UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Photo-conductivity in lead selenide Lee, Peter Arthur 1953

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1953_A1 L3 P4.pdf [ 3.84MB ]
JSON: 831-1.0085953.json
JSON-LD: 831-1.0085953-ld.json
RDF/XML (Pretty): 831-1.0085953-rdf.xml
RDF/JSON: 831-1.0085953-rdf.json
Turtle: 831-1.0085953-turtle.txt
N-Triples: 831-1.0085953-rdf-ntriples.txt
Original Record: 831-1.0085953-source.json
Full Text

Full Text

PHOTO-CONDUCTIVITY IN LEAD SELENIDE by Peter Arthur Lee A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY i n PHYSICS We accept this thesis as conforming to the standard required from candidates for the degree of Doctor of Philosophy. ^embers of"the Department of Physics THE UNIVERSITY OF BRITISH COLUMBIA MAY, 1953 ABSTRACT The method of preparation of photoconductive layers of lead selenidte i s described. Details are given of the construction, of cells using lead selenide as a detector for use i n the infra-red region of the spectrum up to 7 microns*. One such c e l l i s shown to have a minimum detectable power of 1 0 " ^ watts within i t s sensitive spectral range„ Experiments, are described on the conductivity and wavelength sensitivity as a function of the temperature of the layer. The effects of oxygen and heat treatments on such a layer are shown. These observations are used to draw conclusions concerning the mechanism of the photoconductivity i n layers of lead selenide. PREFACE The research described i n this thesis was supported by a contract with the Defense Research Board of Canada,, I wish to thank the University of British Columbia for a Research Fellowship granted to me (1950-1953) to enable me to carry out this work. I am indebted to Dr. A.M. Crooker who suggested and directed the problem, and whose assistance and encouragement are sincerely appreciated, and to Mr.•J. Lees for his help with the construction of the c e l l blanks described i n this thesis. The author also wishes to acknowledge the help received from F.G, Moote and J.P.I, Tyas with the construction of the monochromator and the electronic equipment. P.A. Lee May 1953 TABLE OF CONTENTS -Chapter page I I n t r o d u c t i o n . . . . . . . . . • . . * • • • « • • < > « • 1 II Construction of Cells . . . . . . . . . . . . . . . . . 5 III Test Equipment . . .. .., . . . . . . . . . . . . . . 14 IV Results ....... ., ., . . . .. 26 V Mechanism of Photo-conductivity . . . . .« • . • . • • 41 ¥1 Conclusions • • .. • •• . . .- .. . .- « <• • • 52 Appendix I . . . . . * . . . . . . • » . » » • - • • • 55 Appendix II , • < # . • .- .- ... . ... » •• • . . . . . . 58 References . .. ., * * . . .. • •, • • •- ,» • • • 59 ILLUSTRATIONS To Follow Figure • P a § e 1. Transmission Curves . . . . . . . . . . . . . . « » » • 7 Z Lead Selenide C.ell Blank . ... ., ..• ... .. «. »- •« •> •» 8 3' Schematic Outline of Vacuum System and Oven, .. .. .-. . • 10 4 Lead Telluride Cell, Blank ... ., .. ... ... ... ... .< 12 5: Infra-red Mono-chromator and Associated. Optics , .. 14 6. Black Body Radiator . ... ...... . ... .;. .. ... *, ». .» •- 15 7 Black Body Temperature Control Circuit . ... . ..- .> .. • • 18 8 900 c/s. High Gain Amplifier 21 9 Frequency Response Curves. . ,; ....... .* ...... ...•„ ..- «< 22 10 Multivibrator- ....,.., .. ... . „ . . . . . . . . ...,., .« . . . 2 4 11 Effect of Vacuum Bakes ........... ... ... .,... . ., 27 12 Spectral Response Curves for Lead Selenide . . . . . . . 36 13 Temperature Variation of Threshold Quantum Values .. .. . 36 14 Relative Spectral Response for Lead Telluride . . . . . 37 15 Variation-of. Acr/er with Intensity of. Illumination for Lead Selenide at 90°K . . . . . . . . . . . . . . . . . 37 16 Conductivity of Lead Selenide . . . . . . . . . . . . . 38 17 Energy Levels in Lead Selenide . . . . . . . . . . . . 43 18 Formation of n-p Barriers . . . . . . . . . 0 . . . » 43 19 Raising of Surface Levels, Due to Adsorbed Oxygen .... 49 20 Formation of Barrier Beteeen Two Crystallites . . . . . 49 CHAPTER I INTRODUCTION Many substances are known for •which the conductivity alters on exposure to radiation, i n the visible or near infra-red regions of the spectrum., This phenomenon of photo-conductivity was f i r s t observed by Willoughby-Smith.in 1873 (l) for selenium metal, a fact which is well known for i ts application.in the manufacture of selenium cells.' Thus photo-conductivity has been known for over 70 years;, but i t is only within, the last 10 years, that any detailed investigation, of this phenomenon has been carried out.. Up t i l l that time the spectral limit of such photoconductors was about 1.2 microns, and other more sensitive methods were known for use in infra-red spectroscopy, e.g., thermocouple detectors.. Case (2) in. 1917, showed that a thin layer of lead sulphide when suitably treated mth oxygen was superior as a detector in. the region 1 to 3 microns. He also, showed (3) that thallium sulphide may be made sensitive to infra-red radiation. Much work was done on these, and other substances during the last war, particularly in Germany, and the results of this work have been summarised as far as is known by Oxley (4) and El l i o t (5). In 194-6, (6) a thorough investigation*of the preparation and properties of thallous sulphide was made, but apart from this most of the work on photoconductivity has been concentrated on the three substances lead sulphide, lead selenide and lead telluride. The first two have been prepared by chemical deposition (7) (8), but the method of evaporation i n a vacuum is now generally adopted for a l l three.. This latter technique has 2 been.described in. detail by Sosnowski et al., (9) for the case of lead sul-phide in which they point out the important part played by oxygen in making cells of high sensitivity., The .wavelength cut-off is* defined as. the wavelength, art which, the sensitivity has dropped to half its maxnmum., A l l three of the above sub-stances- show a marked increase i n the value of the wavelength cut-off as the layer is cooled, and for this reason most cells are, made in the form of Dewars. This - is in marked contrast to most other substances (10), for which the wavelength cut.-off generally decreases as the layer is cooled. It has been shown. (7) for instance., that the sensitivity of lead sulphide may .be extended from 2.9 microns to 4*1 microns i f the layer is cooled from room temperature at 3Q0°K, to liquid hydrogen.temperatures at 20°K. Similarly (11) the wavelength cut-off for lead selenide moves from 4.0 microns at 195°K to 5.0 microns at 90°K, and for lead telluride (12) from 4.1 microns at 195^ K to 5.0 microns at 90°K. Similar results have been, obtained by others, until recently, when i t was found (13) that lead selenide could be made sensitive to 4.8 microns at 300°K and to 7.1 microns at 90°K.. This result was; predicted from absorption 'spectra measurements (14) of lead sulphide, lead selenide and'lead telluride for which agreement was: obtained with the photoconductive properties of lead sulphide and lead telluride, but not for the wavelength cut-off values of lead selenide as then known. Further advantages of these cells may. be realised from the tests on. their ultimate sensitivity and time constants. Felgett (15) found for lead sulphide, a. minimum detectable power of 5x10^ 13 watts for a 0.2 cm.? layer at 195°K. while i t has. been found (l6)/that a. lead telluride cell •. could be made sensitive to 2xL0"^ watts. This means- that i t i s possible to make detectors whose performance is more than a hundred times better than conventional infrarred. detectors for comparable conditions of operation within their spectral range. However, reports so far have indicated that i t is much more difficult to obtain'lead selenide cells with such high sensitivity. Milner and Watts: (8) have shown, that a layer made by chemical deposition.' Was sensitive to 5x10"^ watts/mm2 and-a: cell, made by the evaporation technique was found (17) to be superior to a^ 'Schwarz; thermocouple, though the actual sensitivity in the latter case was not quoted. Defining the time constant T T of the cell as; the time taken.for the signal to decay to l/e of the maximum, measurements have shown, that TT is of the order of,10"4 - 10"° seconds for a l l three substances, though generally the value of.T is smallest for lead sulphide and increases as the layer is cooled for a l l three of the substances. These properties indicate the extreme usefulness- of these cells, and'-advantage of them has been taken, particularly the short time constant, in designing fast scanning spectrometers for the investigation of rapidly changing spectra in the infra-red region (18). The present investigation was undertaken with two objects in view. The f i r s t was to find a suitable method of making lead selenide cells sensi-tive to long wavelength radiation, and to investigate the properties of such cells. Previous work has shown that to obtain an efficient photoconductor was almost more of an art than a science. Thus for any given method of producing a lead sulphide c e l l , the same worker seldom obtained cells of the same sensitivity, while workers i n different labora-tories.: often obtained quite contradictory results. Various theories have been outlined (see Chapter ,V) not only to corroborate these different procedures^ but also to explain the properties of these c e l l s . The second object, was therefore to obtain a better understanding of the mechanism of photoconduction i n these substances and i n particular i n lead selenide. It should be-added that i n i t i a l l y several cells were made using lead telluride as the photoconductive substance when this was thought to have the longest wavelength cut-off, and a brief account of their properties w i l l be included. CHAPTER II OQHSTRUOTIOH. OF CELLS (i). Formation of. Lead Selenide The samples of lead selenide and lead telluride used i n this work were both prepared i n a similar manner, and for convenience only the pre-paration of the former w i l l be described. As i s well known, the properties of most semi-conductors 'depend a great deal on the composition of the material, particularly with regard to impurities. Both the lead and selenium were therefore obtained as near pure specimens as possible. The former was supplied i n the form of small pellets, and these were sifted so as to use the lead i n as near powder form as possible. The selenium was a spectroscopically pure specimen, supplied i n the form of a rod. This was scraped, and the chips ground to a fine powder.. Equivalent quantities of lead and selenium, sufficient to make about 1 grm. of lead selenide were placed i n a well cleaned s i l i c a tube, which