PHYSICAL RE VIE% B VOLUME 15, NUMBER 4 15 F EBRUAR Y 1977

Low-temperature photoconductivity in Sro ~T

Jeanette Feldott and Geoffrey P. SummersDepartment of Physics, Oklahoma State University, Stillwater, Oklahoma 74074

{Received 12 April 1976; revised manuscript received 18 June 1976)

Photoconductivity measurements have been made on single-crystal samples of SrO which were "as received, "irradiated at room temperature with y rays, or irradiated at temperatures down to 77 K with 1.S-MeV

electrons or 1.S-MeV protons. The measurements were made over the spectral range from 2 to 6.2 eV and attemperatures from 63 to 180 K. F+ centers introduced into the samples by particle irradiation produce aphotocurrent which measurement indicates is carried by electrons. The free-electron yield from F centers

increased sharply between 63 and 100 K from which it is estimated, using a two-level model, that the thermal

activation energy for release of free electrons is 0.12 +0,01 eV. The two-level model also gives a value ofS X 10' for the ratio 7„/'TD where 1/v0 is the pre-exponential frequency factor for thermal ionization and 7~ is atemperature-independent lifetime for decay from a relaxed excited state of the F+ center. As the incident

photon energy increased towards the band edge there was a sharp rise in the photoresponse in both irradiated

and unirradiated samples. The photoresponse reached a maximum at an energy of S.4 eV, and then decreased

sharply as the incident photon energy increased further.

I. INTRODUCTION

Recently, the alkaline-earth oxides, and in par-ticular, simple defects in these materials, havebeen the subject of many studies. "' %e are con-cerned in this paper with low-temperature photo-conductivity and associated optical-absorptionmeasurements made on SrQ over the spectralrange from 2 eV to the band-gap energy at about6 eV. The temperature dependence and spectralresponse of photoconductivity give a sensitivemethod of investigating both the electronic struc-ture of defect centers and the band edge of a crys-tal, ' and we shall be concerned here with both ofthese aspects.

SrO has the rocksalt structure characteristicof alkali halides. The doubly charged oxygen va-cancies which are produced when currently avail-able SrQ is irradiated with either protons or neu-trons each trap, in general, only a single electronto form E' centers. The vibronic properties ofthe non-Gaussian and asymmetric E' absorptionband, which is located at 3.1 eV, have been in-vestigated by Hughes and %ebb. ' A temperature-dependent luminescence band centered near 2. 5

eV has also been reported for the E' center. "The two-electron center, the F center, has beenproduced by additive coloration of SrQ with excessstrontium. ' The F band is reported near 2.5 eV.

On the basis of charge considerations alone itmight be expected that E centers in alkaline-earthoxides would produce photoconductivity in a man-ner similar to E centers in alkali halides but thatE' centers, being positively charged, would notproduce photoconductivity at ordinary tempera-tures. The experimental results reported so far

indicate, however, that the situation is not as .

straightforward as this viewpoint suggests. InMgO the E band and F'-band overlap at 5.0 eV.Roberts and Crawford' reported that y irradiationjust prior to mounting a sample in the cryostatwas necessary to detect photoconductivity in MgQwhich had been irradiated with electrons or neu-trons. It was tentatively concluded from thesemeasurements and others made on additivelycolored samples that the photocurrent originatedfrom E centers. However, as Roberts and Craw-ford have indicated, and as we shall see below inSec. IV, the magnitude of the photocurrent from acenter depends on the product of the quantum ef-ficiency for the production of free carriers andtheir mean range once released. If either of theseparameters increases the photocurrent will in-crease. From this viewpoint photocurrent origi-nating from F' centers in MgQ cannot yet be ruledout. A similar uncertainty exists for BaQ. Roseand Hensley' have shown recently that the E-bandand E'-band overlap in this material also, so thatthe origin of a photoconductivity band observed byDash" at 2.0 eV in x-irradiated BaQ is still un-certain. Qn the other hand in both CaQ and SrOthe E band and the E' band are separated in en-ergy, and it is possible to investigate each bandindividually. Measurements" on electron-ir-radiated CaO have shown that both E centers andE' centers produce photoconductivity in this ma-terial. Recent experimental results" have alsoshown that E' centers in SrQ can produce photo-conductivity in a way very reminiscent of E centersin alkali halides. A more complete report of theexperiments on SrQ is given here including an ex-tension of the spectral range of the measurements.

