Emission Spectroscopy WILLIAM F. MEGGERS National Bureau of Standards, Washington, D. C.
T
H E principal advances in fundamental research and applications of atomic emission spectrosoopy during six postwar years, 1946-51, were reviewed in three previous reports (87-89). The interest in those reviews encouraged the preparation of this one covering the years 1952-53. As before, new information is listed under standard wave lengths, spectral line intensities, Zeeman effect, term analysis, hyperfine structure, isotope shifts, spectrochemical analysis, and related topics.
perturbing effectsof foreign gases upon J-values was not predicted by theory of pressure shifts IYTENSITIES
STAYDARD WAVE LENGTHS
It was suggested 127 years ago that a wave length of light might serve as an ultimate standard of length. There is hope that this may come to pass eventually. In September 1953, an international Advisory Committee for the Definition of the Meter convened a t the International Bureau of Weights and Measures in SBvres, France. I t was decided that the time has come to favor the establishment of a new definition of the meter based upon a wave length of light, in order to give the fundamental unit of length not only greater precision but also the characteristics of universality and indestructibility. This committee recommended that the meter be defined by a wave length in vacuum from an atom which has a spectrum free from hyperfine structure and all other disturbing influences. Because the committee felt that it did not have sufficient information to choose a standard radiation it suggested that certain laboratories pursue further work that may determine such a choice. In previous reviews (88, 89) it was mentioned that artificial mercury-198, manufactured by transmuting gold, had supplied a new and finer ultimate standard of length than the red radiation of natural cadmium. German metrologists (78) have nominated krypton-84 as a candidate for this honor, and Soviet scientists ( 9 ) have recently proposed that red radiation from enriched cadmium-114 be the unit of length. Comparisons of wave-length standards from Hglss lamps show a reproducibility of ~0.00005A., or i l part in lo8, according to Barrell (7). Relative wave lengths and energy levels of Hg198 and Hg202 were accurately measured by Burns and Adanis (23), who then did the same (24) for HgIggand HgZo0. Vacuum wave lengths ranging from 3555.3204 to 6539.9183 A. were measured for 84 argon radiations relative to red cadmium with two different reflection echelon interferometers by Littlefield and Turnbull (82). At this point the writer wishes to mention the possibility that natural argon-40 might conceivably be a source of the ultimate standard of length; it comprires nearly 1% of the earth’s atmosphere, and requires only the development of an atomic beam lamp to produce lines practically free from Doppler widening. Because standard wave lengths must be free from disturbing effects to be highly reproducible, the effect of some argon gas in Hg’se lamps was investigated by Barrell ( 7 )and by 5Ieggei-s and Bosman (98). The former measured wave lengths emitted by Hg198 lamps containing 3, 5, or 10 mm. (of mercury) of argon, and the latter did likewise with lamps containing 2 or 16 mm. of argon. Both agree that the wave length of the green radiation 5461 A from mercury increases about 0.0001 A. per millimeter of argon. The change in wave length of krypton lines as a function of gas pressure was investigated by Kosters and Engelhard (7’8). The pressure effects of argon or helium on the resonance lines of silver were investigated in absorption up to 83 atmospheres (68); changes in wave length were found to be linear with relative density of gas, but they increase with argon and decrease with helium and are strongly dependent on the total angular momentum of the silver atoms in excited states. The dependence of 54
In astrophysics the determination of temperatures, relative abundance of elements, and interpretation of forbidden spectral lines rests heavily on atomic transition probabilities, oscillator strengths, or intensities of the observed spectral lines. Recent experiments to determine absolute oscillator strength9 are represented by the work of Estabrook (40) on chromium, the results of Huldt and Lagerqvist (61)on chromium and manganese observed in flame spectra, and of Kopfermann and Wessel (77) for resonance lines 3720 and 3737 A. of iron excited in an atomic beam. Quantum-mechanical calculations have been made of line strengths for oxygen 11,neon 11,sulfur 11,and argon I1 spectra (49), for calcium