was pumped out to 10 mm.. Hg pressure. The tube was gently heated for about one hour i n vacuo i n order to remove any adsorbed a i r , and the temperature then gradually raised t i l l the contents flashed. Further heating merely made the lead selenide sublime and condense on the upper parts of the tube, and i t was therefore allowed to cool. The tube' was then sealed off i n vacuo immediately after preparation, and kept u n t i l required. It was found, that the sensitivity of the photoconductive layers depended to a large degree on small departures from the stoichiometric ratio as obtained by the above method. A small amount of lead (about 5%) i n excess of this ratio, introduced at the start of the operation, gave a sample of lead selenide which produced the best results. For comparison* cells made using the material from a single crystal of lead selenide gave no photosensitivity whatsoever.. In the case of lead telluride the cells were made using a sample of stoichiometric proportions, and although the layers responded to infra-red radiation the resulting sensitivies were low in each case. No attempt was made to improve on this, as lead selenide was considered to be of greater interest., (ii) Construction of Cell Blanks In order to be able to obtain high sensitivity using lead selenide and lead telluride, i t is necessary to design the cell in the form of a small Dewar, so as to be able to cool the layers: to liquid air temperatures. Since the layer is deposited on the inner'wall of the Dewar, the outer wall must necessarily be "transparent" to infra-red radiation. .Pyrex glass will transmit-only as far as 3 microns, and i t is therefore necessary to incorporate some form of window in the cell design. The requirements of such a window are as follows: (a) Good transmission from 1-7 microns. (b) Ability to seal directly to glass. (c) This seal to be capable of maintaining a high vacuum. (d) Capable of withstanding temperatures up to 550°C. Such a window may be obtained (16) by making use of pyrex blown 7 into the form of a very thin bubble. Several of these were made, about 3/4" in diameter, and a typical transmission curve i s shown in fi g . 1. Its range of usefulness is seen to be only as far as 4.5 microns, where the „ transmission has fallen to 40$. Moss showed (12) that KRS5 (a crystal containing 42/5 Thallium Bromide and 5$% Thallium Iodide) which has a uniform transmission from 1 to 35 microns, could be used as a window. However, i t does not satisfy condition (b), since the.window had to be waxed to the end of the cell, which i n turn does not satisfy condition (d). Two substances which do meet a l l of the above requirements are sapphire and periclase, and the transmission curves of these are also shown in f i g . 1. The latter is seen to transmit well out to 8 microns and i s readily available in this country from the Norton Abrasive Co., Ontario. It is an isotropic crystal of Magnesium Oxide with a cubic structure, and thin slices (1-2 mm., thick) may be readily cleaved from the crude samples supplied. Young showed in 1951 (19) that periclase could be sealed directly to Z8 type soda glass using an induction heater, and at the same time Rawson (20) showed that i t was possible to seal sapphire to C79 soda glass. Sample sapphire windows were obtained in the form of.polished roundels, 16 mm. i n diameter and 1.5 mm. thick. In view of the convenience of supply and comparative cheapness, i t was decided to use periclase as the window in the lead selenide cells. In order to join the X8 glass to pyrex> graded seals 1 inch in diameter were obtained, consisting of the folloxd.ng glasses with the corresponding expansion coefficients as shown; F i g I T r a n s m i s s i o n C u r v e s 8 G.E.C. Chance Chance Chance Chance Chance Pyrex X8 GWA. GKC GKB GK55 GH2 9.7x10' 8.7 7.5 6.4 5.5 4.7 3.3 Periclase has an expansion coefficient of 10.43 x: 10" °C~ (19). An alternative and more convenient method than Young's was developed for making the glass to periclase seal. The end of the seal was drawn down to about 5/8" in diameter and with i t held vertically, the peri-clase was placed on top and gently heated directly by flame.. The temperature was gradually raised, until the glass, was seen to flow and "wet" the peri-clase, by which time the seal was complete.. This was then annealed in an oven at about 600-700°C for several hours. The whole process had to be very carefully carried out to ensure that small cracks did not develop in either the periclase or the glass. Such faults could be reduced by careful choice of the piece of periclase used, that is to say, one free of small cracks or faults at the edges. window is shown in fig. 2. The graded seal and periclase window constitute the outer wall of the Dewar. The end of the inner wall is made flat, on which are painted Aquadag electrodes. External connections are obtained through 1 mm. diameter tungsten leads which are enclosed in glass sheaths, except for a small length at the end to enable contact to be made with the graphite electrodes. A small V shaped bend i s put in the leads to allow for the different expansions of the tungsten and glass when the cell is cooled. The construction of a lead selenide cell incorporating a periclase 2 The leads are fixed at the lower end by means of small glass beads. (i i i ) Formation of Layer In order to avoid exposing the lead selenide to the air, and thus to ensure a careful control of the impurities in the final layer, the following method was i n i t i a l l y used to form the layer on the electrodes. About 30 milligrams of lead selenide were placed in a small tube sealed to -6 the top of the cell.. The cell was pumped out to 10 mm. Hg pressure, and the sample distilled over to the walls at the top of the Dewar by careful flaming, until a fairly thick film was obtained.. (Lead selenide sublimes at about 4.05°C.) The small side tube was then sealed off. Two small air jets were arranged so as to cool the glass walls immediately behind the glass periclase seal and the oven lowered into position as shown in f i g . 3. The cell was fixed at an inclined angle so as to enable the infra-red radiation to be focussed onto the electrodes, at the same time allowing the cell to be cooled with liquid oxygen whilst s t i l l on the vacuum system. The cell was baked out at 500°C in an oxygen atmosphere of 10"^ mm. Hg pressure.. The air jets which were mounted through the base of the oven, served the; double role of cooling the soft glass to prevent i t collapsing, and also to condense the lead selenide. This pro-cess took about one hour, at which time the oven was removed and the cell allowed to cool.• Liquid oxygen was added to the cell and the lead selenide carefully "flamed" over onto the electrodes, in an oxygen atmosphere at a pressure of 10"3 mm. Hg. This method was used in attempts to make about ten cells, but i t was found in every case except one that small cracks developed in the glass to 10 periclase seal, thus impairing the vacuum.. This was due to the stresses i n the seal as a result of the large difference i n the expansion coeffi-cients of the soft glass and periclase. In view, of the d i f f i c u l t i e s met with i n the above method, the following alternative procedure was adopted... I t had the distinct disadvan-tage that the layer was exposed to the air , but since i t was found possible to obtain photo-sensitivity, the method was used for a l l subsequent c e l l s . I n i t i a l l y the Dewar was made with a pyrex outer wall extending about three inches below, the inner one.. A. small quantity (about thirty milligrams), of lead selenide was placed at the bottom of the outer wall, and the c e l l pumped out -to 10"^ mm., Hg pressure. The lead selenide was heated very gently at f i r s t to remove any adsorbed a i r , and then with the inner wall cooled with li q u i d oxygen, a f i l m was formed by gradually raising the temperature. A very large amount of lead selenide was also deposited on the outer wall, which by reheating could be made to sublime on the elec-trodes. This meant that a large part of the f i l m was made up of lead selenide which had sublimed twice. This appears to improve the chances of obtaining high photo-sensitivity. This process was continued u n t i l a thick layer was. obtained, opaque to a 40 watt tungsten lamp.. Air was l e t into the c e l l which was then removed from the vacuum system and the pyrex outer wall replaced by the graded seal and periclase windovr. The c e l l was re-sealed onto the vacuum system at an inclined angle as seen i n f i g . 3. The layer had a distinct metallic appearance which was generally dark grey by reflexion, with a characteristic purple colour when viewed with transmitted light. Over oxidation, as obtained i n some sensitization processes, changed Mc r c u r y D i f f u s i o n P u P u m p g=5 S c h e m a t i c O u t l i n e of V a c u u m S y s t e m a n d I B a c k i n 9 11 this, to a distinct l i g h t grey, and the layer lost a l o t of i t s mirror-like appearance. (iv) Sensitization of Cells. Fig. 3 gives a schematic outline of the vacuum system used for the formation of the layers and their subsequent sensitization. A mercury diffusion pump i s used to avoid any possible contamination of the layers with o i l vapour, but i n any "case.a li q u i d oxygen trap i s incorporated to remove any unwanted vapours. .. Small quantities of either a i r or oxygen may be l e t into the c e l l by means of the glass taps Tj T2 and T^ T5 .. The oxygen i s obtained by heating potassium chlorate. The pressure i n the "cell i s measured using a Pirani gauge. I t was found necessary to degas the wire element of this gauge periodically due to deposition of lead selenide on. the surface. This was done by connecting the element across a variac, and slowly raising the voltage to 40 volts, thus raising the temperature to white heat, at which point the voltage was quickly reduced to zero. The connection to the c e l l passes through a hole i n an asbestos sheet which serves as the lower side to the oven. The oven i s arranged to move vertically, and when removed, allows the radiation from the infra-red monochromator to be focussed by means of an 8 inch diameter concave mirror onto the electrodes. The temperature of the oven i s measured with an iron-constantan thermocouple and temperatures up to 550°G may be obtained. The heating elements consist of four 125 watt heaters fixed to the asbestos sides of the oven, which i s insulated with at 3 3/4 inch thick layer of rock wool.. The whole i s contained within an aluminium box 17" x 16" x 16"., 12 The layer when i n i t i a l l y formed usually showed no photo-sensi-t i v i t y whatsoever.. A' quick test was obtained by cooling the c e l l vdth liquid a i r , and then observing the change i n resistance of the layer when illuminated with a tungsten lamp. It was found that the a i r admitted into the system, prior to replacing the pyrex wall by the graded seal, made the layer immediately sensitive.. This change i n sensitivity was observed i n greater detail for some of the later cells as shown i n Chapter IV. Sub-sequent illumination with monochromatic radiation, after replacing the periclase window showed that this sensitivity was generally only out ass far as 2.5 microns.. Short vacuum bakes of 10 minutes each at 200°C generally moved this wavelength response out to five or six microns when the c e l l was subsequently cooled to 90°K. This stage of the procedure proved to be very tricky, and great care had to be exercised not to destroy the photo-sensitivity. I f necessary, oxygen bakes would sometimes restore the response, which would lend support to the idea that a vacuum bake effects the removal of oxygen adsorbed i n the layer (8). These oxygen, bjakes were not so c r i t i c a l as the vacuum bakes, and were generally carried out for 1/2 - 1 hour at say twenty microns Hg pressure.. Experience indicated the best values: to adopt i n each particular case. On obtaining the best results the c e l l was sealed off from the vacuum system, which vias kept pumped at 10"^  mm.. Hg pressure. (v) Lead Telluride Cells. These were made incorporating a bubble window/ as shown i n f i g . 