2295

2296 JEANETTE FELDOTT AND GEOFFREY P. SUMMERS

II. PHOTOCONDUCTIVITY MEASUREMENTS

Low -temperature photoconductivity measure-ments were made using a cryostat developed fromone described previously. " Crystals were heldbetween two plane parallel electrodes 8 mm longand 3 mm wide which were located in a copperchamber bolted to the tail of the cryostat. Dryhelium gas could be admitted to the chamber toproduce good thermal contact between the sampleand the cryostat. Temperatures were measuredwith a copper -constantan ther moeouple. Incidentlight fell on the crystal through the front electrodewhich was made of fine phosphor-bronze wiregauze. The back electrode was made of copperfoil and was insulated from ground with a thinsapphire plate. This electrode was eonneeted tothe input of a Cary 401 vibrating-reed electrome-ter which was used in the "rate-of-charge" mode.Both electrodes were blocked with thin sapphireplates to prevent charge entering or leaving thecrystal. The output of the electrometer was fedto a potentiometric recorder. With the systemdescribed above the background-noise level waswell below 10 "A. Measured photocurrents weretypically about 10 "A and individual measurementswere made with the crystal illuminated for a fewseconds at a time. The direction of the appliedelectric field, which was produced by dry cells,could be easily reversed after individual mea-surements to prevent polarization effects in thesample.

The light source used was either a xenon lampor, for ultraviolet measurements, a deuteriumlamp. Light from these lamps was dispersed bya MePherson 218 monochromator with a linear dis-persion of 2.3 nm mm '. The slits were set 1.5mm or 2 mm wide. A front-silvered mirror fo-cused light from the monochromator on the crys-tal sample. A fraction of the light intensity fall-ing on a sample was separated using a MgF, beamsplitter and was monitored by a 62563 photomulti-plier. The photomultiplier was calibrated in aseparate experiment by placing a Molectron pyro-electric radiometer at the sample position. Theradiometer had a calibration traceable to the Natl.Bur. of Stds. (U.S.). It is estimated that the abso-lute measurement of the light intensity falling ona sample could be made to about 8%.

polished on alumina-impregnated lapping disks,which removed surface contamination and whichgave the surfaces a good optical polish. Qpticalabsorption spectra were taken with a Cary 14spectrophotomete r.

The optical density of a typical unirradiatedcrystal at 77 K was similar to that reported byBessent etal. 4 The optical density increased fromabout 0.1 at 700 nm to over 2 at 5.5 eV. Therewas no obvious absorption peaks in the spectrumexcept for a small band which could be sometimesdetected at about 5.7 eV and which seemed to cor-respond to an exciton band. "

The crystals could be irradiated at 77, 195, or300 K with either 1.5-MeV protons or 1.5-MeVelectrons. We show in Fig. 1 the absorption spec-trum at 77 K of a crystal which had been irradiatedat about 195 K with 1.5-MeV protons to a dose ofabout 2 && 10"protons cm '. The irradiation pro-duced the asymmetric E' band with a peak at 3.1eV, which has been seen previously. "' It can beseen that although the optical density at the peakof the F' band is about 0.7, there is no evidenceof the F band at 2. 5 eV. This behavior is in con-trast to that observed in CaO, where the F bandwould be easily observed at this stage. Protonirradiation at 77 K seemed to introduce the E'band at about twice the rate of irradiation at 195K. We did not find it possible, however, to pro-duce a measurable F' absorption band in samplesirradiated with 1.5-MeV electrons at 77 K even upto doses of 5 && 10" electrons cm '. I,ow-incidentcurrent densities were always used to prevent thesamples being heated during irradiation. In these

