FRONT- END REGION

92CM-32312

Fig. 26 - Photomultiplier design with curved faceplate and in-line dynode structure toprovide a minimum transit time and transit-time spread.

PHOTOCATHODE

DYNODE NO. I

TYPICALELECTRONTRAJECTORIES

TYPICALEQUIPOTENTIALLINES

92CM-32313

Fig. 27 - Cross section of a photomultiplier showing equipotential lines and electrontrajectories that were plotted by computer.

31


usually limited to faceplate diameters of twoinches or less because of the excessive thick-ness of the glass at the edge. The plano-concave faceplate may also contribute tosome loss in uniformity of sensitivity be-cause of internal reflection effects near thethick edge of the photocathode.

The performance of the front-end struc-ture shown in Fig. 26 has been determined bythe use of a computer to trace equipotentiallines and electron trajectories which havethen been superimposed on a schematic dia-gram of the tube structure as shown in Fig.27.

A time parameter of interest is the photo-cathode transit-time difference, the time dif-ference between the peak current outputs forsimultaneous small-spot illumination of dif-ferent parts of the photocathode. In aplanarcathode design, the transit time islonger for edge illumination than for centerillumination because of the longer edge tra-jectories and the weaker electric field nearthe edge of the photocathode. The center-to-edge transit-time difference may be as muchas 10 nanoseconds. The spherical-sectionphotocathode affords more uniform time re-sponse than the planar photocathode be-cause all the electron paths are nearly equalin length; however, the transit time is slightlylonger for edge trajectories than for axialtrajectories because of the weaker electricfield at the edge.

The photocathode transit-time differenceis ultimately limited by the initial-velocitydistribution of the photoelectrons; thisdistribution causes time-broadening of theelectron packet during its flight from thephotocathode to the first dynode. Thebroadening effect can be minimized by in-creasing the strength of the electric field atthe surface of the photocathode.

Because the energy spread of secondaryelectrons is even larger than that of photo-electrons, initial-velocity effects are the ma-jor limitation on the time response of theelectron multiplier. Multiplier time responseis usually improved by the use of highelectric-field strengths at the dynode surfacesand compensated design geometries. In acompensated design, such as that shown inFig. 28 and as used in the type of photomul-tiplier shown in Fig. 26, longer electronpaths and weaker fields alternate with

32

shorter electron paths and stronger fieldsfrom dynode to dynode to produce nearlyequal total transit time.

Certain materials have advantages overother materials in providing good multipliertime performance. For example, a standarddynode material, copper beryllium, has amaximum gain per stage of 8 at a 600-volt in-terstage potential difference. In contrast,gallium phosphide exhibits a gain of 60 ormore at a 1200-volt interstage potential dif-ference. The advantage in time performanceof the GaP dynode over one of the CuBetype is that the number of stages, and thusthe total transit time may be reduced, andthat the energy spread of the secondary elec-trons is less.

92cs-32314

Fig. 28 - A compensated-design multiplier.

Continuous-Channel Multiplier StructureThe continuous-channel multiplier struc-

ture35 shown in Fig. 29 is very compact andutilizes a resistive emitter on the inside sur-face of a cylinder rather than a discretenumber of dynodes. The lack of discretedynodes causes the electron-multiplicationstatistics to be poor because of the variablepath lengths and the variable associatedvoltages. The gain of a continuous-channelmultiplier is determined by the ratio of the

channel length to inside diameter; a typicalvalue of this ratio is 50 but it may range from30 to 100.

92cs-32315

Fig. 29 - The continuous-channel multiplierstructure.

Very high gain can be achieved with such asingle-channel multiplier, provided the chan-nel is curved or bent to avoid line-of-sightfeed-back paths. Numerous special purposephotomultiplier tubes have been built usingthis concept.

Recently,36 very-high-speed photomulti-plier tubes have been designed utilizingmicrochannel plates in proximity with thephotocathode and anode. Such a device is il-lustrated in Fig. 30. A microchannel plate isan array of parallel channels, each perhaps

ray is mounted close to the photocathodewith the plane of the channel plate parallel tothe photocathode. A high voltage, 1 kilo-volt, may be applied between photocathodeand microchannel plate, and similar voltagesapplied across the microchannel sandwichand between the plate and the anode. Thesevoltages assure short transit times. Timeresolution for pulses initiated by single

MICROCHANNELPLATE

\


photoelectrons measured at full width at halfmaximum (FWHM) is less than 300 pico-seconds.37 Because of the construction andhigh electric fields employed, the sensitivityto external magnetic fields is much reduced,a fact which is important in some nuclearphysics experimentation.Crossed-Field Multiplier Structure

As mentioned earlier, a crossed-fieldphotomultiplier was first reported in earl1936 by Zworykin, Morton and Malter. 3gTheir tube, shown schematically in Fig. 31,used a combination of electrostatic and mag-netic fields to direct electrons to repeatedstages of secondary emission. Above eachemitter was a field plate whose potential wasset to be equal to that of the next emitterdown the line. As a result of this configura-tion, electrons emitted from the photocath-ode or from one of the secondary emitters,were caused to follow approximately cy-cloidal paths to the next electrode. This earlydevelopment was not carried into large-scalemanufacture because of the criticalmagnetic-field adjustments needed to changethe gain. Also, the rather wide open struc-ture resulted in high dark current because offeedback from ions and light developed nearthe output end of the device.

Recently, however, similar, but improvedcrossed-field photomultipliers39 have beendesigned to provide perhaps the fastest risetime of any photomultiplier. The electrontrajectories are essentially isochronous in thecrossed magnetic and electric fields so that avery short rise time, 250 picoseconds40, is

/PHOTOCATHODE

/FACEPLATE

I 1

ANODE

92cs-32316

Fig. 30 - Microchannel-plate photomultiplier.

33


achieved. Unfortunately, the photocathodeis rather small and inaccessible by the natureof the design. A schematic drawing of the re-cent crossed-field photomultiplier is shownin Fig. 32. Note the single field electrode incontrast to the multiple field the plates of theZworykin tube. The stepped arrangement isrequired in order to provide uniform electricfield. A 50-ohm coaxial output connectorprovides coupling to the high-speed photo-multiplier.

A FIELD PLATES

CATHODE COLLECTOR

92CS - 32317

Fig. 31 - Schematic of the Zworykin crossed-field photomultiplier reported in 1936. A mag-netic field, perpendicular to the plane of thedrawing, and an electrostatic field, producedby the upper fieldplates, combine to bend theelectron paths in the cycloidal trajectories il-Iustrated.

FIRSTDYNODE

92CS-32318

Fig. 32 - Schematic arrangement of amodern static crossed-field photomultiplier.

ANODE CONFIGURATIONSThe primary function of the anode is to

collect secondary electrons from the lastdynode. The anode should exhibit aconstant-current characteristic of the typeshown in Fig. 33. The simplest anode struc-ture, shown in Fig. 34, is a grid-like collectorused in some Venetian-blind structures. Thesecondary electrons from the next-to-lastdynode pass through the grid to the lastdynode. Secondary electrons leaving the lastdynode are then collected on the grid-likeanode.

POTENTIAL BETWEEN LAST DYNODE AND ANODE-VOLTS92cs -32319

Fig. 33 - Constant-current characteristics ofa photomultiplier anode.

GRID-LIKE

TRAJECTORY

Fig. 34 - The simplest anode structure, agrid-like collector.

In applications where large output pulsecurrents are required it is important to pro-vide a design having reasonably high with-drawal fields to avoid space-charge develop-ment which could limit the output current.Space charge may actually limit the output atthe next to the last dynode, which is the casein the circular-cage structure, Fig. 21,because the withdrawal field at the lastdynode is significantly higher than at theeighth dynode. In special applications atapered divider network may be used to pro-vide higher inter-stage voltages in the lastseveral stages of the tube (see Chapter5-Photomultiplier Applications).

Where fast time response is important, thetype of design shown in Fig. 34 may be at adisadvantage because electrons from the lastdynode are not necessarily collected on thefirst pass through the anode grid structure. Itis also important for fast time response thatthe anodes be designed with matched-impedance transmission lines or with shortconnecting support leads. Most high-speedcircuits are designed to utilize a 50-ohm im-

34


pedance, which requires a suitable connectoror lead geometry outside the tube to permitproper impedance matching.

Anode configurations and internal trans-mission lines may be analyzed by use of thestandard methods of cavity and transmis-sion-line analysis. These methods yield ap-proximate design parameters41, which areoptimized experimentally by means of time-domain reflectometry, TDR42’43. TDR pro-vides information about the discontinuitiesin the characteristic impedance of a systemas a function of electrical length and is an ex-tremely useful approach in the design ofvoltage-divider circuits and mating socketswhich do not readily lend themselves tomathematical analysis.

Certain anode structures in fast-rise-timephotomultipliers exhibit a small-amplitudepulse (prepulse) that can be observed ananosecond or two before the true signalpulse. In grod-like anode structures thisprepulse is induced when electrons from thenext-to-last dynode pass through the anodegrid. In more sophisticated photomultiplierdesigns this phenomenon may be suppressedby auxiliary grids that shield the anode fromthe effects of the impinging electron cloud.

REFERENCES

35. K.C. Schmidt and C.F. Hendee,“Continuous-Channel Electron MultiplierOperated in the Pulse-Saturated Mode,”IEEE Trans. Nucl. Sci., Vol. NS-13, No. 3,p. 100 (1966).

36. Ph. Chevalier, J.P. Boutot and G.Pietri, “PM of new design for high speedphysics,” IEEE Trans. Nucl. Sci., Vol.NS-17, No. 3, pp. 75-78 (June, 1970); G.Pietri , “Contribution of the channel elec-tron multiplier to the race of vacuum tubestowards picosecond resolution time,” IEEETrans. Nucl. Sci. , Vol. NS-24, No. 1, pp.228-232, (Feb. , 1977).37. C.C. Lo, P. Lecomte and B. Leskovar,“Performance studies of prototype micro-channel plate photomultipliers," IEEETrans. Nucl. Sci., Vol. NS-24, No. 1 , pp.302-311 , (Feb. , 1977).38. V.K. Zworykin, G.A. Morton, and L.Malter , “The secondary emission multi-plier-a new electronic device”’ Proc.I.R.E. Vol. 24 p. 351-375 (1936).39. R.S. Enck and W.G. Abraham,“Review of high speed communicationsphotomultiplier detectors," Proc. Soc.Photo Optical Instrumentation Engineers,Vol. 150; Laser and Fiber Optics Communi-cations’ San Diego CA; 28-29 Aug. , 1978(Bellingham , WA; Soc. Photo-Optical In-strumentation Engineers , 1978) , p 3 l-8.40. B. Leskovar and C.C. Lo, “Time Res-olution Performance Studies of Contempo-rary High Speed Photomultipliers,” IEEETrans. Nucl. Sci., Vol. NS-25, No. 1, pp.582-590, (Feb. 1978).41. I.A.D. Lewis and F.H. Wells, Milli-Microsecond Pulse Techniques, PergamonPress (1959).42. “Time-Domain Reflectometry,”Hewlett Packard Application Note 62.43. “Time-Domain Reflectometry,”Tektronix Publication 062-0703-00.

35


4. Photomultiplier Characteristics

PHOTOCATHODE-RELATEDCHARACTERISTICS

Current-Voltage CharacteristicsPhotomultiplier tubes may be operated as

photodiodes by utilizing the first dynode andfocusing electrode (if present) tied togetheras an anode. In this mode of operation thephotocurrent is linear with light flux exceptthat, for semitransparent photocathodes, theresistivity of the thin photocathode layerlimits the current which can be drawn. In thecase of a resistive photocathode, when thelight is directed to the center of the sensitivearea, there is a drop in potential between theouter conductive ring and the illuminatedarea. Because the illuminated area becomespositively charged with respect to the photo-cathode contact, only the more energeticphotoelectrons emitted will overcome theslightly repelling electric field near thephotocathode surface.

Fig. 35 shows a pair of current-voltagecharacteristics for 75-mm-diameter photo-multipliers operated as photodiodes. In bothcases the illuminated area is approximately 5mm in diameter and located at the center ofthe photocathode. The multialkali photo-cathode (S-20 response) layer is fairly con-ductive so that even for a photocurrent ofover 200 nanoamperes, good collection effi-ciency is achieved for a collection voltage aslow as 50 volts. On the other hand, thebialkali photocathode , K2CsSb , is veryresistive causing poor collection efficiencyover a wide range of voltage even for photo-emission currents as low as 10 nanoamperes.

For resistive photocathodes it is thusnecessary to limit the maximum photocath-

36

VOLTAGE-VOLTS

92CM-32321

Fig. 35 - Current-voltage characteristics for75-mm diameter photomultiplier tubes oper-ated as photodiodes. The poor collection effi-ciency shown for the tube with the resistivephotocathode is caused by the voltage dropfrom the edge of the photocathode to the cen-ter illuminated spot and the resulting electro-static field distortion between the photocath-ode and the first dynode and focus electrodeused together as an anode.

ode current to avoid non-linear operation.Resistivity characteristics of several types ofphotocathode are shown in Fig. 36. Here,the resistance per square is shown as a func-tion of temperature. The increase in resistivi-ty with decreasing temperature is expectedbecause of the semiconductor nature ofphotocathode materials. Resistance persquare is the resistance of a surface layer be-tween conductors at opposite sides of asquare of the layer. Note that, if the resistivi-

square of side dimension d, and thickness t,

the resistance R is given by

where d l t is the cross-section area. Thus,the resistance per square is independent ofthe side dimension. On photocathode typesthat are very resistive , it is advisable to main-tain operating temperatures above -100° Cdepending upon photocathode diameter , thelight-spot diameter, and the photocathodecurrent. The larger the photocathode diam-eter and the smaller the light-spot diameter,the more severe the effect.

2

Fig. 36 - Resistance per square as a functionof temperature for various semitransparentphotocathodes. These data were ob-tained with special tubes having connectionsto parallel conducting lines or between con-centric conducting circular rings on the Sur-face of the photocathode.

Photomultiplier Characteristics

Photocathode resistive effects can beavoided at manufacture by the use of nearlytransparent conductive undercoatings.Opaque types of photocathode, such asCs3Sb on a solid substrate of nickel, do nothave a resistivity problem. Table II providesguidance as to the maximum dc current thatcan be utilized with various semitransparentphotocathodes at room temperature. Itshould be realized that these data are onlytypical; individual photocathodes may varyconsiderably from the guideline provided intheir capability of delivering current withoutresistive blocking.

Table II - Maximum Recommended Pboto-cathode DC Currents for Various Semitrans-parent Photocathode Types as Determinedby Surface Resistivity.

Photocathode Maximum RecommendedCurrent for T = 22°C

In the case of pulsed photocathode cur-rents, the peak current values may be higherthan those shown in Table II. Average cur-rents, however, are still limited, as shown,by the resistance of the photocathode layer.For a current pulse, the local surface poten-tial of the cathode is sustained for a time bythe electrical charge associated with the pho-tocathode capacitance. Thus, a square cen-timeter of photocathode may have a capaci-tance of about 0.5 picofarad. A 5-voltchange in potential is about the maximumthat could occur without blocking of thephotoemission current. The equivalentcharge is therefore 2.5 picocoulombs. For apulse duration of t seconds, a pulse currentof 2.5/t picoamperes could be maintained.For t = 1 microsecond, the current maximumcould be 2.5 microamperes, as determinedby stored charge alone.

Pulse current maximum can be increasedby the capacitance of the photocathodelayer. For example, if a ground plane is pro-vided in contact with the outside of the

37


photocathode faceplate, the capacitance of asquare centimeter of photocathode may beincreased to about 2 picofarads (assuming adielectric constant 6.75 and a glass thicknessof 3 mm). The ground plane must, of course,be reasonably transparent to the light signal.A conducting mesh is a suitable solution.The ground plane should also be set at ornear photocathode potential, or noisy opera-tion and deterioration of the photocathodemay result. See the section below under“Dark Current and Noise.”

Variations in Spectral Response withTemperature.

Minor variations in the spectral responsecharacteristics of photocathodes occur withchanges in temperature. Data are presentedhere on the typical variations which can beexpected. Deviations from these data may beexpected on individual photocathodes be-cause of variations in thickness and process-ing.

When the temperature is decreased, theresponse of photocathodes usually improvesin the shorter-wavelength region andworsens in the longer-wavelength regionnear the threshold. An illustration of thistrend is provided in Fig. 37 in which the

Fig. 37 - Temperature coefficient of cesium-antimony cathodes as a function of wave-length at 20 “C. Note the large positive effectnear the threshold.

temperature coefficient of responsivity of acesium-antimony photocathode is plotted asa function of wavelength.46 The differencein spectral response from room to liquid-air

temperature is illustrated in Fig. 38, also forCs3Sb. According to Spicer and Wooten47,the increase in response in the blue at lowtemperatures is the result of a decrease inenergy loss from lattice scattering; the de-crease in the red at low temperatures resultsfrom a decrease in occupied defect levels inthe forbidden energy band because of an in-crease in band gap and possibly because ofan unfavorable change in band bending. Itmay also be the case that, in photocathodeshaving significant impurity levels, the ex-tended red sensitivity at higher temperaturesis partly the result of thermally assistedphotoemission. The change of slope at theelevated temperature in Fig. 38 is suggestiveof this mechanism.

0.1 1 I I I 14 0 0 500 6 0 0


92cs-32324

Fig. 38 - Dependence of responsivity of aCs3Sb photocathode on temperature; takenfrom Spicer and Wooten.

The data shown in Fig. 39 is of the sametype as that in Fig. 38, but for a GaAs:Csphotocathode. Note, again, the increase insensitivity with decrease in temperature ex-cept for the shift in the long wavelengththreshold. In GaAs:Cs the photoexcitation isfrom the valence band; there is essentially noexcitation from the forbidden energy band.Note the sharp long-wavelength cutoff.

38


Fig. 39 - Relative spectral responsivity shiftwith temperature for a GaAs:Cs photocath-ode.48

Variation of Photocathode Response withAngle of Incidence and Angle of Polariza-tion.

In a theoretical paper49 Ramberg hasevaluated the optical factors which deter-mine the variation of photoresponse as afunction of polarization angle and angle ofincidence. One interesting conclusion ofRamberg’s work is that for normal incidenceand for a certain regime of cathode thicknessand escape depth, the photoemission “maybe expected to be about 1.4 times as great forillumination from the glass side as for il-lumination from the vacuum side.” Photo-multiplier tubes with semitransparentphotocathodes such as are used in typicalscintillation counting applications are alldesigned for radiation incident on the glasssubstrate which supports the photocathode.The two curves of Fig. 40 are calculatedfrom Ramberg’s formulae for incidencethrough the glass faceplate, and for polariza-tion perpendicular and parallel to the planeof incidence. (i.e. “parallel” implies that theelectric vector lies in the plane defined by theincident ray and the normal to the glass sur-face.) These curves assume an index ofrefraction for the glass of 1.5, for the photo-sensitive layer of 3.25, and the absorption in-dex for the photosensitive layer, K, of 3.25.The photocathode layer is assumed to bethin with respect to the wavelength of light-taken as 550 nanometers for Fig. 40.

Note that Ramberg’s prediction is thatresponse should be greater for perpendicularpolarization, especially at large angles of in-

cidence. This conclusion is at odds withsome reported measurements. Experimentalmeasurements for a semitransparent Cs3Sb

92CM-32326

Fig. 40 - Theoretical photoexcitation deter-mined by optical factors for different polariza-tion orientation as a function of angle ofincidence. Radiation is incident on a semi-transparent photocathode such as Cs3Sbthrough a glass substrate of index of refrac-tion 1.5. Data are taken from Ramberg.

0

2.0

(b) 92cs-32327

Fig. 41 - (a) Photoresponse for a semitrans-parent Cs3Sb photocathode as a function of

(0°) and perpendicular (90°) to the plane of in-cidence. (b) Photoresponse ratio for polariza-

sured at an angle of incidence of 70°(fromHoenig and Cutler50).

39


photocathode on Corning 9741 glass arereported by Hoenig and Cutler50. Theirdata, are shown in Fig. 41 a and b. Theircurve, a, shows the variation of photosen-

for two different angles of polarization. The

parallel to the plane of incidence. The curve,b, shows the ratio of the photoresponse forpolarization angle 0° to the photoresponse

both measured at 70° angle of incidence.Note that the response is greater forpolarization parallel to the plane of in-cidence .

Measurements at RCA support the generalconclusion that the response is higher forparallel polarization for Cs3Sb, K2CsSb andNa2KSb:Cs semitransparent photocathodesat large angles of incidence. On the otherhand, Hora51 finds the response for a“trialkali antimonide” to be higher forperpendicular polarization.

It is likely that the discrepancies notedabove are related to the differences inphotocathode substrates (between the glassand the photocathode), photocathode thick-ness, and photocathode processing whichmay result in non-isotropic and structuredlayers, in contrast, for example, to theassumption by Ramberg of a very thin andisotropic photocathode layer.Optical Devices to Enhance Photoresponse

Numerous optical devices have been de-vised to enhance photoresponse by utilizingmultiple paths in the faceplate and photo-cathode. Experimental data and theoreticalanalyses have been reported by a number ofdifferent authors.52-56 Fig. 42 shows atypical arrangement permitting the lightbeam to enter the glass substrate so thatreflected rays approach the glass-air inter-face at an angle greater than the criticalangle and thus do not permit a refracted rayto escape the glass. In this manner multipleexcitations of the photosensitive layer arepossible. Enhancement of the quantum effi-ciency for an alkali-antimonide photocath-ode may be something less than 2:1 in theblue where the absorption of the photocath-ode layer is high, but may be as high as 4: 1 inthe red where the absorption is low.

An interesting series of measurements andtheoretical predictions have been reported by

40

92CS-32328

Fig. 42 - A glass quadrant in optical contactwith the photocathode window, permitting in-coming radiation to approach the glass-photocathode interface at any angle of in-

critical angle for the glass-air interface (about42°) all of the radiation is reflected, permit-ting multiple excitations of the photosen-sitive surface.

D.P. Jones56 on the behavior of a semitrans-parent Na2KSb:Cs photocathode with angleof incidence, wavelength, and polarizationangle. The optical arrangement is similar tothat shown in Fig. 42 except that a full 180°cylindrical lens is cemented to the faceplateand half of the surface is coated withaluminum to reflect the rays back to thephotocathode. His data and theory favorperpendicular polarization, particularly atlonger wavelengths.Photocathode Uniformity

The uniformity of photocathode responseover its area may be of importance in someapplications. For example, in scintillationcounting, light pulses which excite differentareas of a non-uniform photocathode wouldthen result in variations in measured pulseheight. Such an effect would be more ap-parent in the case where the scintillatingcrystal is thin. In flying-spot scanners, non-uniform photocathode sensitivity wouldresult in picture shading unless precaution istaken to avoid even approximate imaging onthe photocathode surface.

Photocathode non-uniformity may resultfrom a variety of causes. Fig. 43 shows sometests of photocathode uniformity. Thesecurves represent photoresponse for a smallfocused (l/16-inch) spot as it is scannedacross the diameter of a semitransparentphotocathode. Curve (a) represents the


uniformity ( ± 3%) typical of a well-pro-cessed photocathode in a 2-inch photomulti-plier designed for scintillation counting. Incurve (b), the variation (1) across the photo-cathode is the result of a non-uniformevaporation of antimony during the process-ing of the photocathode. Curve (c) showssome peculiar non-uniformities in a 3-inchphotocathode. The peaks (2) are the result oflight entering the rounded corner of the tubeenvelope and increasing emission by reflec-tions in the glass faceplate. The peaks (3)result when light, which is transmitedthrough the photocathode, strikes the re-entry shoulder of the glass envelope. The in-side shoulder of the envelope is aluminizedas part of the photocathode contact. Somelight striking the shoulder may be reflectedback to the photocathode and cause in-

92cs-32329

Fig. 43 - Photocathode uniformity patternsobserved by scanning a 1/16-inch spot acrossa diameter of a semitransparent photocath-ode in 2-inch and 3-inch photomultiplier tubesused in scintillation counting. (a), scan on atypical well-processed 2-inch tube; (b), scanon a tube with non-uniform antimony evapora-tion on the faceplate (1); (c), scan on a typical3-inch tube showing peaks (2) due to entry oflight on the edge of the faceplate and (3)peaks due to reflection or photoemissionfrom transmitted radiation through the photo-cathode onto the shoulder of the bulb.

creased emission, or there may actually besome photosensitivity on the shoulderbecause of the presence of processingmaterials.

Anomalies in the photocathode uniformi-ty such as shown in Fig. 43(c) can be largelyeliminated by a fabrication technique57

creating a diffusing layer on the photocath-ode substrate. The fabrication of a diffusinglayer can be easily accomplished by sand-blasting the inside surface of the photomulti-plier faceplate. The effect of the diffusinglayer is to scatter the incoming light so thatlight transmitted through the photocathodeis spread out in many directions, thusavoiding direct correlation between the posi-tion of the incoming light spot and any inter-nal reflections. The diffusing layer also has adesirable effect of enhancing the photocath-ode sensitivity.

Uniformity of photocathode response,however, is not sufficient to assure auniform output response from the photo-multiplier. All of the emitted photoelectronsmay not be properly directed to the firstdynode by the electron optics. In the case ofa Venetian-blind dynode structure, some ofthe photoelectrons may strike parts of thedynode which do not provide good collec-tion fields to the second dynode. Fig. 44 is a

DISTANCE ACROSS PHOTOCATHODE-INCHES92CS-32330

Fig. 44 - Relative collection efficiency forphotoelectrons in a venetian-blind type photo-multiplier tube.58

plot of collection uniformity in a 3-inchVenetian-blind photomultiplier tube. Thestructure in the pattern for the scan perpen-dicular to the length of the dynode is relatedto the individual slats in the Venetian-blindstructure. By way of contrast, Fig. 45 showssimilar collection uniformity plots for a3-inch “teacup” design where the large

41


open structure of the “tea-cup” first dynodeeliminates the electron-optical effectsobserved in the Venetian-blind type tube.

DISTANCE ACROSS PHOTOCATHODE- INCHES

92CS-32331

Fig. 45 - Relative collection efficiency forphotoelectrons in a “teacup” type RCA pho-tomultiplier tube.58

Uniformity of response of a photomulti-plier may also be affected by the voltage atwhich the focusing electrode is operated.Many photomultipliers are equipped with afocusing electrode, between the photocath-ode and the first dynode to provide optimumcollection of the photoelectrons emittedfrom the photocathode. The focusing-electrode voltage is usually set at the point atwhich maximum anode output current is ob-tained. In some applications, spatial unifor-mity i.e., the variation of anode current withposition of photocathode illumination, maybe more important than maximizing outputcurrent. In such cases, however, the final ad-justment of the focusing-electrode potentialshould not differ significantly from the ad-justment that provides optimized collectionefficiency. Fig. 46 shows a typical focusing-electrode characteristic.Photocathode Stability

Stability and life of a photocathode isusually related approximately inversely tothe current drawn from it. More stable andreliable performance results if small areas ofconcentrated illumination on the cathodesurface are avoided. In normal operation ofphotomultipliers, of course, the photocath-ode current is usually small and damage tothe tube is more likely to be caused by theamplified secondary emission currents in thelatter dynode stages.

If a photomultiplier were to contain anysignificant trace of gas, ionization of the gascould occur , depending upon the appliedvoltages and level of photocurrent. Ion bom-bardment of the photocathode can readilycause damage and loss of sensitivity in pro-portion to the ion current. Photomultipliertubes, however, are generally so well evacu-ated that ion damage is essentially only atheoretical possibility.

Electrolytic effects can bring about seriousfatigue in Cs3Sb photocathodes (Ref. 11, p.79) and presumably in other alkali-antimonide photocathodes. An electrolyticeffect is caused when a potential gradient ismaintained across the cathode surface.Photocathodes generally only have onephysical contact, but a gradient can bedeveloped because of the resistive nature ofthe photocathode layer if a large photocur-rent is drawn, for example, from a centerspot on the cathode. Actual electrolyticdecomposition takes place, which can berecognized by a color change in the cathodematerial.

92CS-32332

Fig. 46 - A typical focusing-electrode charac-teristic.

It is also possible to damage a photocath-ode by maintaining a difference in potentialthrough the faceplate supporting the photo-cathode, particularly at elevated tempera-tures. For example, if the photocathode in aphotomultiplier is maintained at - 1000

42


volts and a ground plane at 0 volts isestablished on the outside surface of theglass, ionic conduction takes place throughthe glass. Photocathode sensitivity will begradually deteriorated by the ionic

Photomultiplier tubes should be stored inthe dark when not in use. Blue and ultra-violet radiation, especially, can cause photo-chemical changes in the photocathode whichresult in changes in sensitivity. It is especiallyimportant to avoid exposure to intense il-lumination such as sunlight even when novoltage is applied to the tube. Permanentdamage may also result if the tube is exposedto radiation so intense that it causes ex-cessive heating of the photocathode. Tubesshould not be stored for long periods attemperatures near the maximum rating ofthe tube; high temperatures almost alwaysresult in loss of sensitivity in the photocath-ode.

A photomultiplier having a multialkaliphotocathode (S-20 spectral response) tendsto lose sensitivity especially in the red por-tion of its spectral response upon extendedexposure to high ambient room lighting; thechange is usually permanent. Contrary tothis behavior, photocathodes of the ex-tended red multi-alkali (ERMA) type ap-parently do not exhibit this loss.

The Ag-O-Cs photocathode (S-l) spectralresponse) also suffers a decrease in sensitivi-ty, particularly during operation, when ex-posed to high radiant-energy levels normallynot harmful to other types of photocathodematerials. The decreased sensitivity occursprimarily in the infrared portion of the spec-trum. Loss of infrared sensitivity may alsooccur following long periods of storage.

The GaAs:Cs photocathode is particularlysensitive to responsivity loss even for

Another effect related to the exposure ofphotocathodes to excessive blue or ultra-violet radiation, as from fluorescent roomlighting, is a temporary increase in photo-cathode dark emission current. The increasemay be as much as three orders of magnitudeeven from relatively short exposure. This in-crease in dark current occurs even thoughvoltage is not applied to the tube and maypersist for a period of from 6 to 24 hoursafter such irradiation. Fig. 47 illustrates therecovery after exposure to fluorescent-lamp

radiation for several photocathode types.Some of this effect is the result of phosphor-escence in the glass faceplate of the photo-multiplier, but a larger part of the effect isapparently an excitation phenomena in the

Fig. 47 - Variation of dark current followingexposure of photocathode to cool white fluo-rescent-lamp radiation. The various photo-cathodes are identified by their spectral-response symbols.

When semitransparent photocathodessuch as Cs3Sb and Ag-Bi-O-Cs (and prob-ably others) are kept in the dark for manyhours, they become very resistive59, especial-ly at reduced temperatures. The effect maybe so great that the photomultiplier may ap-pear to have lost sensitivity. Apparently,passage of current through the photosen-sitive layer as a result of normal operationrestores the photocathode conductivity. Atreduced temperatures, the process may re-quire minutes or hours of operation to re-turn the photocathode to a reasonable con-ductivity.

Photomultiplier tubes have been exposedto gamma and X-radiation to an intensity of1010 roentgens per hour by the Naval

photocathode damage was noted except thatfaceplate discoloration was observed for ex-posure in excess of 104 roentgens. Glassfluorescent effects were also noted duringthe tests. For applications where excessiveradiation may be present, it may be notedthat Corning has developed a non-browningglass, Ce-doped No. 9025, which has beenused for special photomultipliers.

As a result of enhanced dark count ratesobserved in photomultiplier tubes used invarious earth-orbiting satellites, an in-

43


vestigation was made by Viehmann et al61

on fluorescent and phosphorescent effects inwindows used in photomultiplier tubes whenbombarded by beta rays. Source of the betarays was an 0.8 millicurie beta emitter,

to 2.23 MeV (megaelectron volts). A numberof special windows were tested. For Coming9741 glass, two decay constants were ob-served in the phosphorescense: 4.2 and 57minutes. For an exposure of 9.5 x 1010 elec-trons per square centimeter in 30 minutes, an

steradian)- 1 was observed in a photomulti-plier with an S-20 response, closely coupledto the glass plate.

GAIN-RELATED CHARACTERISTICSGain vs. Voltage

When several secondary-emission stagesare coupled together, so that the secondaryelectrons from one become the primary elec-

multiplier phototube is given by

In practice, some of the electrons may skipstages, or become lost to the amplificationprocess by impinging upon nonproductivesecondary-emission areas.

It is customary to describe the gain of themultiplier phototube as a function of the ap-plied voltage. Fig. 48 shows two such curveson a semilog scale. These curves illustrate thewide range of amplification in a multiplierphototube. They also indicate the necessityof providing a well regulated voltage supplyfor the dynode stages.

It is possible to operate a four-to-six stagephotomultiplier so that each stage is at thevoltage required for maximum secondaryemission, as shown in Fig. 19. In such cases,the gain could be made practically indepen-dent of voltage over a small range. However,such a condition would require approximate-ly 500 volts per stage; thus the total voltagerequired would be very high for the amountof gain achieved.

44

In the design or operation of a multiplierphototube having a fixed supply voltage, thenumber of stages can be chosen so that thegain of the tube is maximum. For this pur-pose, the optimum voltage per stage is thatvalue at which a line through the origin (uni-ty gain on the log-gain scale) is tangent to thecurve, as shown in Fig. 48. This point is

92cs-32334

Fig. 48 - Log of gain as a function of voltsper stage for a tube (1P21) with Cs-Sbdynodes and for a tube (6342A) with Cu-Bedynodes.

identified on the graph as the point of max-imum gain per volt. (Note that this argumentneglects the voltage used between the lastdynode and the anode and any discrepancyresulting from nonuniform distribution ofvoltage per stage). In most applications ofmultiplier phototubes, the tubes areoperated above the point of maximum gainper volt. When both the gain and the voltageare presented on a logarithmic scale, theresultant curve is then closely approximatedby a straight line. Fig. 49 describes the anodesensitivity in amperes per lumen and thetypical amplification characteristics of aphotomultiplier as a function of the appliedvoltage. Curves of minimum and maximumsensitivity are also shown.


SUPPLY VOLTS (E) BETWEEN ANODE AND CATHODE

92cs-32335

Fig. 49 - Typical anode sensitivity and ampli-fication characteristics of a photomultipliertube as a function of applied voltage. Notethe log-log scaling.

External Magnetic and Electrostatic FieldsAll photomultipliers are to some extent

sensitive to the presence of external magneticand electrostatic fields. These fields maydeflect electrons from their normal path be-tween stages and cause a loss of gain. Tubesdesigned for scintillation counting aregenerally very sensitive to magnetic fieldsbecause of the relatively long path from thecathode to the first dynode; consequently,such tubes ordinarily require electrostaticand magnetic shielding. Magnetic fields mayeasily reduce the anode current by a much as50 per cent or more of the “no-field” value.

The three curves in Fig. 50 show the effecton anode current of magnetic fields parallelto the main tube axis and perpendicular tothe main axis in the directions parallel andperpendicular to the dynodes. The curves are

I -

D-

92CS-32336

Fig. 50 - Curves for 3/4-inch diameter type4516 photomultiplier showing the effect onanode current of magnetic fields parallel tothe main axis of the tube and perpendicularto the main axis in the directions parallel andperpendicular to the dynodes. (Units for mag-netic field intensity are shown in both Slunits, ampere turns per meter, and conven-tional cgs units, oersteds. Note that 1 oersted

45


plier tubes contain some parts which havemagnetic properties, but if they areneglected, the magnetic induction-or mag-netic flux density B-measured in units ofgauss would be numerically the same as theoersted values. in the case of Sl units formagnetic induction, 1 weber per square meter= 1 tesla = 104 gausses.)

usually provided for one or more values ofover-all applied voltage and indicate therelative anode current in per cent as a func-tion of magnetic field intensity. Fig. 51shows the variation of output current ofseveral photomultiplier tubes as a functionof magnetic-field intensity directly parallelto the major axis of the tube. The magnitudeof the effect depends to a great extent uponthe structure of the tube, the orientation ofthe field, and the operating voltage. Ingeneral, the higher the operating voltage, theless the effect of these fields.

High-mu material in the form of foils orpreformed shields is available commerciallyfor most photomultipliers. When such ashield is used, it must be at cathode poten-tial. The use of an external shield may pre-sent a safety hazard because in many ap-plications the photomultiplier is operatedwith the anode at ground potential and thecathode at a high negative potential. Ade-quate safeguards are therefore required toprevent personnel from coming in contactwith the high potential of the shield.

It is possible to modulate the output cur-rent of a photomultiplier with a magneticfield. The application of a magnetic fieldgenerally causes no permanent damage to aphotomultiplier although it may magnetizethose internal parts of a tube that containferromagnetic materials (tubes are availablewhich contain practically no ferromagneticmaterials). If tube parts do become magne-tized, the performance of the tube may be

`-2400 - 1 6 0 0 - 8 0 0 0 8 0 0 1600 2 4 0 0 3200MAGNETIC FIELD INTENSITY - AMPERE TURNS PER METER

I 1 I I I I I 1- 3 0 - 2 0 - 1 0 0 IO 2 0 3 0 4 0

OERSTEDS

92CM - 32337

Fig. 51- Variation of output current of several photomultiplier tubes as a function ofmagnetic-field intensity directly parallel to the major axis of the tube. Positivevalues of magnetic field are in the direction of the tube base. Operating voltages areindicated. The 931A is a circular-cage, side-on type. The 4902, 8053 and 8575 are2-inch end-on types: teacup; Venetian-blind dynode; and in-line dynode structure,respectively.

46

degraded somewhat; however, the conditionis easily corrected by degaussing, a process inwhich a tube is placed in and then graduallywithdrawn from the center of a coil operatedat an alternating current of 60 Hz with amaximum field strength of 8000 ampereturns per meter (100 oersteds).

LinearityBecause the emission rate of photoelec-

trons is proportional to the incident radiantflux, and the yield of secondary electrons fora given primary electron energy is propor-tional to the number of primary electrons,the anode current of a photomultiplier isproportional to the magnitude of the inci-dent radiant flux. A linearity plot over awide range of light level is shown in Fig. 52

LIGHT FLUX -LUMENS

92CS-32338

Fig. 52 - Range of anode-current linearity asa function of light flux for a 931A photomulti-plier.

for a type 931A. The limit of linearity occurswhen space charge begins to form. Space-charge-limiting effects usually occur in thespace between the last two dynodes. Thevoltage gradient between anode and lastdynode is usually much higher than betweendynodes and, therefore, results in a limita-tion at the previous stage, even though thecurrent is less. The maximum output cur-rent, at the onset of space charge, is propor-tional to the 3/2 power of the voltage gra-dient in the critical dynode region. By use ofan unbalanced dynode-voltage distributionincreasing the interstage voltages near theoutput end of the multiplier, it is possible toincrease the linear range of output current.


Table III provides guidance as to the max-imum pulse current which can be drawnfrom the anode of various types of photo-multipliers before spacecharge effects limitlinearity. Note that the type 8575, which hasa focused dynode structure, provides thehighest withdrawal fields and highest linearoutput current. On the other hand, venetian-blind or box-and-slot dynodes have relative-ly low withdrawal fields. Higher currents canbe obtained from all types of photomulti-pliers by unbalancing the voltage distribu-tion to provide higher fields at the criticallast stages.

Some photomultipliers used in applica-tions requiring high output pulse current usean accelerating grid between the last dynodeand the anode to reduce the effects of space-charge limiting. The potential of such a gridis usually between that of the last and thenext-to-last dynode and is adjusted by ob-serving and maximizing the value of theanode output current.

Another factor that may limit anode-output-current linearity is cathode resistivi-ty; cathode resistivity is a problem only intubes with semitransparent photocathodes,particularly of the Cs-Sb or bialkali type.

Linear behavior is not always obtainedfrom photomultipliers even at low currentlevels. For example, if the test light spot on a931A is not directed close to the center of theactive area of the photocathode, disturbingeffects may arise from the proximity of theceramic end plates. Near the end plates, thefields are not uniform and are affected bycharge patterns on the insulator spacers,which change with the current level. An ex-terior negative shield placed around the bulbwall may improve tube linearity by elimi-nating bulb charging effects. The passage ofexcessive current may change the sensitivityof the tube and cause an apparent non-linearity.

Some photomultipliers exhibit a tem-porary instability in anode current andchange in anode sensitivity for severalseconds after voltage and light are applied.This instability, sometimes called hysteresisbecause of cyclic behavior, may be caused byelectrons striking and charging the dynodesupport spacers and thus slightly changingthe electron optics within the tube. Sensitivi-ty may overshoot or undershoot a few per

47

Tube Type

931A(side-on, circularcage)

4524(3-inch, venetian-blind dynodes)

4900(3 -inch, Tea-cuptype)

8575(2-inch, linear fo-cused dynode struc-ture)

8575

C-31059(1 1/8-inch, box andslot dynodes)

C-31059


cent before reaching a stable value. The timeto reach a stable value is related to the

behavior may be a problem in applications

resistance of the insulator, its surfacesuch as photometry where a photomultiplier

capacitance, and the local photomultiplieris used in a constant-anode-current mode by

current. This instability and non-linearvarying the photomultiplier voltage as thelight input changes.

Table III

Maximum Anode Currents That Can be Utilized in Various Photomultiplier Types WithoutSerious Loss of Linearity from Space-Charge Build-up.

Supply Voltage MaximumVolts Distribution* Output Current (mA)

with Linearity Loss LessThan 10%

1000 1 1

1500

1200

1760

2100

1050

1150

5

5

30

120

4

12

*These number series represent the relative voltage applied between cathode and first dynode,first dynode and second dynode,. . . . . , last dynode and anode. Note that in the case of the 8575and the C-31059, a second voltage distribution is given with increased voltage drop in the laststages providing higher output currents before the onset of space-charge-limiting conditions.

48


Hysteresis has been eliminated in manytubes by coating the dynode spacers with aconductive material in the manufacturingprocess and maintaining the coating at afixed potential. Tubes treated by this methodassume final sensitivity values almost im-mediately upon application of light andvoltage.

Peak linear and saturation currents areusually measured by pulsed methods. Onecommon method makes use of a cathode-raytube with a P15 or P16 phosphor. The grid isdouble-pulsed with pulses of unequal ampli-tudes but fixed amplitude ratio. As theamplitude of the two pulses is increased, apoint is observed at which the amplitude ofthe larger of the anode pulses does not in-crease in the same proportion as the smallerpulse. At this point the tube is assumed tobecome non-linear. The current value at thispoint is then measured by means of anoscilloscope and load resistor. The max-imum saturation current is found when afurther increase in radiation level yields nofurther increase in output.

Although very high anode current can bedrawn from photomultiplier tubes, itshould be emphasized that stability cannotbe expected at such levels. These high cur-rent levels are principally of interest in light-pulse applications. In this case, the stabilityof the tube is approximately that of the tubeoperated at the integrated average currentlevel tGain Variation with Temperature

It is often advisable to reduce the ambienttemperature of a photomultiplier in order toreduce dark emission from the photocathodeand improve the signal-to-noise ratio of themeasurement. Variations of the spectralcharacteristics of photocathodes withtemperature have been noted in the previoussection on Photocathode Stability. There arenumerous reports on the over-all variationof the photomultiplier output current withtemperature including the effect of tempera-ture on photocathode sensitivity62-64. Fig.53 is a typical characteristic for type 8571photomultiplier having Cs-Sb dynodesshowing the variation of gain withtemperature. These data do not include thevariation of photocathode response withtemperature which has been described in thesection on Variations in Spectral Response

with Temperature. The gain data were ob-tained by measuring both the anode responsevariation and the photocathode responsevariation with temperature and dividing outthe latter to obtain only the gain variation. Itis probable that the increase in gain at lowtemperatures is the result of a decrease inenergy loss from lattice scattering. Althoughthe variation in over-all gain is fairly large, itshould be remembered that the variationresults from a small variation in each stagecompounded by the 9 stages. At room tem-perature, the average gain per stage for the8571 is 4.5 at 100 V per stage and the percen-tage change in stage gain with temperature isapproximately -0.06% per °C. Thus thesecondary emission temperature coefficientis generally less that the photoresponsetemperature coefficient, both for Cs-Sb sur-faces. See Fig. 37.

LOX

TEMPERATURE - °C

92cs-32341

Fig. 53 - Typical variation of gain with tem-perature for a 9-stage photomultiplier tube(type 8571) with Cs-Sb dynodes operating at100 volts per stage.

A few words of caution regarding the am-bient temperature of photomultipliers arepertinent. A maximum ambient tempera-ture, and in some instances a minimum tem-perature, i s speci f ied for a l l photo-

49


multipliers. The specification of maximumambient temperatures reduces the possibilityof heat damage to the tube. Cesium, for ex-ample, is very volatile and may beredistributed within the tube causing loss ofsecondary emission gain or loss of cathodesensitivity.

It is recommended that photomultipliersbe operated at or below room temperature sothat the effects of dark current are mini-mized. The variation of dark current, ornoise, is most important because of its effecton ultimate low-light-level sensitivity.Various cryostats and solid-state thermioniccoolers have been designed that reduce darkcurrent at low temperatures in low-light-levelapplications. An important consideration inthe use of these devices is to prevent conden-sation of moisture on the photomultiplierwindow. A controlled low-humidity atmos-phere or special equipment configurationmay be necessary to prevent such condensa-tion.

Another reason for avoiding the operationof photomultipliers at extremely low temper-atures is the possible phase change that thistype of operation may cause in some of themetal parts. These changes are particularlyprobable when Kovar is used in metal-to-glass seals. Tubes utilizing Kovar in theirconstruction should not be operated at tem-peratures below that of liquid nitrogen(-196°C).

In some tubes, particularly those withmultialkali photocathodes, it is sometimesobserved that the noise actually increases asthe temperature of the photocathode is re-duced below about -40°C. The reason forthis noise increase is not understood. How-ever, most of the dark-current reduction hasalready been achieved at temperatures above-40°C.

In general, it is recommended that allwires and connections to the tube be encap-sulated for refrigerated operation. Encap-sulation minimizes breakdown of insulation,especially that caused by moisture condensa-tion.Stability and Gain

All photomultipliers have a maximumanode current rating. The primary reasonfor such a rating is to limit the anode powerdissipation to approximately one-half wattor less. Consequently, the magnitude of the

50

maximum anode current is restricted to afew milliamperes when the tube is operatedat 100 to 200 volts between the last dynodeand the anode. Many photomultipliers arerated for only 0.1 mA or less.

Operating a photomultiplier at an ex-cessively high anode current results in an in-creased fatigue that occurs as the averageanode current increases. The loss in sensitiv-ity occurs as a result of a reduction in thesecondary emission, particularly in the laststages of the photomultiplier where the cur-rents are the highest.

Tube fatigue or loss of anode sensitivity isa function of output-current level, dynodematerials, and previous operating history.The amount of average current that a givenphotomultiplier can withstand varies widely,even among tubes of the same type; conse-quently, only typical patterns of fatigue maybe cited.

The sensitivity changes are thought to bethe result of erosion of the cesium from thedynode surfaces during periods of heavyelectron bombardment, and the subsequentdeposition of the cesium on other areaswithin the photomultiplier. Sensitivity lossesof this type, illustrated in Fig. 54 for a 1P21

TIME-MINUTES92cs-32339

Fig. 54 - Short-time fatigue and recoverycharacteristics of a typical 1P21 operating at100 volts per stage and with a light source ad-justed to give 100 microamperes initial anodecurrent. At the end of 100 minutes the light isturned off and the tube allowed to recoversensitivity. Tubes recover approximately asshown, whether the voltage is on or off.


operated at an output current of 100 micro-amperes, may be reversed during periods ofnon-operation when the cesium may againreturn to the dynode surfaces. This processof recovery may be accelerated by heatingthe photomultiplier during periods of non-operation to a temperature within the max-imum temperature rating of the tube;heating above the maximum rating maycause a permanent loss of sensitivity.