4. The hemispherical shape i s important since i t enables the pressure to be 7~S A -l'/8-V Aqua dag E l e c t r o d e s Pla t inum S t r i p . Tungsten Leads. I mm. diam. F i g . 4 L e a d T e l l u r i d e C e 4 '13 taken up as a pure tensile stress round the circumference of the joint. For the several windows that were made, the optical distortion, was found to be negligably small.. The electrodes are formed by a single strip of Aquadag 5 mm., wide, a gap of 1 mm. being l e f t opposite the window. External connection i s obtained through tungsten leads sealed i n at.the base, contact to the graphite electrodes being made by platinum strips i n order to avoid excessive strain between the inner and outer walls of the c e l l . About five milligrams of lead telluride were used for each film, and were placed i n the c e l l before i t was sealed onto the vacuum system. The c e l l was evacuated to 10"^  mm. Hg pressure, and then heated to 500°C by means of the oven which could be lowered over the c e l l . The vacuum system and oven were the same as that used for the lead selenide except that the c e l l was fixed i n a vertical position. After one hour the oven was. removed and the lead telluride was found to have condensed over the lower parts of the Dewar. (Lead telluride sublimes at 4.50°C.) By heating the outer wall and bubble window very carefully with a flame, a uniform film was formed over the electrode surface and the surrounding parts of the inner wall of the cell.. No detailed investigation of these cells was carried out, though i t was noted that short oxygen bakes improved the sensitivity, but did not change the wavelength cut-off, (about U»5 - 5 microns at 90°K). CHAPTER III TEST EQUIPMENT This chapter is. devoted to a description of the various pieces of apparatus that were assembled in order to determine the characteristics of the cells. The information that was required was the variation of the response of the cell to change in wavelength of the infra-red radiation, the signal to noise ratio and the time constant.. (i) Infra-Red Monochromator <• Fig.. 5 gives a schematic outline of the optical system of the Perkin Elmer monochromator together with the associated optics.. This i s of the Littrow type, and mirrors are used throughout as i s necessary in most infra-red optical systems.. A globar operating at 1400°K is used as a source of black body radiation, and the output from this is focussed onto the entrance s l i t of the monochromator by means of mirrors Ml and M2. The former mirror, which is flat may be rotated so as to allow the light from a mercury arc to be used as an alternative source. The green line in the mercury spectrum forms a very convenient way of lining up the optical system visually. The radiation falling on the entrance s l i t is interrupted by a mechanical chopper at 900 c/s. This consists of a brass disc'with 30 holes, 0.7 inches in diameter drilled round the periphery, and is rotated at 1800 r.p.m.. by a synchronous motor. M3 is an off-axis parabolic mirror which allows a parallel beam to f a l l on the rock-salt prism P. The light i s F i g 5 I n f r a - r e d M o n o c h r o m o t o r a n d Ass o c i a t e d O p t i c s . 15 reflected back through the prism by the Littrow. mirror M4> and is focussed on the exit s l i t by M3 and M5. It is found convenient to test the wave-length response of the cells whilst s t i l l on the vacuum system, and the output from the monochromator is focussed onto the electrodes by mirrors 148, M9 and Mil.. The cell is about 5 feet from the exit s l i t . MS may be placed in the exit beam in order to focus the radiation on the thermocouple T.C., This is assumed to have a. 100$ response to a l l wavelengths and i s therefore used as a standard in calculating the relative spectral response curves of the cells. Rotation of M9 provides an alternative focal point of the output, to enable further tests to be carried out on the cells when completed., The sensitivity of a cell may be defined as the output signal i n microvolts for 1 microwatt of radiation incident on the electrodes. The energy emerging from the exit s l i t obviously depends on the s l i t width, wavelength and dispersion of the prism. This amount of energy may be readily calculated knowing this data, as is shown in Appendix I. (ii) Black Body Radiator In order to attach any significance to the signal/noise ratio i t is necessary to have a standard signal. This is defined as the signal received by a detector placed 20 cms. from an aperture 3 M a s . in diameter in a black body at a temperature of 500°K. The construction of the black body i s shown i n fi g . 6. The radiating cavity consists of a conical hole drilled in a steel cylindrical /2 S c a l e . 16 block 6 cms. i n diameter and 10 cms. long.. The conical hole has an opening of 1 cm. diameter, with a. taper angle of 15° . The steel was originally oxidized at 800°C for several hours, and should thus have an emissivity of about 0.8. I t has been shown (21) that a cavity whose emissivity i s 0.75 and has a length to diameter ratio of 3.8 should give 99$ black body radiation. For above cavity: length - cot 15/2. - 3.79 diameter 2 * A small hole i s d r i l l e d just behind the apex, of the conical hole i n which i s inserted an iron-constantan thermocouple. This i s used for measuring the temperature of the cylinder, which i s heated to 500°K. The cylinder i s cemented into an alundum cylinder, around which i s wrapped a nichrome heater, wound non-induetively. This i s further imbedded i n cement, and on top i s wound a platinum resistance thermometer. This forms part of a bridge network and i s used for controlling the current i n the heater winding as described later. To reduce excessive heat losses, the black body i s well lagged with thermal insulating rock wool and mounted i n a wooden box. A series of aluminium baffle plates i s mounted between the source and the .aperture plate i n order to ensure that the plate remains at room temperature. The baffles are painted black on the side facing the black body, the other side being l e f t polished. .A hole 1 cm. i n diameter i s d r i l l e d through the centre of each of the baffle plates. The radiation incident on the aperture plate may be reduced to room temperature radiation 17 by means of the shutter, which may be moved i n a direction perpendicular to the optic axis of the system as shown. The( radiation is further interrupted by a chopper wheel, operating at 900 c/s in a similar manner to that used in the monochromator. The aperture plate consists of a rotatable disc in which holes are drilled ranging from 1 to 8 mms. i n diameter, t As already indicated, a 3 pn.- aperture is generally used for test purposes. This aperture forms a virtual source of black body radiation with the same temperature as that of the cavity. The detector is placed 20 cms, from the aperture at the end of a brass tube blackened on the inside. This tube screens the cell from stray radiation, e.g.. light, heat from the chopper motor, etc.. .The whole apparatus is mounted on a framework which may be used in either the horizontal or vertical position depending on the design of the cell being tested., The total radiation received by the detector'will'be given by: (f I T4 _ T4 ) A A Q - 2 1 i - 2 — - watts TT d 2 where:. o~ = Stefan's Constant = 5.67xl0~10 watts cms"2 °C"4-A-j_ - aperture area z ~R x(0.3)2 cms2 4 - A 2 s sensitive area of cell T.2 • temperature of black body s- 500°K Tl 3 temperature of chopper wheel a 300°K (say) IB d = distance of c e l l from aperture - 20 cms,: Q - 1.74 x 10'"-' watts cms"2 A2 ( i i i ) Temperature Control Various circuits have been described. (22), (23), (24), for controlling temperatures using a wire resistance thermometer as monitor i n a bridge network. They a l l however, suffer from the disadvantage that they w i l l only control the temperature over a few degrees either side of that required, and cannot therefore be switched on t i l l the temperature i s reached. The circuit shown i n gig. 7 overcomes this d i f f i c u l t y and provides a continuous control from room temperatures up to 500°K. Further, by altering the constants of the bridge network, the temperature may be stabilised anywhere from 400°K to 600°K. Using a potentiometer, the regula-tion was found to be better than * 0.1°K at 500°K, over a period of one hour.. The nichrome heater Rj, for the black body i s connected i n series with a thyratron ( FG57 ). The magnitude of the heater current i s then determined by the relative phase between the anode and grid voltages on the thyratron. The resistance thermometer R^ forms part of a Wheat-stone bridge, the balance of which may be varied by means; of two potentio-meters RJJ. ( 200 and 500 ohms respectively ). The out of balance voltage from the bridge i s mixed with another from a phase shifting network, and the resulting signal i s applied to the grid of the amplifier tube 6SK7. The 5 V _»nonaaoq SOK 6SK 7 nrmnrmr l M -02 56K S6K 1 CH SOK IM I K 20 3 0 H 5 Y 3 M 6K6 VR I OS © VR OS • 8 =F 8 O l IO K F G 5 7 6-3 V 5 V M S 2 \ Red * 6-3V I IO V F i g 7 B l a c k Body T e m p e r a t u r e C o n t r o l C i r c u i t . Grce n 19 