I1 lines (54, 1-99),and for the sharp and diffuee series of neutral potassium (141). The last mentioned calculations, performed with an IBM calculator ired to compute wave functions and transition probabilities. were inspired by studies of energy distribution during a combustion process. Such calculations are greatly desired for fundamental research in complex spectra but are generally too expensive for civilian laboratories. The last review (89) neglected to mention that the most complete compilation of experimental and theoretical spectral-line intensities appears in the well-knoTvn “Landolt-Bornstein” (80). A new record in radiometry appears to have been set by Watanabe and Inn (14-9), who measured the intensities of lines in the principal series of helium from 584 to 516 A. with a thermocouple. ZEEMAN EFFECT
There have not been any recent observations of Zeeman effect, but a number of important papers based mainly on prewar investigations have been published: measurements of the Zeeman effect in manganese I from 2460 to 4825 A (261, asymmetries of Zeeman patterns and g-values in manganese I (2.5), measurement of nearly 1000 patterns in rhenium spectra (SS), a second paper on Zeeman effect and structure in tantalum I ( l l ) ,measurement of Zeeman patterns for 540 ruthenium spark lines (94) photographed in 1940, Zeeman effects in four spectra of tellurium (53) from spectrograms made in 1936, Zeeman effect in chromium 1(7-9) and chromium I1 (74), and first Zeeman data for dyclprosium 11(IS)observed in 1940. FORBIDDEN LINES
Although atomic spectra, in general, obey simple selection rules that govern transitions between atomic energy levels, ‘[forbidden lines” are occasionally detected under special circumstances. In the case of cadmium a comparison of line intensities from enriched isotope samples proves (58) that the observed forbidden lines arise from the nuclear magnetic moment of an isotope having odd integral maw. A preliminary report of this was reviewed in 1950 (88). A new examination of the palladium arc spectrum resulted (126) in the discovery of a nen- type of forbidden transition explained by the theory of “forced” dipole radiation, a type of Stark effect which is due to the interionic fields in thc arc and rauses a sharing of parity of atomic levelq TERM AhALYSIS
The critical compilation and publication of “.4tomic Energy Levels” was mentioned in previous reviews (87-89) This project was begun in 1946 with the ohjcct of publishing a t intervals
V O L U M E 26, N O . 1, J A N U A R Y 1 9 5 4 of about 3 years a series of volumes containing all available information regarding atomic energy levels derived from the analysis of optical spectra. Such information includes relative energy levels, spectral terms, quantum numbers, magnetic splitting factors, electron configurations, and ionization potentials. Volume I, containing data from 205 spectra characteristic of the first 23 atomic numbers (1H to 23V) was published in 1949. Volume 11, presenting similar data from 152 spectra of 18 elements (24C:r to 41Kb) was published in 1952 (100). Volume I11 of this series is now being prepared; it will collect all available atomic data for 34 elements (42Mo to 5iLa, and i2Hf to 89Ac) The fourth and final volume will deal with the two groups of rare earth elements (58Ce to 71Lu and 90Th to 98Cf) most of which are still inadequately investigated spectroscopically. Recent publications on spectral structure are noted below; in many cases the new data on atomic energy levels were supplied in advance of publication for inclusion in the first two volumes of “Atomic Energy Levels.” S e w wave lengths and term tables for spectra of the He I isoelectronic sequence-namely, beryllium 111,boron IV, carbon V, oxygen VII, and fluorine VI11 have been published by E d l h ( 3 7 ) . Sindar data from extreme ultraviolet aic spectia of boron, indium, thallium, lead, and bismuth were published by Clearman (29). Photographic infrared observations have revealed ( 9 5 ) many new lines and some new energy levels of gallium I, indium I, and thallium I. Further infrared observations with lead sulfide detectors have recorded new lines in the spectra of argon (64), magnesium (&), and lanthanum (89). A thermocouple was used by Humphreys (63) to detect a t 12.37 microns the first line of the sixth series of hydrogen. Essentially complete descriptions and analyses of the first two spectra of chromium have been recently published by Kiess (73, 7 4 ) . These papers give full details on more than 1800 chromium I1 lines, and on 4400 chromium I lines; they are excellent examples