WAVE LENGTH (nm)600 500 400 300

2.0

1.5M

LU

O

c 1.0O

0

0.5

III, SAMPLE PREPARATION 02

(

3 4 5 6ENERGY(eV)

Crystals of nominally 99.95%%d-pure SrO were ob-tained from W. and C. Spicer Ltd. and were ap-proximately 10 & 5 && 2 mm in dimensions. Justprior to irradiation, or making optical and photo-eonductivity measurements, the crystals were

FIQ. 1. Optical-absorption spectrum at 77 K of SrOwhich has been irradiated at 195 K with 1.5-3EcV protonsto a dose of approxiinately 2&&10 protons cm '. Themain feature introduced by irradiation is the asymmetricE' band located near 3.1 eV.

LO%- TEMPERATURE PHOTOCONDUCTIVITY IN SrO 2297

experiments, in which the sample was attached toa small cryostat located in the Van de Graaff ex-tension tube, it was necessary to warm the sampleto room temperature before transferring it to theoptical cryostat. This method was used because itwas necessary to transfer an irradiated sample tothe photoconductivity cryostat at room tempera-ture. However, no measurable J' band was pro-duced even in a sample which was irradiated at77 K with 1.5-MeV electrons to a dose of 5x 10"electrons cm ' and then measured without warmup. Nevertheless, as we shall see in Sec. IV, E'centers introduced by electron irradiation producean easily detectable photocurrent because of thesensitivity of the photoconductivity technique. Thereason for using electron-irradiated samples inour experiments was that incident electrons pro-duce mainly individual vacancies throughout thebulk of the crystal, whereas incident protons pro-duce a high density of damage near the surface.For this reason, it was not found possible to makecompletely satisfactory photoconductivity mea-surements on proton-irradiated samples.

&OSO

-11)0

6V

0

-1210

10

I ( i ) l

.3 4 5 6E NERGY (eV)

F~G. 2. Spectral dependence of the photoresponse g&ofor SrO. An unirradiated sample measured at 77 K,curve a; a sample which had been irradiated at 195 Kwith 1.5-MeV electrons to a dose of approximately 1&&10 electrons cm and measured at 77 K, curve b;the same sample as curve b measured at 105 K, curve c.

IV. EXPERIMENTAL RESULTS

In this section we first describe the interpreta-tion of our experimental data. We then describein detail the photoresponse of electron irradiatedsamples and finally we present results obtainedf rom y-irradiated samples and proton-irradiatedsamples.

The photocurrent I„when N„quanta per secondof wavelength X are falling on a sample betweenplane pa, rallel electrodes, is given by the equation'

I~ = (g &@ ) e VN~/d ',as long as the mean charge-carrier range +,V/d,in the direction of the applied electric field V/d,is much less than the thickness of the crystal d.Here, q is the free-carrier yield, thai is, the num-ber of free charge carriers produced per incidentphoton; &, is the mean range of the carriers in thedirection of a unit applied electric field; V is thevoltage applied across the sample; and e is theelectronic charge. Equation (I) can be rearrangedto collect all the measurable quantities on theright-hand side to give

(q(g ) = (I /N„)(d'/eV), (2)

where we call (q&u, )&, the photoresponse of the crys-tal at wavelength X. The requirement that the meancarrier range be small compared to the thicknessof the crystal can be tested by insuring that thephotocurrent is directly proportional to the poten-tial applied across the sample. '