signal i s then passed to the cathode follower 6K6, and the output fed onto the grid of the thyratron. The phase shift network i s supplied from a 6.3vwiniing on the transformer and this can be adjusted to have a phase - 120° with respect to the voltage on the thyratron anode.. This i s set, so that when the bridge i s balanced the resulting signal produced at the grid of the thyratron gives just sufficient anode current to maintain the temperature of the black body constant, i . e . to counter-balance the heat losses. A pre-set gain control i s inserted between the amplifier and the cathode follower and adjusted so as to give a signal below the point at which "hunting" occurs. About 4-5 minutes i s required for the black body to reach an. equilibrium temperature of 500°K. Should the i n i t i a l heating period be required to be reduced, a by-pass switch could be inserted across the thyratron. (iv) Tuned Amplifier As already indicated i n the descriptions of the monochromator and black body, the radiation incident on the c e l l i s interrupted periodic cally at 900 c/s. I f the c e l l i s supplied with a suitable polarising current, then an A.C. voltage A V i s obtained = i A R where AR i s the change i n resistance of the layer due to the incident radiation. The value of this voltage i s usually of the order of microvolts, and i t i s therefore necessary to have a high ggin. amplifier.. Further, since the signal i s of fixed fre-quency, i t i s convenient to use a tuned amplifier, thus reducing the amplification of the noise signal i n the c e l l and i n i t i a l stages of the 20 amplifier.. Two such circuits-have been described in the literature (25), (26).. An amplifier was built on the design of Kuiper et al., (25), but was found to suffer from two serious defects. Firstly, the circuit has only one tuned stage consisting of an LC circuit tuned to 900 c/s in the grid of the second tube of the main amplifier and thus the signal to be measured must always be exactly this frequency in order for the gain to remain con-stant. Since the chopper wheel is driven by a synchronous motor this fre-quency will change for the slight variations always obtained in the fre-quency of the power supply , with consequent variations in the gain of the amplifier. The second difficulty encountered was the lack of stability of the amplifier at high gain. The-, inclusion of a negative feedback loop failed to reduce the oscillations., A small positive feedback loop was included in the original design which was so arranged as to give an effec-tive controllable negative resistance across the LC circuit.. This had the effect of varying the Q of the circuit and consequently the bandwidth of the amplifier. Removal of this also failed to improve the stability at high gain. An attempt was also made to build an amplifier using an RC network of the twin T type as the tuned circuit. In this type of design i t is necessary to use the network in the negative feedback path; at the fre-quency at which the network has a maximum impedance there is no effective negative feedback, and the amplifier gain rises to a maximum. The gain stability is thus no better than that of a simple amplifier, and at high gain was found to be extremely unstable. This was particularly so in this 21 case, since 900 c/s is the 15th harmonic of the frequency of the power supply, and any small signal provided by heater hum or otherwise, produced unstable oscillations.. At very low frequencies however, e.g. less than 10 c/s, such an amplifier is found to be very good (27). The amplifier finally adopted and found to be suitable for this work is shown in fi g . 8. It is seen to consist of three sections of three tubes each., Negative feedback is obtained by connecting together the cathodes of the first and third tubes (VI and YJ, V4 and V6, V7 and V9). Two LG tuned circuits are used, being placed in the anode leads of the third tube (V3 and V6 respectively). They are each tuned to different fre-quencies 890 and 910 c/s respectively, and the output transformer is tuned to 900 c/s in the primary... The effect of this is to give a flat topped response curve over a band of - 5 c/s so that the frequency stability of the chopper whael is not critical.. The output transformer T which has a ratio 1:1 from the primary to each half of the secondary feeds the diode rectifier V10, and the resulting D.C. signal is smoothed by a 1000 micro-farad condenspr. The output may be read on either a 200 micro-amp. meter Aj or a Brown pen recorder. The tuning coils used are 1 henry toroidal inductances and each is screened \-dth double cans, the inner one being insulated from earth. This provides screening from both electrostatic and magnetic pick-up. In both cases i t is found convenient to reduce the Q of the circuits by adding parallel resistances Rl6 and R4I. The output transformer is mounted below the chassis to screen i t from these two coils. Figure 8 . To follow page 21 (For values of components see Appendix II, page 58) R4 :RI C 2 R6 ; R 7 * R 0 C6 C7. RI3 R8 C 3 V I e R2 ?R3 CS R5 : 04 R l l R i 6 i i L i t ° 3 C9 V 2 H I -RIO 7- 7-RI2 R 2 0 CIS V3 p i * CI4 H C 8 i W 4 - | R I 7 R32 R36 CI6 CI 7 C2Q R 33 [R3 4 V 4 C 2 4 r 1 >R42 C2I ;R3.7 R3I - W W * ' -CI8 R 3 « R44 1 , R4I I gl_2 y C 2 5 V6 C 2 7 C 2 3 R 3 5 ?R38 I t ! 5R40 JR43 ' C 2 6 3 0 0 Volts S t a b i l i s e d . To Pe n R ec o r d e r F i g 8 9 0 0 c / s H r g h G a i n A m p l i f i e r . 22 Wirewound resistances have been used throughout in the f i r s t section of the amplifier. To obtain the best results, i t was also found necessary to mount this first section on a separate chassis. The potentiometer Rl, is used to vary the polarising current in the photo-conductive cell which is connected at the input by a screened cable„ Paper condensors, in place of the usual electrolytics are used to reduce the noise, and VI is a 1620 tube which is a low noise 6SJ7.. The f i r s t two sections are each adjusted to have equal gains of 52. db; this equality is obtained by varying slightly the two damping resis-tances R16 and R4I.- The output section has a gain of 42 db. and thus the overall gain i s ; 52 + 52. ±-. 42 a 146 db. This is equal to a voltage gain of 2 x 107 . The amplifier response i s found to be linear for a l l settings of the attenuator, and at maximum gain, f u l l scale deflection (lOmV) of the Brown recorder is obtained for 4 micro-volts input. The frequency response for two different input levels is shown in fig., 9. The almost complete identity of the two curves shows spurious, circuit interactions are negligible.. The bandwidth also remains fairly constant for a l l settings of the attenuator at 30 c/s. The noise signal obtained with the input terminals shorted is equal to- 0.2 microvolts and on open circuit the noise is: 0..95 microvolts, which corresponds practically to that of the Johnson noise of the input circuit (1..8 megohms).. With the input short circuited the equivalent noise is 8 x lo£ ohms; which is very much less than the resistance of most photo-conductive cells1 usually of the order of several megohms. o u t p u t /y~°~°\ 160 -1 4 0 - J \ 120 - 1 3 0 c y c l e s \ 1 OO - / \ %o ^  J \ 6 0 - y I npu t a 3 - 9 8 / j v O , 4 0 -2 0 -O 1 1 1 1 1 1 1 1. 1 1 1 1 8 6 O 8 . 8 0 9 0 0 9 2 b 940 t r e q . c / s o u t p u t iso -• •- \ 1 6 0 ~ 1 4 0 - / V 1.2 O -A 3 2 eye 1 e s \ 1 OO - I \ 8 0 - J V . 6 0 -/ Input =3-16 mv \ ^ 4 0 -20 -O 1 ' „ ' \ « 1 1 1 1 . 1 . ' 1 " "T r— 8 6 O 8 8 O 9 0 0 9 2 0 940 f r e q c / s F i g 9 F r e q u e n c y R e s p o n s e C u r v e s . 23 A' stabilised power pack supplies the H.T. voltage of 300 volts, and to reduce any unwanted signal from the heaters a potentiometer is connected across them,'the slider being connected to earth. To prevent chassis currents producing unwanted coupling between the circuits, a l l the negative H.T. connections are wired to a common bus bar, and the line grounded to the chassis at a point near the output. (v) Time Constant Measurement The time constant ^ of a photo-conductive cell is defined as the time taken for the amplitude of the signal to decay to l/e of its maximum value after the removal of the illumination. The value of tr may range anywhere from a few microseconds to several hundreds of microseconds. In any case i t is out of the question to measure i t using a mechanical chopper. It is not thought that "C varies very much with wavelength, and i t was therefore decided to use a neon lamp as a source. The advantages of this are 'twofold: 1) The neon spectrum is fairly rich with lines in the infra-red region of 1-2 microns (28)., 2) The time constant of a neon tube is usually less than a microsecond, and in any case almost certain to be smaller than that of any photo-conductive cell.. No attempt was made i n this work to distinguish between the rise and decay times, which in any case are known to differ by only small amounts and for practical purposes a knowledge of one of them is sufficient. 24 The method used therefore was to apply a•square wave to the neon tube by means of a multivibrator. The pulses of light emitted were allowed to f a l l on the cell, and the resulting signal was passed through a small pre-amplifier and then displayed on the n plates of an oscilloscope. The square wave applied to the neon tube was connected to the 12 plates, and the time constant deduced by comparison of the square and exponential waves.. The neon tube and the cell were electrostatically screened from each other, using a piece of fine metal gauze to reduce pick-up. The anode to grid couplings of the multivibrator (fig. 10) are made through cathode followers (V3 and V4) in order to maintain fast edges on the square waves, particularly at slow speeds. Two anode catching diodes (V7 and V8) are used and these determine the amplitude of the output signal. In this case 120v is required to s*dtch the neon tube. The frequency of the square wave is determined by the values of Rl, CI and R2,C2 and these are made variable so that signals ranging from 5 microseconds to 10 mi l l i -seconds may be obtained., In each case the edges are of less than 0.5 micro-seconds duration. The output from the multivibrator is passed through two cathode; followers Vi and V6, the signal from the former being connected to the Yl plates of the oscilloscope and the latter to the neon tube via a screened cable. In order to measure f , (29) the square wave is arranged • symmetrically about the exponential wave, and the amplitudes of the two to be equal at low frequencies (i.e. very much less than f ), 4 7 47 47 [SOK VI V 7 Time Base v-4 7 0 I 5K 7W VI , V 6 / » 6 A G 7 V 2 , V 5,sVR 91 V3 , V4 x s 6 A K 7 V 7 , V8,= VR92 V2 SO ,'47 47 A7 O l V3 4 7 0 O l CI -II-250K/47? <47 It <SOK V4 S47 V5 C 2 « H h Rl $ R 2 _ t h _ 4 7 0 V6 1.20 V > 3 O O V — 47 470 15 K 7 W I5K 7.W Neon ISK 7W > -I 5 0 V F i g . 