of thorough and reliable investigations. The spectral terms identified with the low even configurations of molybdenum I have been reported by Trees and Harvey (138). The analysis of the second spectrum of manganese has been extended to 1100 classified lines (SS),and new terms have been found in the second spectrum of iron (120). A second paper on the first -pectrum of tantalum has appeared (11), and an extension of the second spectrum of zirconium in the infrared has been reported ( 7 2 ) . Primarily for purposes of analysis, entirely new descriptions of the arc and spark spectra of rhenium, comprising about 4200 rhenium I, and 1800 rhenium I1 lines, were published by Meggers (90). *4paper on a single intersystem transition in the first spectrum of beryllium gave the connection between singlet and triplet terms and reported (14) the measured intensity ratio of beryllium resonance lines as 3 X lo7. The only recent contribution to regularities in the spectra of rare earth elements is a report ( I S ) on four recurring intervals between dysprosium I1 lines whose Zeeman patterns indicate that the low eneigy levels belong to 6I and 4I terms, among n-hich from 4jlo 6s’ probably represents the ground state of Dy+. Despite the fact that few physicists are now engaged with fine-structure analyses of atomic spectra, there is occasionally duplication of effort. Thus, identical intervals of the (4So) 3p 3P term of oxygen I have been measured interferometrically by Parker and Holmes (115), and by Davis and IIeissner ( S 4 ) . HYPERFINE STRUCTURE
The publications mentioned under term analysis deal only with the fine structures of atomic spectra, excluding hyperfine structures and isotope shifts arising from interaction between optical electrons and atomic nuclei. IIon ever, these interactions yield highly quantitative information about certain properties of atomic nuclei and these will surely continue to be investigated so long as a satisfactory theory of nuc,le.itr structure is still lacking.
55 Hyperfine structures occur only with nuclei possessing magnetic properties, and when resolved, measured, and interpreted, they give numerical values to the mechanical ( I ) , magnetic ( p ) , and electric quadrupole (Q) moments of the nuclei. A fairly complete compilation of these data has been supplied by Klinkenberg (76). Here reference will be made only to contributions from optical spectroscopy published during the past two years. Hyperfine structures of the resonance lines indicate (70) that p = f0.136 nuclear magneton for Aulg7. For Pdlo5hyperfine structure discloses (130) that I = 5 / 2 ( h / 2 ~ )and p = -0.57 nuclear magneton. Both Irlgl and Ir1g3 have Z = 3/2 ( h / 2 ~ ) and , for 1rlg3p = +0.17 nuclear magneton and Q = +1.0 barn (1 barn = square centimeter) (108). Resolved hyperfine structure of arsenic I1 lines gives for APE: p = 1.45 nuclear magnetons and Q = 0.32 barn (110). From hyperfine structure in antimony spectra values of Q = -1.3 barn, and -1.7 barn have been derived for SblZ1and SbI23respectively (129). Results reported for Tala1are I(. = 2.1 nuclear magnetons and Q = +6.5 barn (20). A final report (71) on nuclear moments of Tcgg confirms that I = 9 / 2 ( h / 2 ~ ) ,p = 5.5 & 0.3 nuclear magnetons, and indicates that Q = +0.34 L5 0.17 barn. The fact that no more than six components are found in many Am241 lines proves that Z = 5 / 2 ( h / 2 ~ )for that nucleus (44). The hyperfine structure of the resonante lines of indium has been completely resolved finally with an atomic beam source ( 3 6 ) ; the high spectroscopic accuracy led to values of the hyperfine structure splittings of the 5 *PI;,, 21, terms which agree within L50.0004 K with microwave measurements. In rare earth spectra, partially resolved hyperfine structure of gadolinium lines suggests that for GdIS5and GdIs7Z Q 3/2 (h/2s), p 0.25 nuclear magneton for Gd’M and p 0.3 for Gdls7 (136). Likewise partially resolved hyperfine structure of dysprosium lines suggests that for Dy161 and Dyl63probably Z = 7/2(h/2s), and their nuclear magnetic moments are approximately equal (107). In that paper the eccentricities of odd-proton nuclei, calculated from all existing Q data, are plotted us. the proton number; they appear to attain large values in certain light and heavy elements but the significance of this is not understood. The correlation of nuclear properties with nuclear structures and forces is still incomplete. A summary of 22 outstanding problem nuclei-that is, those naturally occurring isotopes for which the nuclear moments are uncertain or unknown-has been given by McNally (83).
