A DIRECT READING SPECTROPHOTOMETER

By CARL W. KEUFFEL

INTRODUCTION

It is scarcely necessary to point out the need of having availableinstruments with which to measure and specify color. There are twofundamentally different principles which can be used as a basis for acolor measuring instrument. We can either (a) measure the physicalcharacteristics of the stimulus which evokes the sensation of color, or(b) we can use a psychological method and match the color sensationproduced by varying a known stimulus. The Spectrophotometer fallsin class "a"; it measures the spectral distribution of the stimulus andis therefore independent of any psychological factors such as the sensi-tivity of the observer's eye, the color characteristics of the illuminant,etc. Instruments coming under class "b" are called Colorimeters. Theyserve a definite purpose and under the proper conditions will give usefulresults. Since, however, the sensation of color experienced by an ob-server depends not only on the stimulus but upon the individuality ofthe observer and upon many conditions under which the stimulus isobserved, the Colorimeter cannot be considered as a primary instru-ment for color measurement and specification.

It has been pointed out by many authorities that the Spectrophotom-eter must be considered as the fundamental instrument for color meas-urement. I quote from H. E. Ives who in a discussion of methods ofcolorimetry says: "However, it is very questionable whether either ofthe three element (Colorimetry) methods of analysis will ever be satis-factory, beautifully simple though they are in theory. A spectropho-tometric table, derived from at least 25 points, gives the only uniquedescription of color, and it appears probable to the writer that the re-quirements of precision technical color measurement are most likely tobe met by the development of simple and rapid means of plotting andrecording accurate spectrum plots of reflection or transmission charac-teristics." Also, quoting from I. G. Priest: "Of whatever value theso-called 'Colorimeter' may be in special cases, it must be admittedthat the fundamental basis of color specification is spectrophotometry."While it has long been recognized that the spectrophotometer is a

403

[J.O.S.A. & R.S.I., 11

requisite in color measurement, providing, as Priest says, the funda-

mental basis of color specification, the only spectrophotometers hitherto

available have been complicated instruments, elaborate in design,

expensive in original cost, difficult to set up, and not calibrated to read

directly.The instrument which was developed to fill the need for a practical,

direct reading spectrophotometer is called K & E Color Analyzer, and

it consists essentially of:

A. The Constant Deviation Spectrometer with its wave length scale.

B. The Direct Reading Rotating Disc Photometer with its photometer

scale.

C. The Light Source with its holder for the samples when reflection

measurements are made.

D. The stand with adjustable holder for transparent substances.

1. Spherical Light Source.2. Photometer.3. Spectromter.4. Wave Length Scale.5. Photometer Scale.6. Holder for Standard Sample.7 Holder for Reflection Sam 2

pies.8. Holder for Transparent Sam-

ples.9. Field of View thru Eye Slit. l

1o. To Vacuum Ventilator.i1. Plug for Vacuum Ventilator1-2. 400 Watt Lamps.13. Lever for Raising Photo-

meter.14. Sector Discs.15. Universal 110 Volt Motor - .

16. Speed Control Rheostat.17. Entrance Slit.18. Collimator Objectives. 6

19. Dispersion Prisn.20. Bi-Prism.21. Eye Slit.2-. Cast Aluminum Base.

FIG. 1. Diagram of K & E Color Analyzer.

These parts are all rigidly mounted on a cast aluminum base plate

which insures permanent alignment. A sodium flame spirit lamp is fur-

nished for setting the wave length scale. To make the equipment

complete and ready for any kind of color measurement, containers are

also furnished for liquids and powders.

CARL W. KEUFFEL40)4

DIRECT READING SPECTROPHOTOMET-E40

DESCRIPTION OF INSTRUMENT

A. The Constant Deviation Spectrometer has an adjustable entranceslit (17) and a fixed width exit or eyeslit (21). For reflection measure-ments the slit widths are comparatively wide, purity of spectrum beingsacrificed for illumination because few spectral reflection curves havesharp dips. The optical elements consist of the two achromatic colli-mator objectives (18), the dispersion prism of dense flint glass (19) andthe bi-prism (20) which produces the divided field.

The dispersion prism is mounted on a rotating platform having aconical center. This platform is actuated by the drum (4) carrying thewave length scale ranging from 420 to 700 myu, graduated every 5 mn.Blue and Red filters are provided, either of which can be inserted in thebeam at the eyeslit to cut out strdy light when measurements are beingmade at the ends of the spectrum. All parts are constructed rigidlyand fitted accurately. The only adjustments necessary are on theentrance slit, the bi-prism and the wave length scale drum. Theseadjustments when once made are secured rigidly so that under ordinaryconditions the instrument keeps its adjustment for a long time.

