ABSOLUTE ENERGY MEASUREMENTS ON ULTRAVIOLET RADlATION Thesis for the Degree cf M. S. MICHIGAN STATE COLLEGE William Eugene Corbridge I943 a WNW!VVVVHVHHNUVVVVVU{INNVVVlHHVlUVHVVIHHVfl ' $301701 0939 V V LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE / V V 1/98 c:ClRC/DateDue.p65-p 14 It I ll lllll“ [[11 1! 1i " [II‘ I. ' ' 1V C Ver Vl‘lllill'li ABSOLUTE ENERGY MEASUREMENTS 0N ULTRAVIOLET RADIATION by WILLIAM EUGENE CORBRIDGE A THESIS Submitted to the Graduate School of Michigan State College of Agriculture and Applied Science in partial fulfilment of the requirements for the degree of MASTER OF SCIENCE Department of Physics 1943 ( LL \'»4 6""! - CL (3’0 5-13~44 TABLE OF CONTENTS Section 1. Source of Radiation 2. Segregation of Wavelengths 3. Detection Photoelectrically 4. Thermionic Amplifier 6. Curves Obtained 6. Results ABSOLUTE ENERGY MEASUREMENTS 0N ULTRAVIOLET RADIATION When absolute measurements of radiant energy are made it is necessary to have a detecting device whose response is independent of wavelength or one whose response is known over the range of wavelengths to be investigated. the former method is the one usually employed. If the latter method is used, the device must be calibrated, and this by comparison with a standard having a response independent of wavelength. It happens that with this latter method the equipment is generally of greater prac- tical usefulness. Discussions of both methods will be found in the references listed at the end of this article. This paper is concerned with the absolute measurement of ultraviolet radiation by means of a calibrated photocell. Monochromatic radiation, obtained using a grating spec- trograph, when incident upon the photocell caused an emission which.was measured by a thermionic amplifier. This current could then be interpreted in terms of energy by referring to the calibration curve of the photocell. _1_._ Source g_f_ Radiation The source whose output was examined was the Eanovia Sc 2537, a low pressure mercury arc in a quarts tube approximately 10 mm in diameter and with a l mm.wa11 thickness. This tube was Operated by a 110 volt transformer supplying 120 milliamperes at 7500 volts. The ultraviolet source was placed at several different distances from the slit of the spectrograph and the response noted at each position. maximum.light on the photocell, as indicated by its response, was obtained when a 7.5 em diameter quarts lens was used to focus the radiation from a 10 em portion of the tube on the 0.27 mm wide slit of the spectrograph. This occurred in spite of the absorption in the lens and the necessarily greater distance of the source from the slit. Best results were obtained with the lens 27.5 cm from the slit and the tube 58.0 cm from the slit. 3; Segregation'gf Wavelengths The spectrograph.was an Eagle mounting of a concave reflec- tion grating. The grating, an original speculum of 100 cm radius. was ruled with 30,000 lines per inch over a four inch surface. The slit width was adjusted to fimd the optimum separation of the Jaws. When opened the maximum amount a greater quantity of light than necessary to fill the aperture of the frating was admitted. This was de- ereased by closing the slit until the response of the I photocell showed a decrease. At this position the maximum usable amount of light was assumed to be incident upon the grating. The full length of the grating was used. This spectrograph was used as a monochromator by placing a sheet of fiberboard in the position of the plateholder. A one inch diameter hole in this sheet admitted radiation to the photocell which was placed with its window in align- .ment'with and close to the hole. Successive portions of the spectrum were directed upon the cell window by turning the grating to change the angle it made with the cell. §;_Detection Photoelectrically The window of the photocell was masked with black paper to an area 15 mm by 17 mm. This area at the central por- tion of the window was used when one of the calibrations of the cell was performed by its manufacturers. The sen- sitive surface is metallic sodium. This high vacuum cell, the General Idectric ra-coe, has a re-entrant Window whose thickness is so small that ultraviolet radiation may readily ' pass. It was found necessary to shield the photocell electro- statically with an ordinary tube shield. This