10 M u l t i v i b r a t o r . 25 It follows, that! ° l-+exp-t/r •exp-t/ where: A - amplitude of the exponential, wave A 0 - amplitude of the square--wave t = l/2f " f s frequency of the square wave For low frequencies t t • and A S A Q (ii) For high frequencies t « t A N D A s tt ( i i i ) 2; . . . (ii) and ( i i i ) are the equations to two straight lines which intersect at the point tr - t/2 = l/4f. Thus by measuring the variation of A with the frequency f of the square wave, the value of the time constant of the cell may be readily deduced. 26 CHAPTER IV RESULTS. A. Sensitization of Cells The necessary requirements for a good detector are! (i). Long wavelength response,, (ii) High sensitivity (defined as the output of the detector per unit of incident radiation power),, (iii). Low noise. Lead selenide, in the form of a micro—crystalline layer as used in these experiments, is very susceptible to any form of heat treat-ment, and in this way attempts were made to obtain detectors conforming to these three requirementsi. As described in Chapter II, this sensitization, process could be made by means of both oxygen and vacuum bakes at varying temperatures and duration of bakes, and the results of these treatments and the relevant observations on. the above properties for cells 11-28 are noted in Table I,. The i n i t i a l experiments using cells 1-10 were carried out with the first method to deposit the layer, described in Chapter II, and no useful results were obtained due to cracks developing in the graded seal and periclase window. The second method was therefore used in a l l the subsequent cells. The first relevant factor to be observed was that most cells, when in i t i a l l y made were only sensitive out as far as 2,5 microns. 27 Subsequent vacuum bakes generally moved this long x-ravelength response gradually out,.and this effect is shown graphically in f i g . 11 for cell No. 13'. This graph gives the. response of the cell relative to the Eerkin Elmer thermocouple plotted against the wavelength. The resistance of the layer-at both room temperature (R300) and liquid air temperature (R90) and the value obtained when illuminated with a tungsten lamp (Rw) were also noted subsequent to each heat treatment.. The ratio R90/Rw gives a useful indication of the sensitivity of the cell.. Further i t is seen that the effect of a vacuum bake is to increase the value of R300 and R90, and for an oxygen bake the reverse is generally true., This gave rise to a serious difficulty* The signal obtained using the amplifier is dependent on the resistance of the cell not being too high, in order to be able to match the input impedance of the amplifier. It was therefore necessary to obtain the cell with a long wavelength response with the resistance not greater than say a few megohms.. This also helped to reduce the noise signal which would be expected to arise mainly due to Johnson noise. The values, of the sensitivity in volts per watt in Table I were calculated as shown in Appendix I. A - a f t e r f o r m a t i o n of l a y e r B - v a c b a k e f o r 3 0 m i n s at 2 2 5 ° C C - • •> « • I I n H | 6 o° C mi c r o n s F i g . II E f f e c t o f V a c u u m B a k e s . TABIE I. SENSITIZATION OF CELLS (For an explanation of the symbols, used in the following table see page 31) Cell No. Treatment Response Remarks Bake P t T R300 R90 Rw N. Y/fc- Ac 11 Laye V V V Y V V V. 0 r f c 10 rmed 10 15 30 30. 30 30 30 20 200 200 200 200 200 200 250 250 18k 8k 8k 9k 10k 14k 2k 1.7k 4.5k 2M 25k 30k 60k 100k 300k 2.6k 2.2k 7.5k 30k 10k 10k 121c 15k 25k 2.6k '2.2k 7.5k 3 1.8 1.23 1.21 2..06 4.95 7.4 18.5 2.5 3.0 3.2 4.5 4.8 4.8 No response No response Cell removed 12 Laye .0 r f c 10 rmed 10. 30 200 200 4k 6k 130k 2M BI 7M 8k 7k 300k 10.2 12.8 9.5 18..5 58.2 51.4 2.0 3.5 5.5 Response good Cell removed Laye .1: 1 V, r f c rmed 30 30 30 225 180 225 65k 200k 200k 111 3M 10M 20M 50M 56k 400k 450k 50M 66 1.5 2.0 8.0 178 254 568 1.2 2.8 4.5 5.5 Noisy Response good Poor response Cell removed 14 Laye: V r f o rmed 10 250 5k 18k 18k 700k 10k 500k 350 Noisy. Layer flaked badly. Cell removed 15 Laye: V r f o rmed 10 200 28k 85k 10M 20M 70k 100k 33 1.0 26.8 45.8 5.5 5.5 Noisy Response good Cell removed 16 Layei . V V. 0 0 V. c f o 10 10 rmed 20 30 15 15 30 200 300 200 200 225 2.3k 2k Ik 7k . 5k 0.8k 4.2k 6k 1.2k 110k 45k 0.9k 3.0k 5k 1.21c 80k 35k 0.9k 3.5 Poor response No response No response No response No response No response Cell removed 29 Table I continued: Cell No. Treatment. Re si ponse Remarks Bake P t T R300 R90. Rw. N Y/W Ac 17 Laye V r f o rmed 10 200 30k 20k 260k 50k 30k 20k 50 : 6.5 10.2 37.6 5.5 5.5 Response gooc Cell removed 18 Laye V. V V ? r f o rmec 10 .15 30 45 L 200 200 200 250 40k ,. 20k 12k 16k 75k IM 50k 70k 110k 20M 40k 20k 18k 18k 500k 180 5 •3*75 5i5 180 15.2 1.11 4.64 18.8 25.2 2.5 3.0 3.2 3.5 3..5 Poor response Noisy Cell removed 1 19 Laye V V V V r f o rmec 5 15 20 30 I 200 200 200 200 66k 65k 125k 400k 2M 20M . 2M 20M 20M 50M 50k 50k 150k 700k 10M 15 15 6 2.75 41.8 133 220 293 0.34 2.5 3.0 3.2 3.2 Poor response Cell removed 20 Laye v. V r f o rmec 10 5 I 200 200 65k 200k 220k • 1*5M 50M 50M 55k 250k 280k 7.5 •2,7. 2.7 19.9 43.1: 102 2.5 4.5 5.5 Response good Cell removed 21 Laye r f o rmec I Layer flaked badly. Cell removed. 22. Laye V 0 0 V V 0 V r f o 10 10 30C rmec 30 60 60 30 60 60 5 . 200 350 350 250. 350 300 200 20k 150k 22k 18k 20k 100k 12k 18k 700k 50M 10M 51-1 2M 50M 600k 214 20k 300k 150k 24k 20k 50M . 50k 55k 18 4.5 2.2 61.4 149 12..6 2.56 3.17 382 506 2.5 5.5 5.5 2.5 2.5 6.5 6..5 No response Response good Response good Cell removed 30 Table I continued*. Cell Treatment Response Remarks No. Bake p t T. . • R300 R90 Rw. N V/W A t 23 Laye r f o rmec 40k IM 30k 45.6, 2.0 0 200 60 300 100k 50M . 50M No response 0 300 60. 350 150k 50M 50M No response V 60 . 300 20k 55K 55K No response V 30 300 42k 110k 110k No response • 0 id 45 300 42k 110k 110k Cell removed 24 Laye 1 r f o rmec 16k 300k 20k 2.5 150 4.2 V 30 200 300k 5M 5M No response 0 10 60 306 5K 9K 9K • No response V 30 . 200 150k 600k 600k No response 0 5d 60 300 400k 20M- 20M No response 0 300 60 300 350k 50M 50M No response V 60 300 111c: 9k 9k Behaves like metal. Cell removed 25 Laye r fo: rmed 15k 100k 20k 4.5 219 . 5.5 Response good V 10 200 12k 80k 20k 3.5 112 ; 5.5 Response good V 10 200 12k 75k 20k 4.0 48.6 5.0 Response fa l l ing. 7 30 250 200k 50M 50M No response 0 200 30( 300 100k 20M IM 368 5.8 Response good 0 100 30 300 300k 50M 20M 3.0 79.6 ; 5.8 V 10 250 600k 50M 10M 3.0 94 5.8 0 50 60 300 50k 150k 150k : 9.0 1.32 Noisy Poor response V 10 250 100k 1.5M 1.5M 12 2.4 5.5 Poor response 0 50 60 300 50k 300k 300k | 20 0.62 5.0 Poor response; 0 400 60 400 36k 500k 400k 140 Cell removed 126 Laye r f oi *med 23k 500k 80k 200 10.7 2.0 Noisy 0 50 30 300 36k 20M 15M 200 3.5 2.0 0 200 30 300 30k i 10M 10M 35 1.2 3.5 0 200 30 300 80k : IM 211 No response' 0 300 60 450 23k 80k 80k No response V 30 200 10k 600k 200k 12 Poor response V 30 250 50k 2M 800k 12 2.65 5.8 Poor response V i Cell removed 31 Table I continued! • Cell No. Treatment f Response • Remarks Bake P t T R300 R90 •• Rw ' M V/W 27 Laye .V V V r fc >rmed 30 6® 60 300 300 350 14k . 100k 50k 2K 250k : 20M. 20M 3K 22K 214 20M 3K 2.5 4.0 Poor response Poor response No response No response Cell removed 28 Laye V V Y ; r f c >naec 10 10 30 200 250 300 •13k 15k 15k 20k 30k : 120k ; 120k 214 16k 35k , 40k ; 300k 1.68 10.2 8.6 1.2 4.0 :, 5.0. 5.0 5.5 Poor response Poor response .;. Poor response Poor response Cell removed The'following symbols are used in the above tables V - vacuum bake i.e.. pressure is 10"6 mm, Hg.. 0 - oxygen bake.. P a pressure of oxygen bakes in microns. t - time of bake in minutes. T = temperature of bake i n °C. R300 = resistance of layer at 300°K. R90 s resistance of layer at 90°K.. R*j. - resistance of layer at 90°K when illuminated by a 25 watt tungsten lamp. N. =" noise of layer i n micro-volts. V/W. = sensitivity of layer in volts/watt. Xc s estimated value of long wavelength cut-off. 32 It was found for most of the cells that when the layer was ini t i a l l y deposited with the pyrex outer wall in place that i t showed no sensitivity when illuminated with a tungsten lamp. On exposing the layer to a small quantity of oxygen at a few millimetres pressure and room tem-perature, the cell showed a very definite increase in sensitivity, increas-ing s t i l l further with greater exposure to oxygen. This process, is given in detail for cell 27 in Table II. TABLE II. OXIDATION OF LAYER ( symbols as for Table I ) Cell No. Treatment R300 R90 Rw 27 Layer formed 600k 50M 50M Small quantity of oxygen.admitted 150k 20M 350k Small quantity of oxygen admitted 100k 51-1 150k Small quantity of oxygen admitted 75k 2M 100k , Small quantity of oxygen admitted 60k IM 75k Small quantity of oxygen admitted 50k 500k 60k Exposed to air for 3 minutes 33k 200k 45k Pyrex.cover removed and periclase window attached. Exposed to air for 4-5 minutes. 14k 250k 22k 33-This shows the dependence of the sensitivity of lead selenide on oxygen being adsorbed in.the crystalline lattice.. This effect is discussed in greater detail in Chapter V., Further confirmation of this fact was obtained by giving oxygen bakes to cells which had lost their sensitivity e.g. cell 22, (Table I) though i t was not always possible to restore the sensitivity in a l l cases., Should this process f a i l , the cell was removed, the layer cleaned off with nitric acid and the cell blank used for subse-quent experiments:. B.• Properties of Cells (i) Sensitivity Table III gives the results; of tests carried out on certain cells when cooled to 90°K.using the black body and time constant equipment.. TABLE III. SENSITIVIES OF PbSe CELLS,, AT 90°K Cell Area N. S S/N X W No. 2 cms uV uV us watts 10 0.8x0.08 2.2 14.1 6.4 1700 I.l l x l 0 ~ 6 12.7 12 0.7x0.1 7.6 10.4 1.4 1.22xl0"6 8.5 17 0.8x0.06 5.0 8.0 1,6 8.35xl0-7 9.57 20 0.9x0.1 2.4 30.4 12.7 1160 1.56xlO"6 19.5 22 0.8x0.1 2.2 280 125 225 1.39xl0"6 201 (values for V/VJ calculated from black body.) 34 It is seen that cell 22 shows the greatest sensitivity with a signal/noise ratio of 125:1. Using this figure we can calculate the mini-mum detectable power to which this particular cell will respond. -6 Mnimum detectable power = 1,39 x 10— watts 125 a 1.11 x 10"8 watts. This figure is calculated knowing the total energy emitted by the black body., If however, the spectral response of the cell is taken into account, and only the energy in this spectral band used, we obtain, knowing this cell to be sensitive to 6.3 microns: Mnimum detectable, power = 3.33 x 10"9 watts Since we know:. = 30$ for 500°K black body radiation. iotai. energy It is customary in this work to correct this figure for the bandwidth of the amplifier and define the minimum detectable energy as the value obtained for 1 c/s bandwidth. The noise power is proportional to the bandwidth, and thus.the equivalent noise input voltage will be propor-tional to the.square root of the bandwidth.. In this case the bandwidth is 30 c/s. Thus; Minimum detectable power s 6 x 10"^watts. As indicated in the introduction, these cells show two very 3 5 interesting features when the temperature i s lowered viz. the increase i n sensitivity and longer wavelength response.. This is shown in Table IV for cell 22. TABLE IV. SENSITIVITY OF PbSe CELL NO. 22 T 293°K. 193°K 90°K V/W r 4.55x10-2 8.2 675 4.6 5.35 6.3 eV ,±. 