In Fig. 2(a) we show the photoresponse of an un-

treated SrQ sample at 77 K over the spectral rangefrom 2 to 6 eV. It can be seen that the photore-sponse was immeasurably small until the incident-photon energy was at least 4 eV. Then, as thephoton energy increased further, the photoresponseincreased sharply by more than three orders ofmagnitude until the photon energy was about 5.4 eV.The photoresponse then decreased abruptly by twoorders of magnitude and became immeasurablysmall again at 5.8 eV. Although the apparatuscould detect photoconductivity for photon energiesout to about 6.2 eV, no photoconductivity could bedetected in SrQ above 5.8 eV. However, the mini-mum detectable photoresponse was comparativelylarge at 6.0 eV because ihe output of the lightsource and monochromator fell rapidly at ener-gies above 5.0 eV. The resolution of the appara-tus at 6.0 eV was about 0.1 eV. The photocurrentat 5.4 eV was the same for either direction of theapplied electric field and was constant in time forconstant illumination. The photocurrent was pro-portional to the applied electric field up to 1500V cm ', and was also proportional to the incidentlight intensity within an experimental' error of afew percent. The abrupt increase in photoresponseobserved at about 5.3 eV indicated that the inci-dent photon energy was approaching the band gapof the sample. We shall discuss these results inmore detail in Sec. V.

In Figs. 2(b) and 2(c) we show the photqresponseover the spectral range from 2 to 6 eV of a sampleof SrQ which had been irradiated at 195 K with

2298 JEANNETTE FELDOTT A5D GEOFFREY P. SUMMERS

1.5-MeV electrons to a dose of approximately1 && 10"electrons cm '. Curves b and c in Fig. 2show the photoresponse at 77 and 105 K, respec-tively. 'The most prominent feature introduced byelectron irradiation was the temperature-depen-dent and asymmetric band located at 3.1 eV, whichcan be shown to be due to F' centers by comparingit to the E' optical-absorption band, as we demon-strate in Fig. 3. In Fig. 3(n) is shown the photo-response of an electron-irradiated sample in thevicinity of 3.1 eV and in Fig. 3(b) is shown the Eabsorption band measured on a proton-irradiatedsample. Both curves, which were measured at110 K, were normalized at 3.1 eV. It can be seenthat these curves coincide over the whole E' bandexcept near the high-energy tail, where the photo-response from a band located at 3.9 eV, which canbe seen in Fig. 2, overlaps the E' band.

The photocurrent at 3.1 eV was directly propor-tional to the applied electric field up to 1500 Vcm 'within an experimental error of a few percent.Typically an applied electric field of 500 V cm 'was used during the measurements. In additionthe photocurrent at 3.1 eV was also directly pro-portional to the incident light intensity, within afew percent, as the intensity varied by a factor ofabout 50.

It can be seen in Fig. 2 that as the temperatureincreased from 77 to 105 K, the photoresponse ofthe F' band grew by over an order of magnitude,whereas the photoresponse away from the E' bandwas almost independent of temperature over thesame range. The detailed temperature dependenceof the E' photoresponse over the range from 63 to

I I I

tAI-

10—

I-xLI

00X

0

01O r

Vl0XI0

I

0z

I I I I I

3.0 3,4ENERGY (&V)

FIG. 3. Comparison of the spectral dependence of thephotocurrent in electron-irradiated SrO, curve a; andthe optical absorption of the 5 band in proton-irradiatedSrO, curve b. Both sets of measurements were made at110 K.

I I II

I I I II

I I

-'I 210

eIE

3013

C'

1O50

I I I I I I I I L

1OO 150 200TEMPERATURE (K)

FIG. 4. Temperature dependence of the E'-bandphotoresponse in electron-irradiated SrO. The mea-surements were made with an incident photon energyof 3.1 eV. The inset shows the thermally stimulatedcurrent (TSC) in an electron-irradiated crystal whichhad been illuminated with ultraviolet light at 77 K.