-
-
ISOTOPE SHIFTS
Most natural elements consist of a mixture of two to 10 isotopes, and the different isotopes of any given element emit slightly different spectra whose separations are measured as isotope shifts. Many such measurements were referred to earlier (87-89), and a complete compilation of quantitative data on hyperfine structure and isotope shifts which accumulated up to 1951 was published in “LandoltrBornstein” (81). Only recent (1951-1953) literature on isotope shifts is cited here, Details of experimental procedure and results are omitted, and only the isotopes for which spectral shifts have been measured are listed. These are: He3, He4 (46); Li6, Li7 (61); 0l6, 0l8 (114); Mgz4, ;\fg25, MgZ6(104); Mog5 hIog7(130); Pd106,Pd108, Pd1l0(130); Te123, Te125 (118); Eu161, Eu153 ( 1 6 ) ; G c I ~Gd158, ~ ~ , Gd1so(16, 18, 134); Er164, Er166. Er170 (109, 145); Yb170, Yb171, Yb172, Yb173, Yb174, Ybt76 (16, 111); W-‘52. WIa4, W I a 6 (105); Ob186, Os158, Osxq2(1%); Pb204, Pb206,PbZo7,Pb2”, Pb2lO (17, 106, 131); Th23OJ Th232 (13%); u 2 3 3 , u234, u235, U236, U238 ( l d 7 ) ; Pu238, Pu239, Pu240, Pu24* (30).
Although the centers of gravity of spectral terms of different isotopes of an element are usually relatively displaced, the interpretation of these shifts is still far from satisfactory. For light elements this isotope shift is generally explained as a “nuclear mass effect,” whereas for heavy elements i t is usually ascribed to a “nuclear volume effect.” In the latter cases, for more than two isotopes, the term centers of gravity are often not
ANALYTICAL CHEMISTRY
86 equidistant. Some of these irregularities are blamed on perturbations and others on “magic numbers.” Attention has been called ( 1 7 ) to the fact that isotope shifts show a discontinuity for closed shells of 50,82, and 126 neutrons. The isotope shifts for elements of intermediate weight have not been discussed in detail. Coming from many different sources and measured with various degrees of accuracy, the reported isotope shifts leave something to be desired in homogeneity and precision. In some spectra isotope shifts have been measured for only one or two lines while in others for several hundred. In a t least two cases (105, 146) claims were made that it is possible with the aid of isotope shifts to differentiate electron configurations, thus offering hope that in the future they may aid in the fine-structure analysis of complex spectra, as well as in clarification of nuclear physics. Most of the recent researches on hyperfine structures and isotope shifts have been performed by the demilitarized Germans and Japanese. In the preceding review (89) the creation of a Joint Commission for Spectroscopy was mentioned. That commission was organized to collect information on spectroscopic research, equipment, results, and problems of mutual interest th the International Union for Pure and Applied Physics and to the International Astronomical Union. At its Rome meeting in September 1952, the Joint Commission for Spectroscopy received provisional reports of Subcommittees on Institutes and Equipment, Atomic Spectra, Spectra of Diatomic Molecules, Spectra of Polyatomic Molecules, Notation for Atomic Spectra, Notation for Spectra of Diatomic Molecules, and Exchange of Research Problems. These reports, along with minutes of the Rome Meeting, have been published as Transactions of the Joint Commission for Spectroscopy (137); they should be read by all living or active spectroscopists. Four recommendations of that commission are so important that they are quoted here: The Joint Commission for Spectroscopy recommends the use of the tables of E d l h for converting wave lengths in standard air to wave lengths in vacuum. Those tables are based on a new formula (38), (n - 1) lo7 = 643.28 i- 294981(146 - d ) - I 2554.0 (41 - ~ 2 ) - 1 , which represents the index of refraction n of standard air with an accuracy of f1 X 108down to 2000 A. The Joint Commission for Spectroscopy reommends the universal use of the symbol u for wave number instead of Y, in order to avoid confusion with the frequency symbol. The Joint Commission for Spectroscopy recommends that the conversion factor for computing the energy in electron volts from wave number units (kayser) shall be 0.00012395 to conform to the standard book “Atomic Energy Levels.” Various proposed values of that conversion factor made it seem desirable t o advise universal use of one value. The Joint Commission for Spectroscopy recommends that the unit of wave number hitherto designated as cm.-l be named kayser with the natural abbreviation K . This recommendation supports the most logical proposal for a simple symbol and twosyllable word for the unit that has hitherto awkwardly been represented by cm. --I and clumsily called reciprocal centimeter or centimeter to the minus one.