Sensitivity losses for a given operating cur-rent normally occur rather rapidly during in-itial operation and at a much slower rateafter the tube has been in use for some time.Fig. 55 shows this type of behavior for a

0 100 2 0 0 3 0 0 4 0 0 52

TIME -HOURS

92cs-32340

Fig. 55 - Typical sensitivity loss for a 1P21operating at 100 volts per stage for a period of500 hours. Initial anode current is 100 micro-amperes and is readjusted to this operatingvalue at 48, 168, and 360 hours.

1P21 having Cs-Sb dynodes, operating at anoutput current of 100 microamperes. Tubesoperated at lower current levels, of the orderof 10 microamperes or less, experience lessfatigue than those operated at higher cur-rents, and, in fact, may actually recoverfrom high-current operation during periodsof low-current operation.

Fatigue rates are also affected by the typeof dynode materials used in a tube. Copperberyllium or silver magnesium dynodes aregenerally more stable at high operating cur-rents than the cesium antimony types. Thesensitivity for tubes utilizing these dynodesvery often increases during initial hours of

TIME -HOURS

Fig. 56 - Typical responsivity variation on lifefor a photomultiplier having silver-magne-sium dynodes. Initial anode current was 2 mil-liamperes and was readjusted to this operat-ing value at 48, 168, and 360 hours.

operation, after which a very gradual de-crease takes place, as illustrated in Fig. 56.

The operating stability of a photomulti-plier depends on the magnitude of theaverage anode current; when stability is ofprime importance, the use of average anodecurrents of 1 microampere or less is recom-mended.

In addition to the life characteristics,which are probably the result of changes inthe dynode layer itself, other changes of atemporary nature also occur. Not all thesechanges are well understood; some arecharging of insulators in the tube.

Fig. 57 illustrates one of the peculiar in-stabilities which are sometimes observed inphotomultiplier tubes. When the light is firstturned on, the current apparently overshootsand then decays to a steady value. This par-ticular phenomenon is probably the result ofthe charging of the supporting insulator forthe dynodes. The effect is observed to occurmore rapidly at higher currents, presumably

51


b e c a u s e o f t h e g r e a t e r c h a r g i n g c u r r e n t .E l e c t r o n s s t r i k i n g t h e i n s u l a t o r p r o b a b l y

result i n s e c o n d a r y e m i s s i o n a n d a r e s u l t a n tpositive charge. The change in potential af-fects the electron optics in the space betweend y n o d e s . T h e e f f e c t i s o b s e r v e d a s a n i n -c r e a s e i n s o m e t u b e s a n d a s a d e c r e a s e i nothers. When the photomultiplier is designedw i t h m e t a l s h i e l d s o r w i t h c o n d u c t i v ec o a t i n g s o n t h e c r i t i c a l a r e a s o f t h e i n -sulators, this effect is eliminated.

TIME - SECONDS

92cs-32343

Fig. 57 - Sudden shift in anode current prob-ably as the result of insulator spacer charg-ing. Observation was made using an experi-mental photomultiplier in which the effectwas unusually large.

A related phenomenon is the variation ofp u l s e h e i g h t w i t h p u l s e c o u n t r a t e i nscintillationcounting applications. Thus,when a radio-active source is brought closerto a scintillating crystal a greater rate of scin-tillations should be produced, all having thesame magnitude. In a particular photomulti-plier a few per cent change in amplitude mayresult and cause problems in measurement.Fig. 58 shows the typically minor variationof pulse height with pulse-count rate for atype 6342A multiplier phototube.

0.4 0.6COUNTING

0.8RATE

92cs-32344

Fig. 58 - Typical variation of pulse heightwith pulsecount rate for a 6342A. 3(13 Cssource with a Nal:Tl source).

I n o rde r t o i nves t i ga t e t he phenomena o fpulse-height var ia t ion wi th pulsecount ra te ,a p u r p o s e l y e x a g g e r a t e d e x p e r i m e n t w a sdevised. Instead of a scintillating crystal, apulsed cathode-ray tube was used as a l ights o u r c e . T w o p u l s e r a t e s w e r e s t u d i e d : 1 0 0and 10,000 pulses per second. Pulse durationw a s o n e m i c r o s e c o n d ; d e c a y t i m e t o 0 . 1maximum was 0 .1 microsecond. The exper i -ment was devised to s tudy the ra te a t whicht h e p h o t o m u l t i p l i e r t u b e o u t p u t r e s p o n s ec h a n g e d w h e n t h e p u l s e r a t e w a s s u d d e n l ys w i t c h e d b e t w e e n t h e t w o r a t e s . T u b e s s u c has the 6342A, a n d e s p e c i a l l y t h e 8 0 5 3 a n d8054, showed pract ica l ly no ef fec t (of theorder of one per cent or less). Fig. 59 shows

Fig. 59 - Variation of output pulse height asthe rate of pulsing is changed in a poorlydesigned experimental tube. Light pulseswere provided from a cathode-ray tube. At theleft of the graph, which shows the pulse-amplitude envelope with time for the outputof the photomultiplier tube, the pulses are at100 per second. The pulse rate is increasedsuddenly to 10,000 per second and again re-duced as indicated. Changes in amplitude areprobably the result of insulator charging.

the pulses during the switching procedure foran experimental tube. The phenomena werecompletely reversible and were observed (toa lesser extent) on many different tube types.The time-decay period of several secondssuggests the charging of an insulator spacerto a new potential as the result of the in-creased charge flow and the subsequentmodification of interdynode potential fields.

In scintillation counting it is particularlyimportant that the photomultiplier have verygood stability. There are two types of gainstability tests which have been used toevaluate photomultipliers for this applica-

52

tion: (1) a test of long-term drift in pulse-height amplitude measured at a constantcounting rate; and (2) a measure of short-term pulse-height amplitude shift withchange in counting rate.

In the time stability test, a pulse-heightanalyzer, a 137Cs source, and a NaI(T1)crystal are employed to measure the pulseheight. The 137Cs source is located along themajor axis of the tube and crystal so that acount-rate of 1000 counts per second is ob-tained. The entire system is allowed to warmup under operating conditions for a periodof one-half to one hour before readings arerecorded. Following this period of stabiliza-tion, the pulse height is recorded at one-hourintervals for a period of 16 hours. The driftrate in per cent is then calculated as the meangain deviation (MGD) of the series of pulse-height measurements, as follows:

where p is the mean pulse height, pi is thepulse height at the ith reading, and n is thetotal number of readings. Typical maximummean-gain-deviation values for photomulti-pliers with high-stability Cu-Be dynodes areusually less than 1 per cent when measuredunder the conditions specified above. Gainstability becomes particularly importantwhen photopeaks produced by nucleardisintegrations of nearly equal energy are be-ing differentiated.

In the count-rate stability test, the photo-multiplier is first operated at 10,000 countsper second. The photopeak counting rate isthen decreased to 1000 counts per second byincreasing the source-to-crystal distance.The photopeak position is measured andcompared with the last measurement madeat a counting rate of 10,000 per second. Thecount-rate stability is expressed as thepercentage gain shift for the count-ratechange. It should be noted that count-ratestability is related to the hysteresis effectdiscussed above. Photomultipliers designedfor counting stability may be expected tohave a value of no greater than 1 per centgain shift as measured by this count-ratestability test.


Life ExpectancyThe life expectancy of a photomultiplier,

although related to fatigue, is very difficultto predict. Most photomultipliers will func-tion satisfactorily through several thousandhours of conservative operation and propor-tionally less as the severity of operation in-creases. Photomultipliers do not haveelements which “burn out” as in the case ofa filament in a vacuum tube. Furthermore,loss of sensitivity which occurs with opera-tion tends to recover during idle periods orduring conservative operation.

Factors which are known to affect lifeadversely are high-current operation,excessive-voltage operation, high photocath-ode illumination, and high temperature.

Operation of photomultipliers in regionsof intense nuclear radiation or X-rays mayresult in an increase in noise and dark cur-rent as a result of fluorescence and scintilla-tion within the glass portions of the tubes.Continued exposure may cause darkening ofthe glass and a resultant reduction intransmission capability.

DARK CURRENT AND NOISEThe lower limit of light detection for a

photomultiplier tube is determined in manycases by the electrical noise associated withthe anode dark current. There are severalsources of dark current in a photomultiplier.These sources are described below,Sources of Dark Current

Dark current in a photomultiplier tubemay be categorized by origin into threetypes: ohmic leakage, dark or “thermionic”emission of electrons from the cathode andother elements of the tube, and regenerativeeffects .

Ohmic leakage, which results from the im-perfect insulating properties of the glassstem, the supporting members, or the plasticbase, is always present. This type of leakageis usually negligible, but in some tubes it maybecome excessive because of the presence ofresidual metals used in the processing of thephotocathode or the dynodes. Condensationof water vapor, dirt, or grease on the outsideof the tube may increase ohmic leakagebeyond reasonable limits. Simple precau-tions are usually sufficient to eliminate thissort of leakage. In unfavorable environmen-tal conditions, however, it may be necessary

53


to coat the base of the tube with moisture-resisting materials, which may also preventexternal arc-overs resulting from highvoltage.

Ohmic leakage is the predominant sourceof dark current at low-voltage operatingcondition. It can be identified by its propor-tionality with applied voltage. At highervoltages, ohmic leakage is obscured by othersources of dark current.

Fig. 60 shows the typical variation of dark

VOLTS PER STAGE

92CS-32346

Fig. 60 - Typical variation of dark currentwith voltage for a multiplier phototube.

current of a photomultiplier tube as a func-tion of applied voltage. Note that in the mid-range of voltage, the dark current followsthe gain characteristic of the tube. Thesource of the gain-proportional, dark cur-rent is the dark or thermionic emission ofelectrons primarily from the photocathode.Because each electron emitted from thephotocathode is multiplied by the secondary-emission gain of the tube, the result is anoutput pulse having a magnitude equal to thecharge of one electron multiplied by the gainof the tube. (There are statistical amplitudevariations which will be discussed later.)Because the emission of thermionic electronsis random in time, the output dark currentconsists of random unidirectional pulses.The time average of these pulses, which may

54

be measured on a dc meter, is usually theprincipal dc component of the dark currentat normal operating voltages. The limitationto the measurement of very low light levels isthe variable character of the thermionicdark-current component. It is not possible tobalance out this wide-band noise componentof the photomultiplier tube, as it might be tobalance out a steady ohmic-leakage current.Nevertheless, it is usually advantageous tooperate the photomultiplier tube in the rangewhere the thermionic component is domi-nant. In this range, the relationship betweensensitivity and noise is fairly constant as thevoltage is increased because both the photo-electric emission and the thermionic emis-sion are amplified by the same amount.

Typical dark emission current densities forvarious photocathodes are given in Table I(page 16). The resulting anode dark currentmay be estimated by multiplying the darkemission per unit area at the photocathodeby the photocathode area and by the gain ofthe photomultiplier tube at the desiredoperating voltage.

The thermionic component of the darkcurrent varies in a regular way withtemperature as illustrated in Fig. 16, andbecause the thermionic component of thedark current is a source of electronic noise inthe anode circuit, it is frequently advan-tageous to cool the photomultiplier and takeadvantage of the reduced dark current andnoise. Various cryostats have been de-signed62,66 providing low temperatureoperation of photomultipliers. One practicalconsideration is the prevention of condensa-tion of moisture on the window. In a Dewar-type arrangement, condensation may not bea problem; in simpler set-ups moisture con-densation may be prevented by a controlledlow-humidity atmosphere at the externalwindow. On some types of photocathode,too cool a temperature may result in thephotocathode becoming so resistive that thephotoemission is blocked by a drop in poten-tial across the photocathode surface. Seeearlier section on Current-Voltage Charac-teristics. Commercial cryostats or cooledphotomultiplier chambers are availabledesigned especially for photomultiplieroperation67.

At higher dynode voltages, a regenerativetype of dark current develops, as shown in

Fig. 60 . The dark current becomes very er-ratic, and may at times increase to the prac-tical limitations of the circuit. Continuedflow of large dark currents may causedamage to the sensitized surfaces. Somepossible causes of the regenerative behaviorwill be discussed in more detail later. Allphotomultiplier tubes eventually becomeunstable as the gain is increased.Dark-Current Specification

Dark-current values are often specified ata particular value of anode sensitivity ratherthan at a fixed operating voltage. Specifica-tions of dark current in this manner are moreclosely related to the actual application ofthe photomultiplier.

The best operating range for a givenphotomultiplier can usually be predictedfrom the quotient of the anode dark currentand the luminous sensitivity at which thedark current is measured. This quotient isidentified as the Equivalent Anode DarkCurrent Input (EADCI) in the TechnicalData for individual photomultiplier tubes;and is the value of radiant flux incident onthe photocathode required to produce ananode current equal to the dark currentobserved. The units used in specifying EAD-CI are either lumens or watts at the wave-length of maximum cathode responsivity orwatts at a specified wavelength.

The curves in Fig. 61 shows both typicalanode dark current and equivalent anodedark current input (EADCI) as functions ofluminous sensitivity. The optimum operat-ing range occurs in the region of theminimum on the EADCI curve, the region inwhich the signal-to-noise ratio is also near itsmaximum. The increase in the EADCI curveat higher values of sensitivity indicates theonset of a region of unstable and erraticoperation. Many curves of this type also in-clude a scale of anode-to-cathode supplyvoltage corresponding to the sensitivityscale.

Equivalent Noise Input-The dark currentin a photomultiplier is the average currentvalue of the output pulses occurring at ran-dom intervals plus the dc leakage current.Fluctuations or noise associated with thesepulses limit precision of measurement,rather than the particular dark current value.Noise from a photomultiplier may beevaluated in terms of a signal-to-noise-ratio


measurement. If the type of modulation andbandwidth used in the measurement isknown, an equivalent noise input, ENI, canbe calculated from the signal-to-noise ratio,Equivalent noise input is defined as the valueof incident luminous or radiant flux which,when modulated in a stated manner, pro-duces an rms output current equal to the rmsnoise current within a specified bandwidth,usually 1 Hz.

92c3-32347

Fig. 61 - Illustrative data showing the varia-tion of anode dark current and the equivalentanode-dark current input (EADCI) as a func-tion of luminous sensitivity for a type 8575.Operation of the tube at voltages higher thanthat for the minimum of the EADCI character-istic does not provide the best signal-to-noiseratio.

Noise Equivalent Power-Another way tocategorize the limit of detection of aphotomultiplier is by noise equivalent power(NEP) which is essentially the same as EN1except the units are always in watts. NEP isthe radiant flux in watts at a specified wave-length incident on the detector which gives asignal-to-noise ratio of unity. The frequencybandwidth (usually 1 Hz) and the frequencyat which the radiation is chopped must bespecified as well as the spectral content ofthe radiation (most often, monochromaticradiation at the peak of the detectorresponse). It should be noted that NEP is

55


frequently specified in units of wattsHZ-1/2. The numerical value of this for-mulation of NEP is the same as that given inunits of watts but with a specified bandwidthof 1 Hz.

Detectivity-Detectivity (D) is the recipro-cal of NEP; it is expressed in W - 1. Detectiv-ity is a figure of merit providing the same in-formation as NEP but in the reverse sense sothat the lower the radiation level to whichthe photodetector responds, the higher thedetectivity.Regenerative Effects

Dynode Glow. Although photomultipliersare designed to minimize regenerative ef-fects, at some high voltage and gain almostall photomultipliers exhibit breakdownphenomena. One source of regeneration inphotomultipliers is the glowing of thedynodes under electron bombardment.68

The glow has a blue spectral emission and ofcourse is most prominent in the latter stageswhere the current is highest. The regenera-tive effect occurs when the light from thedynode glow is scattered and reflected backto the photocathode. Dynode cage shieldsand opaque support wafers minimize this ef-fect .

Glass Charging Effects. Regenerative pho-tomultiplier currents may also be triggeredby the electrostatic potential of the bulbwalls surrounding the dynode or photocath-ode structure. Particularly when the poten-tial of the bulb is near anode potential, strayelectrons may be attracted to the bulb andcause the emission of light on impact, de-pending upon the nature of the glass surfaceand the presence of contamination. Secon-dary electrons resulting from the impact ofstray electrons on the glass surface are col-lected by the most positive elements in thetube and help maintain the positive potentialof the inner surface of the glass. Under thesecircumstances, it is possible to observe theformation of glowing spots on the inside ofthe glass bulb, provided the eye is darkadapted and the applied voltage is sufficient-ly high. Some of this fluorescent light maybe reflected back to the photocathode andresult in an increase in the photocathodedark (or light) emission.

Shielding. The effect just described can beminimized by controlling the external poten-tial of the glass envelope. Fig. 62 shows the

56

effect of various voltages applied to an exter-nal shield around the tube envelope of a1P21 photomultiplier. The graph shows theequivalent noise input decreasing as theshield is made negative toward the-potentialof the photocathode. It should be noted alsothat an actual contact is not always requiredto produce the effects noted in Fig. 62. Theproximity of a positive potential near theglass bulb can cause a noisy operation.

4

Fig. 62 - Effect of external-shield potentialon the noise of a 1P21 photomultiplier. Notethe desirability of maintaining a negative bulbpotential.

Operation of a photomultiplier tube withan improper external shield may not onlycause an increase in noise or lead to an elec-trical breakdown of the tube, but can resultin damage to the photocathode and reducedtube life. In order to prevent these effects,the envelope wall should be maintained nearphotocathode potential by wrapping orpainting it with conductive material and con-necting this material to cathode potential.The connection is usually made through ahigh impedance to reduce the shock hazard.If a cathode potential shield is not provided,the glass surface in the vicinity of the photo-cathode must be insulated from any sourceof potential difference so that leakage cur-rents to the bulb are less than 10 -12 ampere.

In photomultiplier tubes in which the pho-tocathode is of a transmission type, on theinside surface of the glass bulb, it is par-ticularly important to avoid a positivevoltage contact on the external surface of thephotocathode window. In this case, ioniccurrents can flow through the glass and pro-duce a fluoresence and an accompanyingnoisy photocathode current.

It should also be noted that continuedoperation of a photomultiplier tube with apositive voltage contact to the glass in thephotocathode area can cause a permanentdamage to the photocathode. The damage isreported to be the result of ionic conductionthrough the glass and poisoning of the pho-tocathode by sodium ions.69

Afterpulsing. Afterpulses, which maybe observed when photomultipliers are usedto detect very short light flashes as in scin-tillation counting or in detecting short laserpulses, are identified as minor secondarypulses that follow a main anode-currentpulse. There are two general types of after-pulses; both are characterized by their timeof occurrence in relation to the main pulse.The first type results from light feedbackfrom the area of the anode, or possibly cer-tain dynodes, to the photocathode; the in-tensity of the light is proportional to the tubecurrents. When this light feedback reachesthe cathode, the afterpulse is produced.Afterpulses of this type, characterized by adelay in the order of 40 to 50 nanoseconds,may be a problem in many older photomulti-pliers having open dynode structures. Thetime delay experienced with this type ofafterpulse is equal to the total transit time ofthe signal through the photomultiplier plusthe transit time of the light that is fed back.

The second type of afterpulse has beenshown to be the result of ionization of gas inthe region between the cathode and firstdynode. The time of occurrence of the after-pulse depends upon the tube dimensions, thetype of residual gas involved and the mass ofthe gas ion, but usually ranges from 200nanoseconds to well over 1 microsecondafter the main pulse. When the ion strikesthe photocathode, several secondary elec-trons may be emitted; thus, the resultingafterpulse has an amplitude equal to severalelectron pulse-height equivalents. Thesepulses appear to be identical to the largerdark-current pulses, and it is suspected thatmany of the dark-current pulses are the re-sult of photocathode bombardment by gasions.

Several gases, including N2 + and H2 + ,are known to produce afterpulses. Each gasproduces its own characteristic delay follow-ing the main pulse. The most troublesome,perhaps, is the afterpulse caused by the H2 +


ion; this afterpulse occurs approximately 300nanoseconds after the main pulse in a tubeof type 7850 construction. One source ofhydrogen in the tube is water vapor absorbedby the multiplier section before it is sealed tothe exhaust system. Other gases which maycause afterpulsing may be present as a resultof outgassing of the photomultiplier partsduring processing or operation. Present pho-tomultiplier processing techniques aredesigned to eliminate or at least to minimizethe problem of afterpulsing.

Helium Penetration71,72 Another effectmust be considered in relation to the sourcesof dark current-the penetration of heliumthrough the glass of photomultiplier tubes.When the photomultiplier is operated orstored in an environment where helium ispresent, helium will gradually permeatethrough the glass envelope. Because heliumis inert, it does not react with the photocath-ode or dynode surfaces. But tubes subjectedto such an environment will exhibit a noiseincrease and an increase in afterpulsingbecause of the ionization of the helium byelectron impact. Depending upon the degreeof permeation, a point will be reached atwhich complete ionization and electricalbreakdown occurs making the tube unus-able.Other Noise Sources

Excess noise or dark current can also re-sult from field emission occurring within thetube and from scintillations in the glassenvelope of the tube caused by radioactiveelements within glass (most glasses containsome radioactive 40K). Fused silica issometimes utilized in photomultiplier face-plates to minimize these effects.

Noise in photomultiplier tubes can alsoresult from the proximity of nuclear sourcesor from cosmic rays which result in glassscintillations. Andrew T. Young73, * hasidentified large pulses which originate fromCerenkov light flashes produced by cosmicrays traversing the window of end-on photo-multipliers. The flashes correspond to

*Andrew T. Young also has written a chapter, “Photo-multipliers: Their Cause and Cure,” in Volume 12 ofMethods of Experimental Physics: Astrophysics, L.Marton, Editor, Academic Press, 1974. As well as beinga good general reference on photomultipliers, this chap-ter contains a further discussion of the Cerenkov-light-flash effect.

5 7

PhotomultIplier Handbook

photoemission pulses of 50 electrons andlarger. The number of pulses is greater whenthe face of the photomultiplier is upwardrather than downward because the Cerenkovradiation is emitted in a conical pattern awayfrom the direction of entry. Young reportsthe total number of pulses from this sourceto be about 1.2 min - 1 cm -2.

Another phenomenon which deservesmention in connection with dark current isthe effect of previous exposure to light,especially blue or near ultraviolet. Large in-creases in photocathode dark emission mayoccur as discussed in section on Photocath-ode Stability with rather slow recovery, as il-lustrated in Fig. 47. It is advisable, there-fore, that photomultipliers be kept in thedark at all times, or at least for many hoursbefore they are used for making low-levelmeasurements.Noise Output of a Photomultiplier

The output current of a photomultiplierconsists of a train of unidirectional electricalpulses whether the tube is in the dark or withillumination on its photocathode. Each pulseis the result of an electron emitted from thephotocathode and amplified by the secon-dary emission of the tube. (Some pulses mayoriginate also from light striking the firstdynode or from thermionic emission fromthe first or other dynodes.) In a system hav-ing a wide bandwith, the individual pulsesmay be measured and counted. Their spac-ing in time will have a statistical variation aswill their height. There variations constitutenoise and limit the precision of measure-ments.

If the bandwidth is not sufficient toresolve the individual pulses, particularlywhen the light level is relatively high, theoutput will exhibit noise or rapid variation insignal level about an average value. Thenoise level relative to the signal level willdecrease with increasing signal level or withdecreasing bandwidth. The noise may be de-scribed by giving the rms value of the currentvariation in a specified bandwidth.

Dark-Current and Noise ReductionDark-Current Reduction. If care is taken

to avoid damage to the photomultiplier byoperation with excessive current, the darkcurrent can often be reduced by a process ofoperating the photomultiplier in the dark at

or near the maximum operating voltage.This process, called dark aging, may requireseveral hours to several days. After such aprocess of aging, it is recommended that aphotomultiplier be operated for severalminutes at the reduced voltage before mea-surements are attempted.

Dark Noise Reduction with Cooling.Because the dark emission is reduced as thetemperature of the photocathode is reduced,the dark noise output of the photomultipliermay also be reduced by cooling. Fig. 63

TUBE

Fig. 63 - Equivalent noise input in lumens fora 1P21 photomultiplier as a function of tem-perature.

shows the variation of equivalent noise in-puts, ENI, for a 1P21 (opaque Cs3Sb photo-cathode) over a wide range of temperature.The implication of these data is that by cool-ing from room temperature to - 150 °C thelow light level limit of detection can bereduced by two orders of magnitude.Photomultiplier Noise Characteristics

The following paragraphs describe thenoise and signal levels from both a pulse anda dc point of view. The noise frequency spec-trum is also described and data are presentedshowing the dark noise spectrum.

At very low signal levels, the detectionlimit will be shown to be determined by thedark current of the photomultiplier and itsassociated noise. At high levels, the precisionof measurement is limited by the statisticalvariation in the signal pulses-or the noisewhich is always present to some degree in thedc signal level. The signal-to-noise ratio can


be improved by decreasing the bandwidth indc measurements or by the equivalent of in-creasing the time in pulse-counting applica-tions.

Noise Spectrum. The width of an outputcurrent pulse initiated by an electron fromthe photocathode is determined by the varia-tions in transit time through the tube. Thenoise associated with these pulses of electroncurrent is flat with frequency out to frequen-cies corresponding to the width of the in-dividual pulses. (The frequency spectrum ofa delta function is flat.) A power spectrumof the noise from type 931 has been calcu-lated by R. D. Sard74 and is shown in Fig. 64.Sard’s calculation was done by considering

FREQUENCY - MHz92cs-32349 Fig. 65 - Typical Dark-Pulse Spectrum

Fig. 64 - Spectral energy distribution of thenoise from type 931A operated at 100 volts perstage, as calculated by R. D. Sard.74

the anode current from a single electrontraversing the space between the last twodynodes and anode and the distributionof electron arrival times due to differences intransit time between stages; then, the Fouriertransform of the pulse shape was taken toobtain the frequency spectrum.

The bandwidth of the noise spectrum dif-fers in different tube designs, dependingupon transit-time variations. Usually, high-speed photomultipliers are characterized byrise and fall times of the output pulses ratherthan by bandwidth. This subject is discussedin more detail in a later section.

Dark Noise Pulse Spectrum. Because ofthe statistical nature of the secondary emis-sion gain at each stage of the photomulti-plier, the output pulses have a fairly wide

distribution of heights even though eachpulse is initiated by a single electron. Fig. 65

2

PULSE HEIGHT- PHOTOELECTRON EQUIVALENTS

92CM-32350

shows a differential dark-noise pulse spec-trum-the number of pulses counted perunit of time as a function of their height.

A differential dark-noise spectrum is ob-tained with a multichannel pulse-heightanalyzer. The calibration of the single-photoelectron pulse height is determined byilluminating the photocathode with a lightlevel so low that there is a very low probabili-ty of coincident photoelectron emission. Thedark-pulse distribution is then subtractedfrom the subsequent combination of darkpulses and single-photoelectron pulses, sothat the remainder represents only thatdistribution resulting from single-photoelec-tron events. By adjusting the gain of thepulse-height analyzer, the single-electronphotopeak can be placed in the desired chan-nel to provide a normalized distribution.

The dark-pulse spectrum of Fig. 65 ischaracteristic of photomultipliers intendedfor use in scintillation counting and other

59


low-light-level pulse applications. The curveshown is idealized and represents an averageor typical spectrum. An actual spectrumshows statistical variations depending uponthe length of the count.

The slope of the curve for the pulse-heightregion between 1 and 4 photoelectrons is asexpected for single-electron emission, whenthe statistical nature of secondary-emissionmultiplication is considered. The number ofpulses in this region may be reduced by cool-ing the photomultiplier. Below a pulse heightof one photoelectron equivalent, the curve isdetermined partly by the statistical spreaddue to the multiplication process and partlyfrom emission from some of the dynodesurfaces.

The slope of the curve for the pulse-heightregion greater than 4 photoelectrons ispresumed to be caused by multiple-electron-emission events. These multiple pulses arecaused by processes such as ionic bombard-ment of the photocathode. Other mecha-nisms contributing to the noise spectrum in-clude cosmic rays, field emission, andradioactive contaminants that produce scin-tillations within the glass envelope. Coolinghas little effect upon reduction of thenumber of these multiple electron pulses, butextended operation of the tube may improveperformance. Operation of the tube mayresult in erosion of sharp points and reducethe possible contribution of field effects. Inaddition, improvement occurs becauseresidual gases are absorbed within the tube,ion bombardment of the photocathode isreduced, and the resulting multiple electronemission is lessened.

For many applications it is useful to havea summation of the total number of darkpulses. In Fig. 65, for example, the sum ofdark-pulse counts from 1/8th electronequivalent height to 16 electron equivalentheights is 4 x 104 counts per minute.

Analytical Model of Noise and Signal-to-Noise Ratio-Consider the lower limit oflight detection capability of a photomulti-plier as determined by the fluctuation in thethermionic dark emission. If the dark emis-sion current from the photocathode is id, therms shot noise associated with this current isgiven by

60

.(16)

where e is the charge on the electron and B isthe bandwidth of the observation,

The photocathode dark emission and itsassociated noise are both amplified by the

for the moment that the amplification pro-cess is noise free, and that all of the currentemitted from the photocathode is collectedby the first dynode, the anode dark current isgiven by

(17)

and the anode rms noise current is given by

(18)

Actually, noise is introduced by the secon-dary emission process. (This subject is dis-cussed at length in Appendix G.) If one as-sumes Poisson statistics for the secondaryemission process, the anode noise currentshown in Eq. 18 should be increased by a

dary emission ratio per stage, assumed to bethe same for all stages. If a typical value of

factor of 1.15. Actual measurements indi-cate a somewhat higher figure. On the otherhand, higher voltage on the first stage, or theuse of a GaP:Cs first dynode having a veryhigh secondary-emission ratio would reducethis factor. For the purpose of the presentdiscussion, this increased noise from secon-dary emission will be neglected and thephotomultiplier anode noise will be taken asthe amplified photocathode shot noise asgiven by Eq. 18.

A fundamental advantage of a photomul-tiplier as compared with a photodiode, is thehigh gain and the fact that the secondaryemission process contributes very little to therelative noise output of the tube. The high .gain of the photomultiplier permits the useof a relatively small load resistance (R)without deterioration of the photomultipliersignal-to-noise ratio by the Johnson thermalnoise of the load resistance. The small loadresistance permits an operation of the photo-multiplier with the very high bandwidths in-herent in the photomultiplier design. A largeload resistance shunted by the finite tube andlead capacitance would otherwise seriouslylimit the effective bandwidth of the system.


In order to maintain the fundamental tion and cable used in pulse applica-signal-to-noise capability of the photomulti- tions-the two noise sources are aboutplier, the Johnson noise of the load resis- equal. Therefore, at this point in the exam-tance must be less than the photomultiplier ple chosen there is some, but not a serious,output noise. Johnson noise (rms) is given by reduction of signal-to-noise ratio.

where k is Boltzmann’s constant and T is thetemperature in Kelvin units. In order to com-pare this noise with the photomultiplier out-put noise we may convert it to an equivalentcurrent noise by dividing by the load resis-tance:

I (20)

Fig. 66 illustrates the relative magnitudes

Fig. 66 - Comparison of the magnitudes ofJohnson noise in the load resistance and ofthe output photomultiplier dark noise, both inunits of rms amperes per square root hertz.For the case illustrated, both sources ofnoise are equal for a load resistance of 50ohms.

Noise in Signal. When the photocathodecurrent representing the signal is larger thanthe photocathode dark emission, the domi-nant noise is the noise in the signal. For aphotocathode signal current, is, the anodesignal current Is, is given by

.(21)

and, paralleling equation 18, the anode rmsnoise current associated with the signal cur-rent is given by

Irms (22)

But, one must consider both the dark emis-sion noise and the noise in the signal current.Because there is no correlation between thetwo sources, the noises may be added bysumming their squares, or for total rms noisecurrent:

(23)

An expression for the signal-to-noise ratiomay now be written:*

SNR = iS

Fig. 67 illustrates the variation of signal-to-noise according to equation 24 as a func-tion of bandwidth and photocathode signalcurrent. The cathode dark emission is as-sumed to be 10 - 15 ampere. For signal cur-rents less than the dark emission, the darknoise is the limit to the signal-to-noise ratio.In this case, a narrow bandwidth is a require-ment for detection. When dark noise is thelimit to detection, it is useful to reduce darkemission by cooling. At wider bandwidthsand higher photocathode signal levels, thelimit to the signal-to-noise ratio is in the

*In this discussion, the signal is treated as though it weredc. In an actual application, the signal may bemodulated and signal-to-noise figures would involve therms value of the modulated signal. Note also that thenoise calculation neglects the factor by which the noiseis increased by the secondary emission statistics.

61


noise of the signal so that cooling would notbe advantageous. In Fig. 67, it has been as-sumed that the load resistance is sufficient(greater than 50 ohms) so that Johnson noiseis not a factor.

In applying these signal-to-noise conceptsto an actual case, the photocathode darkcurrent may differ greatly from the 10- 15

ampere figure assumed. (See the data, Fig.16.) The photocathode signal currents maybe calculated from a knowledge of the pho-tocathode responsivity as published in a tubedata sheet. For example, a photocathodemay have a responsivity of 160 micro-amperes per lumen for tungsten irradiation,or 70 milliamperes per watt at the wave-length of maximum responsivity (i.e., 420nanometers). A photocathode current of 1femtoampere would then correspond to

10-15/70x 10-3= 1.43x 10-14 watt fluxincident on the photocathode.

TIME EFFECTSTerminology

Photomultiplier tubes may be character-ized for time response in various ways. The

*These definitions follow the standards appearing in“IEEE Standard Test Procedures for Photomultipliersfor Scintillation Counting and Glossary for ScintillationCounting Field.” ANSI N 42.9 - 1972; IEEE Std 398-1972. Published by The Institute of Electrical and Elec-tronics Engineers, Inc., 345 East 47 St., New York,N.Y. 10017.

62

terms used in these time characterizationsare defined below. In these definitions it isassumed that the photomultiplier is activatedwith a delta-function light pulse. Fig. 68 il-lustrates the definitions of some of the morecommon terms. *

Fig. 68 - The various time relationships in aphotomultiplier output pulse, assuming adelta excitation function. Illustrated are tran-sit time, rise time, fall time, and full width athalf maximum (FWHM).

Transit time† “is the mean time differencebetween the incidence of the light upon thephotocathode (full illumination) and the oc-currence of the half-amplitude point on theoutput-pulse leading edge.”

Rise time “is the mean time difference be-tween the 10- and 90-percent amplitudepoints on the output waveform for fullcathode illumination and delta-function ex-citation.”

Fall time “is the mean time difference be-tween the 90- and 10-percent amplitudepoints on the trailing edge of the output-pulse waveform for full cathode illuminationand delta-function excitation.”

†This definition differs from the definition appearing inthe IEEE Standard Dictionary of Electrical and Elec-tronics Terms (ANSI/IEEE Std. 100-1977: “The timeinterval between the arrival of a delta-function lightpulse at the entrance window of the tube and the time atwhich the output pulse at the anode terminal reachespeak amplitude.”). This latter definition is closer to theactual transit time which might be defined to theaverage time of arrival of the electrons at the output.However, the definition given to the half-amplitudepoint may be more useful in an application where thehalf-amplitude point is used for timing purposes.

Full-width-at-half-maximum (FWHM) isthe mean elapsed time between the half-amplitude points on the output waveformfor full cathode illumination and delta-function excitation.

Delta-function light pulses. In tests relatedto time characterization of photomultipliertubes it is useful to have light pulse sourcesavailable which approach characterization asa delta-function pulse. A delta-functionpulse is one whose duration is significantlyshorter than that of the output pulse to bemeasured. It approaches the mathematicalconcept of a function whose area is finite butwhose width approaches zero.

Various light sources have been used togenerate delta-function pulses for time-testing of photomultipliers.

A reverse-biased light-emitting diode (Fer-ranti type XP-23) has been used byLeskovar75 to obtain light pulse widths of asshort as 200 ps. Mercury-wetted relay sparksources have also been used and have pro-vided pulses having rise times of 500 ps.Pulses of radiation having a duration of 50ps or less can be obtained with a mode-locked Nd:YAG laser. The laser wavelength

radiation at 532 nm by means of a nonlinearcrystal. Random pulses can be producedwith fast scintillators. A rise time of 400 pscan be obtained for Naton 136 and a 60 Cosource.81

Transit TimeTransit Time is expected to increase as the

inverse half power of the applied voltage,provided the effects of initial velocities andsecondary-emission delay are negligible. Fig.69 shows the transit time measurements foran 8053 photomultiplier plotted so as todisplay the reciprocal square root voltagerelationship. The intercept of the line on thetime axis is probably due to the distortion ofthe simple relationship by the magnitude ofthe initial secondary emission velocities.

Fig. 70 shows the transit times for anumber of photomultiplier tubes plottedover a range of operating voltages. Thelarger part of the transit time is just the ac-cumulation of times for electrons to traversefrom stage to stage. Usually, the time for thephotocathode-to-first-dynode transit is thelargest component, especially for photomul-


tipliers with large photocathode areas. Thelarge photocathode necessitates a fairly longpath to the first dynode to provide goodphotoelectron collection from the entirephotocathode.

SUPPLY KILOVOLTS BETWEENANODE AND CATHODE

92CS-32406

Fig. 69 - Transit time as a function of thesquare foot of reciprocal applied voltage for atype 8053 photomultiplier tube.

Fig, 70 - Transit time as a function of supplyvoltage (log scales) for a number of photomul-tiplier tubes.

Photo- and secondary-emission times maybe as short as 10 ps, although for negative-electron-affinity (NEA) materials, the timesmay be as long as 100 ps. One reason for the

63

PhotomultiplIer Handbook

high performance of NEA materials is thatlong diffusion paths for electrons are ob-tained. Although this increase in diffusionpath length also increases the emission time,emission time is not a significant part of thetransit time in most commercial photomulti-pliers. In general, the transit time of a photo-multiplier is not as important as its rise timeor as variations in the transit time whichwould cause uncertainty in time measure-ments made with the photomultiplier. Afixed delay time is easily compensated for bycircuit design.

Rise TimeFig. 71 shows the rise time-from the 10

Fig. 71 - Anode-pulse rise times as a func-tion of anode-to-cathode applied voltage (logscales) for a number of photomultipliers.

to 90% amplitude points-for a number ofphotomultipliers plotted over a range oftypical operating voltages. No correctionwas made for the finite rise time of the lightpulse and measuring equipment, which isestimated to be of the order of 0.8 ns. In us-ing photomultipliers for high-resolution-time spectroscopy, the ideal point on theoutput pulse to use as an indicator would beat the half maximum of the rising character-istic. The rising characteristic is generallyfaster than the fall and thus provides thehighest precision in timing. However,because output pulse heights vary, a fixeddiscriminator level results in a loss of preci-sion. When a fixed discriminator level isused, the highest precision is obtained by useof a discriminator level between 10 and 20%of maximum. A superior method is the useof a constant fraction of the pulse height as atrigger .76

Pulse WidthThe width of the output pulse is deter-

mined by the variation in transit timethrough the secondary emission chain of thephotomultiplier. The variations arise be-cause of variations in emission energies anddirections of the secondary electrons asrelated to the tube structure. The measure-ment of pulse width is generally the fullwidth at half maximum. The output pulsewidth follows the inverse half power of theapplied voltage as does the average transittime. Fig. 72 illustrates the pulse width(FWHM) for an 8053. For timing experi-ments, it is generally desirable to have a nar-row pulse width for good timing precisioncapability.

KILOVOLTS 92CS-32409

Fig. 72 - Pulse width (full width at half max-imum) for a type 8053 as a function of the in-verse half power of the applied voltage.

Pulse Jitter (Time Resolution)Although pulse timing is done on the ris-

ing characteristic of the output pulse and ismore precise for a fast rise time, the ultimatelimit to time measurement is the variation inpulse timing, or pulse jitter. Suppose singlephotoelectrons initiate pulses. Variations intransit time of photoelectrons to the firstdynode will occur because of variations inthe initial velocity and electric field resultingfrom the electrode geometry. The same con-siderations apply to the secondary electrons.If a number of pulses initiated by single elec-trons are observed, a histogram can be devel-oped showing the number of pulses having a

64


given transit-time difference. Such a histo-gram, measured by Birk, Kerns, andTusting,77 is shown in Fig. 73. The time ofeach pulse was measured by using theleading-edge half-height point. The fullwidth at half maximum of this distribution isabout 360 ps and is a measure of the timeresolution capability of this particular tube.

TRANSIT TIME DIFFERENCE - ns

92CS-32410

Fig. 73 - Histogram of transit-time differencefor single-photoelectron pulses from an RCAdevelopmental type photomultiplier. (FromBirk, Kerns, and Tusting77.)

If a number, N, of simultaneous photo-electrons is emitted, the pulse jitter is re-duced by the square root of N, simply by thestatistical averaging process. This relation-ship has been demonstrated by Leskovar andLo75 for a microchannel-plate photomulti-plier as shown in Fig. 74.

Data on time resolution for single photo-electrons for a variety of photomultipliersare given in Table IV. Values are given forfull photocathode illumination as well as forillumination at the center point. The timespread for full photocathode illumination islarger than for a single point because it in-cludes the difference in transit time from thephotocathode to the first dynode for dif-ferent photocathode locations.

92CS-32411

Fig. 74 - Time resolution of a microchannel-plate photomultiplier as a function of thenumber of photoelectrons per pulse, mea-sured with light pulse width of 2.6 ns, for fullphotocathode illumination. Data are fromLeskovar and LO75 The dashed line showingthe inverse square root relationship has beenadded.

Table IV - Time Resolution (FWHM)* for Single Photoelectrons for Various Photomultipliers

65


The improvement of the 8852 over the8575 is the result of the use of the high-gainfirst dynode. The difference between the8850 and the 8852 may be due to the largerphotoelectron emission energies of the multi-alkali photocathode. The poorer time resolu-tion of the 8854 is caused by the very largecathode area and the consequent reducedelectric field strengths and increased pathlengths.

A general treatment of single photoelec-tron detection and timing has been provided

tion for a photomultiplier is very similar tothat for noise in a photomultiplier as given inAppendix G. Variations in transit time in theearly stages of the photomultiplier are mostimportant to the over-all time resolution.The large number of electrons in the latterstages bring about an averaging processwhich reduces the time variation.

Summary of Time-Resolution StatisticsThe summary of time resolution statistics

below follows the work of Gatti and

If the secondary emission of each stage is

electron flight times including all of n stagesis given by

variance of the flight time from the ithdynode to the (i + 1)th dynode. The lastterm represents the variance in flight timefrom the last dynode to the anode. This ex-pression is actually a simplification becauseof the induction effect of the electrons inflight between the last stage and anode.

is negligible. If n is large and the variances offlight times for each stage are assumed to be

follows:

(26)

When variations in secondary emission aretaken into account, assuming all stages have

a Poisson distribution, the total variance in

time of flight for the over-all tube is given by

(27)

Eq. 27 refers to the variance of the centroidof the output pulse from a single photoelec-tron input. If timing is done by a point onthe rising characteristic, an additional timevariance might be included, that for thevariation in pulse width. Gatti and Svelto81have shown that this term would add anegligible amount to the variance as given byEq. 27.

PULSE COUNTINGOne effective way to use a photomultiplier

for measuring very weak signals is to detectand count pulses resulting from singlephotoelectrons-sometimes referred to as“photon counting”. (Actually, the highestquantum efficiencies for photocathodes aresuch that at best one in every three photonswould be detected, assuming a spectralmatch to the peak quantumefficiencywavelength.) Counting of single electronsassumes that the rate of arrival of photons issuch that it would be unusual for more thanone photoelectron to be emitted in a timeequivalent to the output pulse width for asingle electron input to the multiplier.Single-electron pulse counting is an impor-tant technique in applications such asRaman spectroscopy, astronomical photom-etry, and bio-luminescent measurements.

counting has an advantage equivalent to afactor of 1.2 in quantum efficiency overcurrent-measurement techniques.Output Pulse Height Distribution

An important consideration in photoncounting is the distribution of pulse heightsat the output of the photomultiplier. Statis-tics of the variation in pulse heights forsingle electron inputs is discussed in Appen-dix G. (See Fig. G-6). Pulse height distribu-tions are obtained with a multichannel pulse-height analyzer. Fig. 75 shows (1) a differen-tial pulse-height distribution in which thenumber of pulses in a given time interval andin a given channel (between height, h and h

an integral pulse-height distribution in whichthe total number of pulses occurring in thegiven time interval with a height of h or

66

greater is plotted.83 Also shown on thegraph is a rectangle whose intercept on theabscissa is the equivalent of the output pulseassociated with one photoelectron and anaverage photomultiplier gain of G. The rec-tangle is determined by setting its heightequal to the intercept of the integral-pulse-height distribution curve on the ordinateaxis, and by setting the area of the rectangleequal to the area under the integral-pulse-height curve. This area is equal to the totalcharge of all the pulses counted: the totalnumber of pulses multiplied by the averagemultiplier gain and the charge of one elec-tron. This equality may be demonstrated byintegrating the integral-pulse-height curvealong the ordinate axis and comparing theresult with the integral of the product of thedifferential pulse-height-distribution or-dinate (number of pulses) and the abscissavalue (pulse height or charge associated).The pulse-height resolution for single elec-trons in this case is FWHM = 1.6 electronequivalents. Note that the data of Fig. 75

FOR

92CS-32412

Fig. 75 - Single electron (1) differential and (2)integral pulse-height distribution curves. Type4501, photocathode K2CsSb, counting time 10minutes (from G. A. MortonB3). Different scalesfor (1) and (2).

were taken on a tube, 4501, that has copper-beryllium dynodes and a modest secondary-emission gain. The pulse-height resolutionfor single electrons in the case of gallium-phosphide dynodes is much better, as will bediscussed below.

When a very low light flux is measured bythe pulse-counting technique, it is also


necessary to measure the count in the darkand to subtract the rates to obtain that dueto the light alone. The optimum time whichshould be spent on each measurement isdiscussed in Appendix G-see for example,Eq. G-103. If the signal count is much lessthan the dark count, however, about equaltimes should be spent on both measure-ments. At higher signal levels, more timeshould be spent on the signal count.

Copper-Beryllium DynodesDifferential pulse-height distributions are

shown in Fig. 76 on a developmental photo-

- L I G H T

PULSE HEIGHT -ARBITRARY UNITS92cs-32413

Fig. 76 - Differential pulse-height distributionsobtained with type Dev. No. C-70101B. (Simi-lar to type 8575, but having an S-20 response.From R. M. Matheson84.)

multiplier having copper-beryllium dynodes.Note that the dark pulse-height distributiondoes not show a peak representing singleelectrons. The probable explanation is thatdark emission originates not only from thephotocathode, but from the dynodes as well.Electrons originating from the first dynodewould show a distribution similar to that ofthe photoelectrons, but with a pulse heightlower by the secondary-emission ratio of thefirst dynode. Pulse-height distributions alsousually show an extended foot, as in Fig. 76,for heights greater than 7 on the figure’s ar-bitrary scale. These counts represent pulses

67


of more than one electron per pulse and areprobably caused by secondary mechanismssuch as ion feedback, scintillations in theglass envelope initiated by cosmic rays or byradioactive traces in the glass or in the tubeenvironment. Note that the numbers of theselarge pulses are usually several orders ofmagnitude less than the numbers in thesingle-electron distribution. When a single-electron integral count is made, the upperdiscriminator level could be set to excludethese larger pulses.

