Chemical Instrumentation Ediled by GALEN W. EWING, Seton Hall University, So. Orange, N. J. 07079
T h e s e articles are intended to s e m the readers of THIS JOURNAL by calling attention to new developments i n the theory, design, o r availability of chemical laboratory instrumentation, o r by presenting useful insights a n d ezplanations qf topics that are of practical importance to those who use, or teach the use of, modern instrumentalion a n d instrumental techniques. The editor invites correspondence from prospective contributors.
LXVIII. X-rays and Electrons in Analytical Chemistry with Emphasis on Instrumentation
The series of articles in this column by Rudman on x-ray diffraction is the takeoff paint for our attempt to acquaint the reader with the instrumentation that makes x-rays and electrons useful in modern analytical chemistry Rudman has covered his subject comprehensively and included much that is pertinent here. Photoelectrons and Auger electrons are growing in importance as sources of information. As they are intimately related to x-rays, the instruments that make this information available must a t least he mentioned even though x-ray methods based a n absorption, emission, and diffraction are our principal concern. Analytical chemistry as we define it is the eharaeterization (in the laboratory) and control (during processing or manufacture) of materials. Clearly. we cannot do justice to all the available equipment. We begin with one of the most remarkable stories in science. (Continued onpapeA10)
H. A. Liebhafsky, Texas A&M University, College Station. Texas a n d
H. G. Pfeiffer,G e n e r a l Electric C o m p a n y , S c h e n e c t a d y , N. Y.*
In 1967, Dr. Herman A. Liebhafsky elected retirement as Manager, Inorganic
Dr. Heinz G. Pfeiffer is Manager, Technology and Energy Assessment, a t the Pennsylvania Power & Light Company, Allentown. A native of Pforzheim, Germany, Dr. Pteiffer spent his early years in Newark. New Jersey. He received his BA degree from Drew University in 1941, his MA from Syracuse University in 1944, and his Ph.D. in chemistry and physics fiom California Institute of Technology in 1949. He served a s a radio technician with the U. S. Navy from 1944 to 1946. Dr. Pfeiffer joined the staff of General Electric's Research and Development Center in Schenectady. New York, as a research associate in 1948, and has specialized in studies of electric strength, x-ray emission spectroscopy, statistics, and x-ray and elec-
and Electrochemistry Branch, General Electric Research and Development Center, to become Professor of Chemistry a t Texas A&M University, where he is now Professor Emeritus. Born in Zwittau, Austria-Hungary, in 1905, Dr. Liebhafsky came to the United States in 1912 and completed underpradua t e work in chemical engineering at the A. & M. College of Texas in 1926. After receiving a Master's degree in chemistry trom the University of Nebraska in 1927, he was awarded the Ph.D. by the University of California a t Berkeley, where he served subsequently w instructor until he joined the chemistry research staff of the General Electric Company in 1934. During World War 11, he was a technical observer for the armed forces in the European Theater of Operations.
Dr. Liebhafsky has had wide experience in various fields of chemistry. His name appears on over 150 papers in these fields. He was the first chemist employed in industry to receive the ACS Award in Analyt i e d Chemistry (i962). In recent years, his main research interest has been the oscillating reaction on which he first worked a t Berkeley in 1928. He is senior author of three books, all published by Wiley: X-Rm. Absorption ond Ernishion in Anoiytiroi Chemistry; Fuel Cells a n d Fuel Batteries: and X-Rays. Electrons. and Anniytirnl Cherni*tn. Dr. Liebhafsky is a member o f t h e American Chemical Society. which he has served locally and nationally, and of the Electrochemical Society. among others. He is an honorary member of the Society for Applied Spectroscopy.
tron diffraction. He was appointed Manager of Dielectric Studies in 1954. In 1968 he was Director of Educational Technology Planning for the General Learning Corporation and later became Manager of the Educational Technology Branch in the General Electric Research and Development Cen-
"..Dr. .