B. The Direct Reading Rotating Disc Photometer consists of twosectored discs, with two opposite openings, which revolve around thesame axis and in the same direction. The effective angular openingbetween discs is set while they are in motion by means of turning theknurled head to which the photometer scale is attached.

One sector shown by dotted lines Fig. 2) is larger in diameter thanthe other (solid lines) and cuts the upper beam a constant amount whilethe amount of the lower beam passed can be varied from 50 per cent tozero. (The two beams are represented by the open circles, Fig. 2). Theconstant amount cut off by the larger sector is such that the relativechange between the upper nd lower beams can be varied between 0and 110 per cent. The photometer scale reads from 0 to 110.

The fact that the photometer reads bove 100 to 110 per cent isof decided advantage in measuring a sample having very high reflectionor transmission values, as the matching point can still be approachedfrom both sides and the setting made as accurately as with the lowervalues. It also simplifies the calibration of the instrument as 100 percent can be used as a reference point.

When measuring dark samples the center line of the sectors can belowered by means of a lever (13, Fig. 1). (Now the beams are repre-sented by the cross hatched circles). This doubles the illumination andincreases the accuracy of the settings as the sectors are so made that thescale will then read four times the actual value.

Oct., 1925] 40-5

CARL W. KEUFFEL [J.O.S.A. & R.S.I., 11

-Z��i

.s?

Its,5�.2

.1�Q'all,

cn

c-;6;ml

t4

lzt

-4

;z

.2!

-ap

406

.0

DIRECT READING SPECTROPHOTOMETER

The discs are rotated at high speed through an endless rubber beltby a motor which runs on either 110 AC or 110 DC, and the speed iscontrolled by a rheostat. All fast rotating parts run in ball bearings(23) (24) (Fig. 3) and end thrust is also taken up by a ball bearing (25).

The mechanical details of the sector photometer are shown in Fig. 3.The larger sector is attached to the sleeve (26) (27), the smaller sectoris attached to an inner tubular sleeve (28). The plunger (30) carries apin which engages in a straight, lengthwise slot in the inner sleeve andin a cam slot cut into the outer sleeve. The outer sleeve, inner sleeveand plunger are rotated by means of pulley (31). A helical spring at (32)takes out the backlash between the pin and the two slots. The angularrelation between the two sectors is changed by moving the plunger intranslation along the axis of rotation. This is done by turning theknurled ring (33) to which is attached a sleeve (34). As the ring isturned, the sleeve advances, in the cam (35) and through the ballthrust bearing (25), translates the plunger along the axis. The accuracywith which the Photometer performs is dependent upon the excellenceof the design and workmanship of the mechanical parts, and extremecare is taken to build the instrument to withstand long use withoutappreciable wear.

C. The light source consists of a spherical housing (Fig. 1) of castaluminum well ribbed on the outside to radiate the heat generated bythe lamps. Two 400 watt lamps (12) illuminate the inside of this sphere,which is coated with a special white paint to give bright and diffusereflection. Special shields are provided so that no direct light from thelamps reaches the sample. This unit is ventilated and kept cool by amotor driven suction fan (attached at 10), which is furnished with theoutfit. The sample holder is so constructed that a stream of cool airpasses in a thin space between the sphere and the sample holder andover the sample. In this way the sample is kept cool even though it isvery close to the light bulbs. The standard white (magnesium car-bonate) is held in the lower holder (6) and the upper holder (7) carriesthe sample for reflection measurements. The unit as constructed in-sures approximately diffuse illumination for reflection samples. Whentransmission measurements are made both holders carry magnesiumcarbonate blocks.

D. The stand is rigid and the holder (8) for tubes and cells fits thestand so that the tubes or cells can be lined up quickly without muchadjustment.

Oct., 1925] 407

CARL W. KEUFFEL

The tubes for holding liquids (Fig. 4) have a glass liner cementedinto a brass sleeve. The glass end plates are plane polished and heldin place with screw caps. The liquid sample comes in contact withglass only and the tubes are easily cleaned and filled. The lengthsusually used are 100 mm, 30 mm and 10 mm, and they are ground tothese lengths within an accuracy of 1/10 of 1 per cent.

FIG. 4. Sectional view showing construction FIG. 5. Two views showing constructionof tuIbes. of cells.