180 K is shown in Fig. 4. At 63 K the E' band wasalmost indistinguishable from the backgroundphotoconductivity and above 180 K the dark cur-rent became so large that accurate photocurrentmeasurements could not be made. 'The inset inFig. 4 shows how the thermally stimulated cur-rent, or dark current, increased with tempera-ture. For this experiment the sample had beenilluminated at 77 K with ultraviolet light for about90 min to fill available traps in the crystal. It canbe seen that none of these traps empty in the re-gion of the abrupt increase in E' photoresponse atabout 80 K. The sharp increase in photoresponsebetween 63 and 100 K is characteristic of a ther-mally assisted ionization such as occurs, for ex-ample, for the E center in Kcl at about 120 K.'Above 110 K an E' center in SrQ which has ab-sorbed a photon, frees a charge carrier with es-sentially unity probability, and the temperaturedependence of the photoresponse is due mainly tothe temperature dependence of the mean range ofthe carriers. The data in Fig. 4 will be discussedmore fully in Sec. V.

It was found that prolonged illumination in theE'-band region at temperatures above about 90 Kpartially destroyed the photoresponse. It tookabout 1 min to reduce the photoresponse to halfits initial value. However, the photoresponsecould be restored by illuminating the sample withlight of energy greater than 3.7 eV. Low-energyphotons were not effective in restoring the photo-response. Spectral-dependence measurementsabove 90 K, such as those presented in Figs. 2 and

I. O%-TEMPERATURE PHQTQCONOUCTIVITY IN SrO

3 were made by illuminating the crystal for sever-al seconds with ultraviolet light following eachmeasurement made in the region of the E' band.This bleaching effect is similar to experimental

its found by Dash'o in BRO He found that thephotoresponse of the 2.0-eV band could be en-hRnced lf the crystal wRs simultRneously irradi-ated with ultraviolet light. One possible explana-tion of these results is that the ultraviolet lightreleased charge carriers from comparativelydeep traps and that these carriers repopulatedbleached E' centers.

It has been suggested" that the 'A~ ground stateof the E' center lies below the top of the valenceband in the vicinity of the center. If this were thecase, it is possible that the charge carriers ex-cited by illumination into the E' band might bepositive holes. We determined the sign of thecharge carriers to be negative, however, using atechnique due to Peria. " A sample was supportedwith the electric field parallel to the 5-mm dimen-sion and perpendicular to the direction of the in-cident light. A section at the centex of the sample1 mm wide and the height of the sample was illumi-nated, while the right-hand electrode was heldpositive with respect to the left-hand electrode byan external battery. Although the mean range ofthe charge carriers was much less than the thick-ness of the crystal, the net result of illuminatingthe crystal was that a space-charge electric fieldwas established in the crystal going from left toright across the illuminated region. The externalelectric field was removed and the photocurrentwas measured as a function of the position of athin band of light as the band was scanned acrossthe crystal. The data shown in Fig. 5(a) was ob-tained. Note that the photocurrent was a maximumwhen the light was falling on the region of the crys-tal which had been illuminated with 3.1-eV lightRnd which is delineated by the vertical arrows, andfalls rapidly to zero outside this region. The cen-tral region of the crystal was then illuminated with3.i. -eV light again, with no externally applied elec-tric field. During the first illumination period,electrons moved just beyond the right-hand edgeof the illuminated region. During the second illu-mination period, these carriers remained trappedbut carriers originating from centers just insidethe right-hand edge of the illuminated region movetowards the left under the influence of the polar-ization field, with the result that a dipolar space-charge region is established at the right-hand edgeof the illuminated region of the crystal. This di-pole layer could be located experimentally bymonitoring the photocurrent as the crystal wasscanned RgRln with R narrow band of light. Figure5(b) was then obtained. The photocurrent data ot

0—

I

0I

2 3 4poslTION (mm)

FIG. 5. Electric field distribution obtained in the de-termination of the sign of the charge carriers excitedby 3.1-e+ light ln electron irradiated SrO. The mea-surements were made with both electrodes held at thesame potential after illumination with an externallyapplied electric fieM going from right to left, curve a,and after subsequent illumination with no applied electricfield, curve b. Curve a has been reduced by a factor of10 to accommodate it on the figure. The vertical arrowsindicate the region of the crystal which was illuminated.