..t
Pfeiffer is the author of some 40 papers, and is co-author of two books: X Ha>,Absorption and Ernirsion in Anol?tirai Chemist?. and Aduances in X-Ray Mrthodr. He is a member of the American Chemical Society, the American Physical Society, AAAS, and the IEEE, and is a fellow of the American Institute of Chemists. He is a past chairman of the IEEE Committee on Basic Sciences, and a past chairman of the Eastern New York Section ofthe
DR. HElNZ G. P F E I F F E R ACS. In 196i he served as co-chairman of the annual Denver X ~ R a yConference on the Applications of X-Ray Analysis. Volume 5 0 , Number 1. January 7973
/ A9
Chemical Instrumentcrtion
Table 1. A Convenient Classilication of X-Rays Verbal Description
Harda
Wavelength range "Voltaee" classificationb
Up to I A Above 10 kV
Soft
Ultrasoft
From 1 to 10A From 1 to 10 kV
From 10 to 2WA Below 1kV
THE X-RAY STORY IN MINIATURE The story begins on November 8, 1895, when Roentgen noticed a faint light an a bench in his darkened laboratory. He was investigating pulsed cathode rays (electron beams) generated in a Hittorf-Crookes tube that he had covered with black cardboard. The faint light, which appeared synchronously with the pulses, originated from a bit of phosphor that had been left on the bench about a yard from the tube, and that had been excited by x-rays generated when the electron beam struck the tube walls. That day was the end of a n era of classical physics. The story continued with the growth of quantum physics and led eventually to quantum chemistry. The discovery by Beequerel of radioactivity was an impartant sequel. By 1927, work on x-rays had been rewarded with six Nobel prizes, Roentgen's being the first such award in physics. Under proper conditions, x-rays can be absorbed, scattered, or emitted by atoms, free or combined in any state of aggregation. Tables 1 and 2 show how x-rays may conveniently be classified and how they are related to other kinds of radiant energy. This relationship is extended somewhat naively to atomic structure in Figure 1. Much information is implicit in the tables and the figure. Characteristic x-ray lines can he emitted when an electron from a n outer shell fills an electron vacancy in a shell nearer the nucleus, e.g., a K
'This article is based on the fortheaming hook X - R a y Electrons, and Ana!\ticol Chemistn. by H . A. Liebhafsky, H. G. Pfeiffer, E. H. Winslow, and P. D. Zempny, John Wiley & Sons, New York, 1972, I" whlch references to the literature will be found. We thank the publisher for permission to use the book in this way. Much of what it contains has been twice presented in a graduate course a t Texas A&M University. Readers who wish to farm some idea of the enormous literature in the field will do well to examine the series of volumes Advonees in X - R q Analysis. Plenum Press. New York, that records the papers presented a t an annual conference in Denver; and the annual reviews published by Ana/?tical Chemistn.. especially those dealing with fundamentals. Three such were pertinent in 1972: X - R a y Absorption and Emission. which makes no attempt to cover all the published literature: X-Roy Diffraction: and Electron Spectroscopy. II. X-Ra.v I-'hoto~xcitation,a newcomer. References to the six articles entitled X-Ra? Diffraction Analysis by Reuben Rudman are J. Chem. Educ., 44, A i , A99, A187, A289, A399, and A499 (1967); they have been reprinted in "Topics in Chemical Instrumentation." (G. W. Ewing, Ed.), Chemical Education Publishing Ca., Easton, Pa., 1971; p. 75ff. *Present a d d r e s s H. G. Pfeiffer, Pennsylvania Power & Light Company, Allentown, Pa. A10 /Journal o f Chemical Education
Table 2. A Crude Correlation of Various Spectra Soectrum 7-ray x-ray Electronic Vibrational Rotational
Spectral Reeian 7-ray x-ray Ultraviolet Infrared (near) Infrared (far)
Representative Wavelength. A
Quantum E n e r n , erz
0.01 1 loo0 10' IF
2(10-6) 2(10F) 2(10-"1 2(10-'2) 2(10-")