0.268 0.23 0.196 (Values for V/W calculated from monochromator for X =2.4/-- , T-1400°K.) It is seen that the sensitivity in volts per watt increases by a.factor of 10^ when the cell is cooled from 300°K to 90°K and the long wavelength response moves from 4*6 microns to 6..3 microns. The figures for the layer at dry ice temperature 193°K are also given. Three lead telluride cells were made of which only two showed any marked response and the various relevant factors for these two cells are shown in Table V. The values for the signal:noise ratio are much lower than those previously observed for this type of detector, but in view of the greater interest in lead selenide no attempt was made to improve on these figures. 36 TABLE V. SENSITIVITIES OF FbTe CELLS, AT 90°K Cell Area N. S S/N T W 'V/W No.. ( 2 cms^  uV uV UTiS watts 2 0.5x0.1 13 16.0 1.3 1.5 8.7xl0-7 13.9 3 0.5x0*1 35 40 ..9 1.2 2.0 8.7xl0""7 47.0 (ii) Spectral Response Fig. 12 gives a graphical representation of the relative spectral response for cell 22 at three different temperatures, the value for A c being given by the point at which the response has fallen to 50$ of its maximum value. The small peak at 1.5 microns was also found by Moss and Chasmar (10). Attempts were also made to measure the spectral response of lead selenide at liquid helium temperatures (4-°K), but only a limited spectral range could be covered owing to the rapid evaporation of the liquid helium. It was found however that the cell gave 100$ response as far as 7 microns without any noticeable decrease. On plotting the long wavelength cut-offs as threshold quantum values (eV) against the temperature a straight line is obtained as shown in fig. 13. This is in agreement with Moss's observations oh lead sulphide and lead telluride (30) and we can write: E T z E 0 •+ aT as the equation to this graph. 37 The value found for a is 3.58 x 10"^eV°C-1 which is in very good agreement with the value 4- x 10"4- eV°C"l found by Gibson (14.) for the temperature shift of the absorption edges for single crystals of lead selenide. By extrapolating this graph to 4°K we see that X c for this temperature is 7.7 microns which confirms the above observations on the spectral response of lead selenide at liquid helium temperatures. Fig. 14. gives the relative spectral response for a lead telluride cell. It is seen to be sensitive only as far as 5 microns, and the curve is markedly influenced by the absorption of the pyrex bubble window (fig. l ) o With some other type of window this spectral response could no doubt be considerably improved. ( i i i ) Conductivity Results Fig. 15 shows the variation of A<r/<r against intensity of illumi-nation I, where A o~ is the change in the conductivity of the layer and o~ is the "dark" conductivity. In this particular case the radiation was obtained by varying the s l i t width of the monochromator, the wavelength used being 2.5 microns. The value of Ao~ may readily be found from the resulting signal, since: ^<J/(T = AV/V x^ here AY is the input signal- to the amplifier, and V is the voltage across the resistance. The current through the resistance is assumed to remain constant.. c e l l n o . 3 P b T e °l '9 2 3 4 5 6 7 m i c r o n s. F i g . 14 R e l a t i v e S p e c t r a l R e s p o n s e o f P b T e . 38 The resulting graph is seen to give a linear relation between Acr/cr and I: ACT/<r : d where c is a constant.. This is in agreement with Simpson and Sutherland's observations for lead telluride (3l) who founds (<n/<r)2! = l + b l where <5\. is the conductivity under illumination and C the "dark" conduc-tivity.. If 6*) is small i t follows that: Ac / c r = bi/2 where A ( T : <5\ - cT (iv) Phosphorescent Effect It was pointed out at the beginning of this chapter that a tung-sten lamp was used as a preliminary check on the sensitivity of the cells during the sensitization process. It was noticed for every lead selenide cell made, that the resistance of the layer never returned to the former value of R90 on removing the illumination, but remained at some inter-mediate value between Rw and R90. On blowing out the liquid air and warming the layer to room temperature, the former value of ^90 could then once more be obtained on re-cooling the cell to 90°K. It was decided to investigate this effect (called phosphorescent effect for convenience) i n greater detail, and the result is shown in fig.. 16, where the conductivity is plotted on a 39 log scale against the inverse of the temperature.. To obtain the inter-mediate temperatures between 300°K and 90°K a small quantity of mercury was placed in the Dewar in which was immersed a copper-constantan thermo-couple. This was then frozen with liquid air, and then gradually allowed to warm up using a small air jet. The temperature of the layer was assumed to have the same temperature as that of the mercury. The curve ABC is that normally obtained for an impurity simi-conductor, and the first part of AB is seen to be linear conforming to an. equation of the type: <3~ i (TQ exp -From the gradient, the value for A E is found to be 0.11 eV which is in good agreement with that found by other workers (12) for the energy dif-ference between the lead impurity centres and the conduction band. The departure from linearity along BC arises very largely from the sensitivity of the layer to room radiation, as observed by Watts for lead sulphide (33). On illuminating the layer with the radiation from a tungsten lamp, the conductivity changes to the point E. Removing this illumination, the conductivity returns to the point D and not to C as would be expected. Keeping the temperature of the cell constant at 90°K, the conductivity is observed to remain at D for at least one hour., On warming the layer the conductivity follows the curve DB, joining the original curve at the point B. This effect is very marked and has not been observed for lead selenide or lead telluride before, and only once for lead sulphide by Chasmar and Gibson (34-) who ascribed i t as due to transitions to meta— stable levels by the photo-electrons. The curve DB shows the way in which these trapped electrons are thermally released, assuming this to be the correct interpretation of this effect. 41 CHAPTER V MECHANISM OF PHOTOCONDUCTIVITY. No satisfactory theory of this phenomenon has yet been developed, though several ideas have been put forward to explain the various results, obtained, and i t is the purpose of this chapter to review these various; theories in order to see how they compare with the experimental results obtained here for lead selenide., It was Wilson (36) who fir s t pointed out the significant proper-ties, distinguishing semi-conductors from metals and insulators.. We know that the possible energy levels of an electron in a crystal may be divided into bands between which there exist so called forbidden bands. As for insulators Wilson supposes that at the absolute zero of temperature a l l the allowed bands are either f u l l or empty.., For higher temperatures, electrons may be excited thermally to higher allowed levels (called the conduction band). In a semi-conductor the width of the forbidden band is much smaller than that for an insulator, thus giving rise to a larger conductivity, which is one very significant property which distinguishes between these two classes of substances., Wilson divides semi-conductors into two groups:, intrinsic and extrinsic semi-conductors. An intrinsic semi-conductor is one which conducts in the pure state and electrons are raised to the conduction band from the fu l l band. There are very few substances which belong to this class, except at high temperatures and most semi-conductors owe their properties to the presence of impurities thus giving rise to extrinsic conduction. These 42 impurities are usually due to a slight excess of one constituent from the stoichiometric ratio, e.g. in the metallic oxides, either an excess of metal or oxygen.. For these substances, very much less energy is required to bring an electron from an impurity centre than from the f u l l band into the conduction band.. These impurity centres may give rise to either donor or acceptor levels.. In the former case electrons may be excited into the conduction, band, and such semi-conductors are called n-type.. Should the impurity levels li e very close to the f u l l band, electrons may be excited into these levels leaving behind mobile positive holes. These are termed acceptor levels and the conduction p-type. We may distinguish.between these two types of conduction by a measurement of the Hall coefficient, for: R = - 3 TT c/8ne n-type R = + 3 IT c/8ne p-type For either type, the conductivity is given by: <T = nev where: n = number of electrons per c.c. e a electronic charge v = mobility Lead selenide crystallises in a face centred cubic structure with a lattice constant of 6.1 A,U. (35).. Wilman was also able to show from his-, electron diffraction work that this lattice axial dimension remains constant to within 0.1$ for various degrees of oxidation of a layer of lead selenide A3 deposited in vacuum. This indicates that there must be less than 1% by proportion, even for strongly oxidized deposits, of excess lead atoms (or vacant, selenium sites) or of .selenium replaced by oxygen ions in the crystalline deposit. ' He also showed that the mean crystal size varies from 150 A.U.. to 500 A..U. and increases with increasing duration and temperature of the vacuum bakes after the layer is deposited. However, the photo-sensitivity would appear to depend far more on very slight deviations from stoichiometric composition. (i) F Centre Impurities Above 750°K, lead selenide exhibits the properties of an intrinsic semi-conductor, and using this, Putley (4.0) was able to find the energy difference between the f u l l and the empty bands from Hall coefficient and conductivity experiments. The mean of these two results is O.48 eV, and is represented in fig. 17. We can see immediately that the photoelectrons cannot arise from transitions betx^ een these two bands, which would corres-pond to a wavelength cut-off of 2.6 microns. For a sample of lead selenide with an excess of lead, the gap between the donor levels and the conduction band is of the order of 0.1 eV (32), which is far too small to account for the photoconductivity. To overcome this difficulty, we could assume that there must exist other levels, about 0.25 eY below the conduction band which supply the photoelectrons, as proposed by Simpson for lead telluride (31). These electrons are assumed to arise from defects of the F centre type (41) c o n d u c t i o n b a n d T 0> 4 8 e V f u l l b a n d F i g 1 7 E n e r g y L e v e l s i n L e a d S e l e n i d e . I X I c B O O O G 0 F r i C B 0 t p - t ype F B n - t y p e F i g J 8 F o r m a t i o n o f n-p B a r r i e r . 