+
SPECTROCHEMICAL ANALYSIS
Intensive use of atomic emission spectra for quantitative chemical analysis is scarcely 30 years old, but it has expanded with acceleration since 1925. This is shown by the “Index to the Literature of Spectrochemical Analysis” prepared by Meggers and Scribner (96), Part I of which listed 1482 papers from 1920 to 1939, inclusive, Part I1 more than 1000 papers in the interval 1940 to 1945, and Part I11 about 1250 papers from 1945 to 1950. Only a few from 1951 to 1953 have been here selected to illustrate interesting developments. The publication of books dealing with spectrochemical analysis appears to be declining; in the last review (89) eight could be mentioned, but here only two-namely, “Metal Spectroscopy” (140) and “Methods for Emission Spectrochemical Analysis” (3). The latter represents a first approach t o standardization
of empirical procedures for analyzing metals, alloys, and some nonmetals, with special emphasis on exposure conditions, excitation conditions, and analytical line-pair data. Publications designed to present dependable data, “Atomic Energy Levels” (100) and an “Ultraviolet Multiplet Table” (QQ),are now available for spectra of atomic numbers 1 to 41, inclusive; they are in preparation for the remaining elements. The multiplet tables give excitation data for most of the spectral lines likely to be used by spectrographers. The rapid development of the art of spectrochemical analysis has entailed an expansion in scientific language as well as in scientific literature; it has resulted in publication (3, 59) of a suggested nomenclature in applied spectroscopy, including recommended definitions of technical terms that are used in the literature of spectrochemical analysis. Since no practical absolute method of determining chemical composition by means of spectra has been devised all quantitative analyses are relative to standards. The preparation and testing of typical spectrographic standards of steel, aluminum, tin, ores, glass, and minerals has been described by Scribner and Corliss (125). The direct synthesis of spectrochemical standards by powder metallurgy techniques was described by Gerber (50) and Jaycox (66). Three graded series of copper standards were described by Smith (128). With regard to light sources used for spectrochemistry, four improvements or variations of the interrupted direct current arc are noted (6, 91, 22, 119); there is also an improved spark generator ( 2 2 ) and a universal spectroscopic source unit (146). The construction of hollow cathode sources for spectrography (61,98, 112, 124) and of electrodeless high-frequency sources (32, 69, 147) are described. The possibilities and limitations of these sources have not yet been fully explored, but they appear to be especially suited for gas analysis. Noteworthy applications of spectroscopy to gas analysis are found in the determination of fluorine by means of a calcium fluoride band (47, 52) in arc spectra; chlorine, bromine, and iodine in an improved hollow-cathode source (51); chlorine and bromine in engine deposits ( 8 4 ) ; chlorine in stony meteorites (121); oxygen in steel ( 1 1 7 ) ; nitrogen in organic compounds ( 4 6 ) ; deuterium in hydrogen-deuterium mixtures (19); and total water content of a hydrate by dissolving the sample in deuterium oxide and determining the intensities of the Hp and Dp lines in a high-frequency electrodeless discharge (102). The use of photoelectric direct-reading spectrometers in metallurgical production control has become commonplace, but attention is called to two investigations (8, 10)analyzing some sources of