eV, Electron-volts 12.4(10J) 12.4(103) 12.4 1.24 0.0124
Note. (1) It is assumed the atom is combined. Only molecules have vibrational and rotational spectra. (2) The wavelengths are chosen far convenience. The spectral regions are poorly defined: for example, the ultravioleJ region shades into the visible. (31 One electronvolt is 1.6(10-'? . ere. Often an electron with enerw -. Ve electron-volts 1s called a V-electron (see ~ a b l e ' l )Kilovolts . often replace volts in this designation
line when a vacancy in the K shell is thus filled. When a radiationless transition occurs, Auger electrons are emitted instead. The inner (or core) electrons are generally of overriding importance in the absorption and emission of x-rays. This makes for simplicity, and for independence of chemical and physical properties. Of course, as the wavelengths of x-rays increase to the point where their energies became comparable with those of valence-electron transitions, chemical binding increasingly influences x-ray processes. Similarly, as we pass to elements lower and lower in atomic number, the "inner" electrons became fewer and fewer until they have disappeared a t zHe, where the K electrons have become the "outer" electrons. The energy relationships just mentioned dictate the instrumentation needed. Langwavelength x-rays are strongly absorbed: air optical paths cannot be used and windows (or other solids) in the optical path must be thin and highly transparent. Detectors suitable for y-rays may be useless. Electronic noise may bar satisfactory x-ray detection. Bragg reflectors other than crystals may be needed. Pulse-height selection becomes difficult. Generation of long-wavelength characteristic lines by x-ray excitation requires special x-ray sources. One often describes such difficulties as part of the "light-element problem." for even the K lines of the light elements are of long wavelength (e.g., 228A for LiKn); but the problem is fundamentally a problem of low x-ray enern. It is difficult to overstate the importance of x-rays to modern chemistry. That x-ray diffraction, which is the unmodified, coherent scattering (or Bragg reflection) of x-rays, can establish crystal structure is well known ( I ) . Information about less definite structures can also he obtained. X-ray emission spectrograph? is by all odds the best single method for the determination, free or combined, of all but the lightest elements. Other applications of interest to chemists are too numerous far listing here. Perhaps less well known to chemists are certain fundamental matters
such as the fallowing. Compton scattering, the incoherent scattering of x-rays by electrons in which the wavelength is modified (lengthened), is the most direct evidence of the wave-particle duality in matter and leads to Heisenberg's uncertainty princi~ l e Energies . of x-rays absorbed and emitted by atoms are related to the various shells and subshells, and consequently help explain the periodic table. Energy relationships between electrons and x-rays prove the Einstein postulate about energy equivalence in elementary processes. The best values of Avogadro's constant rest on r-ray diffraction data. More could be said,
Figure 1 . Oversimplified model of any atom (combined or uncombined) between ,,Na and S ~ K The ~ . K and L shells are filled. The M shell and the valence band (or shell) need not be. Elements above , B A ~have electrons in the N shell (not shown). The L and M shells divide logically into subshells or energy levels (not shown) on the basis of x-ray evidence, which thus heips to explain the periodic table. The different kinds of radiant energy are shown with Origins indicated. Note that their origins unequivo~allydistinguish y-rays from x-rays. The figure is no guide to energies, although it is true that the energy required to eject an electron from a shell decreases with the distance of the shell from the nucleus.
(Continued onpageAI4)
Chemical Instrumentation
g
'-.