In setting the lower level of thediscriminator, an optimum level is deter-mined by the different character of the light

shown that the minimum error in deter-mining the signal pulse count is obtained byadjusting the lower-level discriminator set-ting so that for an integral count distribution

are much improved. Consequently, suchtubes are particularly advantageous forsingle-electron pulse counting. The improve-ment in multiplication is demonstrated bythe differential pulse-height distribution forsingle electrons shown in Fig. 78. Note the

(28)

where Ns represents the number of signalpulses counted and Nd represents thenumber of dark pulses counted. In Fig. 77,the value of the pulse height that satisfies therelation shown in Eq. 28, is given by hl.

PULSE HEIGHT-PHOTOELECTRON EQUIVALENTS

92CS-32415

Fig. 78 - Resolution of a single electron peakhaving a measured FWHM of 63%. The firstdynode is GaP:Cs. (Data from Morton, Smith,

3

Fig. 77 - Single electron response (I) and darkpulse distribution (2) of tube type 4501 forcounting time of 10 minutes; integral distribu-tion curves. h 1 and h2 are discriminator set-tings (From G. A. MortonB3).

Gallium-Phosphide DynodesFor tubes having the high-gain GaP:Cs

first dynodes, the statistics of multiplication

improvement in resolution as compared withthat of Figs. 75 or 76. When the intensity ofthe light flashes is increased, a pulse-heightspectrum such as illustrated in Fig. 79results, again for a GaP:Cs first dynode. Inthis case the light level has been adjusted sothat the peak counts are nearly equal forone, two, and three electron pulses. Thesmall pulse distribution to the left of thesingle-electron peak distribution is probablyan artifact caused by equipment noise at thelower channel settings.Peak-to-Valley Ratio

Another important consideration for aphotomultiplier used in pulse counting is thepeak-to-valley ratio. Referring to Fig. 79,

68

for single electrons, the peak-to-valley ratiois the value of the peak of the single-electrondistribution divided by the minimum valueof the count distribution between the firstand second peaks-in this case a ratio ofabout 2.3.


Fig. 79 - Pulse-height spectrum showingpeaks corresponding to one, two, and up tofive electrons. The first dynode is GaP:Cs.

SCINTILLATION COUNTINGPulse Magnitude

In scintillation counting, the problems aresimilar to those of counting single photoelec-trons but the numbers of equivalent photo-electrons per pulse can vary from a few to afairly large number. For example, in the caseof a soft-beta emitter, using a coincidenceliquid scintillation counter (for an un-quenched standard, PPO and POPOP intoluene), the total number of photoelectronsdeveloped in the pair of photomultipliers

imately 2.5 per keV of beta-ray energy.Because dark noise from single electrons canbe effectively discriminated against by usingcoincidence techniques, a relatively efficientcount of low-energy beta emission can bemade. On the other hand, in the case of aNaI:Tl crystal coupled to a photomultiplierhaving a K2CsSb photocathode, the numberof photoelectrons developed is approximate-ly 8 per keV of gamma-ray energy. Thus, forthe isotopeenergy is 662 keV, the pulse height would bethe equivalent of 5300 photoelectrons.Isotope 137Cs Sources

Fig. 80 shows the pulse-height distribution

92CS-32417

taken with a NaI:Tl crystal and a two-inch“teacup” photomultiplier (type 4902).

Pulse-height resolution is illustrated for fullwidth at half maximum. The peak at the ex-treme left is the barium K X-ray peak at 32keV. The region between the two peaks is theresult of Compton scattering.

From Appendix G, Eq. G-111, the pulse-height resolution is given by

The following values may be assumed:

mc = number of photons corresponding to

= variance in the number of photonsper photopeak pulse

For the case illustrated in Fig. 80, theFWHM = 7.17%. Substituting the above

325,000. The FWHM related to the crystalstatistics alone is 6.33%. The FWHM relatedto the photomultiplier alone is 3.37%.

Fig. 81 shows a pulse-height distribution

energy of 5.9 keV. Note that for this spec-trum, the FWHM is approximately 40%.Following the same type of analysis as usedabove to evaluate the relative contributions

for the photomultiplier alone is 36% and forthe scintillator, 18%.

69


Narayan and Prescott86 have presenteddata illustrating a trend in which the photo-multiplier statistics dominate the pulse-height resolution at low gamma-ray energiesand the scintillator statistics dominate athigh gamma-ray energies. This trend is il-lustrated in the above analysis of the data inFigs. 80 and 81.

I I I I

Fig. 81 - Pulse-height distribution for 5sFetaken with type 8850 photomultiplier andNal:TI crystal.

As the gamma ray energy is increased, ofcourse, the relative pulse-height resolutionattributable to the photomultiplier decreasesas the square root of the number of photo-electrons per pulse or as the square root ofthe gamma ray energy. The relative pulse-height resolution attributable to the crystalalso decreases with gamma ray energy, butnot as rapidly. The light-pulse statistics ofthe crystal also varies from crystal to crystal.Key Photomultiplier Characteristics forGood Pulse-Height Resolution

Several characteristics of a photomulti-plier tube determine its suitability for obtain-ing good pulse-height resolution. Of primeimportance is the quantum efficiency of thephotocathode, especially as it matches thegenerally blue spectrum of scintillators.Secondly, of course, the photomultipliermust have a suitable electron-optical con-

figuration so that all or nearly all of theemitted photoelectrons effectively impactthe first dynode. The statistics of secondaryemission, especially in the first stage, also af-fect the ultimate resolution. It is desirable,therefore, to provide a fairly high secondary-emission yield either by the nature of thedynode surface material or by applying arather high voltage between the photocath-ode and the first dynode. Finally, it isgenerally advantageous that the photoemis-sion be uniform across the photocathodeand that a uniform collection of all thephotoelectrons be provided. This lattercharacteristic is particularly important in thecase of thin scintillators where the light froma scintillation is more localized on thephotocathode than it would be for thickerscintillators. Light pipes may be used to im-prove the uniformity of light on thephotocathode and therefore improve thepulse-height resolution in the case of thinscintillators. For a crystal whose length iscomparable to its diameter, the output light-flux distribution is generally uniform so thatphotocathode uniformity becomes statisti-cally less significant.

Peak-to-Valley RatioAs in pulse-counting applications, anotherway of classifying the performance of photo-multipliers in scintillation counters is by thepeak-to-valley ratio of the distribution. Thisway is illustrated in Fig. 81. The peak will behigher, of course, if the resolution of thetube is better. Good resolution also results ina lower valley point before the pulse-heightdistribution increases on the low energy sideas a result of photomultiplier dark current.The peak-to-valley ratio is a particularlyvaluable parameter for measurements withsources characterized by low emissionenergy.Plateau Concept

A plateau curve for a scintillation counteris illustrated in Fig. 82. The curve is devel-oped by using a source such as 137Cs andcounting all pulses higher than a fixeddiscriminator level. The total counts are thenplotted as a function of the photomultipliervoltage. The development of the curve inFig. 82 may be understood by referring tothe pulse-height distribution curve of Fig.80. At the lowest value of photomultiplier

70


voltage, the pulse heights are less than theequivalent of the discriminator setting. Asthe voltage is increased, the discriminatorsetting moves in effect from right to left onFig. 80. The rising portion of Fig. 82 thencorresponds to the discriminator movingthrough the photopeak, through the Comp-ton scattering region of the distribution, andthrough the barium X-ray peak. The plateaucorresponds to the dark background87 of thephotomultiplier which would appear at theextreme left of Fig. 80 except for thescale-the barium X-ray peak at 32 keV cor-responds to about 250 photoelectrons. Thedark background pulse distribution of aphotomultiplier is shown in Fig. 76. As thevoltage on the photomultiplier is increased,the dark current of the tube increases be-cause various regenerative effects increase,and the plateau is terminated at the upperend of Fig. 82.

PHOTOMULTlPLlER VOLTAGE-VOLTS92CS-32419

Fig. 82 - Typical plateau is defined as por-tion of integral-bias characteristic in whichchange of counting rate per 100-volt intervalis less than a selected value (Engstrom andWeaver87).

The concept of utilizing a plateau charac-teristic probably stemmed from the plateauof Geiger-Muller tubes. In the case of photo-multipliers, the length of the plateau is deter-mined by the stability of the tube at highgain (a minimum of regenerative effects) andby the characteristic variation of gain withvoltage. That is, a tube having Cs-Sbdynodes would generally have a shorterplateau than one having Cu-Be dynodessimply because of the more rapid increase ingain with voltage for the Cs-Sb dynodes. The

plateau would also be shorter as the numberof stages is increased.

Photomultiplier plateau is of particular in-terest in the application to oil-well logging.A rather intense gamma-ray source such as137Cs is mounted in the sonde near thephotomultiplier and crystal assembly, butshielded from it. The gamma-rays result inCompton scattering in the materials near theprobe. The scattered radiation is detectedand measured with an integral count with thediscriminator set to correspond to a point onthe plateau so that essentially all of the in-tercepted radiation is counted. As the depthof the measurement in the drill hole in-creases, the temperature does also. It is,therefore, important that the counting char-acteristic of the scintillation detector beessentially independent of temperature. But,as the temperature increases, both the darkbackground count and the instability of thephotomultiplier increases and, as a result,the length of the plateau decreases. Fig. 83 il-

92cs-32420

Fig. 83 - Plateau characteristics at roomtemperature and at 150 °C for a photomulti-plier having a Na2KSb photocathode. Thephotomultiplier’s principal application is inoil-well-logging.

lustrates plateau characteristics taken at22°C and 150°C for a photomultiplierhaving a Na2KSb photocathode. The curvesindicate a range of photomultiplier voltagethat would be satisfactory for this range oftemperature.

In scintillation counting applicationswhere gamma-ray energy resolution is im-portant, the plateau concept is not par-ticularly applicable. Direct measurement ofpulse-height resolution would be more use-

71


ful. But in special applications such as theoil-well logging, the plateau characterizes theuseful range of operating voltage.

LIQUID SCINTILLATION COUNTINGLiquid scintillators are most commonly

used for the evaluation of low-energy betaemitters because of the generally short rangeof beta particles and the need to provide anintimate contact between the source and thescintillator .88,89 There are numerous fluorsor solutes which may be used, such as PPO(2,5 diphenyloxazole). A common solvent istoluene which must be miscible with the li-quid radioactive sample. The emission of thetypical liquid scintillator is in the near ultra-violet and blue spectrum so that a photomul-tiplier having a “bialkali” photocathode(K2CsSb) provides a good spectral match.Counting Techniques

When liquid-scintillation counting wasfirst introduced, the technique involved theuse of a single photomultiplier to observe theliquid scintillator . Because of the lowenergies involved and the need to measuresamples of relatively low activity, thebackground count of the photomultiplierwas a severe limitation. An improved ap-proach is the use of a pair of photomultipliertubes facing a common scintillator chamber.Because of the number of photons involvedwhen a scintillation occurs in the sample, it isquite likely that both photomultipliers willsense the flash, thus permitting the use ofcoincidence circuitry. Events that initiatebackground counts in one photomultiplierare only infrequently coincident with thosein the second tube and so a high degree ofdiscrimination against background counts ispossible.

The three most commonly used radioiso-topes in liquid-scintillation counting aretritium, 3H; carbon, 14C; and phosphorus,32P. The beta-ray energies from these radio-isotopes may vary over a fairly broad range;for example, 3H emits betas having energiesvarying from zero to a maximum of 18.6keV. Maximum beta-ray energies for 14Cand 32P are 156 keV and 1.71 MeV respec-tively. Fig. 84 illustrates the distribution ofpulse heights (which are proportional toenergy) for tritium. (The average energy isabout 6 keV.)

72

A block diagram of a liquid-scintillationspectrometer is shown in Fig. 85. If a photo-multiplier having a fast time characteristic isused in the spectrometer, the coincidenceresolving time may be as small as 10 nano-seconds.

Fig. 84 - Pulse-height spectrum representingthe beta energy for tritium (18.6 keV max-imum) from a liquid-scintillation counter.Maximum and minimum discriminator levelsare indicated (From E.D. Bransome88).

92cs-32422

Fig. 85 - Block diagram of a liquid-scintilla-tion spectrometer.


Counter EfficiencyA conventional figure of merit for a liquid

scintillation coincidence counter is the ratio,E2/B, where E is the counting efficiency ex-pressed in per cent, and B is the number ofbackground coincident counts per minute.The efficiency, E, is measured utilizing astandard sample whose disintegration rate isknown. The recorded count rate is comparedwith that of the standard, taking into ac-count the exponential decay in the disinte-gration rate. The background count value,B, is measured by utilizing a blank sample ofthe scintillator .

Note that E2/B is proportional to thesquare of a signal-to-noise figure for thesystem. Thus, the efficiency E is propor-tional to the signal; and the square root of Bis the expected standard deviation in thestatistical count.

Efficiencies of the scintillators and thephotomultipliers in liquid-scintillation-counting equipment (using the most efficientliquid scintillator and photocathodes) aresuch that about 2.5 photoelectrons are emit-ted per keV of energy of the beta ray. Thebest counting efficiency for a given radioiso-tope is obtained when a highly efficient scin-tillator and a photomultiplier having a highquantum efficiency in the spectral emissivityrange of the scintillator are used. The opticalsystem containing the two photomultipliersand the counting vial is also of major impor-tance. The system should be designed sothat, as far as possible, the photons pro-duced in a scintillation are equally dividedbetween the photomultipliers. This divisionassures that a coincidence pulse results fromas many scintillations as possible. In somecases two or more different isotopes may becounted simultaneously. It is desirable,therefore, that the photomultipliers havematched gains and good energy (pulse-height) resolution capability to provide bestisotope separation. A typical value for E isabout 60%. Quantum efficiency of thebialkali photocathode for the scintillationradiation is approximately 25 %.

Sources of Background CountsThere are various sources of the coin-

cident background counts. One source is therandom dark-noise pulses in each of the twophotomultipliers which are occasionally

coincident. The coincident rate may bepredicted by the following equation:

(29)

where C is the chance coincidence rate perminute, N1 is the dark-noise count rate incounts per minute from tube No. 1, N2 is thedark-noise count rate in counts per minute

of the coincidence circuit in seconds. In aliquid-scintillation spectrometer employingtwo tubes each having dark-noise rates of30,000 counts per minute each and a coin-cidence circuit having a resolving time of 10nanoseconds, the number of accidental coin-cidences is approximately 0.3 count perminute.Another source of background count is

cross talk between the two photomultipliersas a result of light flashes in one tube whichare sensed by the other.

Cosmic rays and other natural radiationcan result in flashes in the scintillator orpossibly in the photomultiplier envelope.Shielding the equipment with lead canreduce the background from these sources.However, the photomultipliers themselvesalso contain radioactive isotopes. A com-mon contaminant is 40K, a naturally oc-curring isotope of potassium (0.1 %), that ispresent in many glasses. Photomultipliersdesigned for use in liquid-scintillationcounters may utilize quartz face plates orthin face plates of a glass having a minimumpotassium content and with a low yield ofscintillations from gamma rays.

When a vial filled only with a scintillator isplaced between the photomultipliers, and theoutput from the coincidence circuit is ex-amined by use of a multichannel analyzer, apulse-height distribution such as that shownin Fig. 86 is obtained. Clearly, not many ofthe background pulses shown are caused bythe accidental coincidences of dark-noisepulses from the photomultipliers, but arecaused by cosmic rays of scintillations in thematerial of the vial and photomultiplierenvelope resulting from the presence ofradioisotopes in the materials of which theyare constructed.

Typical values for B, the number ofbackground counts, in coincidence liquid-

73

Photomultlpller Handbook

scintillation counters are in the range 15 to20 counts per minute, and E2/B values aretypically in the range 150 to 250 for tritium.

92CS - 32423

Fig. 86 - Background distribution obtainedfrom a liquid scintillator using a coincidencesystem.

Photomultiplier Selection forLiquid Scintillators

The E2/B figure of merit is, however, notnecessarily the best way of evaluating allsystems. It is valid at very low counting rateswhere the background count is the dominantfactor, but is not of great help at high-counting rates where the background countof the system becomes less important thanthe efficiency in determining the merit of thesystem.

In selecting a photomultiplier tube for aliquid-scintillation application, the followingitems are of major importance: high photo-cathode quantum efficiency, low dark-noisecount rate, minimum internal-light genera-tion, low scintillation-efficiency envelope,fast time response, and good energy resolu-tion. The 4501V3 photomultiplier has beenspecifically designed to meet these require-ments.

ENVIRONMENTAL EFFECTSAlthough photomultiplier tubes are not

overly sensitive to their environment and canbe handled without exercising undue cau-tion, there are various considerations theuser should be aware of in order to maximizetube life, avoid permanent damage, and pro-vide the optimum performance. This sectionprovides guidance relating to specific envi-

74

ronmental limitationstimizing performancetions.

Temperature

or to means for op-under adverse condi-

The maximum temperature to which aphotomultiplier should be exposed duringeither operation or storage is determinedprimarily by the characteristics of the photo-cathode. Temperature ratings vary with tubetype depending upon the particular photo-cathode. For a specific type, its publisheddata should be consulted. In general, how-ever, a temperature of 75 °C should not beexceeded. Excessive exposure to highertemperatures will alter the delicate balanceof chemicals of the photocathode and resultin a loss of responsivity and/or a shift inspectral response. Another reason foravoiding operation of photomultipliers atelevated temperatures is that it causes an in-crease in the dark current. (See Fig. 16.)

There are also hazards in cooling photo-multiplier tubes to reduce the dark emissionfrom the photocathode. Rapid coolingshould be avoided to minimize thermalshock which could result in cracking of thebulb or stem. Temperatures below - 50°Cshould be avoided because the difference inexpansion between the glass and the basematerial could result in cracking of the glass.This hazard is also present for photomulti-plier tubes having metal envelopes or fortubes having stiff lead glass stems whensocketed in a material of different expansioncoefficient.

In the cooling of photomultiplier tubescare must also be taken to avoid moisturecondensation. One method is to provide adry atmosphere for the tube and, especially,its base and socket. Another method is tocover the critical areas of the tube with asilicone rubber such as General ElectricRTV-11. The tube should first be cleanedand then primed (4004 primer for RTV-11)before the silicone rubber coating is applied.

Differential cooling in which only the pho-tocathode end of the tube is cooled is notrecommended. This practice may result inloss of photocathode sensitivity because ofthe vapor transport of alkali metals from thedynodes to the photocathode.

VoltageThe published data for photomultiplier

tubes contain information on the maximumsupply voltage as well as information on thespecific maxima for voltages between adja-cent electrodes such as between dynode No.1 and the photocathode. Too high a voltagebetween closely spaced electrodes can resultin electrical breakdown, particularly acrossinsulator surfaces. The over-all tube voltagemaximum is determined by considerations ofthe maximum gain which a tube can tolerate.Depending upon the construction of thephotomultiplier, at some gain in excess of107 or 10 8, there may be sufficient feed-back-perhaps by light generated in the out-put section of the tube-to result in a sus-tained electrical current. This currentbreakdown, if sufficiently large, can per-manently damage the tube by causing an in-crease in the dark current and a decrease inresponsivity .

Ground potential and shielding of thephotomultiplier can also be a problem. Inorder to minimize regenerative effects it isgenerally recommended that the wall of thephotomultiplier envelope be maintained at apotential at or near cathode potential. Thisrecommendation offers no problem in a cir-cuit in which the cathode is at ground poten-tial. But when the anode is at ground poten-tial, it may be advisable for safety reasons toprovide a double shield: the inner shield atcathode potential, surrounded by an in-sulator, and an outer shield at ground poten-tial. The inner shield should be maintainedat cathode potential through a high resis-tance (usually 5 to 10 megohms) to avoidhazard through accidental contact orbreakdown of the insulation.

It is strongly recommended that no dif-ference of potential be maintained betweenthe semitransparent photocathode layer of aphotomultiplier and its outer glass faceplate.A potential on the outside of the glass that ispositive with respect to that of the photo-cathode can result in damage to the photo-cathode by ionic transport through the glassfaceplate.69 (See section on “Shielding”earlier in this Chapter.)

In supplying the tube voltage by means ofa resistive voltage divider, care must betaken to avoid physical contact between anyof the resistors. Such contact can causeminiature electrostatic discharges resultingin noise spikes in the photomultiplier output.

Photomultlpller Characteristics

Light LevelIt is generally advisable to store photomul-

tiplier tubes in the dark and to avoid ex-cessive exposure of the tubes to any light richin blue or ultraviolet such as that from fluor-escent lamps or sunlight. Exposure of thephotocathode to such radiation will general-ly cause a very substantial increase in thedark current originating from the photocath-ode. Complete recovery from this effect maytake as long as several days. See Fig. 47. Ex-posure of the photocathode to sunlight eventhough no voltage is being applied may alsoresult in photocathode response changessuch as loss of infrared response in a tubehaving a multialkali photocathode.

Before a photomultiplier tube is used forcritical low-light-level measurements, it isrecommended that it be aged for a period of24 hours in the dark with voltage applied.This precaution is particularly recommendedafter a long period of idleness even if thetube has not been exposed to ambientlighting.

When the photomultiplier has the ap-propriate voltage applied, it is of course pru-dent to avoid any excessive light exposure.Particularly, if the voltage divider has asubstantial current flow, an excessive currentmay flow in the photomultiplier which couldresult in loss of gain and an increase in darkcurrent. Photomultiplier data sheets gener-ally contain ratings of the maximum anodecurrent which may be drawn.

Some semitransparent photocathodes arevery resistive. See Fig. 36. Even though thelight flux on the photocathode may berelatively small, it is possible that the drop involtage across the photocathode between theelectrical contact and the light spot may in-hibit the photoelectron current flow andcause a non-linear behavior that could upsetcareful measurements.

When measurements are made at very lowlight levels, it is important that the tube betotally shielded from unwanted stray light toavoid an increase in background noise out-put. A person’s photopic vision is not a goodjudge of the presence of light leakage thatthe photomultiplier might sense readily. Forexample, light may leak through an open-ended coaxial cable connector such as BNCor through certain bases or sockets. It maybe noted that light-tight caps are generally

75


commercially available to terminate unusualcoaxial connectors.

Magnetic FieldsCare should be exercised to keep the

magnetic field environment of a photomulti-plier to a minimum. The electron opticaloperation of a photomultiplier can be con-siderably altered by magnetic fields, asshown in Fig. 51. It is also possible to inducea permanent magnetization of some photo-multiplier parts such as dynodes or dynodeside rods constructed of nickel. Ifmagnetization occurs, degaussing may readi-ly be accomplished by placing the tube at thecenter of a coil operated at an alternatingcurrent of 60 Hz with a maximum fieldstrength of 8000 ampere turns per meter andthen gradually withdrawing the tube fromthe coil.

Magnetic shields are generally availablecommercially for photomultiplier tubes ofvarious constructions. In providing suchprotection, it may be important to use ashield which extends beyond the semitrans-parent photocathode a distance of at leasthalf the diameter of the photocathode.

AtmosphereIn some applications it may be necessary

to operate the photomultiplier in other thannormal atmospheric pressure. Most photo-multipliers will tolerate pressures to three at-mospheres. For the larger tubes, however,this pressure may present an implosionhazard. In these special cases, it would bewell to check with the tube manufacturer.Pressures less than one atmosphere also pre-sent a problem in that electrical breakdowncan more readily occur. In such cases it maybe necessary to coat all exposed connectionsand wiring with an insulator such as siliconerubber.

Corrosive atmospheres must be avoided,especially on photomultipliers having metalenvelopes. Corrosion could destroy theglass-to-metal seal and result in loss ofvacuum.

High-humidity conditions should beavoided if possible because condensation onthe base and socket can result in additionalelectrical leakage or even breakdown. If themoisture situation is unavoidable, it mayagain be advisable to coat exposed connec-tions with an insulator such as silicone rub-

76

ber as described earlier under “Tempera-ture.”

Again, a word of caution about the prob-lem of helium penetration of glass. Even atroom temperature, if there is a significanthelium content in the ambient photomulti-plier atmosphere, some helium will penetratethe glass envelope. The result is an increasein dark current and after pulsing because ofionization of the helium gas. Eventually,with sufficient helium pressure a completeelectrical breakdown can occur with voltageapplied to the photomultiplier tube.

If helium cannot be avoided, it may bedesirable to use a separate enclosure for thephotomultiplier through which a small flowof purging gas such as dry nitrogen is pro-vided. It is also reported90 that heliumpenetration can be blocked by a thin layer ofan epoxy-Epon 828 resin (registered trade-mark of Shell Chemical Company) andBelsamid 125 hardening agent (registeredtrademark of General Mills Incorporated).Shock and Vibration

Most photomultipliers will survive only areasonable amount of shock or vibration(less than 10-g shock, depending on shockduration and direction). Although specialphotomultipliers have been designed to sur-vive in extreme environments (shock valuesfrom 30 to 1500 g), the user should makeevery effort to avoid excessive shock or vi-bration, possibly by the use of specialvibration-isolation fixtures. The photomulti-plier tube should be handled as the delicateinstrument that it is. Excessive shock orvibration can actually cause physical damageto the tube to the point of shorting out someof the elements or even resulting in breakageof the envelope and loss of vacuum. A lesserdegree of shock may cause deformation ofthe tube elements and can result in loss ofgain or deterioration of other performanceparameters. If measurements are being madewhile a tube is vibrated, it is likely that theoutput will be modulated by the vibrationnot only because the light spot may bedeflected to different positions on thephotocathode but also because some of thedynodes may actually vibrate and causemodulation of the secondary-emission gain.

Many photomultipliers have been de-signed for use under severe conditions ofshock and vibration and, in many cases, spe-

cifically for use in missile and rocket applica-tions. Such tubes, however, find uses inmany other applications including oil-welllogging or other industrial control applica-tions where the tube may be subjected torough usage. These tubes are available withmost of the electrical and spectral character-istics typical of the more general-purposetypes. These types differ primarily inmechanical construction in that additionalsupporting members may be employed andan improved cathode connection may beused to assure positive contact when the tubeis subjected to these environments. Rug-gedization of tubes using the glass envelopehas also been accomplished by moving thedynode cage close to the stem (therebydrastically shortening the lead lengths andraising their mechanical resonant frequen-cy), by using heavier leads and extra spacersto hold the dynode cage in place, and by aspecial heavy-duty welding process on themetal-to-metal joints.

Tubes recommended for use under severeenvironmental conditions are usually de-signed to withstand environmental testsequivalent to those specified in the ap-plicable portions of MIL-E-5272C or MIL-STD-810B in which the specified accelera-tions are applied directly to the tubes.Sinusoidal vibration tests are performed onapparatus that applies a variable sinusoidalvibration to the tube. The sinusoidal fre-quency is varied logarithmically with timefrom a minimum to a maximum to a mini-mum value. Each tube is vibrated in each ofthe three orthogonal axes.

Random vibration tests are performed bysubjecting the tube to a specified spectraldensity (g2/Hz) in a specified frequencyband.

Shock tests are performed on an apparat-us that applies a half-wave sinusoidal shockpulse to the photomultiplier tube. The tubeis subjected to the shock in each of the threeorthogonal axes. The shock pulse is ex-pressed in terms of the peak acceleration ofthe pulse and the time duration of the pulse.The tubes may receive more than one shockpulse in each of the orthogonal axes.

Nuclear RadiationIn specialized applications such as the use

of photomultiplier tubes in satellites, the


tubes may be subjected to high levels ofradiation such as occur in the Van Allenbelts. A summary of the effects of radiationon photomultiplier tubes may be found in apaper by S.M. Johnson, Jr.90a. Temporaryeffects of intense radiation are principally anincrease in background current and noise.Continued exposure, however, will alsocause permanent damage to the face-plateglass, and to a much lesser extent the photo-cathode, and the dynodes.

The origin of the increased backgroundcurrent in photomultipliers exposed tonuclear radiation is a fluorescence or scin-tillation in the glass faceplate that causeselectron emission from the photocathode.For example, irradiation with 60Co gammarays produces the equivalent of 10-12

W/cm2 of 420-nm flux for an input of 10 - 3rad/s.

The major damage to a photomultiplier isthe browning of the faceplate glass. Signifi-cant changes in transmission are observedfor 9741 and 7056 glass with an exposure of105 rads of 60Co gamma radiation. Fusedsilica shows much less change for the sameirradiation. Lime glass is reported to be verysusceptible to browning. Sapphire is the leastdegraded by gamma radiation of windowsused in photomultipliers. Of the far-ultra-violet windows, LiF is reported to be verysusceptible to radiation damage. MgF2 isrecommended down to the Lyman-alphawavelength level in the presence of highradiation exposures.

Although most of the damage to photo-multipliers from nuclear radiation is in thefaceplate, some change may be expected inthe photocathode. The damage is not greatcompared with that to the glass, probablybecause of the very small absorption of thevery thin photocathodes. The same may besaid for dynode material where, again, thesurface layer is most important.

REFERENCES44. W.J. Harper and W.J. Choyke, “Theresistance of semitransparent photo-cathodes,” Jour. Appl. Phys., Vol. 27, p.1358 (1956).45. R.W. Engstrom, R.G. Stoudenheimer,H.L. Palmer, and D.A. Bly, “Recent Workon photoemission and dark emission pro-blems,” IRE Trans. Nucl. Sci., Vol. NS 5,

77


No. 3, p. 120-124, (Dec. 1958).46. M. Lontie-Bailliez and A. Messen,“L’influence de la temperature sur lesphotomultiplicateurs,” Annales de laScoiete Scientifique de Bruxelles, Vol. 73,Series 1, p. 390 (1959).47. W.E. Spicer and F. Wooten, “Photo-emission and photomultipliers,” Proc.IEEE, Vol. 51, p. 1119-1126, (Aug. 1963).48. H. Martin, Cooling photomultiplierswith III-V photocathodes,” Electro-OpticalSystems Design, Vol. 8, No. 11, p. 16-20(Nov. 1976).49. E.G. Ramberg, “Optical factors in thephotoemission of thin films,” Appl. Opt.,Vol. 6, No. 12, p. 2163-2170 (Dec. 1967).50. S.A. Hoenig and A. Cutler III,“Polarization sensitivity of the RCA 6903photocathode tube,” Appl. Opt., Vol. 5,No. 6, p. 1091-2 (June 1966).51. H. Hora, “Experimental evidence ofangular dependence of electron excitation inphotoemission”, Phys. Stat. Sol., Vol. (a)5, p. 159-166 (1971).52. W.D. Gunter, Jr., E.F. Erickson, andG.R. Grant, “Enhancement of photomulti-plier sensitivity by total internal reflection,”Appl. Opt., Vol. 4, No. 4, p. 512 (April1965).53. J .R. Sizelove and J.A. Love, III,“Analysis of a multiple reflective translucentphotocathode,” Appl. Opt., Vol. 6, No. 3,p. 443-446, (March 1967).54. J.B. Oke and R.E. Schild, “A practicalmultiple reflection technique for improvingthe quantum efficiency of photomultipliertubes,” Appl. Opt., Vol. 7, No. 4, p.617-622 (Apil 1968).55. W. D. Gunter, Jr., G.R. Grant, andS.A. Shaw, “Optical devices to increasephotocathode quantum efficiency,” Appl.Opt., Vol. 9, No. 2, p. 25 l-257 (Feb. 1970).56. D.P. Jones, “Photomultiplier sensitiv-ity variation with angle of incidence on thephotocathode,” Appl. Opt., Vol. 15, No. 4,p. 910-914 (April 1976).57. F.A. Helvy and R.M. Matheson,“Photosensitive cathode with closely adja-cent light-diffusing layer,” U.S. Patent3 242,626, (Mar. 29, 1966).58. R.W. Engstrom, “Improvement inphotomultiplier and TV camera tubes fornuclear medicine,” IEEE Trans. Nucl. Sci.,Vol. NS-24, No. 2, p. 900-903 (Apil 1977).

78

59. W. Widmaier and R.W. Engstrom,“Variation of the conductivity of the semi-transparent cesium-antimony photo-cathode,” RCA Rev., Vol. 16, No. 1, p.109-115 (March 1955).60. H.A. Zagorites and D.Y. Less, “Gam-ma and X-ray effects in multiplier photo-tubes,” Naval Radiological DefenseLaboratory, USNRDL-TR-763, (7 July1964).61. W. Viehmann, A.G. Eubanks, G.F.Pieper, and J.H. Bredekamp, “Photomulti-plier window materials under electron ir-radiation: fluorescence and phosphores-cence” Appl. Opt., Vol. 14, No. 9, p.2104-2115 (Sept. 1975).62. R.B. Murray and J.J. Manning,“Response of end-window photomultipliertubes as a function of temperature,” IRETrans. Nucl. Sci., Vol. NS-7, No. 2-3, p.80-86 (June-Sept . 1960).63. A.T. Young, “Temperature effects inphotomultipliers and astronomical photo-metry,” Appl. Opt., Vol. 2, No. 1, p.5 l-60-(Jan. 1963).64. R.E. Rohde, “Gain vs. temperature ef-fects in NaI(T1) photomultiplier scintillationdetectors using 10 and 14 stage tubes,” Nucl.Intr. and Methods, Vol. 34, p. 109-115(1965).65. D.F. Cove11 and B.A. Euler, “Gainshift versus counting rate in certainmultiplier phototubes,” USNRDL-TR-521,U . S . N a v a l R a d i o l o g i c a l D e f e n seLaboratory, San Francisco (1961).6 6 . C . S . W i g g i n s a n d K . E a r l e y ,“Photomultiplier refrigerator,” Rev. Sci.Instr., Vol. 33, p. 1057-8 (1962).67. For example, among others: Productsfor Research, Inc., 78 Holten Street,Danvers, Massachusetts 01923(617-774-3250).

Pacific Precision Instruments, 1040Shary Court, Concord, California 945 18(415-827-9010).68. H.R. Krall, “Extraneous light emissionfrom photomultipliers,” IEEE Trans. Nucl.Sci. Vol. NS-14, No. 1, p. 455-9 (Feb.1967).69. Louis Lavoie, “Photomultiplier cath-ode poisoning,” Rev. Sci. Instr., Vol. 38,No. 6, pp 833-4, (1967).70. G.A. Morton, H.M. Smith and R.Wasserman, “Afterpulses in photomulti-

Photomultlpller Characteristics

pliers,” IEEE Trans. Nucl. Sci., Vol. NS-14,No. 1, pp 443-448, (1967).71. V.O. Altemose, “Helium diffusionthrough glass, ” J. Appl. Phys., Vol. 32, No.7, pp 1309-1316, (1961).72. W.C. Paske, “He + afterpulses inphotomultipliers: Their effect on atomic andmolecular lifetime determinations,” Rev.Sci. Instr., Vol. 45, No. 8, p. 1001-3 (1974).73. A.T. Young, “Cosmic ray induceddark current in photomultipliers,” Rev. Sci.Instr., Vol. 37, No. 11, pp 1472-1481 (1966).

74. D. Sard, “Calculated frequency spec-trum of the shot noise from a photo-multiplier tube,” J. Appl. Phys., Vol. 17, p.768-777 (1946).75. B. Leskovar and C.C. Lo, “Transit timespread measurements of microchannel-platephotomultipliers,” European Conference onPrecise Electrical Measurements, Brighton,Sussex, England, 5-9 Sept. 1977 (London,England: IEEE (1977), pp 41-3.)76. D.A. Gedcke and W.J. McDonald,“Design of the constant fraction of pulseheight trigger for optimum time resolution, ”Nucl. Instr. and Meth., Vol. 58, No. 2, pp253-60 (1968).77. M. Birk, Q.A. Kerns, and R.F. Tusting,“Evaluation of the C-70045A high-speedphotomultiplier,” IEEE Trans. Nucl. Sci.Vol. NS-11, No. 3, pp 129-138 (1964)78. F. de la Barre, “Influence of transit timedifferences on photomultiplier time resolu-tion,” IEEE Trans. Nucl Sci., Vol. NS-19,No. 3, pp 119-121 (1972).79. B. Leskovar, C.C. Lo, “Performancestudies of photomultipliers having dynodeswith GaP(Cs) secondary emitting surface,”IEEE Trans. Nucl. Sci., Vol. NS-19, No. 3,pp 50-62 (1972).80. S.K. Poultney, “Single Photon Detec-tion and Timing: Experiments and Techni-ques,” Advances in Electronics and ElectronPhysics, Vol. 31, Academic Press, Inc., New

York, (1972).81. E. Gatti and V. Svelto, “Review oftheories and experiments of resolving timewith scintillation counters,” Nucl. Instr. andMethods, Vol. 43, pp 248-268, (1966).82. W.A. Baum, “The detection and mea-surement of faint astronomical sources,”Astronomical Techniques, Edited by W .A.Hiltner, The University of Chicago Press,(1962).83. G.A. Morton, “Photon Counting,”Appl. Opt., Vol. 7, No. 1, pp l-10, (1968).84. R.M. Matheson, “Recent photomulti-plier developments at RCA,” IEEE Trans.Nucl. Sci. Vol. NS-11, No. 3, pp 64-7 1,(1964).85. G.A. Morton, H.M. Smith, and H.R.Krall, “Pulse-height resolution of high gainfirst dynode photomultipliers,” Appl. Phys.Lett., Vol. 13, No. 10, pp 356-7, (1968).86. G.H. Narayan and J.R. Prescott,“Line-widths in NaI(Tl) scintillation count-ers for low energy gamma-rays,” IEEETrans. Nucl. Sci., Vol. NS-13, No. 3, pp132-7, (1966).87. R.W. Engstrom and J.L. Weaver, “Areplateaus significant in scintillation count-ing?” Nucleonics, Vol. 17, No. 2, pp 70-74,(1959).88. The Current Status of Liquid-Scintillation Counting, Edited by Edwin D.Bransome, Grune and Stratton, New Yorkand London, (1970).89. Liquid-Scintillation Counting, RecentDevelopments, Edited by Philip E. Stanleyand Bruce A. Scoggins, Academic Press(1974).90. Yasunori Nikaido et al, “Electron tubefor photoelectric conversion,” PatentPublication No. SHO 42-16002 (Shimizu Pa-tent Office, Fukiai-Ku, Kobe, Japan).90a. S.M. Johnson, Jr., “Radiation effectson multiplier phototubes,” IEEE Trans.Nucl. Sci., Vol. NS-20, No. 1, pp 113-124,Feb. 1973.

79


5. Photomultiplier Applications

SUMMARY OF SELECTION CRITERIAA comparison of the relative advantages

of photomultipliers and solid-state detectorsis given in the introductory chapter of thismanual. The following is a summary of theapplication requirements that indicate when,in general, a photomultiplier is the mostsuitable detector.

The most important consideration is thelight or radiation level to be detected. Theuse of photomultipliers is recommendedwhen the level of light flux is very low. Iflevels are relatively high, it may be simpler touse a solid-state detector. Furthermore, ahigh level of light flux could overload anddamage the photomultiplier. A photomulti-plier may be used if the light flux is 100microlumens or less or of the order of 0.1microwatt or less at the peak of the spectralresponse.

When a fairly large detector area is re-quired, a photomultiplier is also recom-mended, as is the case in scintillation count-ing where a crystal scintillator may be severalcentimeters in diameter. Silicon avalanchephotodiodes have good low-light-levelcapability but their area is generally limitedto a few square millimeters.

Spectral response of the photomultiplier,of course, must be a reasonable match tothat of the source of radiation. Photomulti-pliers are useful in the range 120 nanometersto 1100 nanometers, depending upon thetype of photocathode and window material.Responses further in the infrared than 1100nanometers require an infrared photocon-ductive detector or some other infrared-sensitive device.

When a fast response time is an importantrequirement, the photomultiplier is usuallythe most suitable detector. Photomultipliertubes have response-time capability down tothe nanosecond range and even better in thecase of specially designed tubes.

80

APPLIED VOLTAGECONSIDERATIONS

Proper operation of a photomultiplierdepends critically upon the applied voltageand the voltage distribution to the dynodestages. A small variation in voltage canresult in a much larger percentage change inthe tube gain. For example, see Fig. 49 whichshows the log of the gain as a function of thelog of the voltage. For a 9-stage tube theslope of the log-gain vs log-voltage curve isapproximately 7; for a 14-stage tube theslope is about 12. (If the secondary emissionratio were linear with voltage, see Fig. 19,one would expect the slope of the log-gain vslog-voltage curves to be equal to the numberof stages.) For a 12-stage tube, a 1% changein voltage results in a 10% change in gain.Thus, there is need for more than ordinaryprecaution to provide a well-regulated powersupply l

Power Supply, Regulation, Polarity, andShielding

Commercial power supplies are generallyavailable that have a line-regulation of0.005% with some available at 0.001% for a10% change in line voltage when working in-to a fixed divider network. Voltage variationwith temperature might be of similar magni-tude for a 1 °C change in temperature. Itwould not be too difficult, therefore, to pro-vide voltage supplies that would result ingain stability of the photomultiplier tube dueto line voltage and temperature changes ofbetter than 0.1%. For requirements not thiscritical, BURLE manufactures a simplifiedcompact power supply that includes a socketand voltage-divider network for 1-l/8inchdiameter, 9-stage, side-on photomultipliers.Also manufactured by BURLE are Inte-grated Photodetection Assemblies (IPA’s)that package a photomultiplier tube, opticalfilter, power supply, signal-conditioning

amplifier, and electrostatic/magneticshielding. The IPA operates from a 12-voltdc supply.

The recommended polarity of the photo-multiplier power-supply voltage with respectto ground depends largely on the applicationintended. Of course, the cathode is alwaysnegative with respect to the anode. In somepulsed application, however, such as scin-tillation counting, the cathode should begrounded and the anode operated at a highpositive potential with a capacitance-coupledoutput. In this case the scintillator and anymagnetic or electrostatic shields should alsobe connected to ground potential.

In applications in which the signal cannotbe passed through a coupling capacitor, thepositive side of the power supply should begrounded. The cathode is then at a highnegative potential with respect to ground.When this arrangement is used, extra pre-cautions must be taken in the mounting andshielding of the photomultiplier. If there is apotential gradient across the tube wall, scin-tillations occurring in the glass will increasedark noise. If this condition continues for asufficiently long period of time, the photo-cathode will be permanently damaged by theionic conduction through the glass. To pre-vent this situation when a shield is used, theshield is connected to photocathode poten-tial. Light-shielding or supporting materialsused in photomultipliers must limit leakagecurrents to 10- 12 ampere or less. Fig. 62shows a curve of the effect of external-shieldpotential on photomultiplier noise.

To reduce the shock hazard to personnel,a very high resistance should be connectedbetween the shield and the negative highvoltage.Voltage Divider Design

The interstage voltage gradients for thephotomultiplier elements may be supplied byindividual voltage sources. The usual source,however, is a resistive voltage divider placedacross a high-voltage source, as shown inFig. 87.

A resistive voltage divider must be de-signed to divide the applied voltage equallyor unequally among the various stages as re-quired by the electrostatic system of thetube. The most common voltage betweenstages is usually referred to as the stage vol-tage and the voltage between other stages as

Photomultiplier Applications

-92CS - 32424

Fig. 87 - Schematic diagram of a resistivevoltage divider.

The following formula may be used tocalculate the voltage between stages:

The voltage-divider resistor values re-quired for each stage can be determinedfrom the value of the total resistance re-quired of the voltage divider and the voltage-divider ratios of the particular tube type.

81


The interstage resistance values are in pro-portion to the voltage-divider ratios asfollows:

where Rj is the resistance between elementsDyj - 1 and Dyj. The recommended resis-tance values for a photomultiplier voltagedivider range from 20,000 ohms per stage to5 megohms per stage; the exact values areusually the result of a compromise. If lowvalues of resistance per stage are utilized, thepower drawn from the regulated power sup-ply may be excessively large. The resistorpower rating should be at least twice thecalculated power dissipation to provide asafety margin and to prevent a shift inresistance values as a result of overheating.The highest suitable value of stage resistance(after consideration has been given toaverage anode current as described below) isdictated by leakage currents in the photo-multiplier and socket wiring.

One criterion for the selection of a suitablerange of voltage-divider resistance values isthe expected maximum anode current thatmay be drawn from the photomultiplier.When the anode current is of the same orderof magnitude as the divider current, non-linear response results. This non-linearity isillustrated in Fig. 88 which shows theresponse of a 931A photomultiplier as afunction of light level using a conventionalvoltage divider such as shown in Fig. 87 withequal voltage per stage.91 (The value of RLwas essentially zero for this measurement.)The anode current is shown relative to thedivider current at zero light level. Thesuperlinearity region is explained by achange which takes place in the voltagedistribution between stages from uniform tonon-uniform as the light level is increased.Thus, the electron current flow from the lastdynode to the anode causes less current toflow through the voltagedivider resistor,

voltage between dynodes and a decrease inthe collection voltage between the lastdynode and the anode. The reduced collec-tion voltage tends to reduce the output cur-rent slightly, but the increased dynodevoltages more than compensate for thisreduction by an increased gain.

82

Fig. 88 - The relative response of a 931A pho-tomultiplier as a function of the light flux us-ing the circuit of Fig. 87 with equal stagevoltage (at zero light level). (From Engstromand Fischer91)

The decrease in sensitivity that occursbeyond region A of Fig. 88 results from theextension of voltage losses to the last two orthree dynode resistors causing defocusingand skipping in the associated dynodestages. In order to prevent this loss andassure a high degree of linearity, the currentthrough the voltage-divider network shouldbe at least ten times the maximum averageanode current required. In calculating thevoltage-divider current, the average anodecurrent must first be estimated; this estimaterequires knowledge of the value of the input(light) signal and the required output (elec-trical) signal.

Photo-multiplier noise or a shift in gainmay result from heat emanating from thevoltage-divider resistors. The divider net-work and other heat-producing components,therefore, should be located so that they willnot increase the tube temperature. Resis-tance values in excess of five megohmsshould be avoided because current leakagebetween the photomultiplier terminals couldcause a variation of the interstage voltage.

The type of resistor used in a dividerdepends on the dynode structure with whichthe divider will be used. Close-toleranceresistors, such as the laser-trimmed metal-film types, are normally required with thefocused structures. On the other hand, inter-dynode voltages in the Venetian-blind struc-tures can vary widely with but little effect on

the photomultiplier. For this reason, theresistors used with a Venetian-blind structurecan be of a less stable variety, such as com-position.

The over-all performance of a tube thathas a cathode-to-first-dynode region essen-tially electrostatically isolated from the re-maining dynode region can be improved bymaintaining a high electric field, in thecathode-to-first-dynode region to reduce thetransit-time spread of photoelectrons arriv-ing at the first dynode and minimize the ef-fect of magnetic fields. A high first-dynodegain, which implies a high

have the disadvantage of reducing the gain in

Intermediate StagesIn applications in which it is desirable to

control the anode sensitivity withoutchanging the over-all voltage, the voltage ofa single dynode may be varied. Fig. 89 showsthe variation of anode current for a 931Aphotomultiplier when one of the dynodevoltages is varied while the total supplyvoltage is held constant. The dynode shouldbe selected from the middle of the dynodestring because a variation of dynode poten-tials near the cathode or anode would have adetrimental effect on photomultiplier opera-tion.Operation with Fewer Stages

A photomultiplier need not be operatedwith all dynode stages, although the designof the tube has been optimized for this con-dition. In some situations where the lightlevel being detected is so high that it would


overload the output stages of the photomul-tiplier, it is possible to eliminate these stagesfrom the circuit. For example, the lastseveral stages of the tube and the anode maybe tied together electrically to form ananode. The tube is then operated with thereduced gain of the remaining stages. In anextreme case, the photomultiplier may beoperated just as a photodiode using only thephotocathode and several or all of the re-maining elements together as an anode. Itmust be realized in operating the photomul-tiplier in such a manner that many photo-cathode types are very resistive (see Figs. 35and 36) and, therefore, the photocurrent willnot be linear with light at the high levelwhich may be available. Photocathodes ofthe opaque type with a conductive substrate,or the semitransparent type with a conduc-tive undercoating will tolerate a much higherlight level without loss of linearity.