error likely in measurements made with such instruments. A new approach to direct-reading spectrochemical analysis (15) has been described, but it is so elaborated with electronics that it is incomprehensible to an old-fashioned spectroscopist. Another contribution has been made (123) toward a description of mutual effects of certain elements on spectral-line intensities in the direct current arc. Three recent papers (4, 60, 122) deal nith the sensitivity of spectral detection of impurities in lead; one reports (4)that the limits of detection of 14 elements range from 0.00003% silver to 0.01% tellurium, another states (60) that the minimum detectable silver is O.OOOOOl%, and the third (122) says it is 0.001%. The disagreement is extraordinary! I n general, the routine spectrochemical analysis of metallic alloys is so accurate and quick that mention of recent applications will here be restricted to some examples in which chemical methods are impractical or wholly inapplicable. Thus the determination of hafnium in zirconium (31, 75, 103) is spectrographically simple but chemically complicated; likewise the separate determination of niobium and tantalum in stainless steel ( 116). Mixtures of rare earths are successfully analyzed (41, 113) only by spectrographic methods. For spectrochemical identification, wave lengths and estimated relative intensities of the stronger lines of artificial technetium
V O L U M E 26, N O . 1, J A N U A R Y 1 9 5 4 and promethium were presented by Meggers (91). Slthough neither element has been detected terrestrially, technetium has been tentatively identified in the solar spectrum (101), and appears to be relatively abundant in certain stars (97). Spectroisotopic analysis of lithium ( I % ) , plutonium (30), and various other elements (83)can now be performed by resolving the isotope shifts and measuring relative intensities. The vital importance of optics in the oil industry (43) is illustrated by the spectrographic determination of metallic constituents of petroleum (48, 66, 67, 68, 86) because some trace metals in petroleum fractions may cause deterioration of cracking catalysts, corrosion of refractories, or may promote chemical reactions resulting in lubricant breakdown and engine wear. Other trace metals are employed as additives to oils or gasolines. rlfter astrophysics, which is universal, the next largest domain of applied spectroscopy is geochemistry. As a tool in geochemistry ( 1) spectrochemical analysis has demonstrated that the potassium-rubidium ratio is constant in all common rock types and in meteorites ( 2 ) , that the abundance of radioactive K40 in dunite is actually about l / 3 0 that resulting from chemical determinations, and that the minimum age of the earth as measured by the Rbg7-Sra7 ratio is 2 X lo9years. A statistical comparison of precision of chemical and spectrographic determinations of the major constituent elements in rocks and minerals shows that the former is better for silica and alumina but the latter is superior for iron and, manganese, magnesium, calcium, sodium, and potassium oxides (35). Some alluring applications of spectrochemical analysis to massive materials are found in determinations of the strontium content of limestones and fossils (79), trace elements in sea water and marine algae ( I $ ) , and rare elements in coal. Thus the germanium content of some 200 British coals (6) and about 80 Japanese coals (65) has been reported, and the industrial significance of metals in coal is suggested by analyses of 596 samples for 38 elements (67) showing that, relative to the earth’s crust, coal ash is 10 