Lsnartd x-?8?s ,"nmod,'iad,
Figure 2. Fate of a monochromatic x-ray beam. TWO types of events-phatoelectr8c absorption and scattering-can occur as x-ray quanta disappear from the beam. The main event is phatoelectric absorption, which leads to the emission of characteristic lines and of Auger electrons; see Fig. 1. As a practical matter, only the transmitted beam need usually be considered in absoiptiometry. Note that soume, sample. and defector are components of many optical systems used in analytical chemistry.
hut this should he enough to indicate that chemical education should include more of the x-ray story than is now usual. ~
~
ABSORPTIOMETRY WITH X-RAYS In the main, x-ray absorption follows Beer's law even though the actual history of an x-ray beam on its progress through matter is quite complex. as Figure 2 shows. Usually, however, x-ray absorption may be regarded as photoelectric absorption that involves inner electrons. Such with wavelength and atomic number between absorption edges. This makes far predictability and simplicity not found in the ultraviolet, visible, or infrared regions ofthe soectrum.
F~gure3 Smple laboratory x-ray photometer A: P h o ~ p h o r ~ p h o t o e e c tdetector l~c in which a photomull~pliertube generates an electric current that measures the ntensity of incident xrays. 5: Sample cell. C: Sample. D: X-ray tuba and housing. E: Milliammeter for readout. F: Amplifier and rectifier vacuum tubes. which might now be replaced by solid-state devices. G: Regulated power supply for amplifier and photamuitiplier. H: Control panel.
A1 4 /Journal of Chemical Education
fb)
Figure 4. Schematic diagram Lo accompany F i g 3.
Absorptiometry is done with both polychromatic and monochromatic x-ray beams. T h e former type, which somewhat resembles calorimetry with white light, is called fluoro.scopy when a fluorescent screen is the detector, and radiography when a photographic plate or film is thus used. This kind of absorptiometry will no doubt soon be introduced for baggage inspection to help forestall skyjacking. Overall, absorptiometry with monochromatic beams is of lesser importance. This technique is the simpler because it entails no
Figure 5 . Stand for comparative absorptiometry. Comparison is accomplished by manual CornmUtaLiOn between standard and unknown samples.
(Continued onpageA16J Volume 50,Number 1. January 1973
/ A15
Causes of error
Chem;cal Instrumentation worry about unsuspected absorption edges or about "filtering," which is the preferential removal of longer wavelengths from a polychromatic beam. Of particular interest as regaxis instrumentation is the use of radiaktive x-ray sources . i n absorptiometry with monochromatic beams. The different kinds of instrumentation needed in the two teehniaues are imdicit in two
quire only simple instrumentation. Monochromatic beams are weak and simple; they can sometimes give specific results but require more elaborate instru-
Signal from ~ h i
standard
++ *l
Error or difference ~ detectors k ~ ~ ~ ~ signal
Amplifier
?
Rolls
I Signal from actual strip
Figure 6 . Schematic diagram to show the flow of information in a servomechanism system (simplified) in which an x-ray thickness gauge automatically controls by absorptiometry of a polychromatic beam the setting and operation of a rolilng mill. Energy sources (such as that for the motor) are not shown.
mentation if an x-ray tube serves as source. (2) Polyehramatic beams are suit.