44 where for an excess sample, there is a missing negative ion, its position being taken by an electron.. The theory of photo-conductivity for F centres in alkali halides is well known.(41) and shows that an absorbed quantum of radiation.first excites an electron without freeing i t from the centre. The electron, is then transferred to the conduction band by thermal energy. Calculations, by Pincherle (42) have shown, however, that the levels associated with F centres in lead sulphide and similar substances are much too deep to account for the photoconductivity. He found that the binding energy for such a centre is 2.67 eV.. should expect the long wavelength cut-off to vary inversely as the fourth power of the refractive index, n , that i s : We may regard the electron trapped at the vacant site, as similar to an. electron in an isolated atom, except that i t is immersed in a medium of dielectric constant K (41). This means that a l l the energy levels of the electron must be scaled down by a factor l/K 2 where K is proportional to n 2. Thus the energy required to raise an electron to an excited state is pro-portional to l/S? , and the above relation immediately follows. For a wide range of substances Moss found that the constant of proportionality had an average value of 77. For lead selenide we haves Using this picture of the F centre, Moss (43) showed that we constant. n 45 4.59 (taken from a paper by Avery (44)) A / n^ c 71 which falls well within the range of values found by Moss-. Assuming then, that the threshold wavelength is determined by the refractive index we may calculate the temperature shift of X Q from the change of the refractive index or dielectric constant with temperature. From, the Lorentz-Lorenz Law: where K-l K+2 N: <* C number of molecules per cc. polari zability constant We obtain on differentiation:.. dK dT where ^  is the linear expansion coefficient. For lead selenide P K 2.67 x 10-5 oc-1 21 (ref. (45)) (ref. (44)) dK , r dT -1.18 x 10"2 46 For cell 22: ^300 = ^ m i c r o n s A 1 9 3 = 5.35 microns, A T, = 100 A K = -1.18 ^193 B (K + A K) 2 A 300 * Kr 2Substituting for A we may find A A -jc}3 = 5.2 microns which agrees tolerably well with the experimentally determined value of 5.35 microns.. Thus, this not only predicts a shift of the long wavelength cut-off, but furthermore of the right order of magnitude. In view of Pincherle's calculations however, i t would seem that this model is not the. correct interpretation of the mechanism.. (ii) n-p Barrier Theory (9) (46) Since the layer consists of a distribution of micro-crystals, i t seems reasonable to assume that there would be a random distribution of both p and n type crystals. The layer is i n i t i a l l y deposited in the form of an n-type semi-conductor i.e.. excess lead forming donor impurity centres. The sample is then treated with oxygen, which presumably converts some of these micro-crystals, into p-type semi-conductors i.e. oxygen impurity centres which form acceptor levels. At the boundary between two such regions there 47 must exist: an n-p barrier.. For equilibrium, there must be an equality in. the density of electrons in any energy state on either side of the barrier, and there will be therefore, a flow of electrons across the barrier from the n-type crystal resulting in the formation of a space charge as shown in fig. 18.. Should such barriers exist, they must be the main factor contri-buting to the resistance of the layer. The temperature dependence of the resistance will be controlled by the variation of the height of the barrier with temperature. Absorption of infra-red radiation increases the number of elec- -trons in the conduction band and of holes in the f u l l band, and this increase in the concentration of electrons which are able to cross the barrier will produce a corresponding decrease in the resistance. It thus> appears that-high photosensitivity is obtained by a careful balancing of . the n- and p-type regions. This theory would account for the absence of photo-conductivity found for lead selenide when the layer is i n i t i a l l y deposited, due to the material being predominantly n-type i.e. excess lead. Table II shows the increase in sensitivity with greater exposure to oxygen, which presumably increases the number of such barriers. Loss of sensitivity could also be explained as due to the formation of a predominently p-type layer and a reduction in the number of these barriers.. However, there are a number of factors which are difficult to explain using this model., (i) Exposure of an n-type sample to oxygen must produce an increase in the number of n-p barriers. We vrould expect therefore, that AS the resistance would increase \ri.th such treatment, whereas the reverse is found to be true. (ii) If we associate the height of the barrier with the long wavelength cut-off, then presumably the change in the value of A c due to heat treatment, as observed in fig. 11, can only arise from a variation, in the height of the barrier. However, this barrier height is dependent only on the relative displacement of the acceptor and donor levels, and the levels would not vary to the extent implied from a change in the impurity concentrations. ( i i i ) Evidence in favour of this model has always been the observed decrease in the A.C. resistance with increasing frequency of the voltage applied across the layer. This is interpreted as due to the capa-citive shorting of these barriers (47).. This argument may now be dis-counted, since i t is possible to show that even a homogeneous resistor must show such a decrease merely due to its distributed capacity (48) . Rittner (4-9) attempted to improve the theory by assuming that instead of sharp barriers being formed, there exists gradual n-p transitions with a so-called quasi-intrinsic region between.the n- and the p-type regions. However, even this does not overcome the above objections. ( i i i ) Surface State Barriers (50) (5l) (52) It is fairly obvious that barriers do play an important part in the properties of these layers, and such barriers can also arise from the adsorption of oxygen on the surfaces of the crystallites in the form of 49 surface oxide films.. Barriers of this type are often referred to as Schottky barriers.. They have been used to explain, for instance, the properties of transistors (53). Since the layer is i n i t i a l l y of the excess or n-type, adsorption of'oxygen on the surface will cause electrons to flow from these donor levels into oxygen surface states (negative ions), leaving behind a positive charge.-. The result of this is to raise the levels at the surface: as shown in fig. 19. Contact between two such crystals may be represented by a barrier of the form shown in f i g . 20. Absorption of infra-red radiation will eject electrons from these surface states, thus reducing the siirface negative charge.. This results in a decrease of the height of the barrier, and the conductivity of the layer is increased.. It is assumed that these surface states l i e at a depth of the order of 0.2 eV below the conduction band. The long wavelength cut-off is then attributed to the energy required to eject an electron from an adsorbed oxygen ion. Smith (32) points out though, that this model will not explain the variation of the long wavelength cut-off with temperature, (fig. 12) . . On this model, as for the n-p barrier, we would expect the resis-tance of the layer to increase on exposing i t to oxygen. For lead selenide, however, the reverse is found, in general, to be true, which makes i t d i f f i -cult to accept the implications of this model to explain .the mechanism of photoconductivity in this substance.. On the other hand, i t does provide a reasonable explanation of the change in the long wavelength cut-off with I c o n d u c t i o n b a n d m p u r i t y l e v e l * f u l l ba n d o x y g e n s u r f a c e i t a t e ». F i g 19 R a i s i n g o f S u r f a c e L e v e l s d u e t o A d s o r b e d O x y g e n . F i g 2 0 F o r m a t i o n of B a r r i e r b e t w e e n t w o C r y s t a l l i t e s . heat treatment as-shown in fig. 11. In accordance with Sosnowski's obser-vations for lead sulphide (9) we assume that a vacuum bake effects the removal of oxygen from these surface states and in consequence reduces the height of the barrier. (iv) Bimolecular Law of Recombination It is easy to show that the linear relation obtained between, the change in conductivity AcT /b~ and the intensity of illumination I, may be explained by the bimolecular law of recombination (36) (31), sometimes referred to as the law of detailed balancing (41). Consider the electrons in the conduction band to be in equilibrium with a single group of unoccupied impurity centres: Rate of recombination Rate of production , Rate of production of excited electrons = of photoelectrons of thermal electrons nj^C nj_ + n j B » n- ( n. + n^B Where: ~n>, a- quantum efficiency B. = recombination coefficient no = excess concentration of trapping centres ni = number of electrons excited by radiation number of electrons excited in the dark <1A = number of quanta of radiation of length \ ( In general, n Q is negligable in comparison with n - and n^ ». 