7 0 0 8 0 0

92CS-32426

Fig. 89 - The output-current variation of a931A when the voltage on one dynode (No. 6)is varied while the total supply voltage re-mains fixed.

Voltage Dividers for Pulsed OperationIn applications in which the input signal is

in the form of pulses, the average anode cur-rent can be determined from the peak pulsecurrent and the duty factor. The total resis-

83


tance of the voltage-divider network iscalculated for the average anode current.

In cases in which the average anode cur-rent is much less than a peak pulse current,dynode potentials can be maintained at anearly constant value during the pulse dura-tion by use of charge-storage capacitors atthe tube socket. The voltage-divider currentneed only be sufficient to provide theaverage anode current for the photomulti-plier. The high peak currents required duringthe large-amplitude light pulses are suppliedby the capacitors.

The capacitor values depend upon thevalue of the output charge associated withthe pulse or train of pulses. The value of thefinal-dynode-to-anode capacitor C is givenby

where C is in farads, q is the total anodecharge per pulse in coulombs, and V is thevoltage across the capacitor. The factor 100is used to limit the voltage change across thecapacitor to a l-per-cent maximum during apulse. Capacitor values for preceding stagesshould take into account the smaller valuesof dynode currents in these stages. Conser-vatively, a factor of approximately 2 perstage is used. Capacitors are not requiredacross those dynode stages at which the peakdynode current is less than 1/10 of theaverage current through the voltage-dividernetwork. For pulse durations in the 1 to 100ns range, consideration should be given tothe inherent stage-to-stage capacitanceswhich are in the order of 1 to 3 pF.

Medium-Speed Pulse ApplicationsIn applications in which the output cur-

rent consists of pulses of short duration, thecapacitance CL of the anode circuit toground becomes very important. The capaci-tance CL is the sum of all capacitances fromthe anode to ground: photomultiplier-anodecapacitance, cable capacitance, and the in-put capacitance of the measuring device. Forpulses having a duration much shorter thanthe anode time constant RLCL, the outputvoltage is equal to the product of the chargeand 1/CL because the anode current is simp-ly charging a capacitor. The capacitor charge

84

then decays exponentially through the anodeload resistor with a time constant of RLCL.To prevent pulses from piling up on eachother, the maximum value of RLCL shouldbe much less than the reciprocal of therepetition rate.

An important example of this type ofoperation is scintillation counting, for exam-ple with NaI:Tl. The time constant of thescintillations is 0.25 microsecond and, be-cause the integral of the output current pulseis a measure of the energy of the incidentradiation, the current pulse is integrated onthe anode-circuit capacitance for a period ofabout 10 microseconds. Because the scin-tillations occur at random, the maximumaverage counting rate is limited to about 10kHz. If circumstances require a highercounting rate, the integration time must bereduced accordingly.Fast-Pulse Applications

In fast-pulse light applications, it isrecommended that the photomultiplier beoperated at negative high voltage with theanode at ground potential. A typical voltage-divider circuit with series-connected capaci-tors is shown in Fig. 90. The parallel con-figuration of capacitors may also be used, asshown in Fig. 91. The parallel arrangementrequires capacitors of higher voltage ratings.Regardless of the configuration, thecapacitors must be located at the socket. Thecapacitor arrangements just described mayalso be applied to negative-ground applica-tions.

Fig. 90 - Series-connectedvoltage-divider circuit usingfor pulse-light applications.

capacitors inpositive ground

The wiring of the anode or dynode “pick-off” circuit is very critical in pulse applica-tions if pulse shape is to be preserved. Mostpulse circuitry uses 50-ohm characteristic im-pedance cables and connectors because of


92cs-32428

Fig. 91 - Parallel-connected capacitors in voltage-divider circuit for pulsed-light ap-plication.

their ready availability, although 75- and92-ohm components are also used. Carefulwiring is required. Figs 92 and 95 illustrateschematically and pictorially the best loca-tion of pulse bypass capacitors that returnthe anode pulse current to Dyn and Dyn _ 1

by paths of minimum residual inductance. Itshould be noted that these bypass capacitorsalso serve the function of charge-storagecapacitors, as shown in Fig. 91.

92cs-32429

Fig. 92 - Bypass capacitors used to makethe last two dynodes appear as a groundplane to a fast-pulsed signal.

Wiring TechniquesGood high-frequency wiring techniques

must be employed in wiring photomultipliersockets and associated voltage dividers ifpulse-shape distortion is to be minimized.Fig. 93 illustrates the pulse shape obtainedfrom an 8575 photomultiplier excited by a

0.5-nanosecond rise-time light pulse. Thepulse shape is most easily seen with the aid ofa repetitive light pulser and a high-speedreal-time oscilloscope. Fig. 94 indicates thegeneral type of distortion encountered withthe use of improperly wired or excessively in-ductive capacitors. The output pulse il-lustrates the ringing that may occur in an im-properly wired socket.

Fig. 93 - Pulse shape obtained from photo-multiplier excited by a light pulse having an0.5-nanosecond rise time.

92cs-32430

Fig. 94 - Pulse shape distortion (ringing) en-countered with improperly wired or excessive-ly inductive capacitors.

85


Fig. 95 shows a socket wired for a negativehigh-voltage application. The disk-typebypass capacitors are mounted in series withminimum lead length because the self-inductance of the lead wires becomes criticalin nanosecond-pulse work. Care should alsobe taken in dressing the bypass capacitorsand coaxial cable. - The resistors for thevoltage divider are shown mounted on thesocket. In applications requiring minimumdark current, the resistors should be remotefrom the photomultiplier to minimizeheating effects.

92cs-32432

Fig. 95 - Anode detail of socket wired for anegative high-voltage application showinglocation of charge-storage capacitors.

Checking Socket and Tube PerformanceSome photomultipliers typically display a

reflected-pulse rise time of the order of 1.5nanoseconds for an initiating pulse of theorder of 50 picoseconds when the tube andsocket are tested with a time-domain reflec-tometer, the instrument used to test theanode-pin region of the socket.

During tube operation, the output pulsecan be viewed with a real-time high-speedoscilloscope. The pulse shape can be in-spected for signs of clipping or ringing.

The effectiveness of the charge-storagecapacitors can be verified with a simplelinearity test. The output of a photomulti-

plot of pulse amplitude as a function of thelogarithm of the voltage value should yield astraight line.

A two-pulse technique may be employedto test linearity at constant operatingvoltage. Consider two sequential pulses hav-ing amplitude ratios in the range from 2:1 to10:1. The ratio of the amplitude of the two

86

photomultiplier current pulses is establishedat low pulse amplitudes where linear opera-tion is assured. Next, the light flux is in-creased in a series of increments. A constant-current pulse amplitude ratio indicates linearoperation. A pair of light-emitting diodesmay be used to provide the two light pulses.The over-all control of the light levels maybe done with a series of neutral-densityfilters. Exact filter calibration is not requiredbecause the system is self-calibrating in thepulse current measurements.Tapered Dividers

Some applications require that photomul-tipliers sustain high signal currents for shorttime intervals, tens of nanoseconds or less.In general, photomultipliers are capable ofsupplying 0.2 ampere or more into a 50-ohmload for short durations. However, the vol-tage divider must be tailored to the applica-tion to allow a photomultiplier to deliverthese high currents.

The principal limitation on current output(into a 50-ohm system) is space charge at thelast few stages. This space charge can beovercome if the potential difference acrossthe last few stages is increased by use of atapered divider rather than an equal-volts-per-stage divider. The tapered divider places3 to 4 times the normal interstage potentialdifference across the last stage. The pro-gression leading to the 4-times potential dif-ference should be gradual to maintain prop-er electrostatic focus between stages; a pro-gression of 1, 1 . . . 1, 1, 1.5, 2.0, 3, 4.2 isrecommended.

Another application of the tapered divideris to provide a large signal voltage across ahigh-impedance load. Voltage excursions of400 to 500 volts can be obtained from photo-multipliers. A possible application is in driv-ing an electro-optical modulator. Becausethe photomultiplier is nearly an ideal con-stant current source, its output voltage signalis limited only by the potential differencebetween the last dynode and the anode. Ifthe potential difference is 100 volts, theanode cannot swing through more than a100-volt excursion. By impressing a muchhigher potential difference between theanode and last dynode by means of a tapereddivider, greater voltage swings can be ob-tained. Tube data sheets should be consultedfor maximum voltage ratings.

Dynamic Compression of Output SignalMost photomultipliers operate linearly

over a dynamic range of six or seven ordersof magnitude, a range few monitor devicescan accommodate without requiring rangechanges. When compression of the dynamicrange is desired, a logarithmic amplifier issometimes used. The photomultiplier mayalso be operated in a compressed-outputmode, however, without the need for addi-tional compression circuitry.

For example, in liquid scintillationcounting where several different isotopesmay be present, the output pulse heightmight normally vary by as much as 100: 1. Bylimiting the potential difference between thelast dynode and anode to, for example, 10volts, space charge will limit the maximumcurrent that can be drawn. The anode pulsecan then be used for the coincidence timingwith a more convenient range of pulseheights. Energy measurements can then bemade using the current at an earlier dynodewhere space-charge saturation is not present.Current Protection of Photomultiplier

If a photomultiplier is accidentally ex-posed to an excessive amount of light, it maybe permanently damaged by the resultanthigh currents. To reduce this possibility, theresistive voltage-divider network may bedesigned to limit the anode current. Theaverage anode current of a photomultipliercannot much exceed the voltage-divider cur-rent; therefore, the zero-light level voltage-divider network serves as an overload pro-

ANODE


tection for the tube. If overexposure is ex-pected frequently, interdynode currents,which can be quite excessive, may cause lossof gain. In some applications it may beworthwhile to protect against dynodedamage by using resistors in series with eachdynode lead .91

Active Divider NetworkAlthough normally, the divider network

current should exceed the maximum photo-multiplier output current by a factor of 10 ormore, it is possible by using the emitter-follower characteristics of transistors to pro-vide a power supply requiring much lessdivider current. Fig. 96 is an example of suchan active divider network devised by C. R.Kerns92 As the dynode current increases,the added current is diverted from the high-beta transistors rather than from the divider-resistor string, thus improving the voltageregulation by a large factor. The capacitorsshown are the usual ones for high-frequencybypassing. The circuit also retains a current-limiting action that prevents damage to thephotomultiplier.

MECHANICAL CONSIDERATIONSHandling

Because most photomultipliers have glassenvelopes, they should be handled with careto avoid damage to the tube seals and otherparts. This caution is especially importantfor tube types having graded-seal envelopeconstruction. The pins or leads of the tubeshould also be treated with care.

Fig. 96 - An active divider network for an 8575 photomultiplier designed to minimizevoltage changes at the dynodes. (From C.R. Kerns92)

87


BasingPhotomultiplier tubes may have either a

temporary or a permanently attached base.Dimensional outline diagrams such as thoseshown in Fig. 97 are provided in the pub-lished technical data for individual tubes. In-dicated on the diagrams are the type of baseemployed, maximum mechanical dimen-sions, radii of curvature where applicable,pin/lead details, location and dimensions ofmagnetic parts used (in tubes utilizing mini-mum number of magnetic materials), andnotes regarding restricted mounting areas,again where applicable.

Photomultiplier tubes intended to besoldered directly to circuit boards or hous-ings are supplied with semiflexible or “fly-ing” leads and a temporary base, intendedfor testing purposes only, that should be re-moved prior to permanent installation.

A lead-terminal diagram that showsphotomultiplier-tube lead orientation withthe temporary base removed, shown in Fig.98, provides a lead indexing reference. Alead-connection diagram, such as the oneshown in Fig. 99, relates terminal to elec-trode. Care must be exercised in interpretingbasing and lead-terminal diagrams to insureagainst possible damage to the photomulti-plier resulting from incorrect connections.Terminal Connections

BURLE photomultipliers are supplied witheither semiflexible leads, semiflexible leadsattached to temporary bases, permanently

PHOTOCATHODE

B

attached bases, or stiff leads. Semiflexibleleads may be soldered, resistance (spot)welded, or crimp connected into theassociated circuitry. When soldering orwelding is employed for such connections,care should be taken to prevent tube destruc-tion due to thermal stress of the seals at thestem. A heat sink, such as locking forceps,should be placed in contact with the semi-flexible leads between the point beingsoldered or welded and the tube seals. Ifsoldering is employed, only a soft solder(e.g., 60% Sn, 40% Pb) should be used.Heat should be applied only long enough topermit the solder to flow freely. By the termsemiflexible, it is implied that excessivebending may break the leads, most common-ly at the stem surface. Some photomulti-pliers are supplied with insulating wafers at-tached to the stem to prevent such an occur-rence. The semiflexible leads are normallymade of dumet or Kovar and are usuallyplated to facilitate soldering.

Photomultipliers supplied with perma-nently attached bases or stiff leads shoulduse only high-grade, low-leakage sockets tominimize leakage currents between adjacentelectrode terminals. Teflon and mica-filledsockets should be used.

Mounting and SupportPhotomultipliers having permanently at-

tached bases normally require no specialmounting arrangements. When specialmounting arrangements are used, however,

9 2 C M - 3 2 4 3 4

Fig. 97 - Typical dimensional-outline drawings showing the type of base suppliedwith each tube and pertinent notes.

88

INDEX

Fig. 98 - Lead-orientation diagram.

BOTTOM VIEWS

PIN I - DYNODE NO. IPIN 2 - DYNODE NO. 3PIN 3 - DYNODE NO.5PIN 4 - DYNODE NO.7PIN 5 - DYNODE NO. 9PIN 6 - ANODEPIN 7 - DYNODE NO.10PIN 8 - DYNODE NO. BPIN 9 - DYNODE NO. 6PIN IO - DYNODE NO 4PIN II - DYNODE NO. 2PIN 12 - PHOTOCATHODE

LEAD I - DYNODE NO. ILEAD 2 - DYNODE NO. 3LEAD 3 - DYNODE NO. 5LEAD 4 - DYNODE NO. 7LEAD 5 - DYNODE NO. 9LEAD6 - ANODELEAD7 - DYNODE NO.10LEAD 8 - DYNODE NO. 8LEAD9 - DYNODE NO.6LEAD 10 - DYNODE NO.4LEAD II - DYNODE NO.2LEAD 12 - PHOTOCATHODE

92CS-32436

Fig. 99 - Lead-connection diagram: (a) withbase connected, (b) with temporary baseremoved.

the envelope, especially that region near thephotocathode, must be maintained atcathode potential. Care should also be takenso that tube performance is not affected byextraneous electrostatic or magnetic fields.Side-on photomultipliers should be mountedto allow rotation of the tube about its majoraxis to obtain maximum anode current for agiven direction of incident radiation. Anangular tolerance with respect to incidentlight direction is normally specified in tubedata sheets.

Direct clamping with non-resilientmaterials to the envelope of tubes not havingpermanently attached bases is not recom-mended nor should clamping be made to any


metal flanges employed in the constructionof a tube. Such flanges, when present, arepart of the tube’s vacuum enclosure and anyundue force or stress applied to them candamage the seals and destroy the tube.

The use of resilient potting compounds orrubber washers is recommended when pho-tomultipliers are clamp-mounted. If a pot-ting compound is used, its characteris-tics-over the temperature range in whichthe tube is to be operated-must be such thatits resilience is maintained at lowtemperature and its expansion, in confinedspace, is not excessive at high temperature.

The electrical insulation properties of anymaterials supporting or shielding the photo-multiplier should be considered. If suchmaterials come into contact with highvoltage with respect to photocathode,minute leakage currents can flow throughthe material and the tube envelope to thephotocathode. Not only does this conditionintroduce excessive noise at the tube outputbut it can also permanently damage thephotocathode sensitivity of the tube throughelectrolysis of the glass envelope. This cau-tion is only true when the tube is operated athigh negative potential with respect toground. Under this operating condition, adecrease in sensitivity can occur if thefaceplate of the tube comes into contact withground. Cathode sensitivity does not recoverafter such an occurrence. Photocathode“poisoning” due to envelope electrolysis candestroy the usefulness of a photomultiplierin a very short time. Therefore, the in-sulating property of materials supporting thetube should be such that leakage current tothe tube envelope is limited to 1 x 10 - 12

ampere, or less.Shielding

Electrostatic and/or magnetic shielding ofmost photomultipliers is usually required.When such shields are used and are in con-tact with the tube envelope, they mustalways be connected to photocathode poten-tial.

In applications where the dc component ofthe signal output is of importance, thecathode is normally operated at high nega-tive voltage with respect to ground. As aresult, the shield is at high negative voltageand precautions must be taken to avoidshorts to ground and to prevent shock

89


hazard to personnel. A 10-megohm resistorshould be placed between the negative highvoltage and shields to avoid such hazards.

In scintillation counting applications, it isrecommended that the photocathode beoperated at ground potential. In this case,the shields should be operated at groundpotential.

Magnetic shielding of most photomulti-pliers is highly desirable. Characteristiccurves showing the effects of magnetic fieldson anode current are provided in many datasheets.Storage

Photomultipliers should be stored in thedark. Storage of tubes in areas where light isincident on the tube results temporarily in ahigher than normal dark-current level whenthe tubes are placed in operation. This in-crease in dark current is primarily due tophosphorescence of the glass and can persistfor about 24 hours. Additionally, storage oftubes designed for operation in the near IRregion of the spectrum (above 700 nano-meters) in illuminated areas may decreasethe “red” sensitivity of the tube.

The phototube should never be stored oroperated in areas where there are concentra-tions of helium because helium readilypermeates glass. The composition of theenvelope material is a major factor govern-ing the rate of helium permeation. As thesilica content in the glass is reduced, the rateof permeation decreases. Accordingly, therate of permeation is greatest for fused silicaand decreases to a minimum in lime glass. Itis also to be noted that the rate of permea-tion is proportional to temperature andvaries directly with pressure.691 92a 92b

Moisture CondensationA very small anode current is observed

when voltage is applied to the electrodes of aphotomultiplier in darkness. Among thecomponents contributing to this dark cur-rent are pulses produced by thermionic emis-sion, ohmic leakage between the anode andadjacent elements, secondary electronsreleased by ionic bombardment of the pho-tocathode, cold emission from the elec-trodes, and light feedback to the photocath-ode. Other conditions contributing to anode

*Distributed by Arthur H. Thomas Company,Vine Street and 3rd, Philadelphia, PA 19105.

90

dark current include external leakage causedby condensation on the tube base and/orsocket when conditions of high humidity ex-ist and contamination of the tube baseand/or socket by handling.

Moisture condensation can be minimizedby potting the tube socket assembly insilicone rubber compounds such as RTV-11,or equivalent. If a tube is suspected ofhaving high ohmic leakage as a result ofhandling, it is recommended that it and itssocket be washed with a solution of alkalinecleaner such as Alconox*, or equivalent, andde-ionized or distilled water having a tem-perature not exceeding 60 °C. The base or thesocket should then be thoroughly rinsed inde-ionized or distilled water (60 °C) and thenair-blown dry.

OPTICAL CONSIDERATIONSIncident Light Flux

Photomultiplier tubes are capable ofoperating usefully over a very wide range ofincident light flux. At the very lowest levels,the limit is determined by the useable signal-to-noise ratio. Upper range of usefulness isdetermined by the saturation of the photo-cathode or by the magnitude of the outputcurrent levels which either result in space-charge limitations or cause damage to thesecondary emission dynodes.

The lower limit of detection is discussed indetail in Chapter 4, Photomultiplier Charac-teristics, in the section “Dark Current andNoise.” See in particular Fig. 67. The lowerlimit in a current-measurement mode is de-termined by dark emission and bandwidth.Photocurrent levels as low as 10 - 16 ampere(625 photoelectrons per second) can bemeasured. This photocurrent level cor-responds to 5 x 10 - 13 lumens for a200-microampere/lumen photocathode. In aphotoelectron-counting mode (see AppendixG), the limit is determined by the statistics ofdiscriminating between photoelectrons andthermally emitted electrons. For a Na2KSbphotocathode, the dark emission from thephotocathode at room temperature may beof the order of 10 - 17 ampere or 62 electronsper second. Counting for one minute in thedark and one minute in the light, one shouldbe able to detect a photocurrent of only afew photoelectrons per second.

For resistive photocathodes (See Table II


in Chapter IV, Photomultiplier Characteris-tics, in the section “Photocathode-RelatedCharacteristics.“) such as K2CsSb, the max-imum recommended photocurrent is only ofthe order of 10 -9 ampere which cor-responds to a flux of the order of 10 -5lumen.

When the photocathode resistivity doesnot limit the input light flux, the maximumlight flux is determined by output currents inthe photomultiplier. For example, the 931Amaximum average anode current is 1.0milliampere which corresponds to an inputlight flux of about 10 -5 lumen with a totalapplied voltage of 1000 volts. If the over-allvoltage were reduced to 500 volts, the tubecould tolerate an input flux of 10 - 3 lumen.

At times it may be useful to reduce thelight level on the photomultiplier by acalibrated amount. Neutral-density filterscan be useful for this purpose. It should beappreciated, however, that such filters arefrequently not neutral and their stateddensity may be only an approximation. Acomparison of the spectral transmittance ofa metalized filter and a gelatin type is shownin Fig. 100. Wratten filters are reasonably

Fig. 100 - Comparison of spectral transmit-tance of metalized- and organic-type neutral-density filters (ND 1.0).

neutral in the visible range but becometransparent in the infrared, which couldcause a problem when they are used with aphotomultiplier having a near-infraredresponse. This effect is much more pro-nounced for the more dense neutral-densityfilters.

Another method for reducing light level isby the use of a mesh or screen. This method

l *670 E. Arques St., Sunnyvale, Calif. 94086

has the advantage of providing a reductionfactor that is essentially independent ofwavelength. Such neutral density screens canbe obtained from the Varian Instrument Di-vision. **

Large reduction factors can be achievedby the use of opal glass, which scatters trans-mitted light in an approximate Lambertian(cosine) distribution (although a bit moreconcentrated near the normal angles than thecosine prediction). Unfortunately, the scat-tered flux is not neutral, providing more redthan blue. See Fig. 101.

WWAVELENGTH - NANOMETERS

92cs-32430

Fig. 101 - The effective spectral transmit-tance (for scattered light) of an opal-glassfilter, 3-mm thickness, into a solid angle ofapproximately 0.01 steradian.

Convex front-surface spherical aluminummirrors can also be used to provide acalibrated light-reduction factor. Aluminummirrors have the advantage of having areasonably flat spectral reflectance-from92.3% at 300 nanometers to 86.7% at 800nanometers.93 If a lamp of CP candelas islocated at a distance a from the surface ofthe mirror having a reflectivity e and aradius r, the flux (L) in lumens through a testaperture of area A located at a distance bfrom the mirror surface is given by94


nent rays on the mirror.

CalibrationFor some purposes it may be useful to pro-

vide a calibration of the photomultiplieranode or photocathode responsivity. A con-venient test source is the tungsten lamp (seeAppendix F). Lamps with candle power andcolor temperature (2856 K) traceable to theNational Bureau of Standards can be ob-tained from BURLE, Lancaster, PA (TypeAJ2239). Some method of reducing the lightflux such as discussed above would normallybe required.

Spectral response measurements generallyrequire a sophisticated monochromator set-up. Reasonably good results, however, canbe obtained by use of calibrated narrow-band-pass color filters and a calibratedtungsten lamp. The power through a par-ticular filter may be obtained by integratingthe product of the filter transmission and thetungsten irradiance over the wavelengthband of the filter. Spectral irradiance for atungsten lamp calibrated to a color tempera-ture of 2856 K may be obtained from Appen-dix F, Table F-I, with a multiplying factorappropriate to the particular total luminousflux. Details are given in the footnote ac-companying Table F-I.

Spot SizeIt is generally advisable to utilize a large

part of the photocathode area rather than tofocus a small spot of the light flux on thephotocathode. If the light flux is fairly high,concentrating it on a small area could causedamage to the photocathode or result innon-linearity effects because of the resistiv-ity of the photocathode layer. Furthermore,variations in photocathode sensitivity acrossthe surface area could cause some uncer-tainty in the measurement if the light spot istoo small. In scintillation-counting applica-tions, spot size is not generally a problembecause of the diffuse nature of the flux andthe size of the crystal. But, in flying-spot-scanner applications it is particularly impor-tant that the image being scanned should notbe in focus on the surface of the photomulti-plier because any non-uniformities ofphotocathode sensitivity would cause fixedpattern modulation.

92

Angle of IncidencePhotocathode response varies somewhat

with the angle of incidence (see Fig. 41). It isalso possible to increase the effective quan-tum efficiency of a photocathode by the useof a specially constructed optical coupling tothe photocathode window that utilizes at theglass-air interface angles of incidence greaterthan the critical angle. (See Fig. 42).Light Modulation

For some purposes, it may be advan-tageous to modulate the light signal with alight chopper. A synchronous motor may beused to rotate a chopper disk which couldthen provide an approximate square-wavemodulation. The output signal could then beanalyzed with a narrow-band-pass amplifiertuned to the chopper frequency. When themodulated signal is in the presence of an un-modulated background whether from thedark current of the tube itself or from an ex-ternal source of light, a significant improve-ment in signal-to-noise may be achieved.

Regarding the testing of photomultipliertubes with delta-function light pulses, see thesection on “Time Effects” in Chapter 4,Photomultiplier Characteristics. Cerenkovradiation also provides very short pulses oflight.

SPECIFIC PHOTOMULTIPLIERAPPLICATIONS

This section discusses various applicationsof photomultipliers and some of the specialconsiderations for each application. Thiscatalogue of applications is not completeeven for presently known applications, andnew ones are being continually devised. Theapplications discussed, however, are some ofthe major ones and the information givencan be readily adapted to other applicationsor to new ones.

Scintillation CountingA scintillation counter is a device used to

detect and register individual light flashescaused by ionizing radiation, usually in theform of an alpha particle, beta particle,gamma ray, or neutron, whose energy maybe in the range from just a few thousandelectron-volts to many million electron-volts. The most common use of scintillationcounters is in gamma-ray detection and spec-troscopy.

The gas, liquid, or solid in which a scin-tillation or light flash occurs is called thescintillator. A photomultiplier mounted incontact with the scintillator provides themeans for detecting and measuring the scin-tillation. Fig. 102 is a diagram of a basic

Fig. 102 - Diagram of a scintillation counter.

scintillation counter. The three most prob-able ways in which incident gamma radiationcan cause a scintillation are by the photoelec-tric effect, Compton scattering, or pair pro-duction. The reaction probabilities associ-ated with each of these-types of interactionare a function of the energy of the incidentradiation as well as the physical size andatomic number of the scintillator material.In general, for a given scintillator, thephotoelectric effect predominates at smallquantum energies, the Compton effect atmedium energies, and pair production atenergies above 1.02 MeV.Scintillation Processes. In the photoelectriceffect, a gamma-ray photon collides with abound electron in the scintillator and im-parts virtually all its energy to the electron.In the Compton effect a gamma-ray photon

tron in the scintillator and transmits onlypart of its energy to the electron, as shown inFig. 103. A scattered photon of lower energy

92CS-32440

Fig. 103 - Compton-effect mechanism.


also results. To satisfy the conditions of con-servation of energy and momentum, there isa maximum energy that can be transferred tothe electron. This maximum energy, knownas the Compton edge, occurs when e in Fig.

given by

the speed of light. The resultant energy im-parted to the electron can then range fromzero to a maximum of TCM.

In pair production, the energy of a gammaray is converted to an electron-positron pairin the field of a nucleus. The gamma raymust have energy at least equal to two timesthe rest-mass-energy equivalent of an elec-

energy is transferred as kinetic energy. Whenthe positron is annihilated, two photons areproduced 180 degrees apart, each with anenergy of 0.51 MeV. The photons are thensubject to the normal probabilities of in-teraction with the scintillator .

In neutron detection, unlike alpha- orbeta-particle or gamma-ray detection, theprimary interaction is with the nuclei of thescintillator atoms rather than its atomic elec-trons. The interaction may consist of scatter-ing or absorption; in either case, some or allof the energy of the neutron is transferred tothe recoil nucleus which then behavessimilarly to an alpha particle.

In each interaction between a form ofionizing radiation and a scintillator, an elec-tron having some kinetic energy is produced.A secondary process follows which is in-dependent of both the kind of ionizing radia-tion incident on the scintillator and the typeof interaction which occurred. In this secon-dary process, the kinetic energy of the ex-cited electron is dissipated by exciting otherelectrons from the valence band in the scin-tillator material into the conduction band.When these excited electrons return to thevalence band, some of them generate light orscintillation photons. The number of pho-tons produced is essentially proportional tothe energy of the incident radiation. In thephotoelectric interaction described above, all

93


of the incident photon energy is transferredto the excited electrons; therefore, thenumber of photons produced in this secon-dary process, and hence the brightness of thescintillation, is proportional to the energy ofthe incident photon.

Scintillation Mechanism. The exact mecha-nism of the scintillation or light-producingprocess is not completely understood in alltypes of materials; however, in an inorganicscintillator, the phenomenon is known to becaused by the absorption of energy by avalence electron in the crystal lattice and itssubsequent return to the valence band. Fig.104 shows a simplified energy-band diagram

IMPURITY - _ -LEVELS

92CS-32441

Fig. 104 - A simplified energy diagram of ascintillation crystal: ionized electron-holepairs; an exciton; an activator center throughwhich an electron may return to the groundstate causing fluorescence; or through whichan electron may return to the ground state bya thermal, non-radiative process; impuritylevels that can trap an electron for a timecausing phosphorescence.

of a scintillation crystal. The presence ofenergy levels or centers between the valenceand conduction bands is the result of im-perfections or impurities in the crystal lat-tice. Three types are important: (1) fluor-escence centers in which an electron, afterexcitation, quickly returns to the valenceband with the emission of a photon; (2)quenching centers in which the excited elec-tron returns to the valence band with thedissipation of heat without emission of light;and (3) phosphorescence centers in which theexcited electron can be trapped in ametastable state until it can absorb some ad-ditional energy and return to the valenceband with the emission of a photon. An im-portant process in the transfer of energy to

the fluorescence, or activator, centers in thegeneration of excitons or bound hole-electron pairs. These pairs behave much likehydrogen atoms and, being electricallyneutral, can wander freely through thecrystal until captured by fluorescencecenters. The emission of a photon from aphosphorescence center is a relatively slowor delayed process. Of the three types ofcenters through which an excited electroncan return to the valence band, the first,fluorescence, is that sought for in thepreparation of scintillators. The secondtype, quenching, tends to lessen the effi-ciency of the scintillator because it does notcause the emission of photons, and the third,phosphorescence, produces an undesirablebackground glow.Scintillator Materials. The most popularscintillator material for gamma-ray energyspectrometry is thallium-activated sodiumiodide, NaI:Tl. This material is particularlygood because its response spectrum containsa well-defined photoelectric peak; i.e., thematerial has a high efficiency or probabilityof photoelectric interaction. In addition, thelight emitted by the material covers a spec-tral range from approximately 350 to 500nanometers with a maximum at about 410nanometers, a range particularly wellmatched to the spectral response of conven-tional photomultipliers, NaI:Tl does not,however, have a fast decay time in com-parison to other scintillators, and, therefore,is generally not used for fast-time resolution.As a comparison, the decay time constantfor NaI:Tl is approximately 250 nano-seconds, while for a fast plastic scintillatorthe dominant decay time constant is in the 1to 4 nanosecond range.

The decay time of a scintillator involvesthe time required for all the light-emittingluminescence centers to return excited elec-trons to the valence band. In some of thebetter scintillators the decay is essentially ex-ponential, with one dominant decay timeconstant. Unfortunately, most scintillatorshave a number of components each with dif-ferent decay time.Collection Considerations. Because scin-tillations can occur anywhere in the bulk ofthe scintillator material and emit photons inall directions, there exists the problem of col-lecting as many of these photons as possible

94

on the faceplate of the photomultiplier. If itis assumed that the fluctuation in thenumber of photons resulting from a singleionizing event follows simple Poisson statis-tics*, the relative standard deviation in thenumber arriving at the face plate is given by

(32)

where Np is the average number of photonsarriving at the faceplate of the photo-multiplier per incident ionizing event. Np is

ber of photons per unit energy for the scin-tillation, E is the energy of the incidentradiation, and t is the fraction of the totalnumber of photons produced which arrive atthe faceplate of the photomultiplier.

Eq. (32) implies that it is highly desirablethat all emitted photons be collected at thephotomultiplier faceplate. This collectionproblem can be simplified by careful selec-tion of the shape and dimensions of the scin-tillator to match the photomultiplier photo-cathode dimensions. The coating of all sidesof the scintillator except that which is to beexposed to the photomultiplier faceplatewith a material that is highly reflective forthe wavelengths of the photons emitted bythe scintillator also proves helpful. BecauseNaI:Tl is damaged by exposure to moist air,it is usually packaged in an aluminum caselined with highly reflective MgO or Al2O3

powder; the NaI:Tl scintillator is providedwith an exit window of glass or quartz. Toavoid total internal reflection, it is importantthat the indices of refraction of the scin-tillator material, its window, any light guide,and the photomultiplier faceplate match asclosely as possible.

If it is not convenient for the photomulti-plier to be directly coupled to the scintillator,as when the photomultiplier entrance win-dow is not flat, light guides can be used.Again, care should be taken in the design ofthe light guide to assure maximum lighttransmission. The outer side or surface ofthe light guide should be polished and coatedwith a highly reflective material. In some

*Actually, the variance in the number of photons ex-iting from the crystal per event is much higher than ex-pected from the simple expression of Eq. (32). See thediscussion in Chapter 4 on “Scintillation Counting”related to Eq. G-l 11.


cases, a flexible fiber-optics bundle can beused. An optically transparent silicone-oilcoupling fluid should be applied at the scin-tillator (-light guide, if used)-photomultiplierinterface regardless of the light-conductionmethod used.

The next important consideration in scin-tillation counting is the conversion of thephotons to photoelectrons from the photo-cathode. The photocathode should have thegreatest quantum efficiency possible over thespectral range defined by the spectralemissivity curve of the scintillator. Themethod of determining the quantum effi-ciency of a photocathode as a function ofwavelength is explained in Appendix E. Ofthe various photocathodes available, bialkalitypes, K2CsSb and Rb2CsSb, having peakquantum efficiencies in excess of 25%, pro-vide the best spectral match to the emissionfrom most scintillators.

The uniformity of the photocathode, i.e.,the variation in quantum efficiency at agiven wavelength as a function of positionon the photocathode, is also important.Because the number of photoelectronsemitted for a constant number of photonsincident on the photocathode is proportionalto the quantum efficiency any variation inquantum efficiency as a function of positionresults in an undesirable variation in thenumber of photoelectrons emitted as a func-tion of position.

When the scintillation is fairly bright, i.e.,when a large number of photons are pro-duced per scintillation, and when the scin-tillator is thick in comparison to itsdiameter, as shown in Fig. 105(a), thephotocathode is approximately uniformly il-luminated during each scintillation. How-ever, if the scintillator is thin in comparisonto its diameter, as shown in Fig. 105(b), or ifthe scintillation is very weak, the illumina-tion of the photocathode as a function ofposition is closely related to the position oforigin of the scintillation; therefore, photo-cathode uniformity is much more importantwhen a thin scintillator is used. In this case,the use of a light guide may be advan-tageous. Photocathode uniformity also be-comes more important as the energy of theincident radiation becomes less and thenumber of photons per disintegration isreduced.

95


REFLECTIVE,-COATING

REFLECTIVE/-COATING

COUPLINGFLUID

(b)

9 2 C S - 3 2 4 4 2

Fig. 105 - Scintillator geometries: (a) thick incomparison to diameter; (6) thin in com-parison to diameter.

In some photomultipliers, the collectionefficiency for photoelectrons decreases nearthe outer edges of the photocathode;therefore, best results are obtained when thescintillator is slightly smaller in diameterthan the photocathode.Significant Photomultiplier Characteristics.In scintillation-counting applications (Seealso the section on “Pulse Counting” inChapter 4, Photomultiplier Characteristics),a photomultiplier should be selected to pro-vide a photocathode diameter matching thatof the crystal to be used. The most importantcharacteristic of the tube to be used in scin-tillation counting is then the effective photo-cathode sensitivity to blue and near-ultra-violet wavelengths. The effective photocath-ode sensitivity includes the basic quantumefficiency and the collection efficiency of theelectron-optical system. Tubes havingVenetian-blind dynodes provide a fairly largeopening to the first dynode area and thusfrequently have better collection efficiencythan tubes having a focused-dynode struc-ture, although the focused structure is muchbetter for time resolution. Photomultipliershaving the recent “tea-cup” type of firstdynode generally have very good collectionefficiency and, in addition, are designed toprovide increased photocathode emission

96

either by reflected light or by excitation ofphotoelectrons from photocathodedeposited on the side walls of the envelope.

At high count rates, tubes having copper-beryllium dynodes generally provide greaterstability than tubes having cesium-antimonydynodes, although for low count rates thelatter prove to be satisfactory. A tube havinggood stability may be expected to shift ingain by no more than 7% in several monthsof continuous operation at a count rate of10,000 per second. Variation of pulse heightor gain with count rate is also of importance.Well-designed tubes should show a variationin pulse height of less than 1% betweencount rates of 1000 per second and of 10,000per second, related to the photopeak, using137Cs and a NaI:Tl crystal.

Photomultiplier dark noise is of particularimportance in scintillation-counting applica-tions when the energy of the ionizing radia-tion is small, or when very little energy istransferred to the scintillation medium; inshort, when the flash per event representsonly a few photons. If a photomultiplier iscoupled to a scintillator and voltage is ap-plied, the composite of all noise pulsescoming from the photomultiplier is referredto as the background of the system. The plotof a frequency distribution of these pulses asa function of energy is shown in Fig. 106.

PULSE HEIGHT

92CS-32443

Fig. 106 - Distribution of radiation back-ground pulses as a function of energy. A cali-bration spectrum showing the 137CS photo-peak is included for reference.

The well-defined peaks seen in the figure arecaused by external radiations and can bereduced by placing the photomultiplier andscintillator in a lead or iron vault to reducebackground radiation.


It is desirable that the background countof the scintillation counter be as low aspossible. Photomultipliers have been devel-oped in which the background count is keptlow by the use of radioactively clean metalcans instead of glass-bulb enclosures andfaceplates of either Lucalox* or sapphire in-stead of glass. As an additional aid in reduc-ing background, the scintillation counter canbe surrounded by another scintillator, suchas a plastic sheet equipped with its ownphotomultipliers. The output from this scin-tillation shield is then fed into an anticoin-cidence gate with the output from the scin-tillation counter. The gating helps to reducethe background contributions from both in-ternal and external radioactive con-taminants.Liquid Scintillation Counting

In liquid scintillation counting, the scin-tillation medium is a liquid and the ionizingradiations to be detected are usually low-energy beta rays. Because low-energy betarays cannot penetrate a scintillation con-tainer, the radioactive material is dissolvedin a solution containing the scintillator. Thescintillator is generally one or more fluores-cent solutes in an organic solvent. The beta-ray energy is transferred by ionization andby excitation that in turn results in the emis-sion of photons in the near ultraviolet. Insome cases, a wavelength shifter is also usedto provide radiation at somewhat longerwavelengths more compatible with the spec-tral responsivity of the photomultiplier.

Important characteristics of the photo-multiplier in this application are again highresponsivity in the spectral region of thescintillation and low background count rate.A useful figure of merit is the square of thecounting efficiency divided by the back-ground coincident count rate in a pairedphotomultiplier arrangement. Furtherdetails of liquid scintillation counting ap-plications are discussed in the section “PulseCounting” in Chapter 4, PhotomultiplierCharacteristics.Cerenkov Radiation Detection

Cerenkov radiation is generated when acharged particle passes through a dielectric

*Registered Trade Name for General Electric Co.material.

with a velocity greater than the velocity oflight in the dielectric (i.e., v greater than c/n,where n is the index of refraction of thedielectric). Polarization of the dielectric bythe particle results in the development of anelectromagnetic wave as the dielectricrelaxes. If the velocity of the particle isgreater than the velocity of light, construc-tive interference occurs and a conical wave-front develops, as illustrated in Fig. 107.

with respect to the direction of the particle isgiven by the expression

(33)

where n is the index of refraction of the

velocity to the velocity of light. The spectralenergy distribution of the radiation increases

limited by the absorption of the medium.Cerenkov radiation is the electromagnetic

counterpart of the shock wave produced in agas by an object traveling faster than sound.It is highly directional and occurs mostly inthe near-ultraviolet part of the elec-tromagnetic spectrum. Because the radiationis propagated in the forward direction ofmotion of the charged particle, as shown inFig. 107, Cerenkov detectors can be made todetect only those particles that enter thesystem from a restricted solid angle.

Cerenkov radiation produced in anaqueous solution by beta emitters can beuseful in radioassay techniques because it isunaffected by chemical quenching andbecause it offers the advantages, over liquid-scintillation counting techniques, of simpli-fied sample preparation and the ability to ac-commodate large-volume samples. Becausea fast particle is required to produceCerenkov radiation, rather high-energy betarays are required; e.g., the threshold forCerenkov radiation is 261 keV for electronsin water. Because the photon yield forCerenkov light is usually very low, the sameconsiderations concerning photomultiplierselection apply for the Cerenkov detection asfor liquid-scintillation counting. The tubeselected should also be equipped with afaceplate capable of good ultraviolet trans-mission. In some experiments it may be im-

97


portant to select photomultipliers for theirspeed of response. Depending upon thedispersion of the medium, the duration ofthe Cerenkov flash can be very short.

Time SpectroscopyIn addition to the energy spectroscopy

described above, there are occasions when itis of advantage to measure time differencessuch as between a pair of gamma rays or acombination of gamma rays and particles incascade de-exciting some level in a nucleus.In time spectroscopy, some special con-siderations must be made in selecting thescintillator, photomultipliers, and techniqueof analyzing the signals from the photomul-tipliers.

In a photomultiplier, time resolution isproportional to (n) - 1/2, where n is theaverage number of photoelectrons per event.It is therefore important to choose a scin-tillator material that provides a high lightyield for a given energy of detected radia-tion. It is also important that the variation ofthe time of interaction of the radiation withthe scintillator be as small as possible. Thisminimum variation is assured by attention toscintillator thickness and source-to-detectorgeometry. The decay time constant of thelight-emitting states in the scintillator should

be as short as possible, and the geometry andreflective coatings of the scintillator shouldbe selected so that variations in path lengthsof photons from the scintillator to thephotocathode of the photomultiplier areminimized. The photocathode of the photo-multiplier selected should have a high quan-tum efficiency. In addition, the transit-timedispersion or jitter (variations in the time re-quired for electrons leaving the photocath-ode to arrive at the anode of the tube) shouldbe small over the entire photocathode area.

The major contribution to transit-timespread occurs in the photocathode-to-first-dynode region and may be a result of the in-itial kinetic energies of the emitted photo-electrons and their angle of emission. Focus-ing aberrations, and the single-electronresponse or rise time, i.e., the output-pulseshape at the anode for a single photoelectronimpinging on the first dynode, may also beof some importance. Although the single-electron response theoretically does not havemuch effect on time resolution, it doeschange the triggering threshold at which thebest time resolution can be obtained.

The most commonly used time-spectro-scopy techniques include leading-edge tim-ing, zero-crossover timing, and constant-fraction-of-pulse-height-trigger timing. Thetechnique used depends on the time resolu-tion and counting efficiency required and therange of the pulse heights encountered. Ablock diagram of a basic time spectrometeris shown in Fig. 108.

Leading-edge timing makes use of a fixedthreshold on the anode-current pulse andprovides good time resolution over a narrowrange of pulse heights. The fractional pulseheight F at which the triggering threshold isset is defined as follows:

(34)

where Vt is the discriminator threshold, andVa is the peak amplitude of the anode-current pulse. The fractional pulse heighthas a considerable effect on the time resolu-tion obtained; best results are usually ob-tained with F equal to 0.2.

In fast zero-crossover timing, the anodepulse is differentiated. This differentiationproduces a bipolar output pulse that triggers

98


DETECTOR I

TIMINGPULSE

NUMBER TIMEOF EVENTS SPECTRUM

SOURCE OFCOINCIDENTRADIATION

t

DETECTOR 2MULTICHANNEL

(PULSE HEIGHT)ANALYZER

Fig. 108 - Generalized block diagram of a time spectrometer.

the timing discriminator at the zerocrossover, the time required to collect ap-proximately 50 per cent of the total charge inthe photomultiplier pulse. Zero-crossovertiming is second to leadingedge timing fortime-resolution work with narrow pulse-height ranges, but is better than the leading-edge method for large pulse-height ranges.

In constant-fraction triggering, the pointon the leading edge of the anode-currentpulse at which leading-edge timing data in-dicate that the best time resolution can beobtained is used regardless of the pulseheight. For this reason, constant-fraction-of-pulse-height timing is the best method forobtaining optimum time resolution nomatter what the pulse-height range.

Oil-Well LoggingLogging is the term given to the method of

determination of the mineral compositionand structure a few miles under the earth’ssurface. Oil-well logging companies gatherdata by means of probes, or sondes, that ex-amine the geological media along very deepbore holes. The probes determine variousphysical and chemical characteristics of thematerial in their vicinity. Measurementsmade by the probes comprise the log.

A variety of sondes are used in selectedcombinations to determine various aspectsof the lithology (the character of a rock for-mation), including density, of the mediaalong the bore hole. The combinations ofsondes used depend very much on the bore-hole media. For example, when a formation-density sonde is used in combination with aneutron sonde in liquid-filled bore holes,both lithology and porosity can be deter-mined. The same pair of sondes allows themeasurement of gas and liquid saturation inbore holes drilled through reservoirs of low-pressure gas. The use of the formation-density sonde, along with either an inductionor a sonic sonde, permits a similar type ofdetermination. The final result of thelogging activity is information concerningthe existence of hydrocarbons and othergeological media of interest in establishingan oil field.

The formation-density sonde is one of themore sophisticated logging devices. Itsoperational elements are encased in a ruggedcylindrical housing, as shown in Fig. 109. Itis designed to withstand the high tempera-ture and shock encountered in probing boreholes miles deep under the earth’s surface.The sonde contains a gamma-ray source,such as radioactive cesium 137, a detectorconsisting of a sodium iodide crystal and aphotomultiplier tube, a gamma-ray shield

99


ELECTRONICS

PRESSURE FOOT

GAMMA SHIELD

Fig. 109 - Formation-density sonde sends gamma rays into rock formation and then detectsreturns with photomultiplier.

for the detector, a pressure foot that pressesthe sonde against the bore-hole wall, andoperating electronics. The objective of thissonde is to determine the bulk density of thematerial in the region of the probe.

Gamma rays from the source interact withthe atoms in the geological medium.Compton-scattered gamma rays are detectedby the crystal and photomultiplier in thesonde. A knowledge of gamma-ray penetra-tion of bulk media, mass absorption data,and the special effects resulting from thechemical nature of various geological mediaallows the information imparted by thepulses to be deciphered by the electronicsinto bulk-density data, which becomes thesubstance of the geological formation-density log.

lustrated in Fig. 110. Most of the decrease isthe result of photocathode sensitivity loss,but some is associated with the crystal. Be-cause of the loss in photocathode responseand the increase in thermionic emission withtemperature, the desired signal is finally lostin the background noise at a temperaturenear 200°C. Permanent damage to the pho-tocathode may also be expected after manycycles of operation at 200°C.