ed to instantaneous process control measurements. For monochromatic beams. quanta must usually be counted and the counts accumulated over a time interval. Figures 3 and 4 show a simple laboratary x-ray photometer that approaches being homemade. Figure 5 shows a homemade sample stand that can accommodate three aluminum cells to contain the samples. With this arrangement, comparative absorptiometry becomes possible. We have here the three elements (source, sample, detector) of Figure 2 and the needed ancillary equipment. This simple photometer has a great many qualitative, semi-quantitative, and quantitative uses that range from instantaneous differentiation between materials (e.g., polythene and chlorinated polymers) to the identification of pure compounds. The instrument can be used for point-to-point explaration of materials to check their uniformity. It can handle solids, liquids. and gases (the last in special cells). It can be used to demonstrate x-ray absorption to students. It will more than earn its keep in any large. diversified laboratory once the staff
Figure 7. The General Electric Raymike" 2000 Thickness Gauge. Steel strip intercepts a poiychromatic x-ray beam (see Fig. 8) as it passes between the jaws of the "C:' which are separated by 84 in. to save the gauge from damage that might otherwise occur under the severe operating conditions. A motorized carriage permits posttioning of the beam and scanning of the strip. (Courtesy of General Electric Campany.1
A1 6 /Journal of Chemical Education
learns what it can do. Reading a n electric current is a quick, simple, and easy way to gain information-at least preliminary information-about a sample. We move now from homemade to spectacular-namely. to the automated thiekness gauging of steel strip. which shows absorptiometry with polychromatic x-ray beams a t its best. The flow of information in such a n automated system is diagrammed in Figure 6. During the thickness measurement, the steel strip may be hot (1500' to 1750'F). moving (about 2000 feet/rninute horizontally with possible vertical vibrations up to several inches in amplitude). and subjected to a spray of cooling water. In 1955. this application was made fully automatic; that is. the error signal was used to readjust tandem cold reduction mills of the U.S. Steel Corporation. Automatic control proved significantly more effective than manual control. Under the drastic operating conditions. the guaranteed aceuracy of the present gauge over its measurement range is: within *2 mils (i.e., thousandths of a n inch) between 140 and 200 mils: within 1% of thickness between 200 and 400 mils; and within 5 4 mils from 400 mils to the upper limit of the range: that is, 1999 mils. Not bad for the conditions! Gauge proper and carriage appear in Figure 7. Figure 8 is a schematic diagram to show how the gauge works. When it is part of an automated system that includes computer and rolling mill. we then have a system that operated in principle like that of Figure 6. Points of interest in Figure 8 are as follows: 1. With two ionization chambers as detectors. the gauge continuously compares unknown (the steel strip) and standard (the wedpe-No. 1 in the figure). The wedge is part of its own s e ~ a m e c h a n i s m system to ensure precise setting and easy rapid calibration. The wedge is set so that a n error signal of zero results when the strip has the desired thickness. 2. Within limits, differences in chemical composition (that is, of mass absorption coefficient) can be allowed far in presetting the wedge. Known standard samples stored in the gauge make possible calibration at different compositions. 3. The ionization chambers give a-c output signals that are preamplified a n d fed through a difference amplifier (No. 6 in the figure), the resultant signal from which is converted to a n analog output proportional in magnitude to the difference between the original signals. There results a d-c error signal that appears on the deviation meter of the operator's control panel. 4. When the gauge is joined to ("interfaced with") computer and rolling mill, the error s i p a l just mentioned is used to accomplish automated control of the mill. 5. Thanks to solid-state circuitry. ~ a u g e response is rapid and background ("noise") is low. T h e gauge is operational after a 3-minute warm-up from a cold start. and 5 scc after the x-ray-tube voltage has been turned on. I t responds to a l ~ most two-thirds of a 5% thickness change within 0.05 sec.
(Continued onpogeA18) Volume 50. Number 1 , January 1973 / A17
Figure 8 . Schematic diagram of the gauge system far attended (not automated) operation of a rolling mtll (Courtesy of General Electric Company.)