shield was covered with black paper to cut out stray light; only the rectangular aperture in front of the cell window remained uncovered. Grounding the tube shield kept that enclosure at the same potential as the ease of the amplifier. Leads to the photocell were tried in various types. The final arrangement used lacquer covered conductors which ‘wcre supported by the binding posts to which their ends 'were attached. All other points were in the open air, which minimised leakage currents. .This arrangement al- lowed electrostatic disturbances to come into full play, but with careful manipulation these were not troublesome. The photocell was designed by its manufacturers to be operated at inter-electrode potentials up to 200 volts, one-third of that being an ordinary working voltage. It was found that for the small currents being measured a potential of 67 volts caused ineonveniently large leak- age currents to flow when the cell was not illuminated. Using only 22} volts caused these currents to be reduced to one-third of their former value. Voltages below as} volts caused instability. _c_,_ Thermionic mlifier With radiation incident upon the cell and a potential applied to make the central collector ring positive with respect to the sensitive surface a small current will flow. This current flowing through a high resistance will create between the ends of the resistor a potential difference of such a magnitude as to be readily measurable. However, the measurement of electromotivc forces in high resistance circuits cannot be done by ordinary methods. A thermionic amplifier, Deeds e lorthrup model 7675, has been developed for such measurements. It uses the Westinghouse Rflb507 cleetrometer tube. The cleetrometer tube was enclosed in a separate compart- ment in the case housing the amplifier. This compartment was kept dry with anhydrous calcium sulfate. The tube was cleaned with absolute alcohol befcre being placed in the compartment in order to remove from its surface any material which might cause current leakage. The thermionic amplifier needs a.high quality resistor for calibration pupposes. The manufacturer of the ampli- fier has designed a shielded resistor with special adapters. Since one of these was not available and even if available would have had to be modified, a high quality molded re- sistor made by the S. S. White company was used. When this resistance in the phptocell circuit is connected to the emf terminals of the amplifier it completes the plate circuit of the cleetrometer tube. Leakage currents across the resistor were minimised by cleaning the resistor with absolute alcohol Just before idsta.were taken. Foreign matter, water vapot or dust :partioles, would affect the value of the resistance and give erratic results. Connections to the resistor were soldered in order to insure good contact. A ferrule was made to fit on the negative emf terminal of the amplifier; this fitted snugly and caused no contact trouble. The other connections in the photocell circuit were made with knurled thumbscrews on binding posts. The potential difference between the ends of the resistor is balanced by one supplied from a potentiometer. The condition of.balance is indicated by sero deflection of a galvanometer. This galvanometer is connected in the grid circuit of the cleetrometer tube so that it may be adjusted for sero deflection at the grid potential existing when the plate is grounded. When the potential of the grid dif- fers from this value, due to the plate being at other than ground potential, the galvanometer shows a deflection. When the potentiometer exactly balances the potential difference between the ends of the resistor, the plate is at ground potential and the galvanometer shows no de- flection. The potentiometer used was a Leeds a Northrup model 7555-8, a self-contained unit including standard cell and work battery. Potentials supplied by it of the order of 0.5 volt could be read to four significant figures. The amplifier is ordinarily used with a grid potential of 4% volts. Twice this value may be used when maximum sensitivity is desired. Since all obtainable sensitivity was needed for these measurements the full 9 volts was used for the grid. A bias adjustment on the amplifier makes it possible to adjust the circuit so that no current will flow between filament and plate when the