Because of the increase in temperaturewith the depth of the bore hole-to perhaps150°C at 15,000 feet-a most importantcharacteristic of the photomultiplier is itsresistance to high temperature. Of thenumerous photocathodes which have beendeveloped, the Na2KSb “bialkali” photo-cathode is the most stable at elevated tem-peratures.95 (See Fig. 83 and related text.)

92cs - 32447

Fig. 110 - Pulse-height resolution and pulseheight as a function of temperature for a pho-tomultiplier having a Na2KSb photocathodeand used with a Nal:TI crystal and a 137Cssource.

As the temperature is increased on a pho-tomultiplier with a Na2KSb photocathodecoupled to a NaI:Tl crystal, the pulse height(137Cs source) gradually decreases, as il-

Gamma-Ray CameraThe gamma-ray camera originally de-

scribed by Anger96is a more sophisticated

100

version of the scintillation counter, and isused for locating tumors or other biologicalabnormalities. The general principle of thegamma camera is illustrated in Fig. 111. A

92CS-32448

Fig. 111 - Structure of gamma-ray camera.

radioactive isotope combined in a suitablecompound is injected into the blood streamof the patient or is ingested orally. Certaincompounds or elements are taken up prefer-entially by tumors or by specific organs ofthe body, such as iodine in the thyroid gland.As the radioactive isotope disintegrates,gamma rays are ejected from the location ofthe concentration.

A lead collimator permits gamma rays topass through it only when they are parallel tothe holes in the lead; gamma rays at otherangles are absorbed in the lead. In this way,the location of the gamma-ray source may bedetermined because gamma rays originating


on the left side of the organ are caused to im-pact the left side of the scintillation crystal,etc. The crystal covers an area about 10inches or more in diameter.

Behind the crystal are, perhaps, 19 photo-multiplier tubes in a hexagonal array. Thelight of the individual scintillation is not col-limated but spreads out to all of the 19 tubes.The location of the point of scintillationorigin is obtained by an algorithm dependingupon the individual signals from each of thephotomultipliers. Resolution is obtained inthis manner to about 1/4 inch. Each scintilla-tion is then correspondingly located by asingle spot on a cathode-ray tube. Countingis continued until several hundred thousandcounts are obtained and the organ in ques-tion is satisfactorily delineated. Fig. 112 is areproduction of such a scintigram. The par-ticular advantage of the gamma camera overother techniques such as the CT scanner(described below) is that the gamma cameraprovides functional information. For exam-ple: Tc-99m polyphosphate is used to revealbone diseases; 123I is used in thyroid studies;127Xe is inhaled to provide information onlung ventilation.

Of great concern to the designer of gammacameras and, of course, to the ultimatecustomer is the inherent resolution capabilityof the device. Generally, with more photo-multiplier tubes sampling the scintillationdistribution, the delineation becomes moreprecise. Most of the original cameras utilized

Fig. 172 - Scintigrams obtained by Lancaster General HospitaI with a gamma camera. The scin-tiphoto on the left shows multiple emboli in the right lung; photo on right shows lungs afterclearing. The isotope technitium-99m was used to tag albumin microspheres-a colloidal formof the albumin protein, with particles ranging in size from 2 to 50 microns. These particles areinjected into the bloodstream and are filtered and trapped in the lung capillary bed; the scancan then determine those areas where the capillary bed is intact. Areas of diminished bloodflow show as “cold spots.”

101


a hexagonal array of 19 tubes. By addinganother circumferential row, an array of 37provided better resolution. In the same man-ner cameras are now available with 61 tubesand even 91. (Note the numerical progres-sion of these hexagonal arrays: 1 + 6 + 12 +18 + 24 + 30.) Different photomultiplierdimensions also are used to provide in-struments with appropriate portability orcoverage. Most common has been the 3-inchphotomultiplier, but a large number of2-inch tubes are also used and there is con-sideration of the use of 1 1/2-inch tubes. Hex-agonal 2- and 3-inch tubes have beendeveloped to provide better space utilization,and a photomultiplier with a squarefaceplate is now available.

Resolution of the gamma-ray camera de-pends fundamentally upon the pulse-heightresolution of the photomultiplier and,hence, the quantum efficiency of the photo-cathode. The most common photocathodeselected for gamma-ray-camera applicationis the bialkali (K2CsSb). Also important inthe determination of good pulse-heightresolution is collection efficiency of thephotoelectrons. Large first dynodes are ap-propriate and the “tea-cup” configuration isused to advantage.

Because each scintillation is sensed by anumber of the photomultipliers in the array,the spatial uniformity of the photomultiplierresponse and its angular response becomeimportant especially as related to thealgorithm of the gamma-ray-camera design.The camera design may also incorporatemodifications in the light pipe such asgrooves or etched patterns to provide im-proved spatial location of the scintillations.

Finally, the stability of the photomulti-plier is important in maintaining the resolu-tion and spatial integrity of the display.Although the photomultiplier currents aregenerally small in this application, changescan occur in the photomultiplier gain whichwould then result in positional errors in theCRT display. Some gamma-ray-cameradesigns include a means of recalibrating thearray of photomultipliers so that the pulse-height “window” is the same for each tube.

"For a more detailed discussion, see the article inScientific American, October, 1975, p 56, “ImageReconstruction from Projections,” R. Gordon, G.T.Herman, and S.A. Johnson.

102

This recalibration could be done periodically(once a day) and would contribute signifi-cantly to consistent, distortion-free opera-tion.Computerized TomographicX-Ray Scanners

The computerized tomographic (CT) scan-ner produces an X-ray image in an entirelydifferent manner from that of conventionalradiography.* The standard X-ray picture isa shadowgraph. Thus, a chest X-ray pro-duces an overlay of shadows from the ribcage and internal body structure. Interpreta-tion is frequently difficult because of the in-terfering images. The CT scanner, on theother hand, provides a density image thatrepresents a cross section of the patient-atomograph. Thus, a CT scan of the headwould show the outer bone structure, thefolds of the brain, and possibly a tumor in-side the skull as though a complete thin slicehad been taken through the middle of thehead. This tomographic image is producedby exposing the head to X-rays at many dif-ferent angles of entrance. The multiplicity ofshadow-type images thus formed are ana-lyzed by a computer, which produces areconstructed cross-section density image.

A diagram of a typical CT scanner isshown in Fig. 113. An X-ray source pro-ducing a fan beam rotates around the patientin a few seconds. (Short periods are desirableto minimize artifacts in the final recon-structed image of the patient’s cross sectioncaused by unavoidable motions.) An arrayof several hundred detectors in an outer cir-cle surrounding the patient provides the datafrom which a computer derives a tomo-graphic view of the patient’s body or head.A typical tomograph is shown in Fig. 114.

Three different X-ray detector systemshave been developed for use in CT scanners.All are in use at present by the various CT-scanner manufacturers. Each detectorsystem has its problems, but all seem com-parable in ultimate performance.

The original CT-scanner developmentutilized a crystal and photomultiplier. Asubsequent development used a tube con-taining xenon gas at several atmospheres.The density of the gas and the high atomicweight of xenon provide sufficient absorp-tion so that a large fraction of the X-rays aredetected by the resulting ion current. More


ARRAY OF STATIONARYDETECTO RS ENCIRCLINGPATIENT

92cS-32662Fig- 113 - Diagram of a typical CT scanner. X-ray source producing a fan beamrotates around the patient. Array of several hundred detectors in outer circle provides the data from which a computer derives a tomographic (cross-section) view ofthe patient’s body or head.

recently, detector packages have beendeveloped using silicon photocells and aCdWO4 crystal. This latter detector packageappears to have cost advantages. Detectionwith the silicon photocell was not feasible inthe early development of CT scanners, but acombination of higher-speed machines andthe improved light output of the CdWO4

crystal, relative to the Bi3Ge4O12 (BGO)crystal, has made this combination aneconomic possibility.

Fig. 114 - Tomograph showing a “slice” ofthe thoracic cavity. Lung tissue, blood ves-sels, air passages, ribs, and spine may be ob-served. (Courtesy Pfizer Medical Systems,Inc.).

The photomultiplier system uses a BGO(bismuth germanate) crystal usually coupledto the photomultiplier by means of a lightpipe. Spectral emission from the crystal is il-lustrated in Fig. 115. It is rich in the blue-

I I I I I300 400 500 600


92CS-32452

Fig. 115 - Fluorescence spectrum ofBi4Ge3O12 (Data from Weber and Mon-champ97).

green region and thus is a good match forbialkali photocathodes. Recently, it has beenfound that the Rb2CsSb photocathode ismore suitable for use with the BGO crystalthan the K2CsSb photocathode because of

103


the somewhat higher response and betterspectral match. See Fig. 10. The Rb2CsSbphotocathode is also more suitable because ithas a lower surface resistance than theK2CsSb photocathode. See Fig. 36.

In the CT-scanner systems, pulse-heightdiscrimination is not used nor are individualscintillations counted. The photomultiplieris used as an integrator of the light flux ineach density measurement. It is importantthat the photomultiplier provide a lineartranslation of the light flux into output cur-rent and also that there be no overshoot orundershoot in photocurrents as the X-raybeam path changes from outside to throughthe patient resulting in a signal variation of1000:1 or more. Typical maximum photo-cathode current (X-ray path through aironly) is of the order of 2 nanoamperes de-pending on operation conditions. The pho-tomultiplier is usually operated at a modestgain figure of the order of 60,000. Highphotocathode sensitivity is important in pro-viding minimum noise in the signal so thatthe inherent signal-to-noise ratio in thetransmitted X-ray beam is retained and theX-ray dose to the patient in minimized. Thephotomultiplier tubes used are either 1/2 inchor 3/4 inch in diameter, depending upon theconfiguration of the particular CT scanner.Positron Camera

vides tomographic presentations based oncoincident gamma-ray emission accompany-ing annihilation of a positron and an elec-tron. Tracer radionuclides such as 11C, 13N,or 15O emit a positron upon disintegration.In the presence of matter such as the brain,the positron interacts almost instantly withan electron resulting in the simultaneousemission of two gamma rays each having anenergy of 511 keV, but moving in nearly op-posite directions. When a pair of detectors,one on either side of the patient, observescoincident events, the point of radionuclidedisintegration lies on a line joining the twodetectors. See Fig. 116.

A number of positron cameras have beendesigned and built, some commercially.Various geometries and reconstruction tech-niques have been utilized. Complications indesign arise when each detector can be incoincidence with any of several detectors onthe opposite side of the patient. More than

104

9 2 C S - 3 2 4 5 3

fig. 116 - Principle of positron coincidencedetection. A disintegration at “a ” results inoppositely directed gamma rays which aredetected in coincidence by the pair of scintil-lator-photomultiplier detectors. An event at"b” results in a count in only one of the detec-tors. (From G.L. Brownell and C.A. Burn-ham98)

one plane may be utilized to provide three-dimensional data.

Reponse time of the photomultiplier is ofparticular importance in positron camerasbecause the discrimination against spuriouscoincidences improves as the resolution timeof the system decreases. Photomultipliertransit time spreads of the order of 2nanoseconds or less are advantageous inthese systems although other time- factors,related to the crystal scintillator (CsF has atime constant of 5 nanoseconds) and cir-cuitry may limit the system discriminationtime to perhaps 10-20 nanoseconds. Photo-multiplier tubes of 3/4-inch and 1 1/2-inchdiameter have been the sizes preferred. Ofcourse, high quantum efficiency matchingthe spectral emission of the scintillators(NaI:Tl, CsF, Bi4Ge3O12) is also important.Photometric and Spectrometric Applications

The side-on photomultiplier has in gener-al, been the most widely used tube in photo-metric and spectrometric applications. Theside-on tube is relatively small and has arectangularly shaped photocathode thatmatches the shape of the light beam from anexit slit. Spectrometric applications requiretubes with good stability, high anode sen-sitivity, low dark current, and broad spectralsensitivity. A high signal-to-noise ratio is im-portant because of the generally small signallevels.

Before the development of the newnegative-electron-affinity type photoemit-ters, such as gallium arsenide, it wasnecessary to use more than one detector inthe measurement of radiant energy from thenear-ultraviolet to the near-infrared part ofthe spectrum. A photomultiplier having agallium arsenide photoemitter can now beused to detect radiant energy from the cutoffpoint of the photomultiplier window in thenear-ultraviolet to 910 nanometers.

Spectrophotometry. Spectrophotometersmeasure the optical density of materials as afunction of wavelength and require photo-multipliers having a broad spectral range.The results of measurements of the absorp-tion characteristics of substances are fre-quently expressed in terms of optical density.This logarithmic method correlates with theway the human eye discriminates differencesin brightness.

The transmission density D is defined by

where PO is the radiant flux incident upon asample, Pt is the radiant flux transmitted bya sample, and T is the transmission figureequal to Pt/Po. Density measurements areuseful in various applications to films andother transparencies in addition to chemicalanalysis where concentration of a solution isstudied as a function of wavelength.

Color-Balancing Photometry. A color-balancing photometer is used to determinecolor balance and exposure times necessaryto produce photographic color prints fromcolor negatives. Such a device is shown inblock diagram form in Fig. 117. It allows thematching of the relative proportions of red,blue, and green light transmitted by a pro-duction negative to those of a master colornegative. As the first step in the matchingprocess, the master negative is used to makean acceptable print. This first print is pro-duced through trial and error by measure-ment of the relative proportions of red, blue,and green light transmitted through a keyarea of the master negative; the key areausually consists of a flesh tone or a grayarea. The amount of light transmitted is a


measure of the density. The exposure timeusing white light and the lens opening usedto obtain the satisfactory print are noted.

Next, an area of the production negativesimilar to that on the master negative ischosen and the relative proportions of thered, blue, and green light transmitted aremeasured again. By use of magenta andyellow correcting filters, the color trans-mitted by the production negative can bebalanced with that of the master negative.When the production negative is used, thelens opening is adjusted so that the exposuretime with white light is the same as it was forexposure of the master negative. When thevalues of color-correcting filters and ex-posure times thus determined are used,prints from the production negative can beobtained which are very nearly as good asthose obtained with the master colornegative.

Most color-balancing photometers employsteady light sources and handle small-amplitude signals. Consequently, the photo-multiplier used must have low values of darkcurrent and good stability. The life expectan-cy of tubes used in this application is longbecause the small signal levels reduce the ef-fects of fatigue which might otherwiseadversely affect measurement accuracy andrepeatability. Low-current operation alsoprovides for linear operation where theanode current is proportional to the inputflux over the range of transmission valuesmeasured. The intensity range of a color-balancing photometer, given in terms of theratio of the radiant flux incident upon asample to the radiant flux transmitted by asample, is usually of the order of 1000 to 1 ormore. It must be remembered that as in mostphotomultiplier applications, the power sup-plies used must be capable of providingvoltages sufficiently regulated and free fromripple to assure minimum variation in sen-sitivity with possible line-voltage variation.

Densitometry. Although techniques suchas those employed in the color-balancingphotometer may be used successfully tomeasure density over an intensity range of 10or 100 to 1 (density 1 to density 2), their usebecomes increasingly difficult as the range isincreased to 1000 to 10,000 to 1 (density 3 todensity 4) or more. The large dynamic rangesencountered in color-film processing place

105


92CS-32454

Fig. 117 - Block diagram of a color-balancing photometer.

severe requirements upon the photomulti-plier because it must be operated in aconstant-voltage mode. Problems develop inthis mode at high-density values at which thedark current may become a significant pro-portion of the signal current. The randomnature of this dark current, which precludesits being “zeroed out”, may lead to output-signal instability. As a result of the type ofoperation needed to produce the dynamicrange required in density measurements, thephotomultiplier anode current is high at lowdensity values. These high currents mayresult in excessive fatigue, and, dependingupon the operating point, perhaps non-linear operation. As with the color-balancingphotometer, a well regulated low-ripplepower supply is needed to assure accuratemeasurements.

Logarithmic Photometry. In the measure-ment of absorption characteristics, thechanges in brightness levels vary over such alarge range that it is advantageous to use aphotometer whose response is approximatelylogarithmic. This response enables the pho-

106

tometer to be equipped with a meter or read-out scale that is linear and provides precisereadings even at high optical densities.

A simplified circuit of a logarithmicphotometer capable of measuring filmdensity with high sensitivity and stability andof providing an appropriate logarithmicelectrical response and linear meter indica-tion of density over three or four densityranges is shown in Fig. 118.

The circuit of Fig. 118, in which thephotomultiplier operates at a constant cur-rent, minimizes photomultiplier fatigue andeliminates the need for a regulated high-voltage supply. The feedback circuit il-lustrated develops a signal across Rl propor-tional to the anode current. This signal con-trols the bias applied to the control deviceand automatically adjusts the current in thevoltage-divider network. By this means, thedynode voltage is maintained at a level suchthat the anode current is held constant at avalue selected to minimize the effects offatigue. At optical densities of 3, the dynodesupply voltage may be 1000 volts; at opticaldensities of less than 1, the dynode voltage


COLOR FILTERS

92CM-32455

Fig. 118 - Block diagram of a logarithmic photometer.

may be as low as 300 volts. Dynode voltageis translated into density by means of thescale on voltmeter V1 .

Because there is an approximately ex-ponential variation of sensitivity of aphotomultiplier with applied voltage (seeFig, 48), in a constant-current mode, thevoltage varies almost linearly with thelogarithm of the input light flux. The voltageis thus close to a linear measure of thedensity. A compensating circuit for the lackof linearity is indicated in Fig. 118 in theform of a variable and automatic shuntacross the voltmeter V1 . As the densityvalues increase, the effective value of theshunt resistance is reduced. This circuit notonly compensates for photomultiplier non-linearity, but also for the non-linearity of theoptical system employed.

SpectrometrySpectrometry, the science of spectrum

analysis, applies the methods of physics andphysical chemistry to chemical analysis.

Spectrometric applications include absorp-tion, emission, Raman, solar, and vacuumspectrometry, and fluorometry.

Absorption Spectrometry. Absorptionspectrometry, used to detect radiant energyin the visible, ultraviolet, and infraredranges, is one of the most important of theinstrumental methods of chemical analysis.It has gained this importance largely as aresult of the development of equipmentemploying photomultipliers as detectors.The principle underlying absorption spec-trometry is the spectrally selective absorp-tion of radiant energy by a substance. Themeasurement of the amount of absorptionaids the scientist in determining the amountof various substances contained in a sample.

The essential components of an absorp-tion spectrometer are a source of radiantenergy, a monochromator for isolating thedesired spectral band, a sample chamber, adetector for converting the radiant energy toelectrical energy, and a meter to measure theelectrical energy. The spectral ranges of the

107


source and detector must be appropriate tothe range in which measurements are to bemade. In some cases this range may includeone wavelength. In others, it may scan allwavelengths between the near-ultravioletand the near-infrared.

Raman Spectroscopy. There are two typesof molecular scattering of light, Rayleighand Raman. Rayleigh scattering is the elasticcollision of photons with the molecules of ahomogeneous medium. Because the scat-tered photons do not gain energy from orlose energy to the molecule, they have the

classic example of Rayleigh scattering oflight from gas molecules is the scattering ofthe sun’s light rays as they pass through theearth’s atmosphere. This scattering accountsfor the brightness and blueness of the sky.

Raman scattering is the inelastic collisionof photons with molecules that producesscattered photons of higher or lower energythan the initial photons. During the collisionthere is a quantized exchange of energy that,depending on the state of the molecules,determines whether the initial photon gainsor loses energy. The differences in energylevels are characteristic of the molecule. If

emitted after the interaction has a frequency

from the molecule that was in the excitedstate. In the reverse case, the initial photonhas given up energy to the molecule in theunexcited state.

Early Raman instruments had a numberof disadvantages and were difficult to use. Itwas difficult to find a stable high-intensitylight source and to discriminate againstRayleigh scattering of the exciting line. Withthe recent development of high-qualitymonochromators and the advent of the laserlight source, a renewed interest in the Ramaneffect as an analytical method of chemicalanalysis has taken place. Raman spec-trophotometers are generally used to in-vestigate the structure of molecules and tosupplement other methods of chemicalanalysis, particularly infrared-absorptionspectrometry.

The scattered photons from a Raman in-

108

teraction are so few in number that only thehighest-quality photomultipliers can be usedas detectors. The tube should have high col-lection efficiency, high gain, good multipli-cation statistics, low noise, and high quan-tum efficiency over the spectral range of in-terest. Fig. 119 shows a Raman spectrum. Toreduce the effects of Rayleigh scattering, asource of noise in Raman spectroscopy, thephotomultiplier is placed at right angles tothe light beam, whose wavelength has beenchosen as long as possible within the rangeof interest.

92CS-32456

Fig. 119 - Typical Raman spectrum.

Fluorometry. The fluorometer is anotherinstrument which utilizes the photomulti-plier’s capability for low-light-level detec-tion, its high gain, and its good signal-to-noise ratio. There are numerous applicationsof this instrument in the fields ofbiochemistry, medical research, and in-dustrial toxicology. Typical applications in-clude the detection and measurement ofminute quantities of air pollutants and ofcomponents of the blood or urine. Somematerials are detected by their ownfluorescence; others by their quenching ef-fects on the fluorescence of other materials.

Illustrative of fluorometer operation is theoptical design of a model designed by G. K.Turner associates103 shown in Fig. 120. Thesample is irradiated by an ultraviolet sourcefiltered to eliminate longer wavelength com-ponents of the lamp irradiation. Fluorescentspectra at a longer wavelength pass througha second filter which eliminates scatteredultraviolet radiation.

A second beam from the ultraviolet sourceprovides a calibrated reference flux. Boththe fluorescent beam and the reference beam


are directed to the photomultiplier and alter-nately sampled by means of the light inter-rupter. The output ac signal from thephotomultiplier is processed to provide anull signal by means of the light cam and theattached dial records the fluorescence level.The null feature of the device eliminates er-rors from changes in the photomultiplier orfrom the ultraviolet source. A third flux in-dicated by the FORWARD LIGHT PATH isprovided so that even for an absence offluorescence a balance is always possible.The BLANK KNOB provides an adjustmentso that the calibration can be adjusted to thezero fluorescence point.Low-Light-Level Detection

Systems for the detection of low lightlevels make use of two basic techniques:charge integration, in which the outputphotocurrent is considered as an integrationof the anode pulses which originate from theindividual photoelectrons, and the digitaltechnique in which individual pulses arecounted.

Charge-Integration Method. In thecharge-integration method, either thetransit-time spread of the photomultiplier orthe time characteristics of the anode circuitcause the anode pulses to overlap and pro-duce a continuous, though perhaps noisy,anode current. The current is modulated byturning the light off and on by means ofsome mechanical device such as a shutter orlight “chopper”. The signal becomes the dif-ference between the current in the light-onand light-off conditions.

Detection is limited by noise in the anodecurrent. At low light levels the noise iscaused primarily by the fluctuations in darkcurrent of the photomultiplier (as discussedin the section, “Dark Current and Noise” inChapter 4 Photomultiplier Characteristicsand in Appendix G, “Statistical Theory ofNoise in Photomultiplier Tubes. ") The noisemay be minimized by reducing the band-width of the measuring system. For example,a dc system may be used with a bandwidth ofonly a few hertz if an appropriate low-levelcurrent meter is selected. Bandwidth can alsobe reduced by some technique of averagingthe current fluctuations over a period oftime.

Another technique of charge integration isto chop the light signal with a motor-drivenchopper disk having uniformly spaced holesor slots. The output current is then fedthrough an amplifier having a narrow band-width tuned to the frequency of chop. Band-widths of the order of 1 Hz are typical.

Digital Method. In the digital method forthe detection of low light levels a series ofoutput pulses, each corresponding to a pho-toelectron leaving the photocathode of aphotomultiplier, appears at the anode. All ofthe output pulses from the tube are shapedby a preamplifier before they enter a pulse-amplitude discrimination circuit. Only thosepulses having amplitudes greater than somepredetermined value and having the properrise-time characteristics pass through to thesignal-processing circuits. The digital tech-nique is superior to charge integration at

109


very low light levels because it eliminates thedc leakage component of the dark current aswell as dark-current components originatingat places other than the photocathode. Fig.121 shows a digital system in block form.

In the special case in which the digitaltechnique is used to count single photons in-cident upon the photocathode of a photo-multiplier, signal pulses appear at the anodewith an average pulse amplitude PH equal toem, where e is the electron charge and m isthe photomultiplier gain. The number ofsignal pulses Na arriving at the anode isgiven by

(36)

where N is the number of photons incidenton the phquantum efficiency of the photocathode atthe photon wavelength including a factor forthe loss of light by reflection and absorption,and a factor for the loss resulting from im-perfect electron-collection efficiency of thefront end of the photomultiplier.

The dark-noise pulses present in additionto the signal pulses originate mainly fromsingle electrons and have a pulse-heightdistribution as shown in the simplified dark-noise pulse-height-distribution spectrum ofFig. 122. Region A of Fig. 122 includescircuit-originated noise, some single-electronpulses, and pulses caused by electronsoriginating at the dynodes in the multipliersection. Pulses originating at the dynodes ex-hibit less gain than the single-electron pulsesfrom the cathode. Region B represents thesingle-electron pulse-height distribution andis the region in which the single-photonsignal pulses appear. Region C is caused bycosmic-ray muons, after-pulsing, and radio-active contaminants in the tube materialsand in the vicinity of the tube. To maximizethe ratio of signal pulses to noise pulses insingle-photon counting, lower- and upper-level discriminators should be located asshown in the figure. For a discussion ofsignal-to-noise ratio and the statistics relatedto pulse counting, see “Pulse Counting Sta-tistics” in Appendix G.

The signal-to-noise ratio may be improvedby the square root of the total count time.Increasing the time of count is analogous todecreasing the bandwidth in the charge-integration method. The signal-to-noise

110

ratio can also be improved by decreasing thenumber of dark-noise pulses. Because thesepulses originate at the photocathode surface,the number can be reduced by reducing thearea of the photocathode or by reducing theeffective photocathode area by using elec-tron optics to image only a small part of thephotocathode on the first dynode. It mayalso be desirable to cool the photomultiplierand thereby reduce the thermionic emissionfrom the photocathode.

In some applications not requiring the useof the full area of the semitransparentphotocathode, it may be useful to restrict theactive area of the photocathode by speciallyarranged magnetic fields104 and thus reducethe collection of thermally emitted electronsfrom non-utilized areas of the photocath-ode. Fig. 123 illustrates the design of sucha magnetic defocusing system type PF-1011 which includes an 8852 photomulti-plier (2-inch diameter, ERMA photocath-ode, and GaP first dynode). The basic ele-ment is a permanent magnet in the form ofan annular ring (1) polarized so that theS-pole or N-pole is toward the photocath-ode. Pole pieces, spacers, and a magnetic-shield cylinder complete the arrangement.The resulting magnetic field is such that onlyelectrons emitted from the central portion ofthe photocathode arrive at the first dynode.Typical variation of photocathode responseas a function of the position of the incidentlight is shown in Fig. 124. Note that the ef-fective diameter of the photocathode hasbeen reduced to about one eighth of an inch.Total dark emission at room temperaturefrom the magnetically controlled photo-cathode is reduced by approximately 80:1. Ifthe tube is then cooled to - 20°C, the darkcount is still further reduced to about twoelectrons per second.

Very-Low-Light-Level Photon-CountingTechnique. Before beginning very-low-light-level photon counting, the following specialprecautions must be taken:

1. The power supply and interconnectingcircuits must have low-noise characteristics.

2. The optical system must be carefullydesigned to minimize photon loss and to pre-vent movement of the image of the object onthe photocathode, a possible cause of errorif the cathode is non-uniform. It is generallygood practice to defocus the image on the


PHOTOMULTIPLIER -ADJUSTABLE

- PULSE AMPLITUDEDISCRIMINATOR

92cs-32458

Fig. 121- Block diagram illustrating a digital system for detection of low light levels.

photocathode, especially in the case of apoint source, to minimize problems whichmay result from a non-uniform photocath-ode.

3. The photomultiplier must be allowedto stabilize before photon counting is begun.The tube should not be exposed toultraviolet radiation before use and should,if possible, be operated for 24 hours at thedesired voltage before the data are taken.

4. The photomultiplier should beoperated with the cathode at ground poten-tial if possible. If the tube is operated withthe photocathode at negative high voltage,care must be taken to prevent the glassenvelope of the tube from coming into con-tact with conductors at ground potential ornoisy insulators such as bakelite or felt.Without this precaution, a very high darknoise may result as well as permanentdamage to the photocathode.

5. Large thermal gradients must not bepermitted across the tube as it is cooling. Inaddition, care must be taken to avoid ex-cessive condensation across the leads of thetube or on the faceplate.

PULSE HEIGHT 92cs-32459

Fig. 122 - Dark-noise pulse-height-distribu-tion spectrum.

92CS-32460

Fig. 123 - Magnetic defocusing system(PF1011) which includes type 8852 photomul-tiplier. This system permits collection of elec-trons only from the central part of the pho-tocathode and thus substantially reduces darkemission. Shown are (7) ceramic magnet, (2)spacer-insulator, (3) magnet pole piece o fblack iron, and (4) magnetic shield cylinder.The assembly is designed to fit over the 8852envelope.

Photomultiplier Selection for PhotonCounting. In the selection of a photomulti-plier for use in photon counting, several im-portant parameters must be considered.First, and most important, the quantum effi-ciency of the photocathode should be as highas possible at the desired wavelength. Tominimize the thermionic dark-noise emis-sion, the photocathode area should be nolarger than necessary for signal collection;the multiplier structure should utilize aslarge a fraction as possible of the electronsfrom the photocathode. The over-all tubeshould have as low a dark noise as possible.In some applications, the rise time and time-resolution capabilities of the tube may alsobe important.

111

PhotomultIplier Handbook

DISTANCE FROM CENTER OF PHOTOCATHODE - INCHES92CS-32461

Fig. 124 - Typical variation of photocathode sensitivity as a function of incident-lightspot position for an 8852 photomultiplier with the magnetic defocusing system ofFig. 123 affixed to the tube.

A number of more recent developments inphotomultiplier design are of considerablesignificance in photon counting. Thesedevelopments fall into two major categories,secondary-emission materials and photo-cathodes. Because of the superior statisticsof gallium phosphide (one of the morerecently developed secondary-emissionmaterials discussed in the section “Secon-dary Emission” in Chapter 2 Photomulti-plier Design), a tight single-electron distribu-tion curve can be obtained, as shown in Fig.79. The tight distribution permits easy loca-tion of the pulse-height discriminator, as isparticularly evident when some of the signalpulses are originated by two or more photo-electrons leaving the cathode simultane-ously. If the average number of photoelec-trons leaving the photocathode per pulsewere three, a pulse-height distributionsimilar to that shown in Fig. 125 would beobtained. Gallium phosphide provides ahigher signal-to-noise ratio than would con-ventional secondary-emitting materials suchas Be0 or Cs3Sb when used in the same tube.

Photocathode developments includeERMA (Extended Red Multi-Alkali), asemitransparent photocathode having aresponse to 940 nanometers, and severalnegative-electron-affinity materials. Perhapsthe most significant of these materials isindium gallium arsenide, whose threshold

wavelength increases with increasing indiumcontent. Spectral-response curves for someof the more recent photocathodes are shownin the section “Photoemission” in Chapter 2Photomultiplier Design.

A

92CS-32462

Fig. 125 - Pulse-height distribution obtainedwith a photomultiplier having a galliumphosphide first dynode. Light level is suchthat three photoelectrons per pulse time isthe most probable number.

Astronomy105. Astronomers were amongthe first users of photomultiplier tubes. Thehigh quantum efficiency, low backgroundnoise, and high gain recommend the use ofphotomultipliers in various astronomical ap-plications. Requirements are often similar to

112

those for low-level photometry or spec-trometry as discussed above. Applicationsinclude the guidance of telescopes, intensitymeasurements, stellar spectrophotometry,Doppler measurement of radial velocities,and the measurement of stellar magneticfields by measuring Zeeman displacements.

Side-on photomultipliers such as the 1P21are frequently used in these applications.Refrigeration may be utilized to reducebackground. Photon counting may beadopted for very low signal measurement.For applications requiring a wide spectralrange, tubes with the GaAs:Cs photocathodeare very useful.Pulsed Photomultipliers

Some years ago, Post106 demonstratedimproved characteristics of photomultipliertubes by operating them with a pulsedvoltage supply. He used side-on photomulti-pliers types 931A and 1P21 with an appliedvoltage of between 4 and 5 kilovolts pulsedfor 2.5 microseconds or less. At thesevoltages, 400 to 500 volts per stage, thesecondary emission for Cs-Sb dynodes is at amaximum (see Fig. 19) so that the gainbecomes relatively insensitive to the varia-tion of the maximum voltage applied. Nor-mally, these tubes could not be operated atsuch high voltages without incurring destruc-


tive run-away dark currents. But, by pulsingthe applied voltage, regenerative effects thatdepend upon transit time of ions areeliminated. Post’s circuit is shown in Fig.126. Not all tubes he tried could tolerate thishigh voltage without field emission andbreak-down effects, but on those tubeswhich were satisfactory initially or which hewas able to stabilize by application ofrepeated high-voltage pulses, he found somevery interesting advantages. Peak pulseamplitudes of 0.7 ampere were possible andprovided a 70-volt pulse to drive anoscilloscope directly. Thus, a singlephotoelectron resulted in a 10-to-15-volt out-put pulse; the tubes were operated at gains inthe neighborhood of 109 The measured risetime he found to be slightly more than onenanosecond. Refer to Fig. 71. Note in Fig.126 the termination of the coaxial outputline at the photomultiplier end which re-duced the output RC time constant. It ispointed out by Post “the loss of a factor oftwo in voltage by terminating at the multi-plier is compensated by voltage doubling atthe unterminated end.”Laser Range Finding

A simplified block diagram of a laserrange finder is shown in Fig. 127. Such adevice may be utilized by the military for

- 4 kVPULSE

Fig. 126 - Post’s voltage divider and bypass capacitor connections of photomulti-plier to output coaxial line for pulsed high-voltage operation.105

113


92CS - 32464

Fig. 127 - Simplified block diagram of a laser range finder.

tank fire control or in various industrialsurveying type applications. Typically, alaser is pulsed in a time range of 25nanoseconds. Measurements are made ofbeam travel time with a photomultiplier pick

may be used with a 3/4-inch diameter photo-multiplier having an S-20 spectral response.

and solid-state junction lasers such as

of the longer wavelengths, the detectors aremore often silicon avalanche diodes. An in-terference filter is used to pass thewavelength of radiant energy of the laserwith a minimum of background radiation.Photomultipliers used in laser range findingmay have a relatively small photocathode,but they must exhibit high quantum efficien-cy, low dark noise, and fast rise time or anequivalent large bandwidth. Most photomul-tipliers can provide bandwidths exceeding100 MHz and at the same time maintainrelatively large output signals. The band-width of a photomultiplier can be limited bythe RC time constant of the anode circuit.

The range of a laser-range-finding systemdepends on system parameters and opera-tional environment. The maximum range ofa given system may be signal-photon limitedor background limited. The photon-limitedcase exists when the background and de-tector noise can be considered negligible.The maximum range in this case is deter-mined by the signal-to-noise ratio in thephotoelectron pulse corresponding to the

114

scattered laser return beam. The signal-to-noise ratio of such a pulse is proportional tothe square root of the product of the numberof incident photons on the photomultiplierand the quantum efficiency of the photoelec-tric conversion. If the atmospheric attenua-tion is neglected and it is assumed that thelaser spot falls entirely within the target, themaximum range in this case would be in-creased as the square root of the quantum ef-ficiency of the photocathode. This increasefollows because the number of photons col-lected by the aperture of the receiving systemvaries inversely as the square of the distancefrom the target.

In cases where the laser pulse must bedetected against the background of adaylight scene, it is said to be backgroundlimited. If atmospheric attenuation again isneglected, the signal is proportional to thenumber of incident photons times the quan-tum efficiency of the photocathode. Thenumber of incident photons from the returnbeam is inversely proportional to the squareof the range, again assuming that the targetis larger than the laser spot. The noise,however, is independent of the range and isdetermined by the square root of the productof the incident background radiation and thequantum efficiency. Thus, to a first approx-imation the range is increased in thebackground-limited case only by the fourthroot of the quantum efficiency. Because thephotomultiplier current caused by thebackground radiation is proportional to thesolid angle of the scene from which the

photomultiplier collects radiation, thebackground current in the photomultipliermay be minimized by the use of a photocath-ode having a small area or by the use of alimiting aperture on the faceplate of thephotomultiplier. No loss in collected laserlight need result because the return beammay generally be considered as originatingfrom a point source. The system aperture, ofcourse, must be large enough to avoid op-tical alignment problems.Scanning Applications

A number of photoelectrically sensedscanning systems have been devised such asfacsimile scanners in which the material tobe transmitted-a photograph or printedmessage-is mounted on a drum that isrotated to provide scan in one dimension. Ascanning head moves parallel to the drumaxis to provide the other dimension. The pic-ture element is illuminated with a focusedlight spot and the scattered light is picked upby a photomultiplier, providing high-speedtransmission utilizing the modulated outputelectrical signal.

Two more recently developed scanningsystems are the flying-spot scanner for thedevelopment of television signals and thesupermarket checkout system which recog-nizes a coded symbol on each product.

Flying-Spot Scanning. The elements of aflying-spot scanning system are shown inFig. 128. A cathode-ray tube, in conjunctionwith its power supplies and deflection cir-cuits, provides a small rapidly moving lightsource which forms a raster on its face. Thisraster is focused by the objective lens in theoptical system onto the object beingscanned, a slide transparency or a motion-picture film. The amount of light passingthrough the film varies with the film density.This modulated light signal is focused uponthe photomultiplier by means of thecondensing-lens system. The photomultiplierconverts the radiant-energy signal into anelectrical video signal. The amplifier and itsassociated equalization circuits increase theamplitude of the video signal as required.

The flying-spot scanning system is capableof providing high-resolution monochromaticperformance. With the addition of (a) ap-propriate dichroic mirrors which selectivelyreflect and transmit the red, blue, and greenwavelengths, (b) two additional photomulti-


pliers with video amplifiers, and (c) ap-propriate filters, color operation is possible.In color operation, the primary wavelengthsare filtered after separation by the light-absorbing filters before being focused uponeach of three photomultipliers, one for eachcolor channel. The output of each photo-multiplier is then fed to a separate videoamplifier.

Flying-spot video-signal generators areused in the television industry primarily forviewing slides, test patterns, motion-picturefilm, and other fixed images. Systems havebeen developed for the home-entertainmentindustry that allow slides and motion-picturefilm to be shown on the picture tube of anytype of commercial television receiver.

A similar system is utilized by thephotographic industry to provide accurateexposure control for film copying. By meansof the three color controls, the operator mayproduce a color television display represent-ing the film and can adjust and measure eachcolor component to produce a visually satis-fying balance.

Several important considerations must betaken into account if the cathode-ray tube inthe system is to produce a light spot capableof providing good resolution. The cathode-ray tube should be operated with as small alight spot as possible. The cathode-ray-tubefaceplate should be as blemish-free aspossible and the tube should employ a fine-grain phosphor. Blemishes adversely affectsignal-to-noise performance and contributeto a loss of resolution.

The spectral output of the cathode-ray-tube phosphor should match the spectralcharacteristic of the photomultiplier. Thismatch can be rather loose in a monochro-matic flying-spot generator. The spectraloutput of the phosphor of the cathode-raytube used in a three-color version, however,must include most of the visible spectrum.Phosphors used in monochromatic systemsmay provide outputs in the ultraviolet regionof the spectrum and still perform satisfac-torily. Phosphors such as P16 and P15,when used with an appropriate ultravioletfilter, display the necessary short persistencerequired in a monochromatic system.

The visible portion of the P15 or P24phosphor is used in color systems. The P15and P24 phosphors are, however, much

115

VERTICALSCANNING

HORIZONTALSCANNING

92CM -32465

Fig. 128 - Elements of a flying-spot scanning system.

slower than the P16 phosphor and cause alag in buildup and decay of output from thescreen.

The phosphor lag results in trailing, a con-dition in which the persistence of energyoutput from the cathode-ray tube causes acontinued and spurious input to the photo-multiplier as the flying spot moves across thepicture being scanned. The result is that alight area may trail into the dark area in thereproduced picture.

Similarly, the lag in buildup of screen out-put causes a dark area to trail over into thelight area. The result of these effects on thereproduced picture is an appearance similarto that produced by a video signal deficientin high frequencies. Consequently, high-frequency equalization is necessary in thevideo amplifier.

The objective lens used in a flying-spotgenerator should be of a high-qualityenlarger type designed for low magnificationand, depending upon the cathode-ray-tubelight output, should be corrected forultraviolet radiation. The diameter of theobjective lens should be adequate to coverthe slide to be scanned. An enlarging f/4.5lens with a focal length of 100 millimeters issuitable for use with 35-millimeter slides.The optics should not image the film on thephotomultiplier because shading effectswould result from non-uniformities of thephotomultiplier response. In some cases itmay be useful to employ a beam splitter

116

between the cathode-ray tube and the lensand to use a second photomultiplier to senseany non-uniformities in the cathode-ray tuberaster display. This sensed signal may thenbe used to eliminate this source of signaldistortion in the primary photomultiplierpickup.

The spectral characteristics of thephotomultiplier (or photomultipliers in thecase of the three-color system) and thecathode-ray-tube phosphor should match.Usually, a photomultiplier having an S-4 orS-l1 (Cs3Sb) spectral response is suitable foruse in a monochromatic system or as thedetector for the blue and green channels. AnS-20 (Na2KSb:Cs) response is very oftenutilized for the red channel. The bialkalicathode (K-CS-Sb) is also well suited for theblue channel. The speed of the detector mustbe sufficient to provide the desired videobandwidth. Most requirements do not ex-ceed 6 to 8 MHz, a figure well within photo-multiplier capabilities.

The anode dark current of the photomulti-plier should be small compared to the usefulsignal current. The signal-to-noise ratio willbe maximized by operation of the photomul-tiplier at the highest light levels possible. Ifnecessary, the over-all photomultiplier gainshould be reduced to prevent excessive anodecurrent and fatigue.

The amplitude of the light input is usuallya compromise between an optimum signal-to-noise ratio and maximum cathode-ray-


tube life. Tube life may be reduced becauseof loss of phosphor efficiency at high beam-current levels. The signal-to-noise ratio canalso be improved by the selection of photo-multipliers having the highest photocathodesensitivities possible. However, because thespread of photocathode sensitivities isseldom greater than two or three to one, theimprovement afforded by such selection islimited to two or three dB, an improvementdifficult to detect during observations of atelevision display but desirable and necessaryin some critical applications.

Photomultiplier gain need only be suffi-cient to provide a signal of the required levelto the succeeding video-amplifier stages.These stages, in addition to providing thenecessary amplification and bandwidth toassure good picture quality, incorporateequalization circuits composed of networkshaving different time constants. Therelatively long decay time of these circuitsgenerally results in appreciable reduction ofthe useful signal-to-noise ratio. Therefore,the use of short-persistence phosphors isrecommended to reduce the required amountof equalization.

In addition to the video amplifier, a

gamma-correction amplifier is required ineach channel. The gammacorrection ampli-fier assures maximum color fidelity by mak-ing the linearity or gamma of the system uni-ty.

Supermarket Checkout Systems. Fig. 129shows the elements of a supermarket check-out system. The Universal Product Code(UPC), which is to be marked on all prod-ucts, is scanned by laser beam by means ofan oscillating mirror and a rotating mirrorsystem. The pattern of the scan is a networkof overlapping sine waves. The reflected pat-tern contains the modulation of the UPCsymbol and is converted by the photomulti-plier to an electronic signal that is thenanalyzed for the content of the code.

The laser used in the supermarketcheckout system is usually a low-power He-Ne type with the principal emission line at633 nanometers. A suitable photomultiplierfor this application is a two-inch end-on typehaving an S-20 (Na2KSb:Cs) spectralresponse or an extended-red multialkalitype. Stability and good signal-to-noise ratioare important characteristics of the photo-multiplier tube.

ROTATING MIRRORPOLYGON

92CM - 32466

Fig. 129 - Point-of-sale supermarket checkout system. At left is Universal ProductCode (UPC) symbol that provides 10-digit product identification. At right is schematic of optical system for scanning the product symbol and detecting the modula-tion with a photomultiplier.

117


REFERENCES91. R.W. Engstrom and E. Fischer, “Ef-

fects of voltage-divider characteristics onmultiplier phototube response,” Rev. Sci.Instr., Vol. 28, pp 525-527, July, 1957.

92. C.R. Kerns, “A high-rate phototubebase,” IEEE Trans. Nucl. Sci., Vol. NS-24,No. 1, pp 353-355, Feb. 1977.92a. V.O. Altemose, “Helium diffusion

through glass,” J. Appl. Phys., Vol. 32, No.7, pp 1309-1316, 1961.92b. F.J. Norton, “Permeation of gases

through solids,” J. Appl. Phys., Vol. 28,No. 1, pp 34-39, 1957.93. American Institute of Physics Hand-

book, Third Edition, McGraw-Hill, 1972.94. R.W. Engstrom, “Luminous Micro-

flux standard,” Rev. Sci. Instrum. Vol. 26,No. 6, pp 622-623, June, 1955.95. D.G. Fisher, A.F. McDonie, and A.H.

Sommer, “Bandbending effects in Na2K Sband K2Cs Sb photocathodes,” J. Appl.Phys., Vol. 45, No. 1, pp 487-8, Jan. 1974.96. H.O. Anger, “Scintillation camera,”

Rev. Sci. Instrum., Vol. 29, pp 27-33, 1958.97. M.J. Weber and R.R. Monchamp,

“Luminescence of Bi4Ge3O12: spectral anddecay properties,” J. Appl. Phys., Vol. 44,No. 12, pp 5495-99, 1973.98. G.L. Brownell and C.A. Burnham,

“Recent developments in positron scin-tigraphy,” Instrumentation in NuclearMedicine, Vol. 2, Ed: G.J. Hine and J.A.Sorenson, Chapter 4, pp 135-159, AcademicPress, 1974.99. M.E. Phelps, E. J. Hoffman, N.A.

Mullani and M.M. Ter-Pogossian, “Ap-

plication of annihilation coincidence detec-tion to transaxial reconstruction tomogra-phy,” J. Nucl. Medicine, Vol. 16, No. 3, pp210-224, 1975.100. S.E. Derenzo, H. Zaklad, and T.F.Budinger , “Analytical study of a high-resolution positron ring detector system fortransaxial reconstruction tomography,” J.Nucl. Med., Vol. 16, No. 12, pp 1166-73,Dec. 1975.101. M.E. Phelps, E.J. Hoffman, Sung-Cheng Huang, and D.E. Kuhl, “ECAT: acomputerized tomographic imaging systemfor positron-emitting radiopharmaceuti-cals,” J. Nucl. Med., Vol. 19, No. 6, pp635-647, June 1978.102. G.L. Brownell, C. Burnham, B.Ahluwalia , N. Alpert, D. Chester, S.Cahavi, J. Correia, L. Deveau, “Positronimaging instrumentation,” IEEE Trans.Nucl. Sci., Vol. 24, No. 2, pp 914-916, Apr.1977.103. Operating and Service Manual, Model111 Fluorometer,” G.K. Turner Associates,2524 Pulgas Ave., Palo Alto, Calif. 94303.104. G.Y. Farkas and P. Varga, “Reduc-tion of dark current in transparent cathodephotomultipliers for use in optical measure-ments,” J. Sci. Instrum., Vol. 41, pp.704-705, 1964.105. Astronomical Techniques, Edited byW.A. Hiltner, The University of ChicagoPress, 1962.106. R.F. Post, “Performance of PulsedPhotomultipliers,” Nucleonics, Vol. 10, No.5, pp 46-50, May 1952.