A18 /Journal 01 Chemical Education
THE SELECTION OF X-RAY WAVELENGTHS The most common ways of selecthg x-ray wavelengths are by filtering and by wavelendh resolution (Braze reflection). . . Reference 1 give> an excellent discussion of filtrring. Though pointed toward x-ray diffraction, the discussion is generally useful and will not be enlarged upon here. The term "wavelength resolution" is introduced as s contrast t o "energy resolution" (pulse-height selection), which will reappear when solid-state detectors are described below. The discussion to follow relates mainly to x-ray emission spectrographs. Collimation is needed in a spectrograph to obtain satisfactory wavelength resolution by a flat Bragg crystal. Collimation selects direction, not wavelength, as is shown by the slits in Figure 10. The figure makes clear that unwanted wavelengths would reach the detector in the spectrograph if collimation were not employed. Soller collimators, or slit systems, are commonly used both in diffractometers and in spectrographs. These are stacks of thin parallel metal plates that form a series of extended slits in the beam direction. In the spectrograph, the Soller collimator makes it possible to use a n entended sample area effectively without sacrificing adequate wavelength resolution. These collimators are often far from ideal; see Figure 11. Figure 12 shows two such collimators in the schematic diagram of an x-ray emission spectrograph. I t was mentioned above that x-rays of low energy (long-wavelength) are troublesome t o deal with. Bragg reflection is impossible a t wavelengths exceeding 2d, where d is the interplanar distance of the reflecting crystal. For the long-wavelength x-rays, such as the K lines of the light elements, reflectors with 2d values in the range -10 to lOOA are needed. Multilayered soap films, which we call Lsngmuir-Blodgett gratings, fill this need. As they are not crystalline, they are properly called Bragg reflectors, an operational name that includes Bragg crystals. Henke7-
Figure 9. Nomenclature and relationships of various x-ray instruments. Only flat Bragg reflectors are shown, and these are assumed to be crystals. Note haw the spectrometer is related to the other instruments, with which it ought not to be confused.
6. The measuring gauge head and carriage weighs about 7000 lb; the equipment control cabinet, about 1500 lh; and the operator's control panel. about 80 ib. The spectacular growth of x-ray emission spectrography is making x-ray absorptiometry in the laboratory into a poor relation among x-ray methods. But x-ray ahsorptiometry is not about to go out of style-after all, i t is still the kind of analytical chemistry that doctors and dentists frequently practice on us!
COMPARISON AND NOMENCLATURE OF X-RAY INSTRUMENTS The current naming of x-ray instruments and methods seems right out of Alice in Wonderland. Figure 9 is our version, arrived a t after matching the various names in current use with the operations that the instruments actually perform. The simple photometer represented in Figure 2 is not included, but it serves as the basis of the other instruments in Figure 9. By adding to i t as in the figure, one ohtains the other instruments. Of course, what happens to the sample depends upon whether an absorption, emission, or diffraction measurement is being made. As the figure is self-explanatory and gives the names of the instruments, only three comments will be made: (1) The spectrometer (no sample) is a component of the spectrograph and o[ the spectrophotometer. (2) The Bragg crystal is often called a monochromafor when it selects a wavelength from a n x-ray source, as i t does when it is placed between source and sample in the spectrophotometer. (3) Monochromatizing a t the source minimizes heating and x-ray damage of the sample.
Figure 10. Diffraction of a beam fmm a point Source by a large crystal. The crystal is positioned far the Bragg reflection of wavelength hr at angle On. Without a dit. Bragg reflection of all wavelengths between At and hl will occur because the crystal receives x-rays at all angles between 8, and &. A slit at A or B will collimate the beam and remove the unwanted wavelengths. (Continued onpogeA20)
Chemical instrumentation
Figure 11. Actuai transmittance of a Soler collimator Consisting of 8ron plates l-mil thick with nominal %mil Openings. Transmittance for xrays recorded on f i l m Ideal pattern would have been lines t-mil thick separated by white spaces 9 mils wide. Note how far the collimator departsfram ideality.
type Bragg reflectors are dual: a lead stea,,te reflects x-rays of long wavel e n a h while the mica substrate reflects the-short. A flat Bragg reflector needs a parallel beam. Curved Bragg reflectors can give excellent resolution, but they need heams that are either convergent or divergent. They require no collimator, which means that they are free of the intensity reduction that collimation entails. They are ideally suited to the spectrograph known as the electron microprohe (see below). For this and other reasons, curved Bragg reflectors were made commercially available over ten years ago, and they have been gaining ground over flat reflectors ever since.