emf and potentiometer terminals are short oircuited. This bias adjustment is in the plate circuit, and when the plate current is adjusted to zero under the conditions stated, any measurements of potential across the resistor are not affected by currents flowing in the amplifier circuit. A constant error in determinations is introduced otherwise. The filament current of the cleetrometer tube is controlled by two resistors. One is a fixed resistor which can be switched in or out of the circuit. The other is a variable resistor of the dial type which incorporates an on-off switch. When first using the tube it was possible to get the rated filament current of 60 milliamperes when a 6 volt source was used. Later it was necessary to increase the voltage to about 8 volts to obtain the rated current. With 5 volts it was impossible to increase the current to its rated value even when both resistors were out of 'the circuit. Since the potentiometer adjustment was checked against its standard cell between each set of data and found to remain nearly constant, it is assumed we be without error. In obtaining measurements of radiation intensity it was necessary to have conditions most Opportune.. The model of galvanometer used, a Leeds & Northrup 2420-0, was that for which the amplifier was designed and better performance was obtained with it than with others. The type of po- tentiometer used greatly affected the ease of handling of the amplifier. All models tried were manufactured by the maker of the amplifier. The enclosed type previously mentioned was used finally. This particular kind appar- ently was the one best suited to the amplifier. The gal- vanometer did not show fluctuations as it did with the other types, and the apparatus was much more manageable. Weather conditions influenced the ability of the appar- atus to give results. On days of heavy rain and high humidity the water vapor on eXposed parts caused varia- 'tions which prohibited the taking of data. The manufac- turers of the photocell state that on mid-summer days when the humidity is high they are unable to make a cal- 1bration. §;_Curves Obtained film the data shown in Tables 1, 2, and 5 a favorable day allowed measurements to be taken, but the apparatus had not been running so had not reached a steady state. The A TABLE 1 Listing resistor voltage caused by photocell current and power supplied to ultraviolet source.. Cold start. Grating Units Resistor Volts Source Volts Source Amps (arbitrary) Primary Primary Slit Covered .5555 4.5 .5461 115.6 9.8 4.6 .5499 115.5 9.9 4.7 .5559 116.5 10.0 4.8 .5625 116.6 10.0 4.9 .5650 116.8 10.0 5.0 .5650 116.7 10.0 5.1 .5842 116.7 10.0 5.3 .6275. 116.8 10.0 5.5 .5700 116.5 10.0 5.4 .5645 116.5 10.0 5.5 .5594 116.5 10.0 5.6 .5651 116.8 10.0 5.77 .5625 116.5 10.0 5.8 .5679 116.0 9.9 5.9 .5701 115.8 9.8 6.0 .5758 115.8 9.9 6.1 .5723 115.5 9.8 6.3. .5880 117.0" 10.0 6.5' .5957 116.3 10.0 6.4 .5742 115.7 9.9 6.5 .5740 115.7 9.8 Slit Covered .5710 TABLE 2 Listing resistor voltage caused by photocell current and power supplied to ultraviolet source. Warming up. Grating Units Resistor Volts Source Volts Source Amps (arbitrary) Primary Primary Slit Covered .3730 4.5 .5854 115.6 9.9 4.6 .5868 115.5 9.8 4.7 .5927 115.8 9.9 4.8 .5975 116.2 10.0 4.9 .5955 116.0 9.9 5.0 .5927 116.0 9.9 5.1 .6330 115.9 9.9 5.2 .5154 115.9 9.9 5.5 .4087 116.0 9.9 5.4 .4040 115.5 9.8 5.5 .5967 115.5 9.8 5.6 .5967 115.9 9.9 5.7 .5967 116.1 9.9 5.8 .5988 116.1 9.9 5.9 .5988. 115.8 9.9 6.0 .5958 115.8 9.9 6.1 .5940 115.5 9.8 6.2. .4122 116.1 9.9 6.5. .4160 116.9 10.0 6.4 .5948 116.8 10.0 6.5 .5955 116.6 10.0 Slit Covered TABLE 5 Listing resistor voltage caused by photocell current and power supplied to ultraviolet source. Near equilibrium. Grating Units Resistor Volts Source Volts Source Amps (arbitrary) Primary Primary Slit Covered .5905 4.5 .5992 116.2 10.0 4.6 .4040 116.8 10.0 4.7 .4091 116.5 10.0 4.8 .4148 116.7 10.0 4.9 .4155 116.5 10.0 5.0 .5754 116.4 10.0 5.1 .6517 116.8 10.0 5.2 .4200 ll6.7 10.0 . 