118

Typical Photomultiplier Applications and Selection Guide

Appendix A-Typical Photomultiplier Applications and Selection Guide

The many and varied requirements ofequipment designers and experimenterspreclude the recommendation of a singlephotomultiplier as the optimum device forany given application category. In most ap-plications, some trade-offs must be made inelectrical characteristics; tube size must beconsidered; the environment in which thedevice is to be operated can be an influentialfactor; and of course, over-all cost is impor-tant. Each of these constraints can be bestevaluated by the individual designer for thespecific application.

This Appendix defines a number of themore common applications of photomulti-pliers and lists tube types which are suitableor are frequently used for the particular ap-plication. The listing is not all-inclusive andis intended to serve only as a general guidefor initial type selection. Other photomulti-pliers may be satisfactory for the specifiedapplications when all system requirementsare considered.

CATALOGUE OF PHOTOMULTIPLIERAPPLICATION CATEGORIES

Astronomy: The guidance of telescopes, in-tensity measurements, stellar spectropho-tometry, and the like.Colorimetry: The quantitative color com-parison of surfaces (reflectance) and solu-tions (transmission).C-T Scanners: A medical X-ray equipmentthat provides a cross section density map(tomograph) of a patient. A photomultiplieris used to detect and measure the light fluxfrom a scintillating crystal.Densitometry: The measurement of opticaldensity of photographic negatives, neutraldensity filters, and similar materials.Gamma-Ray Cameras: A scintillationcounter having a single large crystal and anumber of photomultiplier tubes used in

medical applications to map the location ofisotope disintegrations.High-Temperature Environments: Applica-tions such as the logging of deep oil wells, orgeological exploration, and steel-mill processcontrols.Imaging Devices: A cathode-ray tube ormoving mirrors can be used as a light sourceto sequentially illuminate a film positive ornegative or a printed page. This system isused in (1) optical character recognition, (2)scanning or printed or written material fortransmission by telephone, (3) parts inspec-tion, and (4) reproduction of motion pic-tures, slides, and educational material on atelevision receiver (color or black and white).Inspection, High-Speed: Small objects suchas fruits, vegetables, seeds, candy, toys,paper products and even glass, metal, andother industrial parts can be examined forcolor and defects as they move at high speedpast one or more photomultiplier tubes.Laser Detection: Lasers provide uniquelight sources; they are spectrally pure andproduce very narrow collimated beams.They can be very intense and can be made toproduce light pulses of extremely short dura-tion. The photomultiplier provides timeresolution in the nanosecond and subnano-second ranges and is capable of detectingvery low light levels such as those receivedfrom weak reflected laser light pulses.Photometry: The measurement of illumina-tion or luminance. Levels of light flux varyover a wide range in photography,astronomy, television, and other applica-tions.Photon Counting: A method of detectingphotons by counting single photoelectronsreleased from the photocathode.Pollution Monitoring: The analysis of thelevel and the nature of contaminants in solu-tions, gases, and other waste materials.

119


120

Positron Camera: A scintillation-counter-type device providing tomographic presenta-tions based on coincident gamma-ray emis-sion accompanying annihilation of apositron and an electron . Medical applica-tions utilize tracer radio nuclides such as11C, 13N, and 15O.Process Control: The measurement oftransmitted or reflected light in continuousflow processes using solids, liquids, or gases.Detects flaws, improper marking, andchanges in color and optical density. Byusing radioactive sources and scintillators,photomultipliers can be used for the controlof the weight and thickness of opaquematerials.Radioimmunoassay (RIA): A techniquethat enables the measurement of minutequantities of substances in biological fluidsas small as 10-12 gram. Compounds aretagged with radioactive isotopes and aremeasured with a liquid scintillation counterfor beta-emitting isotopes or a solid crystalscintillation counter for gamma-emittingisotopes.Radiometry: The measurement of irra-diance or radiance.Raman Spectrometry: Measurement of thewavelength shift of scattered photons from ahighly monochromatic source such as a laserprovides information on molecular structureand bonding energy.Scintillation Counting: The measurement ofnuclear radiation by detecting light or singlescintillations emitted from a scintillationmaterial receiving nuclear radiation.Severe Physical Environments: Applica-tions such as oil-well logging, satellites, andmilitary vehicles subject to severe shock andvibration.Thermoluminescent Dosimetry (TLD): Certain phosphors emit light when they areheated after having been exposed to ionizingradiation. Devices using this principle affordpersonnel protection by determining dosagelevels in medical and biological treatmentsand studies. Energy stored in TLD’s is proportional to dosage over a very wide range.Time Measurement: In nuclear experimentsthe “time of flight” of nuclear particles isimportant. Photomultipliers permit timemeasurements down to a fraction of anonosecond.

Spectrophotometry

BURLEType

Photo-cathodeMaterial

WindowMaterial1

No.ofStages

Diameter2

1P21 Cs3Sb 0080 9 1-1/8"s1P28 Cs3Sb 8337 9 1-1/8"s1P28A Cs3Sb 8337 9 1-1/8"s1P28B K2CsSb 8337 9 1-1/8"s931A GaAs 0080 9 1-1/8"s931B GaAs 0080 9 1-1/8"s4526A Na2KCsSb 0080 10 1-1/2"dC31034 GaAs 8337 11 2"eC31034A GaAs 8337 11 2"eS83063E Na2KCsSb 7056 10 1-1/8"eS83068E Rb2CsSb 7056 10 1-1/8"e83089-600 Na2KCsSb 7056 10 1-1/8"e83101-600 Rb2CsSb B270 10 3/4"e1 0080, B270 – lime glass; 8337 - Schott, 7056 – borosilicate2 e – end-window; s – side-window; d – dormer window

TLD

BURLEType


WindowMaterial1

No.ofStages

Diameter2

8575 K2CsSb 7740 12 2"e8850 K2CsSb 7740 12 2"eS83062E Rb2CsSb 7056 10 1"e83087-100 Rb2CsSb B270 10 3/4"e83101-600 Rb2CsSb B270 10 3/4"e1 B270 – lime glass; 7740 – Pyrex ; 7056 - borosilicate2 e – end-window

Gamma -Ray Cameras

BURLEType


WindowMaterial1

No.ofStages

Diameter2

4900 K2CsSb B270 10 3" e6199 Cs3Sb 0080 10 1-1/2" eS83010E Rb2CsSb 0080 10 1-1/2" eS83013F K2CsSb 0080 10 3-1/2" eS83019F K2CsSb B270 10 2" eS83020F K2CsSb B270 10 60mm h,eS83021E K2CsSb B270 10 3" eS83022F K2CsSb B270 10 2" h,eS83025F K2CsSb B270 10 3" h,eS83049F K2CsSb B270 8 3" eS83053F K2CsSb B270 8 60mm h,eS83054F K2CsSb B270 8 2" eS83056F K2CsSb B270 8 3" h,eS83069E K2CsSb B270 8 35x46.5mm, mhS83079E K2CsSb B270 8 3" sq1 0080, B270 – lime glass2 e – end-window; h – hexagonal; mh – modified hexagonal; sq - square


121

High Temperature Environments

BURLEType


WindowMaterial1

No.ofStages

Diameter2

C31000AP Na2KSb 7056 12 2"eC31016G Na2KSb 7056 10 1"eC83051 Na2KSb sa 10 1"eC83060 Na2KSb sa 10 1-1/4"eC83065 Na2KSb 7056 10 1"e83103 100 Na2KSb 7056 10 3/4"e1 sa – sapphire, 7056 – borosilicate2 e – end-window

Flying-Spot Scanners

BURLEType


WindowMaterial1

No.ofStages

Diameter2

4552 K2CsSb 0080 9 1-1/8"s1P21 Cs3Sb 0080 9 1-1/8"s931A Cs3Sb 0080 9 1-1/8"s931B K2CsSb 0080 9 1-1/8"s6199 Cs3Sb 0080 10 1-1/2"e83087-100 Rb2CsSb B270 10 3/4"e83090-600 Rb2CsSb B270 9 3/4"e83101-600 Rb2CsSb B270 10 3/4"eS83010E Rb2CsSb 0080 10 1-1/2"e1 0080 – lime glass; 8337 - Schott2 e – end-window; s – side-window

Inspection, High Speed

BURLEType


WindowMaterial1

No.ofStages

Diameter2

1P21 Cs3Sb 0080 9 1-1/8" s4552 K2CsSb 0080 9 1-1/8" s6199 Cs3Sb 0080 10 1-1/2" e931A Cs3Sb 0080 9 1-1/8" s931B K2CsSb 0080 9 1-1/8" sS83062E Rb2CsSb 7056 10 1" eS83010E Rb2CsSb 0080 10 1-1/2" e83087-100 Rb2CsSb B270 10 3/4"e83090-600 Rb2CsSb B270 9 3/4" e83101-600 Rb2CsSb B270 10 3/4" e1 0080, B270 – lime glass, 7056 – borosilicate2 e – end-window; s – side-window

Raman Spectroscopy

BURLEType


WindowMaterial1

No.ofStages

Diameter2

8850 K2CsSb 7740 12 2”e8852 Na2KCsSb 7740 12 2”eC31034 GaAs 8337 11 2”eC31034A GaAs 8337 11 2”e1 8337 - Schott; 7740 – Pyrex2 e – end-window

Laser Detection

BURLEType


WindowMaterial1

No.ofStages

Diameter2

4526A Na2KCsSb 0080 10 1-1/2"d8575 K2CsSb 7740 12 2"e8850 K2CsSb 7740 12 2"e8852 Na2KCsSb 7740 12 2"eS83063E Na2KCsSb 7056 10 1-1/8"e83087-100 Rb2CsSb B270 9 3/4"e83101-600 Rb2CsSb B270 10 3/4"eC31034 GaAs 8337 11 2"eC31034A GaAs 8337 11 2"e1 0080, B270 – lime glass; 8337 - Schott; 7740 – Pyrex 7056 – borosilicate2 e – end-window; d – dormer window

Photometry

BURLEType


WindowMaterial1

No.ofStages

Diameter2

1P21 Cs3Sb 0080 9 1-1/8”s931A Cs3Sb 0080 9 1-1/8”s931B K2CsSb 0080 9 1-1/8”s4552 K2CsSb 0080 9 1-1/8”s1 0080 – lime glass2 s – side-window

Photon Counting

BURLEType


WindowMaterial1

No.ofStages

Diameter2

8575 K2CsSb 7740 12 2"e8850 K2CsSb 7740 12 2"e8852 Na2KCsSb 7740 12 2"e8854 K2CsSb 8337 14 5"eC31034 GaAs 8337 11 2"eC31034A GaAs 8337 11 2"eS83062E Rb2CsSb 7056 10 1-1/2"e83089-600 Na2KCsSb 7056 10 1-1/8"e83090-600 Rb2CsSb B270 9 3/4"e83101-600 Rb2CsSb B270 10 3/4"e1 0080, B270 – lime glass; 8337 - Schott; 7740 – Pyrex 7056 – borosilicate2 e – end-window

Positron Cameras

BURLEType


WindowMaterial1

No.ofStages

Diameter2

83087-100 Rb2CsSb B270 10 3/4"e83090-600 Rb2CsSb B270 9 3/4"e83101-600 Rb2CsSb B270 10 3/4"e83102 100 Rb2CsSb 0080 9 1"eS83062E Rb2CsSb 7056 10 1"e1 0080, B270 – lime glass, 7056 – borosilicate2 e – end-window


122

Process Control

BURLEType


WindowMaterial1

No.ofStages

Diameter2

1P21 Cs3Sb 0080 9 1-1/8"s1P28 Cs3Sb 8337 9 1-1/8"s1P28A Cs3Sb 8337 9 1-1/8"s1P28B K2CsSb 8337 9 1-1/8"s931A Cs3Sb 0080 9 1-1/8"s931B K2CsSb 0080 9 1-1/8"s2060 Cs3Sb 0080 10 1-1/2"e4856 K2CsSb 0080 10 2"e4900 K2CsSb 0080 10 3"e6199 Cs3Sb 0080 10 1-1/2"eS83010E Rb2CsSb 0080 10 1-1/2"eS83019F K2CsSb B270 10 2"eS83021E K2CsSb B270 10 3"eS83049F K2CsSb B270 8 3"eS83054F K2CsSb B270 8 2"eS83079E K2CsSb B270 8 3"sq,e1 0080, B270 – lime glass; 8337 - Schott2 e – end-window; s – side-window

Severe Physical Environments

BURLEType


WindowMaterial1

No.ofStages

Diameter2

83101 100 Rb2CsSb B270 10 3/4"eC31016G Na2KSb 0080 10 1"eC83051 Na2KSb sa 10 1"eC83060 Na2KSb sa 10 1-1/4"e1 0080, B270 – lime glass; sa - sapphire2 e – end-window

Radiometry

BURLEType


WindowMaterial1

No.ofStages

Diameter2

1P21 Cs3Sb 0080 9 1-1/8"s1P28 Cs3Sb 8337 9 1-1/8"s1P28A Cs3Sb 8337 9 1-1/8"s1P28B K2CsSb 8337 9 1-1/8"s931A Cs3Sb 0080 9 1-1/8"s931B K2CsSb 0080 9 1-1/8"s4526A Na2KCsSb 0080 9 1-1.2"d2060 Cs3Sb 0080 10 1-1/2"e6199 Cs3Sb 0080 10 1-1/2"e8575 K2CsSb 7740 12 2"e8850 K2CsSb 7740 12 2"e8852 Na2KCsSb 7740 12 2"eS83010E Rb2CsSb 0080 10 1-1/2"eS83062E Rb2CsSb 7056 10 1"e83087-100 Rb2CsSb B270 10 3/4"e83101-600 Rb2CsSb B270 10 3/4"e1 0080, B270 – lime glass; 8337 - Schott; 7740 - Pyrex 7056 – borosilicate2 e – end-window; s – side-window; d – dormer window

Time Measurement

BURLEType


WindowMaterial1

No.ofStages

Diameter2

8575 K2CsSb 7740 12 2"e8850 K2CsSb 7740 12 2"e8852 Na2KSb 7740 12 2"e8854 K2CsSb 8337 14 5"eS83062E Rb2CsSb 7056 10 1"eS83068E Rb2CsSb 7056 10 1-1/8"e83087-100 Rb2CsSb B270 10 3/4"e83101-600 Rb2CsSb B270 10 3/4"e1 0080, B270 – lime glass; 8337 - Schott; 7740 – Pyrex, 7056 – borosilicate2 e – end-window

Scintillation Counting

BURLEType


WindowMaterial1

No.ofStages

Diameter2

2060 Cs3Sb 0080 10 1-1/2"e4900 K2CsSb 0080 10 3"e6199 Cs3Sb 0080 10 1-1/2"e83103 100 Na2KSb 7056 10 3/4"e8575 K2CsSb 7740 12 2"e8850 K2CsSb 7740 12 2"e8852 Na2KCsSb 7740 12 2"e8854 K2CsSb 8337 14 5"eC31000AP Na2KSb 7056 12 2"eC31016G Na2KSb 0080 10 1"eC31016H Na2KSb 0080 10 1"eC31034 GaAs 8337 11 2"eC31034A GaAs 8337 11 2"eS83006F K2CsSb 0080 10 5"eS83062E Rb2CsSb 7056 10 1"eS83010E Rb2CsSb 0080 10 1-1/2"e83087-100 Rb2CsSb B270 10 3/4"e83092-500 Na2KSb 0080 10 1"e83101-600 Rb2CsSb B270 10 3/4"e1 0080, B270 – lime glass; 8337 - Schott; 7740 – Pyrex 7056 – borosilicate2 e – end-window

Radioimmunoassay

BURLEType


WindowMaterial1

No.ofStages

Diameter2

4856 K2CsSb 0080 10 2"e4900 K2CsSb 0080 10 3"e6199 Cs3Sb 0080 10 1-1/2"eS83068E Rb2CsSb 7056 10 1-1/8"eS83010E Rb2CsSb 0080 10 1-1/2"eS83019F K2CsSb B270 10 2"eS83054F K2CsSb B270 8 2"e1 0080, B270 – lime glass; 7056 – borosilicate

2 e – end-window


123

Densitometry

BURLEType


WindowMaterial1

No.ofStages

Diameter2

1P21 Cs3Sb 0080 9 1-1/8"s1P28 Cs3Sb 8337 9 1-1/8"s931A GaAs 0080 9 1-1/8"s931B GaAs 0080 9 1-1/8"s4552 K2CsSb 0080 9 1-1/8"s1 0080 – lime glass; 8337 - Schott2 s – side-window

Colorimetry

BURLEType


WindowMaterial1

No.ofStages

Diameter2

1P21 Cs3Sb 0080 9 1-1/8"s1P28 Cs3Sb 8337 9 1-1/8"s931A GaAs 0080 9 1-1/8"s931B GaAs 0080 9 1-1/8"s83089-600 Na2KCsSb 7056 10 1-1/8”e4552 K2CsSb 0080 9 1-1/8"s1 0080 – lime glass; 8337 – Schott; 7056 – borosilicate2 s – side-window; e – end-window

Astronomy

BURLEType


WindowMaterial1

No.ofStages

Diameter2

1P21 Cs3Sb 0080 9 1-1/8"s1P28 Cs3Sb 8337 9 1-1/8"s1P28B K2CsSb 8337 9 1-1/8"s8575 K2CsSb 7740 12 2"e8850 K2CsSb 7740 12 2"e8852 Na2KCsSb 7740 12 2"eC31034 GaAs 8337 11 2"eC31034A GaAs 8337 11 2"e1 0080 – lime glass; 8337 - Schott; 7740 - Pyrex2 e – end-window; s – side-window

Pollution Monitoring

BURLEType


WindowMaterial1

No.ofStages

Diameter2

4526A Na2KCsSb 0080 9 1-1/2"dS83063E Rb2CsSb 7056 10 1-1/8"e8850 K2CsSb 7740 12 2"e8852 Na2KCsSb 7740 12 2"e83089-600 Na2KCsSb 7056 10 1-1/8"e1 0080 – lime glass; 7740 – Pyrex, 7056 – borosilicate2 e – end-window; d – dormer window

Cerenkov Radiation

BURLEType


WindowMaterial1

No.ofStages

Diameter2

8850 K2CsSb 7740 12 2"e8854 K2CsSb 8337 14 5"eS83062E Rb2CsSb 7056 10 1"e83090-600 Rb2CsSb B270 9 3/4"e83101-600 Rb2CsSb B270 10 3/4"e1 0080, B270 – lime glass; 7740 – Pyrex, 8337 - Schott 7056 – borosilicate2 e – end-window


124

This page is blank.

Glossary of Terms

Glossary of Termsand

Appendix B-Related to Photomultiplier TubesTheir Applications

This Appendix contains a glossary ofterms frequently used in connection withphotomultiplier tubes and their applications.Where suitable definitions were availablefrom established standards or other reliablesources, they were used, possibly with minormodifications. For these cases, the sourcesof the definitions are indicated by thereference number at the end of the defini-tion. In many cases, however, the definitionswere prepared by the writers of this Manual.In these latter cases, no source reference isindicated after the definition.Absorptance: The ratio of the radiant fluxabsorbed in a body of material to that inci-dent upon it. If the absorptance is a, the

Acceptor: An impurity element in a p-typesemiconductor that may become ionized bytaking an electron from the valence bandand induce conduction by holes. For exam-ple, boron with a valence of three is a possi-ble acceptor impurity in silicon (valence,four).Afterpulse: A spurious pulse induced in aphotomultiplier by a previous pulse. (112)Angle of incidence: The angle between a rayof light striking a surface and the normal tothat surface.Angstrom unit, Å : 1 0-10 meter, or 0.1nanometer. In the International System ofUnits (SI units), the nanometer or microm-eter is the preferred unit for use in specifica-tion of wavelengths of light.Anode: An electrode through which a prin-cipal stream of electrons leaves the inter-electrode space. In a photomultiplier, the

anode is operated at a voltage positive withrespect to that of the last dynode and collectsthe secondary electrons emitted from the lastdynode. (112)Anticoincidence circuit: A circuit that pro-duces a specified output signal when one(frequently predesignated) of two inputsreceives a signal and the other receives nosignal within an assigned time interval. (112)Background counts (in radiation counters):Counts caused by ionizing radiation comingfrom sources other than that to be measured.(112)Bandwidth: In electrical measurements, thedifference between limiting frequencies in afrequency band, expressed in hertz (cyclesper second). The noise equivalent bandwidthis the bandwidth of a rectangular equivalentspectrum and may be defined as

circuit, A is the maximum absolute value of

angular frequency. (110)Bleeder: Resistive voltage divider.Cage: Part of a photomultiplier includingthe dynodes, focusing structure, anode, andsupport members.Candela, cd: The SI unit of luminous in-tensity .Cathode: See Photocathode.CAT-Scanner or CT-Scanner: Computer-ized Axial Tomography - scanner; a medicalX-ray equipment which provides a cross-sectional density map of a patient. In atypical device, an X-ray fan beam incident

125


on and rotating around the patient isdetected by a large array of scintillationcrystals and photomultiplier tubes.Cerenkov radiation: Radiation generated bya high-energy charged particle movingthrough a dielectric with a velocity greaterthan the velocity of light in the dielectric(i.e., greater than c/n where n is the index ofrefraction of the dielectric).Channel, channel number: In scintillationcounting, a number proportional to pulseheight specifying a generally narrow range ofpulse heights.Channel multiplier: A tubular electron-multiplier having a continuous interior sur-face of secondary-electron emissive material.(107)Coincidence circuit: A circuit that produces

a specified output signal when and onlywhen a specified number (two or more) or aspecified combination of input terminalsreceives signals within an assigned time inter-val. (112)-- --Collection efficiency: The fraction of elec-trons emitted by the photocathode of aphotomultiplier that lands on the firstdynode. Or more generally, the fraction ofelectrons emitted by one electrode that landson the next electrode (dynode or anode).Color temperature: The temperature of ablack body radiator such that its chromati-city is the same as that of the light under con-sideration. (109)Conduction band: A partially filled energyband in which the electrons can move freely,allowing the material to carry an electric cur-rent. The term is usually restricted tosemiconductors and insulators, where theconduction band is normally empty and isseparated by an energy-gap from the fullbands below it. (109)Count (in radiation counters): A singleresponse of the counting system. (112)Counting efficiency (scintillation counters):The ratio of (1) the average number ofphotons or particles of ionizing radiationthat produce counts to (2) the averagenumber incident on the sensitive area. (112)Crosstalk: As applied to photomultipliersused in liquid scintillation counting, lightoriginating internally in one photomultiplierand transmitted to another, causing coinci-dent background pulses.

Crystal: See Scintillator.Curie, Ci: A unit of radioactivity defined asthe quantity of an radioactive nuclide inwhich the number of disintegrations persecond is 3.7 x 1010. (B4)Current amplification (photomultipliers):The ratio of (1) the output signal current to(2) the photoelectric signal current from thephotocathode. (112)

Dark current: That current flowing in thecathode circuit (cathode dark current) or inthe anode circuit (anode dark current) in theabsence of light or radiation in the spectrumto which the photomultiplier is sensitive.

Delay line: A transmission line for intro-ducing signal time delay. (112)

Delta function (Dirac Delta Function): A

1, when the integration iscarried out over the full range of thevariable. Pulsed light sources are sometimesreferred to as delta light sources when thelength of the pulse is short with respect to theresponse time of the photomultiplier ordetecting instrument.

Detectivity, D: Reciprocal noise equivalentpower, NEP; it is expressed in W -1. Detec-tivity is a figure of merit providing the sameinformation as NEP but describes thecharacteristic such that the lower the radia-tion level to which the photodetector canrespond, the higher the detectivity. See NoiseEquivalent Power. (108)Discriminator, constant-fraction pulse-height: A pulse-height discriminator inwhich the threshold changes with inputamplitude in such a way that the triggeringpoint corresponds to a constant fraction ofthe input pulse height. (112)

Discriminator, pulse-height: A circuit thatproduces a specified output signal if andonly if it receives an input pulse whoseamplitude exceeds in one case or is less thanan assigned value in another case. (112)

Donor: An impurity element in an n-typesemiconductor that may become ionized bylosing an electron to the conduction bandand induce conduction by electrons. For ex-ample, phosphorus with a valence of five is apossible donor impurity in silicon (valence,four).

126

Dynode: An electrode that performs auseful function, such as current amplifica-tion, by means of secondary emission. (107)EADCI, Equivalent Anode Dark CurrentInput: The input flux in lumens or watts at aspecific wavelength which results in an in-crease in the anode current of a photomulti-plier tube just equal to the anode dark cur-rent.E2/B: A figure of merit used to evaluateperformance of liquid scintillation counters.E is the counting efficiency in per cent. B isthe number of background coincident countsper minute.Electron affinity, EA: The energy, usuallyexpressed in electron volts, required to movean electron from the bottom of the conduc-tion band to the vacuum level.Electron multiplier: That portion of thephotomultiplier consisting of dynodes thatproduce current amplification by secondaryelectron emission. (112)Electron resolution: The ability of the elec-tron multiplier section of the photomulti-plier to resolve inputs consisting of n andn + 1 electrons. This ability may be expressedas a fractional FWHM of the nth peak, or asthe peak-to-valley ratio of the nth peak to thevalley between the n th and the (n+ 1)thpeaks. (112)Electron volt: The energy received by anelectron in falling through a potential dif-ference of one volt. (108)

Equivalent Noise Input, ENI: That value ofinput radiant or luminous flux that producesan rms signal current that is just equal to therms value of the noise current in a specifiedbandwidth (Usually 1 Hz). See NoiseEquivalent Power. (108)Exitance: The density of radiant flux emit-ted from a surface. Radiant exitance is theintegral of radiant flux over all wavelengthswith units of watt per square meter. Lumi-nous exitance is the total of all luminous fluxfrom a surface with units of lumen persquare meter. Spectral radiant exitance is theexitance at a particular wavelength for aspecified wavelength interval; units are wattper square meter and micrometer. The termemittance (now deprecated) is synonomouswith exitance. Sometimes the term exitanceis used to indicate the density of radiant flux

Glossary of Terms

incident upon a surface. It is recommended,however, that the use of the term exitance berestricted to emission from a surface becauseterms such as irradiance, spectral irradiance,and illuminance are commonly used to in-dicate flux density incident on a surface.Extended Red Multi-Alkali, ERMA: Adesignation of a Na2K Sb:Cs photocathodeprocessed to obtain increased red-near-infrared response.Fall time: The mean time difference of thetrailing edge of a pulse between the 90- and10-per cent amplitude points.Fatigue: The tendency of a photomultiplierresponsivity to decrease during operation.Most commonly, responsivity loss is theresult of a lowering of secondary emission,particularly in the latter stages of a photo-multiplier. Recovery may or may not occurduring a period of idleness.Fermi level: The value of the electron energyat which the Fermi distribution function hasthe value one-half. (107)Focusing electrode: An electrode whosepotential is adjusted to control the cross-sectional area of the electron beam. (112)Flying-Spot Scanner: A system for generat-ing video signals for a television display. In atypical conception, the scanned raster of acathode-ray tube is focused onto a photo-graphic transparency. The modulated trans-mitted light signal is directed onto the photo-cathode of a photomultiplier tube whichprovides the electrical video signal.Footcandle: A unit of illuminance equal toone lumen per square foot. The SI unit of il-uminance, the lux (lumen per square meter),is preferred. (108)Footlambert: A unit of luminance equal to

the candela per square meter (nit), is pre-ferred. (108)Forbidden band: In the band theory ofsolids, a range of energies in which there areno electronic levels. (109)Full Width at Half Maximum, FWHM: Thefull width of a distribution measured at halfthe maximum ordinate. For a normal distri-bution it is equal to 2(2 ln 2)1/2 times the stan-

Gain (photomultipliers). See Current Ampli-fication.

127


Gamma-ray camera: A device used innuclear medicine to image distributions ofgamma-ray emitters. In a typical instrument,gamma rays emanating from tracer elementsintroduced into the patient are collimatedand cause scintillations in a single large, thinsodium iodide crystal. An array of photo-multiplier tubes views the crystal and pro-vides addressing information with which anoutput image of dots is constructed on acathode-ray tube.Hertz: Hz, the SI unit of frequencyequivalent to cycle per second. (108)Hysteresis: A borrowed term to describe acyclic gain variation in photomultiplierssometimes observed as a result of insulatorcharging and discharging.Illuminance: The density of the luminousflux on a surface; it is the quotient of theflux by the area of the surface when the lat-ter is uniformly illuminated. The SI unit isthe lux, lumen per square meter. (109)Irradiance: The density of radiant flux on asurface; it is the quotient of the flux by thearea of the surface when the latter isuniformly irradiated. The SI unit is the wattper square meter.Lambert’s cosine law: A law stating that theflux per solid angle in any direction from aplane surface varies as the cosine of the anglebetween that direction and the perpendicularto the surface. (107)Light pipe: An optical transmission elementthat utilizes unfocused transmission andreflection to reduce photon losses. (112)

Liquid scintillation counter: The combina-tion of a liquid scintillator and one or more(usually two) photomultiplier tubes used tomeasure or count radioactive disintegra-tions. The most common application is thecounting of beta rays emanating from asample mixed with the liquid scintillator.Lumen, lm: The SI unit of luminous flux. Itis equal to the flux through a unit solid angle(steradian) from a uniform point source ofone candela. (107, 108)Luminance: The luminous intensity per pro-jected area normal to the line of observation.Formerly called photometric brightness orbrightness. (108)Luminous efficacy: The quotient of thetotal luminous flux by the total radiant flux.

It is expressed in lumens per watt. For exam-ple, the maximum luminous efficacy of ablack body (which occurs at about 6600 K) is95 lumens per watt. (108)Luminous efficiency: The ratio of theluminous efficacy of a given source to themaximum spectral luminous efficacy. Forexample, the maximum luminous efficiencyof a black body (which occurs at about 6600K) is 95 lumens per watt/680 lumens per watt= 0.14.Luminous intensity: The luminous flux perunit solid angle in the direction in question.It is expressed in candelas (lumens per stera-dian). (107)Lux, lx: The SI unit of illuminance equal tothe flux of one lumen uniformly distributedover an area of one square meter.Microchannel plate: An array of smallaligned channel multipliers usually used forintensification. (107)Multichannel Analyzer, MCA: See Pulse-height analyzer.Negative Electron Affinity, NEA: A termreferring to an electron emitter whose sur-face has been treated with an electropositivematerial in such a way that the conductionband minimum lies above the vacuum level.Nit: The name recommended by the Inter-national Commission on Illumination forthe unit of luminance equal to one candelaper square meter. Note: Candela per squaremeter is the unit of luminance in the Interna-tional System of Units (SI). (107)Noise (photomultiplier tubes): The randomoutput that limits the minimum observablesignal from the photomultiplier tube. (112)Noise Equivalent Power, NEP: The radiantflux in watts incident on a detector whichgives a signal-to-noise ratio of unity. Thebandwidth and the manner in which theradiation is chopped must be specified aswell as the spectral content of the radiation.The most common spectral specification isfor monochromatic radiation at the peak ofthe detector response. (Some detectormanufacturers rate their detectors in termsof an NEP having units of watts Hz - 1/2.Assuming that the noise spectrum is flatwithin the range of the specification and thatNEP is normally specified for a bandwidthof one hertz, the two forms of NEP arenumerically equal.) (108)

Glossary of Terms

Noise&signal, (photomultiplier tubes): Thenoise output resulting from the statisticalvariation in the signal current itself as con-trasted with that which may be present whenthe detector is in the dark.Opaque photocathode: A photocathodewherein photoelectrons are emitted from thesame surface as that on which the photonsare incident. (Also called reflective photo-cathode.) (112)Peak-to-valley ratio: In a pulse-height-distribution characteristic, the ratio of thecounting rate at the maximum to that at theminimum-usually preceding the maximum.Photocathode: An electrode used for ob-taining photoelectric emission when ir-radiated. (112)Photocell: A solid-state photosensitive elec-tron device in which use is made of the varia-tion of the current-voltage characteristic as afunction of incident radiation. (112)

Photomultiplier, PMT: A phototube withone or more dynodes between its photocath-ode and output electrode. (112)

Photon counting: The technique, using aphotomultiplier, of counting output pulsesoriginating from single photoelectrons.Photopic vision: Vision mediated essential-ly or exclusively by the cones. It is generallyassociated with adaptation to luminance ofat least 3 candelas per square meter. (107)

Phototube: An electron tube that contains aphotocathode and has an output dependingat every instant on the total photoelectricemission from the irradiated area of thephotocathode. (112)Plateau: (counter): The portion of thecounting-rate-versus-voltage characteristiccurve in which the counting rate is substan-tially independent of the applied voltage.(109)Pulse-height analyzer, PHA: An instrumentcapable of indicating the number or rate ofoccurrence of pulses falling within each ofone or more specified amplitude ranges.(112)Pulse-height distribution: A histogram dis-playing the pulse count versus channel num-ber as obtained with a multichannel ana-lyzer, particularly as applied to scintillationcounting.

Pulse-height resolution, PHR: The ratio ofthe full width at half maximum of the pulse-height-distribution curve to the pulse heightcorresponding to the maximum of the distri-bution curve. In scintillation spectroscopy, itis customary to state pulse-height resolutionas a percentage. (112)Pulse jitter: A relatively small variation ofthe pulse spacing in a pulse train. Inphotomultipliers, pulse jitter is the result ofelectron transit time variations. (109)

Pulse width: The time interval between thefirst and last instants at which the instan-taneous amplitude reaches a stated fractionof peak pulse amplitude. (109)

Quantum efficiency (photocathodes): Theaverage number of electrons photoelectric-ally emitted from the photocathode per inci-dent photon of a given wavelength. (107)

Rad: A unit of absorbed radiation equal to100 ergs per gram-O.01 J/kg in SI units.(109)Radiance: The radiant flux per unit solidangle per unit of projected area of thesource. The SI unit is the watt per steradianand square meter, Wsr - 1 m - 2. (109)

Radiant intensity: The radiant flux pro-ceeding from the source per unit solid anglein the direction considered. The SI unit iswatt per steradian. (107)Reflectance: The ratio of the radiant fluxreflected from a body of material to that in-cident upon it. (See Absorptance.)Reflective photocathode: A photocathodewherein photoelectrons are emitted from thesame surface as that on which the photonsare incident. (Also called opaque photocath-ode.)Rem: Abbreviation for roentgen equivalentman. (1) In older usage, the dose (absorbed)of any ionizing radiation that will producethe same biological effect as that producedby one roentgen of high voltage x-radiation.(2) The unit of the RBE (relative biologicaleffectiveness) dose that is equal to the ab-sorbed dose in rads times the RBE. (109)

Resistance per square: The resistance of asquare of a thin conductive coating mea-sured between opposite sides of the square.The value is independent of the size of thesquare.

129


Responsivity: The ratio of the output cur-rent or voltage to the input flux in watts orlumens. For example, as applied to photo-multipliers: radiant responsivity expressed inmA W - 1 at a specific wavelength or

lm - 1. (108)Rise time: The mean time difference of theleading edge of a pulse between the 10- and90-percent amplitude points.

Roentgen: A unit of X- or gamma-radiationexposure such that the associated secondaryionizing particles produce, in air, ions carry-ing one electrostatic unit of charge of eithersign per 0.001293 gram of air. This quantityis the equivalent of 2.58 x 10 -4 coulomb perkilogram of air. (109)

Scintillation counter: The combination ofscintillator , photomultiplier, and associatedcircuitry for detection and measurement ofionizing radiation. (112)Scintillator: The body of scintillatormaterial together with its container. (107)Scotopic vision: Vision mediated essentiallyor exclusively by the rods. It is generallyassociated with adaptation to luminancebelow about 0.03 candela per square meter.(107)Secondary emission: Electron emissionfrom solids or liquids due directly to bom-bardment of their surfaces by electrons orions. (107)Secondary emission ratio (electrons): Theaverage number of electrons emitted from asurface per incident primary electron.Note: The result of a sufficiently largenumber of events should be averaged to en-sure that statistical fluctuations are negli-gible. (107)Semitransparent photocathode: A photo-cathode in which radiant flux incident onone side produces photoelectric emissionfrom the opposite side. Synonymous withTransmission-mode photocathode. (112)

Sensitivity: See Responsivity, the preferredterm.Spectral luminous efficacy (radiant flux):The quotient of the luminous flux at a givenwavelength by the radiant flux at thatwavelength. It is expressed in lumens perwatt. The maximum spectral luminous ef-

is 680 lumens per watt at awavelength of 555 nm. (107)Spectral luminous efficiency (radiant flux):The ratio of the spectral luminous efficacyfor a given wavelength to the maximumspectral luminous efficacy. Accordingly, thespectral luminous efficiency, V(X), is theratio V(X) = K(X)/680 and is identical to thestandard visibility factor of the photopichuman eye and to the y-tristimulus value

tristimulus system. (107, 111)Spectral radiant intensity: Radiant intensityper unit wavelength interval; for example,watts per (steradian-nanometer). (107)Stage: One step of a multiplier, as onedynode stage.

,

Stem: The portion of a photomultiplierenvelope containing the leads to electrodes.Steradian: The unit of solid angle whichsubtends an area equal to the square of theradius. (107)Tea-cup: Descriptive term for an RCAphotomultiplier type having a large cup-shaped first dynode.Time jitter: See Transit-time spread.Time-to-amplitude converter, TAC: An in-strument producing an output pulse whoseamplitude is proportional to the time dif-ference between start and stop pulses. (112)Traceability: Process by which the assignedvalue of a measurement is compared, di-rectly or indirectly, through a series ofcalibrations to the value established by theU.S. national standard. (107)Transit time: For a discussion of the severaldefinitions of this term, refer to the section“Time Effects” in Chapter 4. Photomulti-plier Characteristics.Transit-time spread: The FWHM (full-width-at-half maximum) of the time distri-bution of a set of pulses each of which cor-responds to the photomultiplier transit timefor that individual event. (112)Transmission-mode photocathode: A pho-tocathode in which radiant flux incident onone side produces photoelectric emissionfrom the opposite side. Synonymous withSemitransparent photocathode. (112)Transmittance: The ratio of the radiant fluxtransmitted through a body of material to

Glossary of Terms

that incident upon it. See Absorptance.Vacuum level: The minimum potential ener-gy level an electron must reach to escape en-tirely from the attraction of a solid or anatom.Valence band: The range of energy states inthe spectrum of a solid crystal in which liethe energies of the valence electrons thatbind the crystal together. (107)Venetian-blind dynode: A descriptive termfor a photomultiplier dynode structure. Thedynode is constructed with a number ofparallel slats with space between permittingsecondary electrons to be directed to the nextstage.Voltage divider (photomultiplier): A seriesstring of resistors across which a voltage isapplied providing an appropriate voltagedrop per stage.Work function: The minimum energy re-quired to remove an electron from the Fermilevel of a material into field-free space. (107)

REFERENCES107. IEEE Standard Dictionary of Electricaland Electronics Terms, Wiley-Interscience.108. RCA Electro-Optics Handbook,EOH-11.109. The International Dictionary ofPhysics and Electronics, D. Van Nostrand.110. Information Transmission, Modula-tion, and Noise, Mischa Schwartz, McGraw-Hill, 1959.111. Leo Levi, Applied Optics, Vol. 1, JohnWiley and Sons, Inc., 1968.112. An American National Standard,IEEE Standard Test Procedures for Photo-multipliers for Scintillation Counting andGlossary for Scintillation Counting Field.ANSI N42.9-1972; IEEE Std 398-1972.

131


Appendix C-Spectral Response Designation Systems

The spectral response of a photomultipliertube depends primarily on the chemical com-ponents of the photocathodes. Differencesin spectral response, however, can resultfrom variations in the processing ofphotocathodes even for the same chemicals.Some photocathodes are of the “opaque” or“reflective” type where the photoemission isfrom the same side as the excitation. Others,having essentially the same chemistry but ofthe “semitransparent” type where thephotoemission is from the side oppositefrom that of the excitation, may have asomewhat different spectral response. Theshort-wavelength-cutoff characteristic of thespectral response is the result of thetransmission characteristic of the particularglass used for the photocathode window.S-Designation System

In the early 1940’s, the JEDEC (JointElectron Devices Engineering Council) in-dustry committee on photosensitive devicesdeveloped the “W-system of designatingspectral responses. The philosophy includedthe idea that the product user need only beconcerned about the response of the device,not by how it might be fabricated. And, infact, the chemical compositions of some ofthe photocathodes and dynodes were con-sidered proprietary by various manufac-turers. S-numbers were registered from S-lthrough S-40. See Table C-I. Subsequently,the lack of activity of the industry committeeon photosensitive devices resulted in in-dependent assignment of codes for spectralresponses. In other cases, the spectralresponse was described by a curve and an ac-tual description of the photocathode com-position and of the window material.

The following material is a brief accountof the different designation systems andtheir relationship and meaning. This infor-

132

mation is of value to the photomultiplieruser because it will help in the interpretationof the published spectral characteristics datafor a specific tube type regardless of whichresponse designation system is used.

Numerical SystemIn 1971 RCA/BURLE introduced a

numerical system of spectral response des-ignations supplementing and overlappingthe JEDEC S-designation system. The rea-son for this extended system was the prolif-eration of spectral responses resulting fromthe combinations of different glasses, newphotocathode types, and processing varia-tions. Table C-II is a catalogue of this 1971numbering system providing an identifica-tion and description of each number.

Alphanumeric Coded SystemIn 1976, RCA/BURLE changed its number

system for spectral-response characteristics toan alphanumeric combination coded system.The new designations were combinations ofalphanumerics based on (1) the photocath-ode material, (2) the window material, and(3) the photocathode operating mode. TableC-III provides the code for the spectralresponse designation in four columns. Thefirst two digits in the designation number(Column I) indicate the photocathodematerial; the following alphabetic character(Column II) indicates the window material;the next alphabetic character (Column III)indicates the photocathode operating mode.Where required, the letter “X” is used as asuffix to the designation to indicate an ex-tended response in the red or near infrared.

As an example of the usage of this system,tube type 931A has a spectral response thatwas previously designated as 102 (S-4) inTable C-II. This tube type has a Cs3Sbphotocathode, a 0080 lime glass window,

Spectral Response Designation Systems

and a reflection-type photocathode. Itsdesignation according to the 1976 codesystem is 20AR.

Similarly, a tube type having a Cs3Sbphotocathode, a 0080 glass window, and atransmission-type photocathode is desig-nated 20AT. This response was previouslydesignated 107 (S-l 1).Current Practices

When the coded spectral response designa-tion system was devised in 1976, it was an-ticipated that all the information providedby the code would be useful to customers inspecifying the type of photomultiplierneeded. Experience has indicated that very

few are sufficiently acquainted with the codeto make good use of it. Instead, mostknowledgeable customers prefer to be in-formed directly of the nature of thephotocathode and the window. As a result,therefore, BURLE has discontinued the use ofcoded spectral response designations exceptfor occasional reference to some of the morecommon JEDEC S-numbers, which havebecome well established. Instead, referenceis made to the photocathode material, thewindow material, and any special processinginformation. In addition, typical spectralresponse and other related data are pro-vided .

133


Table C-I Spectral Response Designations as Specified by JEDEC

S-DesignationS-l

S-2S-3S-4S-5S-6S-7S-8S-9S-10S-11S-12S-13S-14S-15S-16S-17

PhotocathodeComposition Window*

Ag-O-Cs lime glass

(Obsolete; formerly similar to S-l)Ag-O-Rb lime glass

Cs3Sb lime glassCs3Sb Corning 9741

Na uv transmittingAg-O-Rb-Cs borosilicate

Cs3Bi lime glass(Obsolete; formerly similar to S-l1)

Ag-Bi-O-Cs lime glassCs3Sb lime glass

(CdS - a photoconductive crystal)Cs3Sb fused silica

(Ge - photovoltaic)(CdS-CdSe - a photoconductor)(CdSe - a photoconductor)

Cs3Sb lime glass

S-18 (Sb-S - photoconductor, camera tubes)S-19 Cs3Sb fused silicaS-20 Na2KSb:Cs lime glassS-21 Cs3Sb Corning 9741S-22 (Not used)S-23 Rb-Te fused silicaS-24 (Not used)S-25 Na2KSb:Cs lime glass

S-26S-27S-28S-29S-30S-31S-32S-33S-34S-35S-36S-37S-38S-39S-40

(InSb - photovoltaic)(Ge:Au - photoconductive)(InAs - photovoltaic)(PbSe - photoconductive)(Ge:Cu - photoconductive)(PbS - photoconductive)(PbS - photoconductive)(PbS - photoconductive)(InAs - photovoltaic)(InSb - photoconductive)(GaAs - photovoltaic)(Si - photovoltaic)(PbSe - photoconductive)(PbSe - photoconductive)(Ge:Hg - photoconductive)

*Window may be of material other than thatspecified, but it will have equivalent spectralcharacteristics.

References: 113, 114

134

Notessemitransparent or

opaque

opaqueopaqueopaqueopaqueopaqueopaque

semitransparentsemitransparent

semitransparent

opaque withreflecting substrate

opaquesemitransparentsemitransparent

semitransparent

semitransparent,processed for ex-

tended red response

Number101102103104105106107108109110111112113114115

116 K2CsSb117 K2CsSb118 K2CsSb119 Na2KSb:Cs

120 K2CsSb121 Cs-Te122 K2CsSb123 Cs3Sb124 Cs3Sb125 Cs-Te126 K2CsSb

127 Ag-Bi-O-Cs128 Ga-As129 Ga-As-P130 Na2KSb:Cs

131 Na2KSb:Cs

132 Na2KSb:Cs

133 K2CsSb134 Ga-As135 Ga-As-P136 K2CsSb


Table C-II RCA/BURLE 1971 Spectral Response Numbering Code

JEDECS-Designation

S-lS-4

S-5S-8

S-10S-l 1S-13S-19S-20

PhotocathodeComposition

Ag-o-csCs3SbCs3SbCs3SbCs-Bi

Ag-Bi-O-CsCs3SbCs3SbCs3Sb

Na2KSb:CsNa2KSb:CsNa2KSb:CsNa2KSb:CsNa2KSb:Cs

K2CsSb

Windowlime glasslime glass

Corning 9741 glassCorning 9741 glass

lime glasslime glasslime glass

SiO2SiO2

Borosilicate glasslime glass

Corning 9823 glassPyrexSiO2

Lime or boro-silicate glass

PyrexCorning 9823 glassCorning 9741 glass

Pyrex

Sapphire, uv gradeSiO2

Al203

Sapphire, uv gradeCorning 9741 glass

LiFBorosilicate or lime

glassCorning 9741 glassCorning 9741 glassCorning 9741 glassBorosilicate or lime

glassBorosilicate glass

Lime orborosilicate glass

SiO2Sapphire, uv gradeSapphire, uv grade

Lime glass

uvuv

Notes

transmittingtransmitting

uv transmitting

uv transmittinguv transmittingextended red:

ERMA III

extended red:ERMA III

extended red:ERMA II

135


Table C-II RCA/BURLE 1971 Spectral Response Numbering Code (cont’d)

NumberJEDEC Photocathode

S-Designation CompositionWindow Notes

137

138

139140141142

Na2KSb:Cs Corning 9741 glass extended red:ERMA II

Na2KSb:Cs Corning 9741 glass extended red:ERMA I

Na2KSb Borosilicate glass high temperatureIn.06-Ga.94-As Corning 9741 glass Type IIn.12-Ga- .88-As Corning 9741 glass Type IIIn.18- G a.82- A s Corning 9741 glass Type III

Table C-III RCA/BURLE 1976 Coded System for Spectral Response Designation

Column I10 = AgOCs15 = AgBiOCs20 = CsSb25 = CsBi30 = CsTe35 = KCsSb (Bialkali)40 = NaKSb (High

temperature bialkali)45 = RbCsSb50 = NaKCsSb (Multialkali)51 = NaKCsSb (ERMA I)52 = NaKCsSb (ERMA II)53 = NaKCsSb (ERMA III)60 = GaAs71 = InGaAs (Type I)72 = InGaAs (Type II)73 = InGaAs (Type III)

Column II Column IIIA = 0080 (lime glass) or D = Dormer-window type

7056 (Borosilicate glass) R = Reflection TypeC = 7740 (Pyrex) T = Transmission TypeE = 9741 (UV transmitting

glass)G = 9823 (UV transmitting

glass)J = SiO2 (Fused silica)M = UV-grade SapphireP = LiF

Column IVX = Extended Response

REFERENCES113. “Relative spectral response data forphotosensitive devices (“S” curves)“,JEDEC Publication No. 50, Electronic In-dustries Association, Engineering Depart-ment, 2001 Eye Street, N.W., Washington,D.C. 20006 (1964)

114. “Typical characteristics of photosen-sitive surfaces,” JEDEC Publication No. 61,Electronic Industries Association, Engineer-ing Department, 2001 Eye Street, N. W.,Washington, D.C. 20006. (1966)

136


Appendix D-Photometric Units and Photometric-to-Radiometric Conversion

Photometry is concerned with the mea-surement of light. Because the origins of thephotoelectric industry were associated withthe visible spectrum, the units first used forevaluating photosensitive devices were pho-tometric. Today, however, even thoughmany of the applications of photosensitivedevices are for radiation outside the visiblespectrum, the photometric units are still re-tained for many purposes. Because theseunits are based on the characteristics of theeye, this discussion begins with a considera-tion of some of these characteristics.