curve a, has been reduced by a factor of 10 to a.c-commodate it on the figure. 'The maximum in thephotocurrent can be seen to have occurred Rt theright-hand edge of the region which had been illu-minated with 3.1-eV light. This measurement lo-cated the dipole layer and indicated that the chargecarriers were negative, i.e., electrons. If thecarriers had been positive holes, Rn analysis sim-ilar to that above shows that the dipole layer wouldhave been located at the left-hand edge of the re-gion which had been illuminated with 3.1-eV light. "The experiment was repeated with the initial ex-ternally applied electric field in the opposite di-rection, in order to determine whether any geo-metrical effects in the apparatus were Rffectingthe result. 'The sign of the carriers was again de-termined to be negative.

The performance and interpretation of the typeof experiments described above are somewhat dif-ficult, and more direct confirmation of the resultwill be required, such as would come, fox ex-ample, from Hall-effect measurements. It isworth noting, however, that similar experimentsperformed on the 400-nm E band in CaO indicatedthat the charge carriers were electrons in thiscase too, as expected.

The photocurrent originating from E' centers inSrQ was found to go to zero within a fraction of asecond of the incident light being removed. Therewas no evidence of the long-lived dark currentfound by g,oberts and Crawford' after illuminatingthe 5.0-eV band in MgO. We also found that the

2300 J E A IV E T 'r E F E- I. 0 0 'r T A N 0 GEOFFREY P. SUMMERS

1010

-1110EO

1 012

102 4 5

ENERGY (eY)

FIG. 6. Spectral dependence of the photoresponse 7~coo

for SrO at 77 K after y irradiation at room temperature,curve a; after irradiation with 1.5 MeV protons to adose of approximately 2xlo' protons cm, curve b.

E' absorption band in proton-irradiated SrQ couldnot be optically bleached at 120 K, although fromthe temperature dependence of the photoresponse,Fig. 4, E' centers which have absorbed a photonare expected to be thermally ionized with almostunity probability at this temperature. If we as-sume that E"' centers produced in proton-irradiatedSrQ are mutually independent, these results sug-gest that in high concentrations the oxygen vacancyis an efficient trap in SrQ. However, the fact thatthere is no sign of E centers being formed in SrQwhich has been heavily irradiated and which con-tains a comparatively large concentration of F'centers, Fig. 1, suggests that F' centers on theother hand are not effective electron traps in ourmaterial.

When an electron-irradiated sample was subse-quently y irradiated for 30 min at room tempera-ture just prior to being mounted in the cryostat,the photoresponse was found to have increaseduniformly by a factor of about 2 across the spec-tral range from 2 to 4 eV. A "Co y source wasused with a strength of about 5x 10' Rh '. TheE' photoresponse was unchanged relative to thebackground and no band was introduced in the re-gion of 2.5 eV, where the E absorption has beenreported. ' This result again suggests that the E'center is not an efficient electron trap comparedto other traps in the crystal.

We show in Fig. 6(a) the photoresponse at 77 Kof a crystal which was y irradiated at room tem-perature for about 30 min and measured with nofurther treatment. y irradiation increased thephotoresponse generally between 2 and 5 eV, and

introduced a band at 2.65 eV, which is also ob-served in electron-irradiated crystals. The in-flexion in the photoresponse at about 4.8 eV, whichis observed in unirradiated crystals, is furtherenhanced by the y irradiation. y irradiation alonedid not produce any E photoresponse.