5.5 .4174 116.8 10.0 5.4 .4114 116.8 10.0 5.5 .4059 116.8 10.0 5.6 .4059 117.1 10.0 5.7 .4020 117.0 10.0 5.8 .4045 116.5 10.0 5.9 .4050 116.8 10.0 6.0 .4020 116.8 10.0 6.1 .4012 117.0 10.0 6.2 .4162 116.8 10.0 6.5 .4185 116.5 10.0 6.4 .5990 116.4 10.0 6.5 .5982 115.8 9.9 Slit Covered j 2-5 vans g ' POSITION or GRATING, ARBITRAR‘! arm's 3°15 no as so as Posmou or swarms, ARBITRARY UNITS "Tar e . . . . . c . e a , a, C O O p. I- c . < . o ‘4 "315 so 5.5 so: 6.5 FIGURE 1 Upper plot is original data.of Table 1 including dark current. Lower plot is power supplied to transformer for ultra» violet source. Cold start. ‘1 5 so as so as POSITION or GRATING, ARBITRARY owns "7 0 e WATTS 4:5 50 5.5 6.0 6.5 POSITION OF’ GRATING, ARBITRARY UNITS .FIGURE 2 Upper plot is original data. of Table 2 including dark current. Lower plot is power supplied to transformer for ultra? violet source. Warming up. 55) L 5 VOLTS ram 3H! “"75 so :5 so as POSITION or GRATING, aseirsssv umrs "10.... e.mec...es°o. on' ° ' h- 5‘: . 3 "#5 5.0 5.5 6.0 6.5 POSFTHDN CHEISRATU‘Q ARBFTRARYVURHTS F1GURE Upper plot is original data of Table 3 including dark current. Lower plot is power supplied to transformer for ultra— violet sourcc. Near equilibrium. three sets of data were taken in succession and in the order of the number of the tables. It will be noticed that the background or dark current when the slit of the spectrograph was covered increased in each table and from one table to the next. The dark measurements before and ' after each set show a trend toward constancy, the values obtained for the last set being nearly equal at the begin- ning and end of the run. These data are plotted in Figures 1, 2, and 5. The lower part of each figure shows the power supplied to the trans- former operating the light source, obtained by multiplying source volts by source amperes, plotted against grating position. The upper part shows the voltage across the resistor, as balanced by the potentiometer, plotted against grating position. The straight line beneath these points connects the two values obtained as dark current when the slit was covered at the beginning and end of the run.. Since each of these sets of data was obtained under dif- ferent background conditions, they cannot be compared in their original form. Using the lines connecting the dark currents as new hor- isontal axes, these points were replotted to give Figures 4, 5, and 6. In these figures an abrupt rise and fall will be noticed in the left part of the curve with a smaller E VOLTS M 50 5.5 so as Posmou or GRATING, ARBITRARY mute FIGURE 4 Curve obtained from Figure 1 when corrected for dark current. .200 § VOLTS / 50 55 an 5 POSITION or alumna, ARBITRARY QNITS 5 FIGURE 5 Curve obtained from Figure 2 when corrected for dark current. "s’ vans E ‘50 85 CD 85 POSITION OF GRATIUO, ARBITRARY UNITS FIGURE 6 Curve obtained from Figure 5 when corrected. for dark current. 10 but definite peak in the right part. However, these peaks do not occur at exactly the same grating settings nor do they rise to the same height on the ordinate in each of the figures. Since the voltage values could be read on the potentiometer more accurately than the curve can be read, and the setting on the spectrograph was accurate only to the nearest tenth of the unit used, both of these discrep- ancies may be attributed to the error in setting the epoc- trograph. To minimize this error in setting, the three curves were superposed and a new curve drawn having as its ordinates the averages of the ordinates of the other curves. This averaged curve is given in Figure 7. It shows the same characteristics as the other curves, the only apparent difference being a more abrupt rise in the central portion of the left peak. Photographs at intervals over the range investigated in- dicate that the high peak at the left is due to the intense 2537 A. line of the mercury spectrum, as was expected. The smaller peak on the right is dueto the 3151 A. doublet and the 3126 A. line. The intervening ripples are possibly due to weaker lines which showed in the photograph. The ordinate at any point on the curve in Figure 7 may j A vol." 1 \ :1 so as en es Posmou or charms, ARBITRARY umre FIGURE 7 Error in grating setting minimized by the superposition of Figures 4, 5, and. 6. 