CHARACTERISTICS OFTHE EYE

The sensors in the retina of the human eyeare of two kinds: “cones” which predomi-nate the central (or foveal) vision and“rods” which provide peripheral vision. Thecones are responsible for our color vision;rods provide no color information but in thedark-adapted state are more sensitive thanthe cones and thus provide the basis of dark-adapted vision. Because there are no rods inthe fovea1 region, faint objects can morereadily be observed at night when the eye isnot exactly directed toward the faint object.The response of the light-adapted eye (conevision) is referred to as the Photopic eyeresponse; the response of the dark-adaptedeye (rod vision) is referred to as Scotopic eyeresponse.

Although characteristics of the human eyevary from person to person, standardluminosity coefficients for the eye weredefined by the Commission Internationaled’Eclairage (International Commission onIllumination) in 193 1. These standard C . I. E.luminosity coefficients for photopic visionare given in Table D-I. They represent therelative luminous equivalents of an equal-

energy spectrum for each wavelength in thevisible range, assuming foveal vision. An ab-solute “sensitivity” figure established forthe standard eye relates photometric unitsand radiant power units. At 555 nanometers,the wavelength of maximum sensitivity ofthe eye, one watt of radiant power cor-responds to approximately 680 lumens. Thequotient of the luminous flux at a givenwavelength by the radiant flux at thatwavelength is referred to as the SpectralLuminous Efficacy ,

D-l

Various determinations of the maximumvalue, K (555 nanometers), have variedsomewhat from the nominal value of 680lumens per watt.

For the dark-adapted eye, the peak sen-sitivity increases and is shifted toward theviolet end of the spectrum. A tabulation ofthe relative scotopic vision is also given inTable D-I. The peak luminosity for scotopicvision occurs at 511 nanometers and is theequivalent of 1746 lumens/watt. Fig. D-lshows the comparison of the absoluteluminosity curves for scotopic and photopicvision as a function of wavelength.

The sensitivity of the eye outside thewavelength limits shown in Table D-I is verylow, but not actually zero. Studies with in-tense infrared sources have shown that theeye is sensitive to radiation of wavelength atleast as long as 1050 nanometers. Fig. D-2shows a composite curve given by Griffin,Hubbard, and Wald115 for the sensitivity ofthe eye for both foveal and peripheral visionfrom 360 to 1050 nanometers. According toGoodeve116 the ultraviolet sensitivity of theeye extends to between 302.3 and 312.5nanometers. Below this level the absorption

137


Table D-IRelative Luminosity Values for Photopic and Scotopic Vision

Wavelength(nm)

410420430

450460470480490500510520530540550560570580590

610620630640650660670680690700710720730740750760770

(B>3 cd m-2) ( B < 3 x 1 0- 5c d m- 2)

---

0.000040.000120.00040.00120.00400.01160.0230.0380.0600.0910.1390.2080.3230.5030.7100.8620.9540.9950.9950.9520.8700.7570.6310.5030.3810.2650.1750.1070.0610.0320.0170.00820.00410.00210.001050.000520.000250.000120.000060.00003

0.00030.00080.00220.00550.01270.02700.05300.09500.1570.2390.3390.4560.5760.7130.8420.9480.9990.9530.8490.6970.5310.3650.2430.1550.09420.05610.03240.01880.01050.00580.00320.00170.00090.00050.00020.0001-------

138

Photometric Units and Photometric-to-Radiometric Conversion

of radiation by the proteins of the eye lensapparently limits further extension of visioninto the ultraviolet. Light having a wave-length of 302 nanometers is detected by itsfluorescent effect in the front part of the eye.


92CS- 32467

Fig. D-7 - Absolute luminosity curves forscotopic and photopic eye response.

PHOTOMETRIC UNITSLuminous intensity (or candlepower) de-

scribes luminous flux per unit solid angle in aparticular direction from a light source. Themeasure of luminous intensity is the fun-damental standard from which all other pho-tometric units are derived. The standard ofluminous intensity is the candela; the olderterm candle is sometimes still used, butrefers to the new candle or candela.

The candela is defined by the radiationfrom a black body at the temperature ofsolidification of platinum. A candela is one-sixtieth of the luminous intensity of onesquare centimeter of such a radiator.

2 I-PERIPHERY

- 8

-10.3 0 0 5 0 0 7 0 0 9 0 0 1100


92CS- 32468

Fig. D-2 - Relative spectral sensitivity of thedark-adapted foveal and peripheral retina.

A suitable standard for practical photo-electric measurements is the AJ2239 cali-brated lamp, which operates at a currentof about 4.5 amperes and a voltage of 7 to 10volts. A typical lamp calibrated at a colortemperature of 2856 K provides a luminousintensity of 55 candelas. The luminous inten-sity of a tungsten lamp measured in candelasis usually numerically somewhat greater thanthe power delivered to the lamp in watts.

A color temperature of 2870 K served asthe basic test standard in this country forabout 30 years. A change had been made toagree with C.I.E. illuminant A, a more wide-spread standard that at first required a colortemperature of 2854 K, but has more recent-ly been adjusted to 2856 K to accommodateto the international practical temperaturescale of 1968. See also Appendix F. The dif-ference between the old lamp standard at2870 K and at the new temperature isgenerally negligible. 117

Luminous flux is the rate of flow of lightenergy, the characteristic of radiant energythat produces visual sensation. The unit ofluminous flux is the lumen, which is the fluxemitted per unit solid angle by a uniformpoint source of one candela. Such a source

A radiant source may be evaluated interms of luminous flux if the radiant-energy

139


is the total radiant power in watts per unitwavelength, total radiant power over all

as follows:

D-2

efficiency. The lumen is the most widelyused unit in the rating of photoemissivedevices. For photomultipliers, the typicaltest levels of luminous flux range from 10 -7to 10-5 lumen (0.1 to 10 microlumens).

Illuminance (or illumination) is the densityof luminous flux incident on a surface. Acommon unit of illuminance is the footcan-dle, the illumination produced by one lumenuniformly distributed over an area of onesquare foot. It follows that one candela pro-duces an illuminance of one footcandle at adistance of one foot. The preferred SI unit(International System of Units) of illumi-nance is the lux, which is the illuminationproduced by one lumen uniformly distri-buted over an area of one square meter. (1 lx= 1 lm m - 2) It also follows that onecandela produces an illuminance of one luxat a distance of one meter.

1 lux = 0.0929 footcandle

Table D-II lists some common values of il-luminance. Further information concerningnatural radiation is shown in Fig. D-3 whichindicates the change in natural illuminationat ground level during, before, and aftersunset for a condition of clear sky and no

Photometric luminance (or brightness) is ameasure of the luminous flux per unit solidangle leaving a surface at a given point in agiven direction, per unit of projected area.The term photometric luminance is used todistinguish a physically measured luminancefrom a subjective brightness. The lattervaries with illuminance because of the shiftin spectral response of the eye toward theblue region at lower levels of illuminance.The term luminance describes the light emis-sion from a surface, whether the surface isself-luminous or receives its light from someexternal luminous body.

Table D-IITypical Values of Natural Scene Illuminance

Sky ConditionApprox. Levels of

Illuminance-lux (lm m-2)

Direct sunlight . . . . . . . . . . . . . . . l-l .3 x 105Full daylight (not direct sunlight) . l-2 x 104

Overcast day . . . . . . . . . . . . . . . . . . . . . . . .103

Very dark day . . . . . . . . . . . . . . . . . . . . . . . 102

Twilight . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10Deep twilight ........................ 1Full moon . . . . . . . . . . . . . . . . . . . . . . . . 10-1

Quarter moon ..................... 10 -2Moonless, clear night sky. ........... 10 -3Moonless, overcast night sky ......... 10 -4

92CS-32469

Fig. D-3 - Natural illuminance on the earthfor the hours immediately before and aftersunset with a clear sky and no moon.

For a surface that is uniformly diffusing,luminance is the same regardless of the anglefrom which the surface is viewed. This con-dition results from the fact that a uniformlydiffusing surface obeys Lambert’s Law (thecosine law) of emission. Thus, both the emis-sion per unit solid angle and the projectedarea are proportional to the cosine of theangle between the direction of observationand the surface normal.

140

Photometric Units and Photometric-to-Radiometric Conversion

The SI unit of luminance is the candela persquare meter or a lumen per steradian andsquare meter. This unit is called the nit. Acommonly used unit of luminance is the

per square foot.

1 nit = 0.2919 footlambert

The relationship between luminance andtotal luminous flux from a uniform diffuseris illustrated by the use of Fig. D-4.

Fig. D-4 - Diagram illustrating Lambert’sand the calculation of total luminousfrom a diffuse radiator.

lawflux

Consider an elementary portion of the dif-fusing surface having an area, A, and aluminance of L. The projected area at the

sented by the differential area on the surface

luminance, L, represents the luminous fluxper unit solid angle leaving the surface perunit of projected area. Therefore, the flux

and the total flux from the area, A, is givenby

On the other hand if a perfectly diffusingand reflecting surface is illuminated with 1

square meter (nits).Table D-III provides data on luminance

values of various sources. Table D-IV is aconversion table for various photometricunits.

Table D-IIILuminance Values for Various Sources

SourceLuminance

(Footlamberts)

Sun, as observed from Earth’s surface at meridian . . . . . 4.7 x 108 . . . . .Moon, bright spot, as observed from Earth’s surface. . . 730 . . . . . . . . . .Clear blue sky . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2300. . . . . . . . .Lightning flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 x 1010 . . . . . .Atomic fission bomb, 0.1 millisecond after firing,

90-feet diameter ball . . . . . . . . . . . . . . . . . . . . . . . . . . 6 x 1011 . . . . . .Tungsten filament lamp, gas-filled, 16 lumen/watt . . . . 6 x 10 6 . . . . .Plain carbon arc, positive crater . . . . . . . . . . . . . . . . . . . . 4.7 x 106 . . . . .Fluorescent lamp, T-12 bulb, cool white, 430 mA,

medium loading . . . . . . . . . . . . . . . . . . . . . . . . . . . 2000 . . . . . . . . . .Color television screen, average brightness . . . . . . . . . . . 50 . . . . . . . . . . . .

Luminance(Candelas m -2 )

1.6 x 109

25007900

7 x 1010

2 x 1012

9 x 106

1.6 x 107

7000170

141


CALCULATION OF RADIANTRESPONSIVITY FROM

LUMINOUS RESPONSITIVITYSpecification of photocathode responsi-

vity is most frequently given in terms oflumens from a tungsten source at a colortemperature of 2856 K. If a relative spectralresponse is known, it is possible to calculatethe absolute radiant responsivity of the

Table F-I in Appendix F. The integrationsindicated in Eq. D-7 have been performedfor most photocathode spectral responses. A

watt, is provided in Table I. This factorrepresents the ratio of the peak radiantresponsitivity in amperes per watt to theluminous responsivity in amperes per lumen.

photocathode as follows.Let the relative spectral response (with a

maximum value of unity) of the photocath-Table D-IV

Conversion Table for

absolute radiant response at the peak of theVarious Photometric Units

response characteristic is then given by SI Units Other Units

lamp radiation striking the photocathode beLuminous Intensity (I)

1 candela (cd) = 1 lumen/steradian

The response of the photocathode (in am- (lm sr - 1)

peres) to the total radiation is then given by

D-4 uniform point source of 1 candela

The integration is done over the completerange of wavelengths either as limited by

The light flux (in lumens) represented bythe total radiant flux is given by

Illuminance (E)1 lux (lx) = 1 lumen/ 1 footcandle (fc) =

1 lux = 0.0929 footcandle

D-5Luminance (L)

1 nit (nt) = 1 candela/ 1 footlambert (fL) =

0.ciency as given in Table D-I and 680 lumensper watt is taken as the maximum spectral

1 nit = 0.2919 footlambert

luminous efficacy.The luminous responsivity of the photo-

cathode in amperes per lumen is then givenby the ratio of expressions D-4 and D-5: REFERENCES

D-6

From Eq. D-6, the maximum radiant

amperes per watt:

Note that the absolute magnitude of the

lamp operated at a color temperature of

115. D.R. Griffin, R. Hubbard, and G.Wald, “The sensitivity of the human eye toinfrared radiation,” JOSA, Vol. 37, No. 7,pp 546, (1947).116. C.F. Goodeve, “Vision in the ultra-violet,” Nature (1934)117. R.W. Engstrom and A.L. Morehead,“Standard test lamp temperature forphotosensitive devices-relationship of ab-solute and luminous sensitivities,” RCAReview, Vol. 28, pp 419-423 (1967)118. I.E.S. Lighting Handbook, Illuminat-ing Engineering Society, New York, NewYork (1959).

142

Spectral Response and Source-Detector Matching

Appendix E-Spectral Response and Source-Detector Matching

This appendix covers the significance ofthe spectral response of photomultipliertubes; describes some of the methods usedfor measuring this response; and discusses insome detail the calculations and other con-siderations useful for matching the radiationsource and the photomultiplier tube type fora specific application.

SPECTRAL-RESPONSECHARACTERISTICS

A spectral-response characteristic is adisplay of the response of a photosensitivedevice as a function of the wavelength of theexciting radiation. Such curves may be on anabsolute or a relative basis. In the latter casethe curves are usually normalized to unity atthe peak of the spectral-response curve. Fora photocathode the absolute radiant sen-sitivity is expressed in amperes per watt.Curves of absolute spectral response mayalso be expressed in terms of the quantum ef-ficiency at the particular wavelength. If thecurve is presented in terms of amperes perwatt, lines of equal quantum efficiency maybe indicated for convenient reference.Typical curves are usually included in thepublished tube data. It should be understoodthat because of variations in processing,deviations from these typical curves may beexpected. It would not be unusual for thewavelength for peak response to vary by 30nanometers from the typical value. The samevariation may be expected in the long-wavelength cutoff as judged by the wave-length for which the response is 10% of themaximum. On the other hand, the short-wavelength cutoff is more closely held by theglass transmission characteristic of theenvelope.

The relationship between the spectral

of a photon isE-l

quency of the incident radiation, c is the

the incident radiation. If the quantum effi-

of emitted photoelectrons to the number ofincident photons), the responsivity is givenby

= - - h c E-2

in units of coulombs per joule or amperesper watt where e is the charge on the elec-tron. Solving for the quantum efficiency:

E-3

MEASUREMENT TECHNIQUESIn order to determine a spectral-response

characteristic, (1) a source of essentiallymonochromatic radiation, (2) a current-sensitive instrument to measure the outputof the photocathode, and (3) a method ofcalibrating the monochromatic radiation forits magnitude in units of power are required.

Monochromatic radiation is often pro-vided by a prism or a grating type of mono-chromator. Interference filters may also beused to isolate narrow spectral bands.

143


Although interference filters do not providethe flexibility of a monochromator, theymay be indicated in situations in whichrepeated measurements are required in a par-ticular region of the spectrum.

The width of the spectral transmissionband in these measurements must be narrowenough to delineate the spectral-responsecharacteristic in the required detail. How-ever, for the most part, spectral-responsecharacteristics do not require fine detail andgenerally have broad peaks with exponentialcutoff characteristics at the long-wavelengthlimit and rather sharp cutoffs at the short-wavelength end. For spectral measurements,therefore, a reasonably wide band is used.Such a band has the following important ad-vantages: (1) because the level of radiation ishigher, measurements are easier and moreprecise; and (2) spectral leakage in otherparts of the spectrum is relatively less impor-tant. Spectral leakage is a problem in anymonochromator because of scattered radia-tion, and in any filter because there is sometransmission outside the desired pass band.A double monochromator may be used andwill greatly reduce the spectral leakage out-side the pass band. The double monochro-mator is at a disadvantage in cost and com-plexity. If a pass band of 10 nanometers isused, spectral leakage can be insignificantfor most of the spectral measurements. Atthe same time, this pass band is narrowenough to avoid distortion in the measuredspectral-response characteristic. It is oftenadvisable to vary the pass band dependingupon what part of the curve is beingmeasured. For example, at the long-wavelength cutoff where the response of thephotocathode may be very small, the leakagespectrum may play an important part; thus itis advisable to increase the spectral bandpassof the measurement. Wide pass-band colorfilters that exclude the wavelength of themeasurement and include the suspected spec-tral leakage region, or vice versa, are used inchecking the magnitude of the possibleleakage spectrum.

ENERGY SOURCESVarious radiant-energy sources are used to

advantage in spectral-response measure-ments. A tungsten halogen lamp is usefulfrom 350 nanometers to wavelengths much

greater than 1000 nanometers because of itsuniform and stable spectral-emission charac-teristic. A mercury vapor discharge lampprovides a high concentration in specificradiation lines and thus minimizes thebackground scattered-radiation problem.The mercury lamp is particularly useful inthe ultraviolet end of the spectrum where thetungsten lamp fails. Another useful sourcefor the ultraviolet is the deuterium lamp.(See the discussion on radiant energy sourcesin Appendix F.)

MEASUREMENT OFRADIANT POWER OUTPUT

A radiation thermocouple or thermopilehaving a black absorbing surface is com-monly used to measure the radiation poweroutput at a specific wavelength. Althoughthese devices are relatively low in sensitivity,they do provide a reasonably reliable meansof measuring radiation independent of thewavelength. The limitation to their accuracyis the flatness of the spectral absorptioncharacteristic of the black coating on thedetector. Throughout the visible and near-infrared regions there is usually no problem.There is some question, however, as to theflatness of the response in the ultravioletpart of the spectrum.

The output of the thermocouple orthermopile is a voltage proportional to theinput radiation power. This voltage is con-verted by means of a suitable sensitivevoltmeter to a calibrated measure of powerin watts. The calibration may be accom-plished by means of standard radiationlamps obtained from the National Bureau ofStandards. It is theoretically possible tocalculate the monochromatic power from aknowledge of the emission characteristic ofthe source, the dispersion characteristic ofthe monochromator or the transmissioncharacteristic of the filter, and from thetransmission characteristic of the variouslenses. This procedure is difficult, subject toerror, and is not recommended exceptperhaps in the case of a tungsten lamp sourcecombined with a narrow-band filter system.

Another useful reference standard is thepyroelectric detector. In this case, the radia-tion to be measured must be interrupted bymeans of a light chopper at about 15 hertz.Absolute calibration might proceed in a

144


manner similar to that of the thermopile.Special pyroelectric detectors have beenfabricated that can be self-calibrated bymeans of electric power input. 119

MEASUREMENT OFPHOTOCATHODE OUTPUT

For measuring the spectral characteristicof a photocathode, a very sensitive ammeteris required. When the output of the photo-cathode of a photomultiplier is measured,the tube is usually operated as a photodiodeby connecting all elements other than thephotocathode together to serve as the anode.When the photomultiplier is operated as aconventional photomultiplier, the output isvery easy to measure. It is necessary, how-ever, to be careful to avoid fatigue effectswhich could distort the spectral-responsemeasurement. It should be noted that thespectral response of a photomultiplier maybe somewhat different from that of thephotocathode alone because of the effect ofinitial velocities on collection efficiency atthe first dynode and because of the possibili-ty of a photoeffect on the first dynode bylight transmitted through the photocathode,especially if the first dynode is a photosen-sitive material such as cesium-antimony.

SOURCE AND DETECTORMATCHING

One of the most important parameters tobe considered in the selection of a photo-multiplier type for a specific application isthe photocathode spectral response. Thespectral response of BURLE photomultipliertubes covers the spectrum from the ultra-violet to the near-infrared region. In thisrange there are a large variety of spectralresponses to choose from. Some cover nar-row ranges of the spectrum while otherscover a very broad range. The publisheddata for each photomultiplier type show therelative and absolute spectral-responsecurves for a typical tube of that particulartype. The relative typical spectral-responsecurves published may be used for matchingthe detector to a light source for all but themost exacting applications. The matching ofdetector to source consists of choosing thephotomultiplier tube type that has a spectralresponse providing maximum overlap of thespectral distributions of detector and lightsource.

MATCHING CALCULATIONSThe average power radiating from a light

source may be expressed as follows:

E-4where PO is the incident power in watts perunit wavelength at the peak of the relative

which is normalized to unity.If the absolute spectral distribution for the

light source and the absolute spectralresponse of the photomultiplier tube areknown, the resulting photocathode currentIk when the light is incident on the detectorcan be expressed as follows:

E-5

photocathode in amperes per watt at the

sents the relative photocathode spectralresponse as a function of wavelength nor-malized to unity at the peak. When Eq. E-4is solved for the peak power per unit wave-length, PO, and this solution is substituted in-to Eq. E-5, the cathode current is expressedas follows:

The ratio of the dimensionless integrals canbe defined as the matching factor, M. Thematching factor is the ratio of the area underthe curve defined by the product of therelative source and detector spectral curvesto the area under the relative spectral sourcecurve.

Fig. E-l shows an example of the data in-volved in the evaluation of the matching fac-tor, M, as given in Eq. E-7.

If the input light distribution incident onthe detector is modified with a filter or anyother optical device, the matching-factorformulas must be changed accordingly. Ifthe transmission of the filter or optical

145


W A V E L E N G T H - N A N O M E T E R S

9 2 c s - 3 2 4 7 0

Fig. E-l - Graphic example of factors used inevaluation of matching factor, M.

Table E-I shows a number of matching fac-tors calculated for various light sources andspectral response characteristics. When thespectral range of a source exceeded 1200nanometers, the integration was terminatedat this wavelength. Because none of thephotoresponses exceed 1200 nanometers,conclusions as to the relative merit ofvarious combinations are still valid.

It should also be noted that since thesedata were originally published there has beena proliferation of photocathode develop-ment. Many of the new photocathodes, how-ever, have spectral responses similar to thoselisted in Table E-l. As a result, the spectralmatching factors given could also be usedfor many of the new photocathodes. Forexample, the data on S-l 1 could well be sub-stituted for photocathodes such as Rb2CsSb,K2CsSb or Na2KSb, with only moderateerror.

When M is substituted for the integralratio in Eq. E-6, the photocathode currentbecomes

E-10In any photomultiplier application it is

desirable to choose a detector having aphotocathode spectral response that willmaximize the photocathode current, Ik, fora given light source. Maximizing the cathodecurrent is important to maximize the signal-to-noise ratio. From Eq. E-9, it can be seenthat the product of the matching factor Mand the peak absolute photocathode sensitiv-

cathode current.The importance of taking into account the

absolute photocathode sensitivity, as well asthe matching factor, is illustrated by a com-parison of the S-l and S-20 photocathodeswith a tungsten light source operating at acolor temperature of 2856 K. The S-l andthe S-20 matching factors are 0.516 and0.112, respectively. From the matching fac-tors alone it appears that the S-l is the bestchoice of photocathode spectral response.The S-l has a peak absolute sensitivity of 2.3milliamperes per watt and the S-20 has apeak absolute sensitivity of 64 milliamperesper watt. Then, from Eq. E-9 the expectedphotocathode currents are

Ik(S-1)=0.0023 x 0.516 P=0.00119 PIk(S-20)=0.064 x 0.112 P=0.00717 P

These calculations show that the S-20photocathode will provide a response to thetungsten lamp six times that of the S-lphotocathode.

REFERENCES119. W.M. Doyle and B.C. McIntosh (LaserPrecision Corp., Irvine, California) and JonGeist (National Bureau of Standards, Wash-ington, D.C.), “Implementation of a systemof optical calibration based on pyroelectricradiometry,” Optical Engineering, Vol. 15,No. 6, pp 541-548, (Nov.-Dec. 1976)120. E.H. Eberhardt, “Source-detectorspectral matching factors,” Applied Optics,Vol. 7, p 2037 (1968)

146


Light

SourcePhosphors

P1P4P7P11P15P16P20P22BP22GP22RP24P31NaI

Lamps2870/2856

stdFluorescent

SunIn space+2 air

massesDay sky

Blackbodies6000K3000K2870 K2856 K2810 K2042 K

Notes

a 0.278 0.498 0.807 0.687 0.892 0.700 0.853 0.768 0.743a,b 0.310 0.549 0.767 0.661 0.734 0.724 0.861 0.402 0.452

a 0.312 0.611 0.805 0.709 0.773 0.771 0.882 0.411 0.388a 0.217 0.816 0.949 0.914 0.954 0.877 0.953 0.201 0.601a 0.385 0.701 0.855 0.787 0.871 0.802 0.904 0.376 0.495a 0.830 0.970 0.853 0.880 0.855 0.902 0.922 0.003 0.042a 0.395 0.284 0.612 0.427 0.563 0.583 0.782 0.707 0.354C 0.217 0.893 0.974 0.960 0.948 0.927 0.979 0.808 0.477C 0.278 0.495 0.807 0.686 0.896 0.699 0.855 0.784 0.747C 0.632 0.036 0.264 0.055 0.077 0.368 0.623 0.225 0.008a 0.279 0.545 0.806 0.696 0.827 0.725 0.869 0.540 0.621

a,d 0.276 0.533 0.811 0.698 0.853 0.722 0.868 0.626 0.65 1e 0.534 0.923 0.885 0.889 0.889 0.900 0.933 0.046 0.224

fg

h

------

Table E-1Spectral Matching Factors**

OtherDetectors

Photocathodes Pho- Sco-topic topic

S1 S4 S1O S11 S17 S20 S25 eye eyek k k k k k k 1 m

0.516* 0.046* 0.095* 0.060* 0.072* 0.112* 0.227* 0.071†* 0.040*0.395 0.390 0.650 0.496 0.575 0.635 0.805 0.502 0.314

0.535* 0.308* 0.388* 0.328* 0.380* 0.406* 0.547* 0.179* 0.172*

0.536* 0.236* 0.348* 0.277* 0.315* 0.360* 0.513* 0.197* 0.175*0.537* 0.520* 0.556* 0.508* 0.589* 0.581* 0.700* 0.170* 0.218*

0.533* 0.308* 0.376* 0.320* 0.375* 0.397* 0.521* 0.167* 0.159*0.512* 0.053* 0.102* 0.067* 0.080* 0.120* 0.232* 0.075* 0.044*0.504* 0.044* 0.090* 0.057* 0.069* 0.106* 0.216* 0.067* 0.038*0.500* 0.042* 0.088* 0.055* 0.068* 0.103* 0.211* 0.065* 0.037*0.493* 0.039* 0.081* 0.051* 0.062* 0.097* 0.150* 0.061* 0.034*0.401* 0.008* 0.023* 0.011* 0.014* 0.033* 0.090* 0.018* 0.007*

*Entry valid only for 300-1200-nm wavelengthinterval.

†For the total wavelength spectrum this entrywould be 0.0294.

Notes: a Registered spectral distribution.Data extrapolated as required.

b Sulfide type.c BURLE data.d Low brightness type.e Harshaw Chemical Co. data.f Standard test lamp distribution.g General Electric Co. data.h From Handbook of Geophysics.i Approximately noon sea level flux at

60° latitude.

j From Gates (Science, Vol. 151, p 523(1966)) between 300 nm and 530 nm. bet-ween 300 nm and 530 nm.A 12,000 K blackbody spectra) dis-tribution was assumed between 530 nmand 1200 nm.

k Registered spectral distribution. Dataextrapolated as required.

l Standard tabulatedphotopic visibilitydistribution.

m Standard tabulatedscotopic visibilitydistribution.

** Data from E.H. Eberhardt, 120.

147


Appendix F -Radiant Energy and Sources

Radiant energy is energy traveling in theform of electromagnetic waves. It is mea-sured in joules, ergs, or calories. The rate offlow of radiant energy is called radiant flux,and it is expressed in watts (joules persecond).

Planck’s equation for the spectral radiantexitance of a black body in a vacuum is

F-l

(in W m-2m-1) ,where

h = Planck’s constant (J s)c = velocity of light (m s -1)k = Boltzmann’s constant (J K-1)T = absolute temperature (K)

BLACK-BODY RADIATIONAs a body is raised in temperature, it first

emits radiation primarily in the invisible in-frared region. Then, as the temperature is in-creased, the radiation shifts toward theshorter wavelengths. A certain type of radia-tion called black-body radiation is used as astandard for the infrared region. Othersources may be described in terms of theblack body.

A black body is one which absorbs all inci-dent radiation; none is transmitted and noneis reflected. Because, in thermal equilibrium,thermal radiation balances absorption, itfollows that a black body is the most effi-cient thermal radiator possible. A blackbody radiates more total power and morepower at a particular wavelength than anyother thermally radiating source at the sametemperature. Although no material is ideallyblack, the equivalent of a theoretical blackbody can be achieved in the laboratory byproviding a hollow radiator with a small exit

hole. The radiation from the hole ap-proaches that from a theoretical blackradiator if the cross-sectional area of thecavity is large compared with the area of theexit hole. The characteristic of 100-per-centabsorption is achieved because any radiationentering the hole is reflected many times in-side the cavity.

The radiation distribution for a sourcewhich is not black may be calculated fromthe black-body radiation laws provided theemissivity as a function of wavelength isknown. Spectral emissivity is defined as theratio of the output of a radiator at a specificwavelength to that of a black body at thesame temperature. Tungsten sources, forwhich tables of emissivity data areavailable,121 are widely used as practicalstandards, particularly for the visible range.Tungsten radiation standards for the visiblerange are frequently given in terms of colortemperature, instead of true temperature.The color temperature of a selective radiatoris determined by comparison with a blackbody. When the outputs of the selectiveradiator and a black body are the closestpossible approximation to a perfect colormatch in the range of visual sensitivity, thecolor temperature of the selective radiator isnumerically the same as the black-body truetemperature. For a tungsten source, the rela-tive distribution of radiant energy in the visi-ble spectral range is very close to that of ablack body, although the absolute tempera-tures differ. However, the match of energydistribution becomes progressively worse inthe ultraviolet and infrared spectral regions.

TUNGSTEN SOURCESTungsten Lamps

Tungsten lamps are probably the most im-portant type of radiation source because of

their availability, reliability, and constancyof operating characteristics. Commercialphotomultiplier design has been consider-ably influenced by the characteristics of thetungsten lamp. A relative spectral-emissioncharacteristic for a tungsten lamp at 2856 Kcolor temperature is shown in Fig. F-l.

W A V E L E N G T H - N A N O M E T E R S9 2 c s - 3 2 4 7 1

Fig. F-l - Relative spectra/emission charac-teristic for a tungsten lamp at a color temper-ature of 2856 K.

As a result of the work of industry com-mittees, virtually the entire photosensitive-device industry in the United States uses thetungsten lamp at 2856 K color temperature*as a general test source. The lamp is cali-brated in lumens and is utilized in the infra-red spectrum as well as the visible. Typical aswell as maximum and minimum photosensi-tivities are quoted in microamperes perlumen.

The principal disadvantages of using thetungsten lamp as an industry standard testare that it does not provide a direct measureof radiant sensitivity as a function of wave-length and that it is a somewhat misleadingterm when the response of the photomulti-plier lies outside the visible range. To assistthe scientist in using photomultipliers,technical specifications for BURLE photomul-tiplier types include photocathode spectral-response curves which give the sensitivity inabsolute terms such as amperes per watt andquantum efficiency as a function of wave-

*Formerly 2870 K, but changed to agree with C.I.E.designated Illuminant A at 2854 K, and again morerecently to 2856 K because of the adoption of the inter-national practical temperature scale of 1968.122

Radiant Energy and Sources

length. Methods of computing the responseof a given photodetector to a particularradiation source are outlined in Appendix E,Spectral Response and Source-DetectorMatching.

The relative spectral irradiance from astandard tungsten test lamp operated at 2870K or 2854 K color temperature has been cal-culated by Engstrom and Morehead.123

Their data utilized the black body character-istic, the tungsten spectral emissivity, andthe transmission of the lamp envelope.Tabulated data (ST-3340) are available fromBURLE Application Engineering. Theirdata in the wavelength range 300 to 1200nanometers has been adjusted to a tempera-ture of 2856 K color temperature and areshown in Table F-I. The data are normalizedto unity at the maximum. The conversionwas made by multiplying by the appropriatePlanckian functions with a value for the se-cond radiation constant, C2 = 14,387.86 µmK 124.

Tungsten-Halogen LampsA variation of the tungsten lamp is the

tungsten-halogen lamp which is remarkablein that it can be operated at relatively hightemperatures with increased life and withpractically constant light output until theend of life. 125 Darkening of the envelope isvirtually prevented in these lamps by a reac-tion of the halogen gas and the evaporatedtungsten. In a typical application the lampmay contain between 0.2 and 0.3 micro-moles/per cubic centimeter of I2 which isbroken down to atomic iodine by the heat ofthe filament. The atomic iodine reacts withthe evaporated tungsten on the envelope wallto form WI,. The volatile WI2 diffuses to thefilament where it is decomposed, depositingW on the filament and freeing I to repeat thecycle.

The temperature of the envelope wall mustbe in the range 250 to 1200°C for the iodinecycle to operate successfully. It is common,therefore, for the envelopes to be made ofquartz. Because the envelopes are small,even for high power lamps, the wall attainsthe proper temperature. If the envelope wallis at too low a temperature, the WI2 will notbe desorbed and a brown deposit of WI2 willbe formed. At too high a wall temperature,the reaction, WI2 - W + 2I will occur and

149


Wavelength,nm300310320330340350360370380390

410420430440450460470480490500510520530540550560570580590600

Table F-IRelative Spectral Irradiance from aTungsten Test Lamp Operated at a

Color Temperature of 2856 KRelativespectral Wavelength,Irradiance nm

.0004 610

.0017 620

.0044 630

.0078 640

.0117 650

.0164 660

.0221 670

.0285 680

.0361 690

.0450 700

.0551 710

.0663 720

.0789 730

.0928 740

.1082 750

.1249 760

.1428 770

.1623 780

.1830 790

.2048 800

.2278 810

.2517 820

.2766 830

.3025 840

.3286 850

..3826 870

.4097 880

.4368 890

.4636 900

.4906

Relativespectral Wavelength,Irradiance nm

.5179 910

.5449 920

.5717 930

.5984 940

.6249 950

.6500 960

.6747 970

.6988 980

.7219 990

.7438 1000

.7649 1010

.7860 1020

.8054 1030

.8233 1040

.8402 1050

.8557 1060

.8719 1070

.8861 1080

.8993 1090

.9114 1100

.9228 1110

.9346 1120

.9444 1130

.9545 1140

.9624 1150

.9692 1160

.9751 1170

.9809 1180

.9862 1190

.9893 1200

= 680 lumens per watt, the luminous flux repre-sented by the tabular values is 2797 lumens. (There issome uncertainty about the value 680 lumens per watt;

relative spectral luminous efficiency values. If the max- various slightly differing values have been reported inimum luminous efficacy at 555 nanometers is recent years.)

RelativespectralIrradiance

.9916

.9945

.9965

.9977

.99901.0000.9985.9963.9961.9934.9909.9865.9823.9787.9747.9692.9647.9582.95299457.9412.9321.9257.9174.9111.9037.8948.8868.8796.8704

150


the W will not be removed from the wall.Because of the higher operating tempera-

tures and the long life with minimumenvelope darkening, the tungsten-halogenlamps are useful as standard test lamps. 126

Robert Saunders, Jr., of the OpticalRadiation Section of the National Bureau ofStandards has developed an empirical for-mula* representing the spectral irradiance of1000-watt quartz-halogen type DXW lamps.Saunders formula is

The fit of this formula to the actual data isof the order of 0.1% at each wavelength inthe range of 350 to 900 nanometers for eachof four lamps tested. Rounded-off averagesof the constants in Eq. F-2 are A=0.867;

pressed in nanometers. Ts is the apparentblack-body temperature. The term ea repre-sents a magnitude. Saunder’s data weretaken at a temperature, Ts , of approximate-ly 3025 K.

For various purposes it is useful to have atabulation of the relative radiant spectralflux from a tungsten-halogen lamp operatingat 3200 K color temperature. Saunder’s datawere used to extrapolate to this temperature.A value of Ts = 3184 K was found to providea spectral distribution most closely fitting ablack body at 3200 K and thus providing acolor temperature of 3200 K. Using a valueof 1.4388 x 107 nm K for the second radia-tion constant, C2, the relative spectral ra-diant flux values from such a lamp weredetermined from Eq. F-2 and are tabulatedin Table F-II.

ARC AND GAS-DISCHARGE SOURCES.Mercury Lamps

Of the various types of electrical dischargethat have been used as radiation sources, themercury arc is one of the most useful. The

*Private Communication

@More detailed information on arc and gas-dischargesources may be found in Handbook of Optics,sponsored by the Optical Society of America, W. G.Driscoll and W. Vaughan, editors, McGraw-Hill,1978.

0 ’4 0 0 5 0 0 6 0 0 7 0 0

W A V E L E N G T H - N A N O M E T E R S92CS-32472

Fig. F-2 - Typical spectra/emission curve fora water-cooled mercury-arc lamp at apressure of 130 atmospheres.

character of the light emitted from a mer-cury arc varies with pressure and operationconditions. At low pressure, the spectral out-put consists of sharp lines, which are veryuseful as reference spectra. Table F-III pro-vides a list of some of the mercury lines inthe visible and near-visible spectral range. Inthe case of germicidal lamps, most of theenergy radiated is in one spectral line, 253.7nanometers.

At increasing pressures, the spectral-energy distribution from the arc changesfrom the typical mercury-line spectral char-acteristic to an almost continuous spectrumof high intensity in the near-infrared, visible,and ultraviolet regions. Fig. F-2 shows thespectral-energy distribution from a water-cooled mercury arc at a pressure of 130 at-mospheres.Deuterium Lamps

These lamps provide a continuous, line-free spectrum in the ultraviolet range. Thelamps are useful in spectrophotometry andrelated applications. A typical relative spec-tral energy distribution from a deuteriumlamp is shown in Fig. F-3. Deuterium lampsmay also be obtained with calibrated spectralirradiance over the wavelength range of 180to 400 nanometers.Zirconium Concentrated-Arc Lamps

A very useful point source is the zir-conium concentrated-arc lamp. Concen-trated-arc lamps are available with ratings

These lamps may be obtained from Cenco Company.

151


Table F-IIRelative Spectral Irradiance from a Tungsten-Halogen Lamp in a

Quartz Envelope Operated at a Color Temperature of 3200 K

Wavelength,nm

300310320330

350360370380390

410420430440450460470480490500510520530540550

570580590600

RelativeSpectralIrradiance

.0109

.0151

.0204

.0269

.0347

.0440

.0548

.O671

.0811

.0967

.1139

.1327

.1531

.1749

.1981

.2224

.2480

.2745

.3018

.3299

.3585

.3875

.4169

.4463

.4757

.5048 860

.5338

.5625

.5908

.6183

.6450

Wavelength,nm610620630

650660670680690700710720730740750760770780790800810820830840850

870880890900

sent watts per 10-nanometer interval, the sum,

relative spectral luminous efficiency values. If the max-imum luminous efficacy at 555 nanometers is


Wavelength,nm


.6714

.6966

.7211

.7444

.7669

.7883

.8086

.8280

.8464

.8636

.8796

.8945

.9083

.9213

.9327

.9434

.9530

.9618

.9694

.9763

.9820

.9870

.9912

.9943

.9969

.9985

.99961.0000.9996.9985

910920930940950960970980990100010101020103010401050106010701080109011001110112011301140115011601170118011901200

.9969

.9950

.9924

.9893

.9859

.9820

.9778

.9733

.9683

.9629

.9572

.9515

.9454

.9392

.9327

.9259

.9194

.9125

.9052

.8984

.8911

.8838

.8766

.8693

.8621

.8544

.8472

.83998326.8254

=680 lumens per watt, the luminous fluxrepresented by the tabular values is 3861 lumens. (Thereis some uncertainty about the value 680 lumens perwatt; various slightly differing values have beenreported in recent years.)

152


Table F-IIISome of the Principal Spectral Lines

Characteristic of a Low-PressureMercury Discharge (in nanometers)

237.8 365.0253.7 404.7265.3 435.8296.7 546.1302.1 577.0313.2 579.0334.1

(A more complete table may be found inAmerican Institute of Physics Handbook,Third edition, 1972, McGraw-Hill, pp 7-92-7-96.)

from 2 to 300 watts, and in point diametersfrom 0.08 mm to 2.9 mm. These lamps re-quire one special circuit to provide a highstarting voltage and another well-filteredand ballasted circuit for operation.Arc Lamps

The carbon arc is a source of great inten-sity and high color temperature. A typicalenergy-distribution spectrum of a dc high-intensity arc is shown in Fig. F-4. Figs. F-5and F-6 show relative spectral-emissioncharacteristic curves for xenon and argon

DEUTERIUM L A M P

W A V E L E N G T H - N A N O M E T E R S92cs-32473

Fig. F-3 - Typical relative spectral energydistribution from a deuterium lamp. (From Op-tronic Laboratories, Inc., Silver Spring, Md.)

C A R B O N - A R C L A M P8

92cs-32474

Fig. F-4 - Typical spectra/emission curve fora dc high-intensity carbon-arc lamp.

I

1 0 0 0 1 2 0 0WAVELENGTH - N A N O M E T E R S

92cs-32475

Fig. F-5 - Typical spectralemission curve fora xenon-arc lamp.

n

W A V E L E N G T H - N A N O M E T E R S92CS-32476

Fig. F-6 - Typical spectral-emission curve foran argon-arc lamp.

153


Fluorescent LampsThe common fluorescent lamp, a very ef-

ficient light source, consists of an argon-mercury glow discharge in a glass envelopeinternally coated with a phosphor that con-verts ultraviolet radiation from the dischargeinto useful light output. There are numeroustypes of fluorescent lamps, each with a dif-ferent output spectral distribution dependingupon the phosphor and gas filling. The spec-tral response shown in Fig. F-7 is a typicalcurve for a fluorescent lamp of the daylighttype.

W A V E L E N G T H - N A N O M E T E R S9 2 c s - 3 2 4 7 7

Fig. F-7 - Typical spectra/emission curve fora daylight-type fluorescent lamp.

SUMMARY OF TYPICAL SOURCESTable F-IV provides typical parameters

for the most commonly used radiant energysources.

LASERS AND LIGHT-EMITTING DIODES

In recent years the development of varioustypes of lasers and p-n light-emitting diodeswith very high modulation frequencies andshort rise times has increased the types ofsources that photomultipliers are calledupon to detect. Although many of these in-teresting devices have their principal wave-lengths of emission in the infrared beyondthe sensitivity range of photomultipliertubes, some do not. Because of the growingimportance of laser applications and the useof photomultipliers for detecting their radia-tion, Tables F-V through F-IX are providedas reference data on crystalline, gas, and li-quid lasers, and on p-n junction light-emitting diodes.

Table F-IVSummary of Typical Sources/Parameters for the Most Commonly Used

Radiant Energy SourcesLamp Type

Mercury Short Arc(high pressure)Xenon Short ArcXenon Short Arc

Zirconium Arc 100

Vortex-StabilizedArgon ArcTungstenLightBulbsFluorescent LampStandard Warm WhiteCarbon Arc

Non-Rotating

40

2,000

Rotating 15,800Deuterium Lamp 40

DCInputPower(watts)

Arc Luminous Luminous AverageDimensions F l u x Efficiency Luminance

(mm) (lm) (lm W-1) (cd mm- 2)

200 2.5 x 1.8 9500150 1.3 x 1.0 3200

20,000 12.5 x6 1,150,000

1.5(diam.)

250

3x10---

422,00079

163021,500

- 2,560

=5x5 36,800

= 8 x 8 350,0001.0 (Nominal irradiance at 250 nm at

(diam.)

47.5 25021 30057 3000 (in 3

mmx6mm)2.5 100

17 14007.9 1016.321.5 25

64 -

18.4 175

22.2 800

154


Table F-VTypical Characteristics of a Number of

Useful Crystal Laser SystemsDopant Wavelength

of Laser

0.05% 0.6934Cr3 + 0.69290.5% 0.7009Cr3 + 0.7041

0.76701% 1.6220

Ni2 +1% 1.7500

Co2 + 1.80301% 2.6113

Co2+1% 1.0580

Nd3 + 0.91451.3392

1% 1.0460Nd3 +1.8% 1.0610Nd3 +Nd3 + 1.0648

1%Nd3 +

1%Pr3 +0.5%Pr3 +5%

Eu3 +Ho3 +0.5%Ho3 +Ho3 +

1.0633

0.5985

1.0468

0.6113

0.55122.0460

2.0975

1%Er3 +Er3 +Er3 +

Tm3+Tm3+

Tm3+Yb3+

P = Pulse

1.6120

1.61001.66021.91102.0132

1.93401.0296

Host

Al2O3

Al2O3

MgF2

MgF2

ZnF2

CaWO4

CaF2

CaMoO4

Y3Al5O12

LaF3

LaF3

CaWO4

Y2033

CaF2

CaWO4

Y3Al5O12

CaWO4

Ca(NbO3)2

Y3Al5O12CaWO4

Y3Al5O12

CW = Continuous

Mode and HighestTemperature of .Operation (K)

CW,P 350P 300P 77P 77P 300P 77

P 77P 77P 77

CW 300P 77P 300P 77

CW 300

CW 360P 440P 300

P 77

P 77

P 220

P 77P 77

CW 77P 300P 77

P 77P 77P 77

CW 77P 300

CW 77P 77

155


Host

CaF2

SrF2

CaF2

SrF2

CaF2

CaF2

CW = Continuous

MaterialActive System

Nd3+Y Al 03 5 12Nd3+Y Al 03 5 12Nd3+Y Al 03 5 12

- W 2.36 - 0.06- Hg 0.69 0.1- W 1.06 0.2

1.06 0.6- Plasma Arc 1.06 0.2

Cr3+ Na Doped HgHg

1.06 0.4

1.2 771.0 3002 30015 300

200 3000.5 30010 300

Table F-VTypical Characteristics of a Number ofUseful Crystal Laser Systems (Cont’d)

Dopant Wavelength Mode and Highestof Laser Temperature of

Operation (K)0.05 % 2.6130 P 300U3+ CW 77U3+ 2.4070 P 90

0.01 % 0.7083 P 20Sm2+0.01% 0.6969 P 4.2Sm2+0.01% 2.3588 CW 77Dy2 + P 1450.01% 1.1160 P 27Tm2+ CW 4.2

P = Pulse

Table F-VIComparison of Characteristics of Continuous

Crystalline LasersSensitizer Optical Pump Wave-

Length Eff.(%)Power Operating(Watts) Temp.(K)

Table F-VIITypical Characteristics for Two

Liquid LasersLiquid Principal Pulse Energy(J)

Wavelength (Pulse Width)

Eu(O-ClBTFA)4 DMA* 0.61175 0.1 JNd + 3:SeOCl ** * 3 2 1.056+dimethylammonium salt of tetrakis europium-ortho-chloro-benzoyltrifluoracetonate.**trivalent neodymium in selenium oxychloride.