Figure 6(b) shows the photoresponse at 77 K ofa crystal which was proton irradiated at 77 K toa dose of about 2 && 10"protons cm '. It can beseen that the most prominent feature introduced byti16 ll radlatlon ls again 'the asymmetric F+ bandlocated at 3.1 eV. The irradiation also introducedphotoresponse bands centered at 2.65 and 3.9 eVas well as an inflexion at 2.5 eV. The 2.65-eVband was introduced by electron irradiation aswell as by y irradiation and is presumably due toan impurity in the crystal. The band at 3.9 eV,which was also produced by electron irradiation,may be associated with the same type of centerwhich is produced when SrQ is plastically de-fol med

V. DISCUSSION

(l/q )-l = (r,/r„) exp(E, /kT). (3)

Since photoconductivity originating from E' cen-ters is unexpected, it is of interest to compare itsmagnitude to that measured from E centers in, forexample, MgF, ." For a useful comparison, it isnecessary to compute the value of p~&, for thetwo centers, where r)~ is the free-electron yieldper absorbed photon. If we estimate a maximumvalue of 0.002 for the optical density of the E' bandfor the sample used for Fig. 4, we obtain a lowerlimit of 4x 10"cm'V ' for g~&0 at 120 K. Forlow-concentration E centers in MgF„q~&, isabout 4 && 10 "cm'V ' at room temperature. It isworth recalling, however, that the magnitude ofthe photoresponse depends on (d, which is closelydependent on the purity of the crystals.

No detailed calculation has been reported to dateof the electronic structure of the relaxed F' centerin SrQ which can be compared with our photocon-ductivity measurements. However, the tempera-ture dependence of the photoresponse between 70a,nd 110 K can be explained quite well by a two-level model. ' In this model, an electron in. therelaxed excited state of the center has a tempera-ture-dependent probability 1/rr ——(1/v, ) exp(-E, /kT)of being thermally ionized across an energy gapE,. 1/r, is the preexponential frequency factor,k is Boltzmann's constant, and T is the tempera-ture. 'There is also a temperature-independentprobability l/vs of decay to a lower energy level.The yield of free carriers per absorbed photon q~is then expected to follow

LOW- TEMPERATURE PHOTOCONOUCTIVITY IN SrO 230I

In order to compare this equation with our ex-perimental results, Fig. 4, which give directly theproduct q&, as a function of temperature, we makethe approximation that &, and the oscillatorstrength of the F' band remain independent oftemperature between 70 and 110 K. The tempera-ture dependence of q&, is thus assumed to be dueentirely to the change in q~ over this range. Sinceit appears that F' centers are effectively fullyionized at 115 K, g~ is taken as unity at this tem-perature and we can then plot log„[(l/qr) ljagainst 1/T using the data of Fig. 4. The result-ing curve is shown in Fig. 7, where it can be seenthat over several orders of magnitude a straightline is obtained. From the data in Fig. 7, we ob-tain a value for E, of 0.12*0.01 eV and a valuefor v„/vo of 5 && 10'. lt is interesting that thesevalues are close to those found for F centers inalkali halides, " and estimating a reasonable valuefor ~„we obtain a value for v~ of the order of 1p,sec." If this were the radiative lifetime for anallowed decay from the 'T,„excited state to the'A. ~ ground state, it would be surprisingly long foran F' center. However, without detailed knowledgeabout the electronic structure such as that ob-tained by Wood and Wilson' for F centers in alka-line-earth oxides, it is not possible to determineto which transition the parameter ~~ corresponds.Nevertheless, an increase in the free-electronyield with increasing temperature is expected tobe accompanied by a decrease in the fluorescenceyield, and Hughes and Webb' have found that theintensity of the F' luminescence decreases mark-edly between V'7 K and room temperature.