51 B nf = ^ x q x 1- n2B. ( n./ n )2 : 1 + ( W B n2 ) This may be compared with: ( <r±/' O" ) 2 = 1 + bl. ( page 38) from which.we foundl A cr / cr r bl/2 Many other substances have been shown to obey this law, and lead selenide does not appear to be an exception to this rule. Using this law. as the basis, attempts have been made to formulate so-called "numbers", theories (37) which although successful for thallous sxilphide (38), will not in general explain the results for the lead series (51). An attempt was made recently (39) to interpret the photoconductivity in silicon using this method,, 5 2 CHAPTER VI CONCLUSIONS-The method described here for making lead selenide cells, as far as is known, has hot been used before, and in view of the fact that sensitivities have been obtained which approach those previously reported, would appear to be satisfactory. However, i t was 'not possible to obtain, a cell with a sensitivity comparable to that for lead sulphide and lead telluride cells. The "phosphorescent" effect has not been observed in lead selenide before, and i t is conceivable that this may account for the lower sensi-tivity obtained with this type of cell in contrast to lead sulphide and lead telluride cells., It is not possible to measure this effect for low intensities of illumination, since i t may only be observed using a D.C., method of measurement, but i t was found down to the smallest intensity possible using this method. The experiments confirm that lead selenide can be made sensitive as far as 7 microns when cooled"to 90°K, and to 8 microns i f cooled with liquid helium at 4°K.. The cell design was not originally made for the layer to be cooled to 4°K, and a change in this design, to enable further experiments to be carried out at this temperature should provide useful additional data on the properties of lead selenide. 53 The results show that i t is very difficult to accept any of the known theories on photo-conductivity.. It has been possible to explain these results only in terms of different aspects of these theories, thus: indicating the need for a more satisfactory description of the probable mechanism of photo-conductivity in lead selenide., A better knowledge of the properties of lead selenide which produce photo-conductivity could be obtained, i f a simultaneous study of the thermo-electric power and the Hall coefficient were made after each heat treatment. It was not possible to do this in this case, and i t is difficult to see how such experiments could be carried out.. A study of the former was made by Sosnowski et al. (9) for lead sulphide and they were able to show that maximum sensitivity is obtained when the thermo-electric power is a minimum.. Some information about the distribution of energy levels in a semi-conductor can also be obtained from the dependence of the conductivity on the temperature, and Sosnowski further observed that this conductivity was also a minimum for maximum sensitivity for lead sulphide. The results obtained here for lead selenide would also appear to indicate this, but not so conclusively as Sosnowski indicates. What is fairly obvious for lead selenide is the fact that the conductivity decreases as the long wave-length cut-off increases. No such change was observed for lead sulphide.. It seems fairly definite that photo-conductivity in lead selenide is only obtained for a sample containing an excess of lead, together with oxygen as an additional impurity. This is in contrast with lead telluride, which may be made sensitive without the addition of oxygen (31).. On the other hand sensitivity could not be obtained for lead sulphide"with only 54 one type of impurity present (9). No attempt was made to investigate i n detail, how the photo-conductivity': varied i f the amount of excess lead was changed. It was found in i t i a l l y that a sample containing lead and selenium in the correct stoichiometric ratio produced no sensitivity whatsoever. The lead selenide used for cells 10 - 28 contained 5% excess lead. It i s known that the amount of excess lead contained in the final deposit depends very much oh the rate and temperature at which the deposition takes place (31). The rate of deposition also affects the size of the crystallites which are formed (35).. A slight loss of the electro-negative constituent (in this case selenium), always takes place during such a process.,. Any detailed investigation of the properties of lead selenide due to changes i n the amount of excess lead are therefore somewhat complicated by this effect. It was for this reason that i t was decided to study the properties of lead selenide with this one sample, though i t seems quite possible that greater sensitivity could be obtained by slightly varying this amount of excess lead. A^PPENDIX I 55 For the purposes of comparing the sensitivities of the cells during the sensitization process, i t is convenient to know the power emerging from the exit s l i t of the monochromator and subsequently focussed on the cell (see fig., 5 ) . This may be calculated as follows: The power passing the. entrance s l i t and incident on the prism is given by ( 5 4 ) : W..XdX = -A-Sl x _E*_dX_ w a t t s . TT where:. A.. s effective area, of prism in cms2 f = focal length of off-axis parabola:in. cms-s- - entrance s l i t width in cms. 1 : length of s l i t in cms. E> d> = 2 -V/f V * 1 0 ' 7 ^ t t s <**~Z exp he/ X kt - 1 h. = 6..54 x. 10 2 7 ergs sees, c = 3.00 x 10 1 0 cms sees."1 k a 1.37 x 10" 1 6 ergs deg °C-1 S and d.X in cms. 56 Assume that the dispersion of the prism at the exit s l i t is D\ wavelength units cm"-*- i..e,. cans cms"-1-. Thus the energy W\ dX will be dispersed over an area = 1, d X where we have assumed the dimensions of the exit s l i t to be the same as those of the entrance s l i t . The s l i t jaws in the Perkin Elmer monochroiaator are so adjusted for this equality to hold,. Power per unit area at the exit s l i t W,\ dX Dx l d X . Power passing exit s l i t 3- h C 2 . x 1 0 " 7 watts exp 1 , 4 3 8 / X T - 1 For the monochromator used we have the following constants: A ~ 6 x 4 « 4 cms2 f = 2 7 cms, 1 Z 1 , 2 cms. W'JS dK - ,.D e 2 A l r 57 It was convenient in most instances to choose the peak of the black body curve at"2.4. microns as the wavelength at which to compare the sensitivities.. Using an optical pyrometer, the temperature of the globar was found to be 1400°K. We may therefore substitute the following values in the above formula: X r 2.4 x 10-4 cms T. = 1400°K Dx = 4500 °A mm"1 for NaCl at 2.4 microns ' = 4.5 x 10"*^  cms cms-1 Wx dX = 0.404 s 2 watts This should be further multiplied by a loss factor Ii, due to absorption in'the prism, reflection at the mirrors etc. Wx dX = 0.404 Xs 2 watts By comparison with sensitivity measurements on the black body X is known to be about 40$ which agrees very well with the calculated value. Choosing a typical value for the s l i t width as an example: s = 0.1 mm .'. w'x dX : 1.616 x 10-5 watts 58 APPENDIX II Components for 900 c/s amplifier (see f i g . 8) HI 500K W.W.. pot.. R2. 100K W.W*. R3 1 meg.. W.W. R4 10KW.W.. R5 3.3 meg. R6 2.2K W.W, R7 10K 1 watt R8 33K W.W. R9 150K.W.W.. RlO 330K W.W.. Rll 50K W.W. R12 680 ohms. W.W.. R13 180K W.W.. R 1 4 330K W.W. R15 50K W.W_ R16 680 ohms W.W. R17 250 ohms W.W. R18 4.7K W.W. R19 47K W.W., R20 39KW.W. R21 680K R22 280K R23 75K R24 22K R25 8.2K R26 2..7K R27 800 ohms: R28 300 ohms R29 100 ohms R30 50 ohms R31 400 ohms R32 10K R33 39K R34 180K R35 330K R36,10K R37 47K R38 680 ohms R39 150K R40 330K R41 380K R42-10K R43 250 ohms R44 47K R45 1 meg. R46 12K R47 47K R48 180K R49 330K > R50 12K R51 47K R52 680 ohms R53 150K R54 330K R55 4.3K W.W. R56 460 ohms R57 150 ohms R58 27K R59 27K R60 10K R6l 15K R62 250 ohms R63 25 ohms Cl 1 uf C2 8 uf paper C3 0.001 uf C4 2 uf paper C5 0.1 uf C6 8 uf paper C7 8 uf paper C8 0.005 uf C9 0.005 uf C10 8 uf paper Oil 8 uf paper C12 8 uf paper C13 adjusted C14 0.1 uf C15 0.001 uf C16 8 uf elect, C17 0.001 uf. CIS 0.01 uf C19 0 . 1 uf C20 8 uf elect. 7 C21 0 . 0 0 1 uf C22 0 . 0 0 5 uf C23 2 uf C24 8 uf elect. C25 adjusted C26S0.1 uf C27 0.0047 uf. C28 8 uf elect. C29 0.001 uf C30 0 . 1 uf C31 0 . 0 0 5 uf C32 8 uf elect. C33 0.005 uf C34 2 uf C35 adjusted C36 0 . 1 uf. C37 8 uf elect. C38 1000 uf elect. V I 1620 V 2 6 S J 7 V 3 C V 1 3 8 V4 6 S J 7 V 5 6 S J 7 V 6 C V 1 3 8 V 7 6 S J 7 V 8 6 S J 7 V9 C V 1 3 8 V 1 0 6H6 LI 1 henry toroidal L2 1 henry toroidal A 200 micro-ammeter REFERENCES Smith, W.., 1873, Am. Jour. Sci., j>, 301. Case,•T.W.j 1917, Phys. Rev., % 305. Case,TiW.., 1920, Phys. Rev., 1J5, 289. Oxley, C.L., 1946, J. Opt. Soc. Am., J56, 356. Elliott, A.., 1947, "Electronics", London: Pilot Press. Yon Hippel, A. et al., 194-6, J. Chem.. Phys., 1£, 355. MOSSJ T.S., 194-7, Nature, 1$%, 4-76. Miner, C.J. and Watts, B.N-., 194-9, Nature, 16J, 322. Sosnowski, L. et al.., 194-7, Nature, 159, 818.. Moss, TiS..., 194-9, Proc. Phys.. Soc., B 62. 741.-Mossj T.S.. and Chasmar, R.P., 194-8, Nature, 161, 244. Moss, T.S., 1948, Nature, 161, 766. Gibson, A.F. et al.., 1951, Proc. Phys. Soc, A 64., 1054-. Gibson, A.F.., 1952, Proc. Phys.. Soc, B. 65, 378. Fellgett, P., 194-9, J. Opt. Soc. Am., 970. Simpson, 0. et al., 1948, Nature, 161, 281. Blackwell, D.E. et al., 1947, Nature, 160, 793. Brown, D.A.H. and Roberts, V., 1953, J. Sci. Instr., JO, 5. Young, A.S.., 1951, J. Sci. Instr., 28, 207. Rawson, H., 1951, J. Sci., Instr., 28, 208. Buckley, H., 1934, Phil. Mag., 17, 576. Benedict, M., 1937, Rev. S'ci. Instr., 8, 252. Sturtevant, J.M., 1938, Rev. Sci. Instr., % 276. McFee, R.H.., 1952, Rev. Sci.. Instr., 23., 52. 60 Kuiper, G.P. et al., 1947, Astro. Jour., 106, 243. Brown, D.A.H., 1952, J., Sd.. Instr., 22, 292. Brown, D.A.H., 1949,* J.. Sd. Instr., 26, 194. Handbook of Chemistry and Physics. Brown, D.A.H., T.R.F. Memorandum Number T2125. Moss, T.S.., 1949, Proc.' Phys. Soc.., B 62, 741. Simpson, 0.''and Sutherland, G.B.B.M., 1951 Phil, Trans. Roy. Soc. 243 , 547., Smith, R.A., 1950, Proceedings of the Conference on Properties of Semi-Conducting Materials, 198. Watts, B.N.., 1951, Proc. Phys. Soc., A 62. 456. Chasmar, R.P. and Gibson, A.F., 1951, Proc. Phys. Soc., B. 64. 595. Wilman, H., 1948, Proc. Phys. Soc., 60, 117. Wilson, A.H., 1939, Semi-Conductors and Metals, Cambridge University Press. Pick, H., .19.48, Ann. der Phys., j}, 255. Von Hippel, A. and Rittner, E.S., 1946, J. Chem, Phys., IA, 370. Rollins, B.V. and Simmons, E.L., 1953, Proc. Phys. Soc. B 66, 162. Putley, E.H., 1952,, Proc, Phys. Soc., B. 65, 993. Mott, N.F. and Journey,. R.W., Electronic Processes in Ionic Crystals, Osford, 1948, Pincherle, L., 1951, Proc. Phys. Soc., A 64, 648, ! Moss, T.S.,'1950, Proc. Phys. Soc., B 63, 167. Avery, D.G., 1953, Proc. Phys. S0'c., B 66. 134. Simpson, 0., 1947, Nature, 160, 791. James, H.M., 1949, Science, 110, 254. Chasmar, R,P., 1948, Nature, 161. 281. Rittner, E.S. and Grace, F., 1952, Phys. Rev., 86, 955. Rittner, E.S., 1950, Science, 111, 685. Schwarz, E., 1949, Proc. Phys. Soc., A 62. 530. Gibson, A.F., 1951, Proc.. Phys.. Soc., B 64. 603. Bardeen, J.., 1947, Phys.. Rev.., 71, 717. Bardeen, J., and Brattain, W.H», 1949, Phys.. Rev.., 21, 1208. Daly, EJT.. and Sutherland, G.B.-B.M*, 1949, Proc. Phys.. Soc., A 62. 205... • • " 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items