156

Gas

Ne(ionized)Ne(unionized)He-Ne(unionized)

Xe(ionized)

Xe(unionized)

A(ionized)

N2(ionizedKr(ionized)

CO2(molecularexcitation)CF3I

H2O(molecularexcitation)CN(molecularexcitation)


Table F-VIIITypical Characteristics of a Number of Useful Gas Lasers

PrincipalWavelengths

0.3324

Output Power

0.5401

0.5944-0.61430.63281.15233.3913

0.4603-0.62710.5419-0.62710.4965 - 0.5971

2.0263.5075.5759.007

0.48800.5145

0.4545 to0.5287

0.33

0.35070.5208-0.6871

0.568210.55210.57210.5921.31527.9118118337

Typical-

-

1 mWl-5 mW<1 mW5 mW10 mW1 mW1 mW

-0.5 mW0.5 mW0.5 W0.5 W1.5 W

-----

50 W

-----

Maximum10 mW10 mW1 kW

Pulsed orContinuous

PulsedCW

Pulsed

-150 mW25 mW10 mW

-1 W1 W

10 mW1 mW5 mW5 mW5W5 W

40 W

Pulsed

CWCW

PulsedCW

PulsedCWCWCWCWCWCWCW

200 kW Pulsed100 mW CW300 mW CW

3 W CW100 mW CW

3.2 kW CW

10 kW Pulsed1.2 W Pulsed1 mW Pulsed

CW50 mW Pulsed

157


Table F-IXTypical Characteristics of p-n Junction

Light-Emitting Diodescrystal

PbSePbTeInSbPbS

In PxAs1-x)GaSbInPGaAs

CdTe(ZnxCd l-x)TeCdTe-ZnTeBPCu2Se-ZnSeZn(SexTe1-x)ZnTeGaPGaPSiC

LIGHT SOURCES FOR TESTINGMonochromatic sources of many wave-

lengths may be produced by narrow-bandfilters or monochromators. Narrow-bandfilters are more practical for productiontesting, but, at best, such tests are time-consuming and subject to error. Monochro-matic sources are not used in general-purpose testing because most applicationsinvolve broader-band light sources; amonochromatic test might grossly misrepre-sent the situation because of spectral-response variations.

A broad-band source is probably moreuseful as a single test because it tends to in-tegrate out irregularities in the spectral-response characteristic and more nearly rep-resents the typical application. The tungstenlamp has been used for many years because

Wavelength LaserAction

8.5 Yes6.5 Yes5.2 Yes4.3 Yes3.15 Yes

0.85-3.15 Yes0.91-3.15 Yes

1.6 No0.91 Yes0.90 Yes

0.80-0.90 Yes0.55-0.90 Yes

0.855 No0.59-0.83 No0.56-0.66 No

0.64 No0.40-0.63 No

0.627 No0.62 No0.565 No0.68 No

0.456 ?

it is relatively simple, stable, and inexpen-sive, and maintains its calibration. Thetungsten lamp emits a broad band of energywith relatively smooth transitions from oneend of the spectrum to the other. Its prin-cipal disadvantage as a general source is itslack of ultraviolet output and relatively lowblue output.

Sources such as arcs and glow dischargesare difficult to calibrate and show serioustime variations.

Filters are frequently used to narrow thespectral range for specific purposes; how-ever, they sometimes contribute to errorsbecause of significant transmission outsidethe band of interest. Filters are also subjectto change in transmission with time and arevery difficult to reproduce with identicalcharacteristics.

158


In many applications it is appropriate totest photomultipliers in the same manner inwhich they are to be utilized in the final ap-plication. For example, photomultipliers tobe used in scintillation counting may betested by means of an NaI:Tl crystal and a137Cs source or a simulated NaI light sourceutilizing a tungsten lamp whose light passesthrough a one-half stock-thickness CorningCS5-58 filter (5113 glass). Interference-typefilters are becoming increasingly importantin isolating specific wavelengths for testingphotomultiplier tubes for laser applications.

When a photomultiplier is manufacturedfor a variety of purposes, including scientificapplications, it would be highly desirable ifsensitivity were specified by a complete spec-tral response in terms of quantum efficiencyor radiant sensitivity. This informationcould then be utilized with the known spec-tral emission of any source to compute theresponse of the photomultiplier to thatsource. Complete spectral sensitivity data,however, are rarely provided because it isunnecessary for most practical situationsand would considerably increase devicecosts.

REFERENCES121. J.C. DeVos, “A new determination ofthe emissivity of tungsten ribbon,” Physica,Vol. 20, pp 107-131, (1954).122. “Color temperature, luminous efficacyand the international practical temperaturescale of 1968,” National Bureau of Stan-dards Technical News Bulletin, Vol. 54, No.9, pp 206-7, Sept. 1970.123. R.W. Engstrom and A.L. Morehead,“Standard test-lamp temperature forphotosensitive devices-relationship of ab-solute and luminous sensitivities,” RCARev., Vol. 28, pp 419-423 (1967).124. E.R. Cohen, CODATA Bulletin 11,Dec. 1973. See also D.G. Fink and H.W.Beaty, Standard Handbook for ElectricalEngineers, 11th Edition, McGraw-Hill,1978.125. E.G. Zubler and F.A. Mosby, “Aniodine incandescent lamp with virtually 100per cent lumen maintenance,” IlluminatingEngineering, Vol. 54, No. 12, pp 734-740,Dec. 1959.126. R. Stair, W.E. Schneider and J.K.Jackson, “A new standard of spectral irra-diance,” Appl. Opt., Vol. 2, p 1151, (1963).

159


Appendix G -Statistical Theory of Noise in Photomultiplier Tubes

Generating FunctionsIn this treatment of noise and signal-to-

noise ratio in photomultiplier operation,generating functions 127 will be used to devel-op the statistics of the different processesand their combinations. This approach hasbeen selected because it provides a general,straightforward, and relatively simplemethod of solving problems of this nature.Those not acquainted with the use ofgenerating functions may find it worthwhileto investigate some of the sources inreference127. A short summary of generatingfunctions and their use is provided in thissection. Others may use this Appendix forthe summarized expressions relating tophotomultiplier statistics.

A generating function may be defined as

and

160

The variance is defined as

Substituting G-3 and G-8 in G-6 establishesthe identity of G-6 and G-7, and thus of G-6and G-5.

Additive EventsNow, suppose there are two independent

events whose scores are to be added. An ex-ample is the roll of two dice, or in the case ofa photomultiplier, the photocathode currentgenerated thermally and photoelectrically.Assume two corresponding generating func-tions: QA(s) and QB(s). It may be shown thatthe generating function for the sum is

Statistical Theory of Noise in Photomultiplier Tubes

The case of subtraction is just a modifica-tion of the addition. In this case

G-13and, as before,

G-13a

Cascade EventsConsider a pair of independent sequential

events such as the emission of photoelec-trons and the multiplication that occurs atthe first dynode of a photomultiplier. Eachphotoelectron is multiplied by a secondaryemission factor that has an average valueand a variance. Assume two generating func-tions: A(s) for the primary event, and B(s)for the secondary or multiplying event. Itmay then be shown that the generating func-tion for the cascaded event is

Q AB(s) = A[B(s)] G-14

In the case of the cascaded event relatingto the photomultiplier, each electron fromthe preceding stage is acted on independentlyby the secondary emission gain. In contrast,consider the case of multiplication of two in-dependent scores: i.e., a die is rolled and thescore noted, then rolled again and scoresmultiplied. The order of the events is notsignificant. The generating function for oneevent may be

G-17

G-18

PHOTON NOISE

. G-22

G-25Thus;the signal-to-noise ratio of the photon

There are a few cases in which the fluctua-tion in the photon flux has an additionalterm. 128 In broad-band thermal sources,however, Eq. G-24 yields the proper expres-sion for the noise in the input photon flux.

PHOTOEMISSION NOISEThe physics of photoemission and

photocathodes is discussed in Chapter 2 onPhotomultiplier Design. In the followingdiscussion it is assumed that the time be-tween the absorption of a photon and thesubsequent emission of an electron, whenemission occurs, is short (about 10-1 2

second). In addition, it is assumed that allthe statistical processes of absorption, elec-tron transport within the photocathode, andphotoemission can be described andcharacterized by one number, the quantum

function of the photon wavelength.

161


A simple model of photoemission will aidin understanding the noise contributions ofthe photocathode. For each photon thatstrikes the photocathode, an electron is

reflection effects at the various cathode in-

The chance that no electron is emitted is

162

G-30

Thus, for quantum efficiencies less thanunity, the statistics of the photoemissionprocess result in a finite signal-to-noise ratiofor the photoelectron number even thoughthere is no assumed noise in the photonstream.

In practice, the signal-to-noise ratio of theinput photon flux is never infinite; the fluxalways contains some noise. In most sources,the signal-to-noise ratio of the photon flux isgiven by Eq. G-25.

The photoelectron signal-to-noise ratio isG-33

A quantum efficiency of 40 per cent reducesthe signal-to-noise ratio of the photoelectronflux to about 63 per cent of that of thephoton flux.

It is important to realize that the degrada-

being less than unity is irreversible in that noamount of noise-free amplification can im-prove the photoelectron signal-to-noiseratio.

In some applications, multiphoton pulsesform the input signal. In these applications,integral numbers of photoelectrons areemitted from the photocathode within a time

Statistical Theory of Noise In Photomultiplier Tubes

that is short with respect to the resolutiontime of the photomultiplier. Examples ofthis type of input can be found in radioactivetracer scintillations, such as those observedin tritium and carbon spectroscopy.

or

Eq. G-39 is just the coefficient of sr in the ex-

the chance of (m-r) failures to photoemitand the binomial coeffi-

ber of ways such an output can occur amongthe m-photon inputs.

9 2 c s - 3 2 4 8 3

G-l - Photoelectron output pulse spec-trum resulting from a flux of I-photon input

A

Fig. G-2 - Photoelectron output pulsesresulting from a flux of 4-photon input pulses

is no photoemission only 13 per cent of thetime; most of the photoemission is dividedbetween 1- and (2-, 3-, of 4-) photoelectronpulses in the ratio of 13 to 20. In sharp con-

occur 96 per cent of the time. Of the timeswhen photoelectrons are emitted, single-electron pulses occur 70 times more oftenthan any others; 3- and 4-photoelectronpulses almost never occur. It should be clearthat in multiphoton-pulse spectroscopy it is

163


The signal-to-noise ratio expected in thephotoelectron pulse distribution generatedfrom an m-photon input depends upon thequantum efficiency. From Eqs. G-34 andG-35, the relation

G-40

THERMAL EMISSION ADDED TOPHOTOEMISSION NOISE

Another source of photocathode noise isthe thermionic emission of single electrons.The strength of the emission varies withphotocathode type. In the bialkali cathode,for example, thermionic emission is virtuallyabsent; on the other hand, Ag-O-Csphotocathodes exhibit relatively large darkcurrents as a result of thermionic emission.These currents can be eliminated to some ex-tent by cooling the tube.

Randomly emitted thermionic electronsadd a term to the fluctuation in thephotoelectron current proportional to theirnumber. Because of their independence withrespect to any usual signal current, the termadds to that of the signal noise inquadrature. (See Eq. G-l1.) That is,

STATISTICS RELATED TOSECONDARY EMISSION

Although the amplification (multiplica-tion) of the photoelectron current in thedynode chain of a photomultiplier is oftenreferred to as noise-free, careful examina-tion of the statistics of the gain mechanismsinvolved shows that this statement is not en-tirely correct. Approximate noise-free opera-tion can be attained, however, with the useof proper electron optics and newlydeveloped dynode materials. In the fol-lowing discussion the statistical gain pro-cesses in the individual dynodes are ex-amined and then combined to yield thestatistical properties of the entire multiplierchain. Much of the work presented was ac-complished at a very early stage inphotomultiplier history. 129

164

For a given primary energy, it is possibleto obtain any number of secondaries ns fromzero to a maximum ns(max). The maximumnumber is given by the quotient of theprimary energy EP and the energy required toproduce a hole-electron pair within the

G-42Over many repeated measurements using

primary electrons of the same energy, a trun-cated distribution is obtained for nS. Amodel that describes the observed distribu-tions from most practical dynodes follows.

The observed number distributions forsecondary electrons vary among the dif-ferent types of dynodes used commercially.Nearly all the distributions fall within theclass limited by a Poisson distribution130 atone extreme and an exponential distributionat the other.131 To describe this wide varietyof distributions the Polya, or compoundPoisson, distribution is employed.132

Through the adjustment of one parameter,the distribution runs from purely Poisson toexponential. Therefore, this one distributioncan be used to describe and to interpret thebulk of the observed secondary-emissionstatistics.

The distribution has the following form:

where P(n,b) is the probability of observing

distribution, and b is the parameter con-trolling the shape of the distribution. Withb = 0, the distribution is Poisson, as follows:

G-44

1 distribution

G-45Fig. G.3 shows a family of distributions forvarious values of b.

The generating function for P(n,b) isgiven by

G-47


Fig. G-3 - Single-particle output distributionfor a dynode displaying Polya statistics. Avalue of b = gives an exponential distribu-tion; b = 0 gives a Poisson distribution; b = 0.2is intermediate between the two extremevalues.

Fig. G-4 shows a log-log plot of the signal-

parameter. The signal-to-noise ratio im-

distribution, but approaches unity with large

any non-zero value of b, the signal-to-noise

in Fig. G-4, even small departures from

Fig. G-4 - A comparison of the SNR as a

Polya distributions. For Poisson statistics(b = 0), the SNR increases as the square root

square root of b - for large mean gains.

Poisson statistics significantly reduce thesignal-to-noise ratio at moderately highgains of 10 to 20. It can be anticipated thatdepartures from Poisson statistics degradethe single-electron pulse-height resolution.

The Polya distribution has an interestinginterpretation with respect to secondary-emission statistics. 132 For non-zero values ofb, the distributions described by Eq. G-43can be shown to be composed of a numberof different Poisson processes, each with adifferent mean value. The mean values are,in turn, distributed according to the Laplacedistribution. When b equals 0, the distribu-tion of the mean values collapses to a delta

Poisson distribution. For b equals 1, thedistribution of the mean values is exponen-tial. The physical interpretation for a dynodedisplaying non-Poisson statistics is thatphysical non-uniformities on the dynode sur-face cause each element of the surface tohave a different mean value for emission.Although each small element exhibitsPoisson statistics with respect to emission,the total emission from the entire dynode isnon-Poisson because it comprises a distribu-tion of Poisson distributions. It is possiblethat the basic emission process from a givendynode is not a Poisson process. However,

165

high-gain GaP dynodes exhibit nearlyPoisson statistics130 and at present it isbelieved that departures from this norm arecaused, to a large extent, by dynode non-uniformities.

The departure from Poisson statistics af-fects the single-particle pulse-height resolu-tion. Fig. G-5 shows the output-pulsedistribution from a single dynode for anumber of multiple-particle inputs. Theresolution is clearly degraded in passingfrom a Poisson distribution to an exponen-tial one. With the exponential distributionshown in Fig. G-5, it would be difficult todistinguish among one-, three-, and five-particle input pulses, whereas the problemvirtually disappears for a Poisson distribu-

0

9 2 C S - 3 2 4 7 8

Fig. G-5 - A comparison of multiple-particledistributions from a single dynode havingPoisson and exponential distribution. Clearly,

the particle resolution characteristics of theexponential distribution are much poorer.

Statistics for a Series of DynodesAssume that one primary electron im-

pinging on the first dynode releases, on the

The output of the first dynode striking thesecond dynode produces an average gain at

events, the average gain and its variance maybe related to the individual dynode statisticsas follows:

G-50and

2 G-51

emission and variance for the second dynodefor a single-input electron. Continuing inthis manner, the gain and fluctuation fromthe third stage are given by

G-52and

2 G-53Eq. G-53 can be rearranged to read

Eq. G-55 states the expected results: thatthe total average gain for a series of kdynodes is the product of the secondary-emission yields of the individual dynodes inthe series. Eq. G-56 shows that the relativecontribution of any state to the total fluctua-tion decreases with the proximity of thedynodes to the output end of the chain. Thefirst stage contributes most to the totalvariance. The higher the first-stage gain, theless each subsequent stage contributes to thetotal variance. This property is an importantfeature of the high-gain GaP first-dynodephotomultipliers.

166


The signal-to-noise ratio for the multiplierchain is given by

For large first-stage gains, the multipliersignal-to-noise ratio is high. Most of thenoise contribution is from the first stage. If,in addition to a large gain, the first stage ex-hibits Poisson statistics, as explained above,the signal-to-noise ratio becomes

G-58

The noise added to the input signal is verysmall. It is in this sense that the multipli-cation chain is said to provide noise-freegain.Multiple-Particle Inputs

The output-pulse distribution formultiple-particle inputs is obtained from thegenerating function for the multiplier chain.

By an extension of Eq. G-14 for thegenerating function for a cascaded event, thegenerating functions for a chain of k dy-nodes may be obtained:

G - 5 9

The probability Pk(n) of observing n elec-trons (for a one-electron input) at the outputof a k-stage chain is derived from Eq. G-4 asfollows:

The generating function for a k-stagemultiplier chain for multiple-particle inputsis given by extension of Eq. G-9 as follows:

Qk(s,m)=[Qk(s)]m G-61The probability of observing n output elec-trons from m input particles is thereforegiven by

9 2 C S - 3 2 4 8 0

When identical dynodes are used, the output Fig. G-7 - Theoretical pulse-height distribu-distribution for single-electron input pulses tion,

evolves toward a steady-state distributionafter four or five stages, and exhibits littlechange thereafter. Fig. G-6 shows somesingle-electron distribution132 computed byuse of the Polya statistics for each stage inthe chain as explained above. As b ap-proaches 1, the distribution becomes moresharply peaked.

0 I 2 3 4

P U L S E H E I G H T

9 2 C S - 3 2 4 7 9

Fig. G-6 - Computed singleelectron distribu-tion for a range of values of parameter b.Parameter b is defined in Appendix B.

Computer values of multiple-particle out-puts for a two-stage structure are shown inFig. G-7. 131 The two curves relate to two dif-ferent structures that have the same dynodeas a first stage. The solid-line curve showsthe output when the first stage is followed by

nearly pure Poisson statistics (b = 0.01). Theoutput peaks are sharply defined, and pulsesup to a ten-electron input pulse are clearlyresolvable. The dashed line describes thefinal output distribution when the first stage

N U M B E R O F S E C O N D A R Y E L E C T R O N S I N AP U L S E F R O M D Y N O D E S U R F A C E

167


having an exponential output distribution.Individual peaks are no longer discernible;the large variance associated with the ex-ponential statistics of the second stageeliminates all the structure in the output ofthe first stage.

BURLE has developed a high-gain galliumphosphide dynode which, when used as thefirst stage in a conventional copperberyllium multiplier chain, greatly increasesthe pulse-height resolut ion of thephotomultiplier. The high-gain first stage ina photomultiplier having multiple photoelec-tron events originating from the photo-cathode is similar to the case illustrated inFig. G-7 for a multiplier where the high-gainsecond dynode amplifies the multiple pulsesoriginating from the first dynode which inturn are initiated by single electrons. Typi-cal gains for the gallium phosphide dyn-odes are 30 to 45, and their statistics arenearly Poisson. Fig. G-8 shows the multiple-particle pulse-height distribution for thetube, and Fig. G-9 shows the pulse-heightcurves for a tritium scintillation input for aconventional tube using the standardcopper-beryllium first dynode along withthat for a tube with a gallium phosphide firstdynode. The increased resolution of thegallium phosphide dynode is clearly shown.Note: The first photoelectron peak of the

8850 spectra includes dark noise fromthe photomultiplier, chemilumi-nescence and phosphorescence fromthe vial and cocktail, as well as H3disintegration.

RESOLUTION = 4 0 %

0 I 2 3 4 5 6P U L S E H E I G H T - P H O T O E L E C T R O N E Q U I V A L E N T S

92cs-32481

Fig. G-8 - Typical photoelectron pulse-heightspectrum for a photomultiplier having a GaPfirst dynode.

92CS-32482

Fig. G-Q - A comparison of the tritium scin-tillation pulse-height spectra obtained usinga conventional photomultiplier having ailCuBe dynodes and a photomultiplier having aGaP first stage. Note: The first photoelectronpeak of the 8850 spectra includes dark noisefrom the photomultiplier, chemiluminescenceand phosphorescence from the vial andcocktail, as well as 3H disintegration.

FLUCTUATIONS IN THE TUBEAS A WHOLE

In the previous sections the noise con-tributions from the photocathode and themultiplier chain were considered. Theseresults can be combined to obtain the signal-to-noise ratio for the photomultiplier as awhole.

The average number of photoelectrons

G-63The variance is given by

2 G-64

flux of photons displays Poisson statistics. If

process, Eq. G-l6 must be used to obtain2

Using these expressions to describe the in-put to the photomultiplier chain, the averagenumber of electrons collected at the anodecan be stated as follows:

G-65

average gain of a k-stage multiplier. Thevariance for the output electron stream isgiven by

168


variance in the average gain of a k-stagemultiplier chain.

Eq. G-66 can be rearranged as follows:G-67

For equal-gain stages described by Poissonstatistics in the multiplier chain, Eq. G-56becomes

G-68

In this case

G-69

G-70

much less than 1 and hence thatG-7 1

The signal-to-noise ratio at the anode isgiven by

G-72

For high-gain dynodes exhibiting Poissonstatistics, therefore, SNRa is essentially thatof the photoelectrons, SNRp.e., as given inEq. G-33.

In a photomultiplier in which the dynodegain is not high but still exhibits Poissonstatistics, SNRa is given by

G-73

decreased by a factor of 0.87 from its value

changes the degradation factor to 0.94. Fur-

very much.In the case of fully exponential dynode

statistics, the variance for each dynode in achain of identical dynodes is given by Eq.G-48; i.e.,

G-74

total anode fluctuation then becomes

For large dynode gains,G-76

Even for large dynode gains, exponential

drop in SNR, is accompanied by a severe lossin single- and multiple-electron pulse-heightresolution, as shown in Fig. G-7. To resolvesingle-photoelectron pulses, the multiplierchain must exhibit both high gain and good(i.e., Poisson) statistics.

A significant improvement in SNR, resultswhen the photocathode quantum efficiency

0.7 would improve SNRa by a factor of 1.4.

would have an SNR, equal to the square root

the ideal SNRa by a factor of 0.59. Con-siderable improvement can be expected withthe development of photocathode materialsof increased sensitivity.

The application of these equations to pre-sent photomultipliers indicates that theavailable photocathode quantum efficiencyis the principal degrading influence on

the multiple-photon input-pulse resolution

greater than 6) do not significantly degradethe input signal-to-noise ratio of photoelec-trons provided the dynode statistics arenearly Poisson.

OTHER SOURCES OF NOISEIN PHOTOMULTIPLIER TUBES

Within a photomultiplier there are sourcesof noise that are not associated directly withthe processes of photoelectric conversionand electron multiplication. These sourcescan, in general, be separated into twogroups: (1) those that are not correlatedwith and (2) those that are correlated withthe signal pulse.Non-Correlated Noise Sources

The materials used to fabricate the in-ternal structure and the glass envelope of a

169


photomultiplier may contain amounts ofcertain radioactive elements that decay andgive off gamma rays or other high-energyparticles. If one of these emitted particlesstrikes the photocathode or one of the firstfew dynodes, it will produce an anode darkpulse. The size of the anode pulse may beequivalent to one or more photoelectronsemitted at the photocathode. These pulsesare randomly emitted. In tube manufacturethis type of emission is minimized throughthe careful selection of materials.

Analytically, the fluctuations resultingfrom non-correlated random sources add inquadrature to those of the signal. Becausethey are random, the dark-pulse variances

-

The total anode variances are given by:

are the average single-photoelectron and the average single-dark-emitted electron numbers observed in a time

anode signal-to-noise ratio. Again, in thoseinstances where the incoming radiationsignal comprises more than one photon,coincidence techniques can be employed toreduce the effect of randomly emitted elec-trons originating at the cathode.

Electrons may originate at dynodes wellalong in the multiplier chain. At the anode,these electrons appear as fractional-photoelectron pulses. Such pulses also resultfrom interstage skipping, generally near thebeginning of the chain. With good statisticsin the chain, the fractional pulses may bediscriminated against because the single-electron peak in the pulse-height spectrumstands out sharply.

Correlated Noise SourcesA gas atom or molecule within the

photomultiplier may be ionized by aphotoelectron pulse, This ionization mayoccur at the first-dynode surface or in thevacuum between the photocathode and the

170

first dynode. The positive ion thus createdtravels backward to the cathode where itmay release one or more electrons from thephotocathode. Because there is a time delayin the ion-emitted electron pulse equal to thetime of flight of the ion to the photocathode,the resulting pulse, usually referred to as anafterpulse, occurs after the true signal pulseat the anode. Afterpulses are caused mainlyby hydrogen ions and their occurrence canbe minimized in tube processing.

Primary electrons produce photons as wellas secondary electrons within the dynodes ofthe multiplier chain. Despite the low effi-ciency of this process, some of the emittedphotons may eventually reach the photo-cathode and release additional electrons. Atime delay is observed corresponding to thetransit time for the regenerated electronpulse to reach the point of origin of the light.Depending upon the type of dynodemultiplier cage, this time may be of the orderof 20 nanoseconds. Most of the photonscomprising the light feedback originate inthe region of the last few dynodes or of theanode.

If the voltage across the tube is increased,the dark-pulse rate also increases and usuallyproduces some observable light near theanode region, a fraction of which is fed backto the photocathode. The result of thispositive feedback. is that, at a certainvoltage, the photomultiplier becomesunstable and allows the output dark-pulserate to increase to an intolerably high level.The voltage at which this increase occurs isgenerally above the recommended maximumoperating voltage.

Not every signal pulse initiates an after-pulse. Therefore, coincidence techniques,using more than one photomultiplier, can beemployed in some instances to eliminate thissource of noise as well as the uncorrelatedsources discussed above.

If correlated noise sources cannot beeliminated by time discrimination or othermeans, an analytical treatment for the totalvariance would follow the form given by Eq.G-12.

NOISE AND THE BANDWIDTHOF THE OBSERVATION

At high counting rates, noise calculationsare performed with the average and variance


of the photoelectron current rather than withindividual photoelectron pulses. The expres-sions which have been developed for signal-to-noise ratio by consideration of the

variance from average may be converted toexpressions of signal-to-noise ratio involvingcurrents and bandwidth, B, by consideringthe reciprocal nature of the observation timeand the bandwidth.

Noise equivalent bandwidth B may bedefined133 as follows:

G-79

transfer response of the circuit, and Am is

"circuit” in this case counts pulses in a time

Laplace transform of the impulse response.The impulse response is a rectangular pulse

case,

and Am is found to be equal to 7. The noiseequivalent bandwidth is then readily foundto be

G-81

It is of interest to compare this relation withthe equivalent noise bandwidth for exponen-tial impulse response as in an RC circuit; inthis case

B=1/4RC G-82

From Eq. G-25, the SNR for a randomphoton flux is given by

G-83

where I, is the average photon arrival rateand 7 is the time interval of the count. When

In the case of the photocathode electron cur-rent,

where i is the photocathode emission currentin amperes and e is the charge of the elec-tron. If the signal current is considered as i,the noise in the bandwidth B for thephotocurrent is the familiar shot noiseformula (2eiB)1/2.

The equation for the photon noise squaredis given by

G-86

The photocurrent noise squared is given by

Both these equations involve the photon“current” I,.

If the photons are not randomly emitted,Eq. G-86 must be modified. In the case inwhich the variance in the photon current is

For a randomly emitted photon, the noisecurrent squared at the anode is given by

where the multiplier chain is assumed to becomposed of k high-gain dynodes exhibitingPoisson statistics, each with an average gain

At the anode, the input resistance and ca-pacitance of a preamplifier generate a noisecurrent squared given by

where g = l/R is the shunt conductance inthe anode lead, C is the shunt capacitance, Bis the bandwidth, k is Boltzmann’s constant,T is the absolute temperature, and Rn is theequivalent noise resistance of the preamplifi-er input. The total noise current squared


The value of I, shown in Eq. G-92 is thelower limit for the average photon currentand makes the squared anode-current fluc-tuation greater than the squared noise cur-rent in the photomultiplier preamplifier in-put by a factor of ten. The resulting value ofthe average photon current corresponds to aphotoelectric current of about 10-14 am-pere, or about 6 x 104 photoelectrons persecond. In cases in which the dark current iseffectively higher than 6 x 104 photoelec-trons per second, the photomultiplier sen-sitivity limit is set by the dark current andnot by the preamplifier noise. It is the noise-free gain of the multiplier chain which in-creases the rms photocurrent shot noise by a

operation. See also Fig. 65 in Chapter 4,Photomultiplier Characteristics.

PULSE COUNTING STATISTICS*In utilizing a photomultiplier in a

“photon” counting mode, individual anodepulses initiated by electrons from thephotocathode are counted. Because thecount at very low light levels originates fromboth photoemission and thermionic emis-sion, two separate counts are required: onein the light and one in the dark. In deter-mining the photoelectron count rate, the

*Measured signal-to-noise ratios are similar for pulse-counting or for current -measurement techniques.134

Baum135 has shown that in some cases, pulse countingcan have an advantage of the equivalent of a factor of1.2 in quantum efficiency.

statistics of the measurement are improvedby increasing the count time. The presentdiscussion recommends the optimum timedivision between the dark and the lightcounts.

G-93

G-94

The variances for the light and dark mea-surements are given by

G-95

G-96

The final “signal” which is a measure of I,is given by

G-97

Assume that the times of the counts are ac-curately determined so that the respective

use of Eq. G-21 for multiplication (by thereciprocal of the count times), the varianceof each term in G-97 is then

G-98

G-99

For the variance of a difference, using Eq.G-11,

G-100

The signal-to-noise ratio in the determina-tion of I, is then given by

G-101

172


to be minimized with respect to the division

nd from G-93 and G-94:

G-102

sidering the other quantities as invariant.

imum SNR condition:

G-103

The implication of Eq. G-103 is asfollows: For example, if the signal count ismuch less than the dark count, about equaltimes should be spent for both readings. Ifthe signal count and dark count are about

PULSE-HEIGHT RESOLUTION INSCINTILLATION COUNTING

In scintillation counting, for the case of aK2CsSb photocathode and a NaI:Tl scin-tillator, the typical photoemission yield isapproximately 8 photoelectrons per keV ofgamma-ray energy absorbed in the crystal-or 125 eV per photoelectron. The peak of theemission spectrum for NaI:Tl is approx-imately at 415 nm which corresponds to aphoton energy of 3 eV. The quantum effi-ciency of the bialkali photocathode for theblue spectrum of the scintillator is approx-imately 25 per cent. If the crystal were 100per cent efficient in converting gamma-rayenergy to light energy, one would expect aphotoelectron for every 12 eV. Thus, theconversion efficiency of the crystal is about10 per cent.

The statistics of the pulse-height distribu-tion, therefore, involve both the crystal andthe photomultiplier processes. On the otherhand, if the conversion efficiency of thecrystal were 100 per cent, the photoelectricconversion of the gamma-ray would yieldessentially a constant number of photons perconversion and the crystal would then notcontribute to the statistical process.

Let the average number of photons exitingfrom the scintillator onto the photocathodeper photoelectrically converted gamma-ray

average quantum efficiency of the photo-

emission process is as before (Eq. G-28):

G-104

Now, using G-15 and G-16 for a cascadedevent, the average number of photoelectronsper pulse is

G-105

and the variance in thetrons per pulse is

number of photoelec-

It is instructive to consider the signal-to-noise ratio at each stage of the photo-multiplier. For the photocurrent in the pulse,using G-105 and G-106:

For the current pulse leaving the firstdynode, assuming a secondary emission

statistics), again using G-15 and G-16:

Variance in this number =

SNR (pulse out of first dynode) =

G-108

Similarly, for the pulse out of the seconddynode:

Variance in this number =


SNR (pulse out of second dynode) =

If a tube having k stages of equal secon-

G-l10

In pulse-height resolution measurements itis common to refer to the Full-Width-Half-Maximum (FWHM) which may be related tothe SR as follows:

FWHM =2.355/SNR

G-111

The first term in the brackets represents therelative variance of the photomultiplier con-tribution and the second term is the relativevariance of the crystal contribution.

the number of photonsper pulse. In this case,

Actually, the crystal statistics are worse thanthe above assumption. This question isdiscussed in more detail in Chapter 4,Photomultiplier Characteristics in the sec-tion on “Pulse Counting.”

SUMMARYThe more important expressions relating

to signal and noise discussed in this Appen-dix are summarized below. Beginning at theinput to the photomultiplier, the signal isfollowed through the tube, and the varianceand the SNR associated with the signal arenoted.

Photon flux: Conditions: random emis-sion, Poisson statistics. Terminology: av-

Dark emission from a photocathode: Ter-minology: average number of thermionic

Combined dark and photoemission fromthe photocathode: Terminology: average

Photoemission determined from totalemission: Conditions: dark emission de-

so that the variance in the measurement isnegligible; dark emission to be subtractedfrom total emission to determine photoemis-sion .

174


(Also, see section above on Pulse CountingStatistics for cases in which the dark emis-sion count is not determined over a long timeinterval.)

Multiplier chain with single electron input:

2

Photomultiplier tube: Terminology: av-erage number of electrons collected at the

Photomultiplier tube: Conditions: Eachstage is identical and has Poisson statistics.

Photomultiplier tube: Conditions: Eachstage identical with exponential statistics.

gain dynodes assumed, high photon-counting rates, output circuit noise contribu-tions included. Terminology: charge on theelectron, e; bandwidth, B; Boltzmann’s con-stant, k; temperature, T, (degrees Kelvin);equivalent noise resistance of the input ofthe preamplifier that processes the anodesignal, Rn; shunt conductance in the anodelead, g; shunt capacitance in the anode lead,C.

Photomultiplier tube: Conditions: high-gain dynodes assumed, gain and output cur-rent sufficient so that circuit noise com-ponents can be neglected. Terminology:photocathode emission current, i.

Photomultiplier tube and scintillator:Conditions: scintillations resulting fromgamma-ray photoelectric excitation in thecrystal; equal gain per stage assumed. Ter-minology: full width half maximum,FWHM; photocathode quantum efficiency,

average number of photons incident on the

variance in the number of photons in each

FWHM = 2.355/SNR

The first term inside the brackets is therelative variance associated with the photo-multiplier; the second term is the relativevariance associated with the scintillator.


REFERENCES127. T. Jorgensen, “On Probability Gener-ating Functions,” Am. J. Phys., Vol. 16, p285 (1948); E. Breitenberger , “ScintillationSpectrometer Statistics,” Prog. Nuc. Phys.,Vol. 4, (Ed. O.R. Frisch, Pergamon Press,London, p. 56 (1955); William Feller, An In-troduction to Probability Theory and Its Ap-plications, Vol. 1, John Wiley and Sons,Third Edition (1968).128. R.H. Brown and R.Q. Twiss, “Inter-ferometry of the intensity fluctuations inlight,” Proc. Royal Soc. London, Vol.243A, p 291 (1958).129. W. Shockley and J.R. Pierce, “Atheory of noise for electron multipliers,”Proc. IRE, Vol. 26, p 321 (1938); V.K.Zworkykin, G.A. Morton, L. Malter, “Thesecondary emission multiplier-a new elec-tron device,” Proc. IRE, Vol. 24, p 351(1936).130. G.A. Morton, H.M. Smith, and H.R.Krall, “Pulse height resolution of high gain

first dynode photomultipliers,” Appl. Phys.Lett., Vol. 13, p 356 (1968).131. L.A. Dietz, L.R. Hanrahan and A.B.Hance, “Single-electron response of aporous KCl transmission dynode and appli-cation of Polya statistics to particle countingin an electron multiplier,” Rev. Sci. Inst.,Vol. 38, p 176 (1967).132. J.R. Prescott, Nuc. Instr. Methods,“A statistical model for photomultipliersingle-electron statistics,” Vol. 39, p 173(1966).133. M. Schwartz, Information Transmis-sion, Modulation, and Noise, McGraw-Hill,p. 207 (1959).134. F. Robben, “Noise in the measure-ment of light with photomultipliers,” Appl.Opt., Vol. 10, No. 4, pp 776-796, (1971).135. W.A. Baum, “The detection andmeasurement of faint astronomical sources,”Astronomical Techniques, Edited by W.A.Hiltner, The University of Chicago Press,(1962).

176

Index

Index

A bsorptance . . . . . . . . . . . . . . . . . . . . . . . . . 18,125Absorption coefficient . . . . . . . . . . . . . . . . . . . . . 18Acceptor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125Acceptor levels . . . . . . . . . . . . . . . . . . . . . . . . . . . 14After pulse . . . . . . . . . . . . . . . . . . . . . . . . . . . 57, 125Aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 58, 75Air pressure and photomultiplier operation . . . . 76Angle of incidence and photoemission . . . . . 39,92Angle of polarization and photoemission . . . 39,40Angstrom unit . . . . . . . . . . . . . . . . . . . . . . . . . . . 125Anode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3,34,125Anode current maximum . . . . . . . . . . . . . . . . . . . 50Anode pulse current maximum . . . . . . . . . . . . . .47Anticoincidence circuit. . . . . . . . . . . . . . . . . . . .125Applications list. . . . . . . . . . . . . . . . . . Appendix AArgon arc lamp. . . . . . . . . . . . . . . . . . . . . . . . . . .153Astronomy, application to . . . . . . . . . . . . . . .5,112Avalanche photodiodes . . . . . . . . . . . . . . . . . . . . .8

B ackground counts . . . . . . . . . . . . . . . . . . . . ..12 5Band bending . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14Bandwidth and noise spectrum . .59,125,170,171Basing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...88Beta radiation, effect on photomultiplier . . . . . .44Black-body radiation . . . . . . . . . . . . . . . . . . . . . .148B l e e d e r . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 5Box-andgrid dynode structure . . . . . . . . . . . . . .28

Ca g e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 5Calibration, standard lamp. . . . . . . . . . . . . . . . . .92Candela . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125,139Capacitors, by pass. . . . . . . . . . . . . . . . . . . . . . . .84Carbon-arc lamp . . . . . . . . . . . . . . . . . . . . . . . . .153CAT scanner . . . . . . . . . . . . . . . . . . See CT-scannerCerencov radiation . . . . . . . . . . . .19,57,92,97,126Channel multiplier . . . . . . . . . . . . . . . . . . . . .32,126Channel number . . . . . . . . . . . . . . . . . . . . . . . . .126Circular cage . . . . . . . . . . . . . . . . . . . . . . . . . . .4,27Coincidence counter., . . . . . . . . . . . . . . . . . . . . .72Collection efficiency . . . . . . . . . . . . . . . . . . . . . . . 27Collection uniformity . . . . . . . . . . . . . . . . . . . . . .41Color temperature . . . . . . . . . . . . . . . . . . . .126,139Color-temperature standard. . . . . . . . . . . . . . . .149Compton effect . . . . . . . . . . . . . . . . . . . . . . . . . . .93Conduction band . . . . . . . . . . . . . . . . . . . . . . . . . 126Connections, terminal . . . . . . . . . . . . . . . . . . . . . .88Constant-fraction triggering . . . . . . . . . . . . . . . .99Cooling photomultipliers . . . . . . . . . . . . . . . . . . .74Count . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..12 6Counting efficiency . . . . . . . . . . . . . . . . . . . . . . .126Count-rate stability . . . . . . . . . . . . . . . . . . . . .52,96Crossed-field photomultiplier . . . . . . . . . . . . . . .33Cross ta l k . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 6Cross talk in liquid scintillation counting . . . . . .73Cryostats for photomultipliers. . . . . . . . . . . . . . .54CT-scanner . . . . . . . . . . . . . . . . . . . . . .6,7,102,125C u r i e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 6Current-voltage characteristic . . . . . . . . . . . . . . .36

Darkcurrent . . . . . . . . . . . . . . . . . . . . . .4,8,16,126Dark current and noise . . . . . . . . . . . . . . . . . . . . .54Dark current and temperature . . . . . . . . . . . . . . . .8Dark current, increase following exposure

to fluorescent lamps . . . . . . . . . . . . . . . .43,58Dark current, sources of . . . . . . . . . . . . . . . . .19,53Dark noise pulse spectrum . . . . . . . . . . . . . . . . . .59Dark noise reduction with cooling . . . . . .50,54,58Degaussing . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..46Delayline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..12 6Delta function . . . . . . . . . . . . . . . . . . . . . .59,63,126Densitometry . . . . . . . . . . . . . . . . . . . . . . . . . . . .105Detectivity . . . . . . . . . . . . . . . . . . . . . . . . .56,90, 126Deuterium lamp . . . . . . . . . . . . . . . . . . . . . . . . . .151Differential cooling . . . . . . . . . . . . . . . . . . . . . . .74Discriminator, pulse height . . . . . . . . . . . . . . . 126Discriminator setting . . . . . . . . . . . . . . . . . . . . . . . 68Donor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126Dynamic compression of output signal . . . . . . .87Dynode.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Dynode, Cs-Sb . . . . . . . . . . . . . . . . . . . . . . . . . . . .96Dynode, Cu-Be . . . . . . . . . . . . . . . . . . . . . . . . . ...98Dynode, GaP:Cs . . . . . . . . . . . . . . . . . . . . . . . .68,83Dynode glow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56Dynode voltage control . . . . . . . . . . . . . . . . . . . . .83

EADCI (Equivalent Anode DarkCurrent Input). . . . . . . . . . . . . . . . . . . .

E2B ..55, 127

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Einstein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10Electron affinity . . . . . . . . . . . . . . . . . . . . . . .13,127Electron multiplier. . . . . . . . . . . . . . . . . . . . . . . .127Electron optics . . . . . . . . . . . . . . . . . . . . . . . . .26-35Electron resolution (See also Pulse-height

resolution, single electron) . . . . . . . . . . . . .127Electron volt . . . . . . . . . . . . . . . . . . . . . . . . . .2O, 127Electrostatic focus . . . . . . . . . . . . . . . . . . . . . . . . .4Energy diagram . . . . . . . . . . . . . . . . . .10, 13,14,94Energy distribution, photoelectrons . . . . . . . . . .12ENI (Equivalent Noise Input) . . . . . . . . .8,9,55,127Environmental effects . . . . . . . . . . . . . . . . . . . . . .74Environment, high humidity . . . . . . . . . . . . . . . . .76Environment, pressure . . . . . . . . . . . . . . . . . . . . .76ERMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15,16,127Escape depth . . . . . . . . . . . . . . . . . . . . . . . . . . 13,18

Fall time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62,127Fatigue . . . . . . . . . . . . . . . . . . . . . . . . . . . 50,51,127Fatigue and dynode materials . . . . . . . . . . . . . . .51Feed back. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Fermi-Dirac energy distribution function . . 11,13Fermilevel . . . . . . . . . . . . . . . . . . . . . .10,11,13,127Fiber-optic communication systems. . . . . . . . . . .8Field emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60Filters, narrow bandpass color . . . . . . . . . .92,158Filters, neutral density . . . . . . . . . . . . . . . . . . . . .91Fluorescent lamp . . . . . . . . . . . . . . . . . . . . . . . . .154

177


Fluorometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106Flying-spot scanner. . . . . . . . . . . . . . . . . . .115,127Focussing electrode . . . . . . . . . . . . . . . . . . . . . .127Focussing-electrode voltage . . . . . . . . . . . . . . . .42Foot candle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127Foot Lambert . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127Forbidden band . . . . . . . . . . . . . . . . . . . . . . .13,127FWHM (Full Width at Half Maximum) . .63,64,127

G ain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3,9,127Gain control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83Gain, maximum . . . . . . . . . . . . . . . . . . . . . . . . . . . 75Gain, variation with temperature . . . . . . . . . . . . .49Gain, variation with voltage . . . . . . . . . . . . . . . . .44Gamma radiation and glass discoloration . . . . .43Gamma-ray camera . . . . . . . . . . . . . . . . .6,100,128Generating functions . . . . . . . . . . . . . . . . . . . . .160Glass, browning of . . . . . . . . . . . . . . . . . . . . . .19,77Glass charging effects . . . . . . . . . . . . . . . . . . . . .56Glass, properties of. . . . . . . . . . . . . . . . . . . . . . . .19Glass transmission . . . . . . . . . . . . . . . . . . . . . . . .19Ground potential . . . . . . . . . . . . . . . . . . . . . . . . . . 75

H eadlight dimmer . . . . . . . . . . . . . . . . . . . . . . . . . 6Helium penetration . . . . . . . . . . . . . . . . . .57,76,90Hertz (Hz) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128History of photomultiplier development . . . . . .3-6Hofstadter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5“Hysteresis” (cyclic instability) . . . . . . .47,52,127

Illuminance . . . . . . . . . . . . . . . . . . . . . . . . .128,140Illuminated area, photocathode. . . . . . . . . . . . . .92Insulator charging . . . . . . . . . . . . . . . . . . . . . .47,52IPA’s (Integrated Photodetection Assemblies) .80Irradiance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..13 2Johnson noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

L ambert’s cosine law . . . . . . . . . . . . . . . . .128,141Laplace’s equation . . . . . . . . . . . . . . . . . . . . . . . .27Laser range finding . . . . . . . . . . . . . . . . . . . . .8,113Lasers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155-157Lasers, Nd:YAG . . . . . . . . . . . . . . . . . . . . . . . . . . . .8Leading edge timing . . . . . . . . . . . . . . . . . . . . . ..9 6Leakage, ohmic . . . . . . . . . . . . . . . . . . . . . . . . . . .53L i fe expec tancy . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Light chopper. . . . . . . . . . . . . . . . . . . . . . . . . . . . .92Light emitting diodes, (LED’s) . . . . . . . . . . . .8,158Light level for photomultiplier use. . . . . . . . .89,90Light level reduction methods . . . . . . . . . . . .90,91Light pipe . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95,128Light shielding . . . . . . . . . . . . . . . . . . . . . . . . . . . .75Light sources for testing. . . . . . . . . . . . . . . . . . .158L i n e a r i t y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 7Linearity and photocathode resistivity . . . . . . . .47Linearity and voltage divider current. . . . . . . . . .82Linearity measurement. . . . . . . . . . . . . . . . . . . . .49Linearity, pulse measurement of . . . . . . . . . . . . .86Liquid scintillation counting . . . .69,72,87,97,128Lithium fluoride . . . . . . . . . . . . . . . . . . . . . . . . . . .19L u m e n ( l m ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 8Luminance.. . . . . . . . . . . . . . . . . . . . . . . ..128.14 0Luminous efficacy . . . . . . . . . . . . . . . . . . . . . . . .128Luminous Intensity . . . . . . . . . . . . . . . . . . .128,129L u x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 8

178

Index

179


180

Documents

Photo Multiplier Handbook