The main feature of the photoresponse of SrQ athigh photon energies, Fig. 2, was the relativelyabrupt increase in the vicinity of 5.3 eV. A steeprise in the photoresponse is expected when the in-cident photon. energy approaches the band edge ofthe crystal. " A threshold for intrinsic photocon-ductivity can then occur. Photons of higher energydo not penetrate far into the surface because of thestrong absorption of the crystal and the photore-sponse usually decreases. However, photoconduc-tivity can be found at energies close to the bandedge due to an indirect process. Under these cir-

100

0.18 12

&/T ( &0 K')

FIG. 7. Semilogarithmic plot of (1/gg —1 vs inversetemperature 1/T. Data points were obtained from theexperimental results shown in Fig. 4 as described inthe text.

cumstances excitons, which are produced by theincident photons, give up their energy to ionizablecenters, such as, for example, F centers, andphotoconductivity can result. ' ~ Thj.s mechanismcan be very efficient. Since there is an excitonband in SrQ at 5.V eV,"it is possible that a simi-lar mechanism was occurring here. 'The samephotoresponse peak occurred, however, in bothirradiated and unirradiated samples, so that thephotoconductivity would have to be produced mainlyby impurity or defect centers other than F' cen-ters, although a mechanism similar to that ob-served by Apker and 'Taft in KI cannot be ruledout. From the evidence available so far, we pre-fer to interpret the steep rise in photoresponse inSrQ at high photon energies as due to the onset ofband to band transitions. Qur results, then, sug-gest that the photoconductivity band gap of SrQ isabout 5.3 eV, which is somewhat lower than hasbeen given previously. " In addition, the excitonbands at 5.7 and" 5.8 eV would, then, be placedabout 0.5 eV above the bottom of the conductionband.

~Besearch supported by the U.S. Energy Besearch andDevelopment Administration under Contract No. AT-(40-1)-4837.

)Paper based in part on a thesis to be submitted byJeanette Feldott to Oklahoma State University in partialfulfillment of the requirements for the degree of Doc-tor of Philosophy.

'B. Henderson and J. E. Wertz, Adv. Phys. 17, 749(1968).

2A. E. Hughes and B. Henderson, in Point Defects inSolids, edited by J. H. Crawford and L. Slifkin (Plenum,New York, 1972).

3N. F. Mott and B. W. Gurney, E/ect~onic Processes inIonic Crystals (Oxford U.P., New York, 1940).B. G. Bessent, B. C. Cavenett, and I. C. Hunter, J.Phys. Chem. Solids 29, 1523 (1968).

A. E. Hughes and A. P. Webb, Solid State Commun. 13,167 (1973).

2302 JEANNETTE FELDOTT AND GEOFFREY P. SUMMERS

68. D. Evans and J. C. Kemp, Phys. Rev. B 2, 4179(1970).

B. P. Johnson and E. B. Hensley, Phys. Rev. 180, 931(1969).

R. W. Roberts and J. H. Crawford, J. Nonmetals 2,133 (1974).

B. H. Rose and E. B. Hensley, Phys. Rev. Lett. 29,861 (1972).

' W. C. Dash, Phys. Rev. 92, 68 (1953).'~J. Feldott and G. P. Summers, Bull. Am. Phys. Soc.

21, 345 (1976).' J. Feldott and G. P. Summers, Solid State Commun. 18,

347 (1976).~3G. P. Summers, J. Phys. C 8, 3621 (1975).'4R. J. Zollweg, Phys. Rev. 111, 113 (1958).' J. C. Kemp and V. I. Neeley, Phys. Rev. 132, 215

(1963).'6W. T. Peria, Phys. Rev. 112, 423 (1958).~~T. J. Turner, N. N. eisenhower, and P. K. Tse, Solid

State Commun. 7, 1661 (1969).H. Fedders, M. Hunger, and F. Luty, J. Phys. Chem.Solids 22, 299 (1961).R. K. Swank and F. C. Brown, Phys. Rev. 130, 34(1963).R. F. Wood and T. M. Wilson, State Commun. 16, 545(1975).

'Y. Nakai and K. Teegarden, J. Phys. Chem. Solids22, 327 (1961).

~L. Apker and E. Taft, Phys. Rev. 79, 964 (1950).3N. Inchauspe, Phys. Rev. 106, 898 (1957).American Institute of Physics Handbook (McGraw-Hill,New York, 1957).