The cells for holding liquids (Fig. 5) consist of three plane parallelglass blocks, held together with clamps. A cylindrical hole 10 mm indiameter in the middle block, forms the cavity which holds the liquid.These cells are used in the following thicknesses:

3 mm plus or minus 0.003 mm and mm plus or minus 0.002 mm.Both types of container have proven very practical in actual use.

ADJUSTMENT OF INSTRUMENT

Although the Color Analyzer is constructed rigidly and will not getout of adjustment easily, yet it is necessary and easy to check all adjust-ments frequently.

The adjustment of the wamde length scale is checked by means of thesodium flame lamp, which gives the spectral D line. The wave lengthscale is marked with an extra graduation at 589 mu to facilitate theadjustment. Any other monochromatic light source can of course beused to check the wave length scale. The cylinder carrying the wavelength scale is moveable on its shaft and can be locked when properlypositioned with respect to its index. This provides the means for ad-justing the wave length scale.

The photometer scale must read zero when the photometer discs cutout all the light of the lower beam, i.e., when one-half of the divided fieldappears black. The drum carrying the photometer scale is moveableabout its shaft and can be locked when properly adjusted.

[J.O.S.A. & R.S.T., 11408S

DIRECT READING SPECTROPHOTOMETER

The other major adjustment which must be checked is the equalityof brightness of the two beams of light. To do this it is necessary toplace a block of magnesium carbonate in each sample holder and seethat these blocks are properly illuminated with diffuse light, i.e., theshields between the lamp filaments and samples must be in place.

When the two fields appear matched in brightness the photometerscale should read 100 per cent. If it does not read 100 per cent theadjustment is made at the entrance slit by changing the relative slitwidth of the upper and lower beams. The equality of brightness of thetwo beams must be checked throughout the spectrum and if it is foundthat the photometer reads high at the red end of the spectrum and lowat the blue, or vice versa, the correction must be made by adjustingthe bi-prism angularly about the axis of the telescope.

ACCURACY OF INSTRUMENT

The instrument gives accurate results when properly adjusted.Transmission measurements on standard glasses with various individualinstruments have checked well within the limits of the precision ofreading.

The absolute accuracy of reflection measurements depends to acertain extent upon the illumination of the sample as provided by thelight box. Reflection measurements taken on the same samples on dif-ferent Model E Color Analyzers agree very well. However, since theillumination of the samples is not perfectly diffuse, the results on glossysamples will not check with results obtained where the sample has beenilluminated with perfectly diffuse light.

This factor of the method of illuminating the sample has not beenproperly taken care of in making spectral reflection measurements inthe past, and probably accounts for the difficulty experienced by ob-servers with different instruments who have attempted to check meas-urements on the same set of samples. It is difficult to devise means forilluminating a sample with perfectly diffuse light and at the same timeget enough brightness with which to make spectral reflection meas-urements. It may be better practice to specify some other arbitraryway of illuminating the sample for spectral reflection measurementsand in addition get some arbitrary measure of gloss in order to ccm-pletely specify the reflecting properties of a sample.

PRECISION OF MEASUREMENT

The precision with which an average observer takes measurementscan be judged from the following:

Oct., 1925] 409

[J.O.S.A. & R.S.I., 11

The probable error for samples reading between 10 per cent and90 per cent for an average of five settings at any one wave length be-

tween 480 mgu and 700 mgt is about ±0.8 per cent of the reading. Theprobable error below 480 mu is greater than this due to the lower bright-ness at this end of the spectrum and is about ± 1.2 per cent of the read-ing.

An average of 10 readings reduces the probable error considerably.The precision of reading depends somewhat on the experience of theobserver, but the results given above can be obtained with very littlepractice.

SUMMARY

The K & E Color Analyzer is accurately made, rigidly constructed,easy to adjust and convenient to use.

It is a complete spectrophotometer in one unit and provides a rapidand precise means of obtaining a complete color analysis of any sample,liquid, or solid, transparent or opaque.

It is being used for color standardization and factory control work inmany of the industries on the following products: paints and varnishes,dyes and dyed fabrics, oils, both mineral and vegetable, soaps, sugars,honeys, etc., even artificial teeth are graded with the Color Analyzer.A few of our progressive hospitals are making use of this instrument in

their research work and it is part of the equipment in the physics,chemistry, physiology and psychology laboratories in many univer-sities.

TECHNICAL DEPARTMENT,KEUFFEL & ESSER CO.,

HOBOKEN, N. J.

410 CARL W. KEUFFEL