ENERGY RESOLUTION. SOLID IONIZATION DETECTORS As solid ionization detectors, pmpor-
Figure 12. Schematic views of spectrograph components. (a) Side view. 9-plane Ito page. Motion in @-planeat angular velocities shown. (b) View of same parts from above. 9-plane coincides with page. Arrows show motian and angular velocity during Scanning from 29 = 0; (initial value). X-ray sourcenot shown in lb).
tional detectors, and scintillation detectors all give pulses of heights proportional to the energies of absorbed x-ray quanta, they offer the possibility of energy resolution to replace wavelengh (Bragg) resolution. The replacement has much to offer: notably, the simultaneous recording of the intensities of all analytical lines, and the elimination of the serious intensity lasses that accompany Bragg reflection. Energy resolution is accomplished by pulse-height selection in a multichannel analyzer, a technique highly developed in nuclear physics for situations generally simpler than those in analytical chemistry. The pulse-height distributions for proportional and scintillation detectors are so wide as to make these detectors unpromising for quantitative determinations by unaided energy resolution. Foriunately, the solid ionization detector is another story. With the other detectors, energy resolution combined with wavelength resolution is now the accepted way to eliminate interference by lines reflected a t orders higher than the first and to reduce background. The improvement and simplification of complex electronic circuitry by the semiconductor industry have greatly henefited the instrumentation systems under discussion here. The development of the solid ionization detectors mentioned above is not yet complete, but it 'promises far-
Figure 13. Visualized operation of Si(Li) x-ray detector. An x-ray quantum of t A wavelength strikes the sensitive zone (also called here the intrinsic zone and the electrolyte) and creates a b u t 3.540 hole-electron pairs. These highly mobile charge-carriers move rapidly to the electrodes and are collected before recombination can occur. For comparison, note that the same x-ray would produce about 400 ion-pairs in a gas-filled ionizaion detector.
A20 /Journal o f Chemical Education
Analog to pulse-height converter 400-channel P.H.A.
Specimen,
l l.; ",. , ~1
Lead shield recorder
Channel/number
(b) Figure 14. (a) Diagram of an x-ray emission spectrograph system with unaided energy resolution. Broken lines enclose the low-temperature region. (b) Gamma and x-ray spectra of Am2" obtained with this system. A 2-mm plastic cover on the radioactive source barred alpha particles from the Si(Li) detector. After Bawmanetal., Science, 151. 362 (1966).
reaching results because these detectors should eventually make Bragg reflectors obsolete in many determinations by x-ray emission spectrography of all but the lightest elements. For reasons clear in the physics and chemistry of semiconductors, only silicon and germanium presently qualify as materials for making the new detectors. At present, most such detectors are "lithiumdrifter-that is, lithium is introduced as "dopant" to combat the traces of residual impurities; they are known as Si(Li)"silly"-and Ge(Li)-"jelly"-detectors. Germanium, first made by R. N. Hall with a n impurity content near 1 in 1012,is now offered by the General Electric Company as a detector that needs cooling only during use, and that matches the best Si(Li) in performance. How solid ionization detectors work is shown simply in Figure 13. An absorbed x-ray quantum produces a number of charge carriers (hole-electron pairs) that is proportional to its energy. There is no amplification. The bundle of charge carriers
thus produced is counted as anelectronic pulse after i t has been sorted according to size by a pulse-height selector (multichannel analyzer). X-rays of different wavelengths (energies) are thus simultaneously recorded as to wavelength and intensity. Solid ionization detectors need field-effect transistors (FET's) as first preamplifiers. The combination must be cooled t o 77'K during operation; Si(Li) and Ge(Li) -but not Ge-detectors must be cooled thus also during storage. Because the cryostat needs a window, coaling during use means that the detectors lose the advantage of being windowless themselves. Figure 14 shows a schematic diagram for an early solid-ionization-detector system and a recorded spectrum therefrom. Resolution much better than this is obtainable today. These detectors are particularly suited to the electron micropmbe. Acknowledgment: We thank the Robryt A. Welch Foundation for partial support.
(Tobe condudedin the February
issue.1