11 be taken as an indication of the energy incident upon the photocell for the grating setting given by the correSponding abscissa when the cell is illuminated under the specified conditions. The curve is not an energy contour of a portion of the spectrum because the aperture of the photocell was very wide in comparison with.the width of the spectrum lines. Thus it is not the area of the-region under a peak which is indicative of the intensity of a line, but the maximum height of the peak, since at the abscissa corresponding to the maximum point the greatest energy was incident. At any point ot either side of the maximum the response may be less because the most intense part of the line may be falling on a thicker portion of the cell window or be partially obstructed by the mask or internal elements of the cell. §;,Results The maximum for the 2537 A. line rises to a value of 0.245 volts, which was effective over a resistance of 9820 meg- ohms. The sensitivity of the photocell at that wavelength is 8.75 microamperes per milliwatt according to the cal- ibration curve obtained from the manufacturer. This gives 9 an intensity of 9.07xlo' watts incident upon the photocell. A maximum of 0.0275 volts across the 9820 megohm resistor with a cell sensitivity of 1.825 microamperes per milliwatt 12 for the 3131 A. line gives an intensity of 1.537110"9 watts. In both of these cases nearly all of the energy was con- fined to an area of only a few square millimeters as con- firmed photographically, but the area of the cell window. 'exposed was greater than two hundred square millimeters. Thus any determination of the energy incident per unit area is likely to be quite misleading. However, the re- sults obtained give the energy under the conditions of operation. The experiment which has been described gives numerical answers, correct to about 5 percent, to the question, "How many watts of monochromatic radiation in the region 2300 A. to 5200 A. are falling in bands about 100 A. wide on the central rectangular area, 13 mm by 17 mm, of the PJ-405 photocell?" The writer wishes to express his appreciation of the in- terest shown in this work and of the loan of equipment by many members of the Physics Department and by Professor earl P. Swanson of the Botany Department of liehigsm State college. Professor Thomas H. Osgood and Professor 0. D. Reuse of the Physics Department contributed many ideas and suggestions. Their guidance and inspiration throughout the course of this work is gratefully acknowledged. 152.??? o_zo_2~_m=._. m manor. magnet. «so m EEEEZfiE .m . w .. ‘ .> 0% arm :85 - ¢ , noes 6 Film . 500-1”. Gown >._..>_._._mzmw nqmuohbzm a 552.... «Sextei z. zeezunuiz. e8» 8e» 88 30a 00%“ 3 a a .LLVM I‘I‘llw 83d. 93!! HJWVOU SIN 23 6n ream 12.21...." m J. S. Anderson, Editor, Photo-eleetrie Cells'g Their.Ap- lications (The Physical and OptIcaI Societies, Eondon, N. R. Campbell and D. Ritchie, Photoelectric Cells (Sir Isaac Pitman & Sons, Ltd., London, 4 . W. W. Coblentz,,5 Portable Vacuum Thermo ile (Bureau of Itandards Research Paper S4I5,’vol. I7, I520-1922). W. W. Coblents, Constants'gg Spectral Radiation'gf Uni- forml Heated IncIosure, or Spree e ack o ,._ (Bureau of Standards Sullztin 8504, voI. 10, 19 4). W. W. Coblents, Measurements 0n Standards of Radiation i3 Absolute Value (Bureau of Sfandards SuIIEtIn SSS7, vol. II, I§l . W. W. Coblentz and R. Stair, A Standard Source of Ultra- violet Radiation for Calibrating Photoelectrfc'53sa e {Etensit"Mbters‘TRSsearch‘Paper‘158, Bureau of Standards SournaI of Research, vol. 16, Feb. 1956). W. W. Coblents and R. Stair, Evaluation‘gf Ultraviolet Solar Radiation of Short Wavelen the (Research Paper S77, Bureau of standards ourna 0 Research, vol. 16, Apr. 1936 e W. W. chlents and R. Stair, The Present Statuslgf the Standards of Thermal Radiation maintained ¥y the Bureau of Standards (Research Paper 575, Sureau 0 Standards '33urnaI of Research, vol. 11, July 1955). W. I. Coblents, R. Stair, and J. M. Hogue, Tests of'g Balanced Thermocouple'ggg Filter Radiometer as‘g'Standard UItravIole osage ntensit Ester ( Research—Paper 450, Sureau of Standards Journa of Research, vol. 8, 1952). W. E. Forsythe, editor, Measurement 23 Radiant Energy (MbGraw-Hill Book Company, Inc., New York and London, 1937 . A1 D. Hughes and L. A. DuBridge, Photoelectric Phenomena (McGraw-Hill Book Company, Inc., Few York and—London, IS52). M. Duckiesh, Ultraviolet Radiation (D. Van Nostrand Company, Inc., New York, 27 . John Strong, with others,.Prccednresuin.Experimsnia1.2hzsica (Prentice-Hall, Inc., New York, 1941). V. K. Zworykin and E. D. Wilson, Photocells and Their A 11°?t10n (John Wiley & Sons, Inc., New York, 1952). pHYSIcs-MATH L'- 11111)) (IN) “(“11 ((1)1111) 111111)) 3129 93017