to 185 times richer in lithium, strontium, silver, arsenic, bismuth, boron, gallium, germanium, lanthanum, mercury, lead, antimony, tin, zinc, and zirconium. Salts of the Dead Sea have been analyzed spectrographically for their rubidium content (55). Since silver has been found spectroscopically in normal human hair of all age groups ( I C $ ) , the expression “silvery locks” can now be applied to any normal human with hair. A direct spectrographic method for detecting minute amounts of obnoxious beryllium in air has been devised (27), and applications of spectrography to criminology (86, 144) are illustrated by the determination of metallic paints and poisons, and the identification of safebreakers and murderers. These examples suffice to show that spectrochemical analysis not only protects our health but also improves our morals. LITERATURE CITED
(1) Ahrens, L. H., A p p l . Spectroscopy, 6,No. 5, 11-13 (1952). (2) Ahrens, L.H., Pinson, W.H., and Kearns, M.AI., Grochim. et Cosmochim. Acta, 2, 229-42 (1592). (3) Am. Soc. Testing Materials, Philadelphia, Pa., Committee E-2, “Methods for Emission Spectrochemical Analysis,” 1953. (4) hreghini,E.,and Songa, T., Spectrochim. Acta, 5,114-23 (1952). (5) ilubrey, K.V., Fuel, 31, 429-37 (1952). (6) Bard6ca, .&., Acta P h y s . Acad. Sci. H u n g . , 1, 247-60 (1952). (7) Barrell, H., Proc. R o y . Sac. (London), 209A, 132-41 (1952). (8) Bartoli, L., andBecchio, C., Spectrochim. Acta, 5,184-200 (1952). (9) Batarchukova, N.R.,Kartashev, -4.I., and Romanova, &I.F., Doklady A k a d . N a u k . S.S.S.R., 90, 153 (1953). (10) Benussi, L., and Caroli, A., Spectrochim. Acta, 5 , 97-113 (1952). (11) Berg, G. J. van den, and Klinkenberg, P. F. A.,Physica, 18, 221-48 (1952). (12) Black, W. A. P., and Mitchell, R. L., J . M a r i n e Biol. Assoc. United Kingdom, 30,575-84(1952). (13) Blank, J., thesis, Massachusetts Institute of Technology, 1953. (14) Bozman, W. R.,Corliss, C. H., Meggers, W. F., and Trees, R. E., J . Research Natl. Bur. Standards, 50,131-2 (1953).
57 (15) Brehm, R. K., and Fassel, V.A., J . Opt. SOC.A m e r . , 43,886-9 (1953). (16) Brix, P.,2. Physik., 132,579-607 (1952). (17) Brix, P., Buttlar, H., Houtermans, F. G., and Kopfermann, H., Ibid., 133,192-200(1952). (18) Brix, P., and Engler, H. D., Ibid., 133,362-9 (1952). (19) Broida, H. P., and Rloyer, J. W., J . Opt. SOC.Ampr., 42,37-41 (1952). (20) Brown, B. RT., and Tomboulian, D. H., Phys. Rev., 88, 115862 (1952). (21) Buckert, H., Spectrochim. Acta, 4,525-32 (1952). (22) Ibid., 5, 5-8 (1952). (23) Burns, I.,and Adams, K. E., J . O p t . Sac. Amer., 42,56-9 (1952). (24) Ibid., pp. 716-21. (25) CatalLn, M.A., J.ResearchNat2.Bur. Stand.,47,502-24(1951). (26) CatalAn, h l . A , and Riquelme, 0. G., A n a l . real. S O C . espafi.fis. y quim. ( M a d r i d ) , 47A,173-80 (1951). (27) Churchill, W.L.,and Gillieson, -4.H. C. P., Spectrochim. Acta, 5, 238-50 (1952). (28) Clayton, E. D., and Ch’en, S-Y, Phgs. Rer., 85, 68-70 (1952). (29) Clearman, H. E., J . Opt. SOC.Anier., 42, 373-9 (1952). (30) Conway, J. G.,and Fred, PI., Ibid., 43,216 (1953). (31) Cooley, R. A., Martin, A. V., Feldman, C., and Gillespie, J., Geochim. et Cosmochim. Acta, 3,30-3 (1952). (32) Corliss, C.H., Bozman, W. R., and Westfall, F. O., J . Opt. SOC. Amer., 43, 398-400 (1953). (33) Curtis, C. W., Ibid., 42, 300-5 (1952). (34) Davis. D. 0.. and Meissner. K. W.. Ibid.. 43. 510-11 (1953). (35) Dennem, W.’H., Ahrens, L. H., and Fairbairn, H. W., U . S. Geol. Survey Bull., No. 980,25-52 (1951). (36) Deverall, G.V., Meissner, K. W., and Zissis, G. J., Phys. Rev., 91,297-9 (1953). (37) EdlBn, B., Arkiu. Fysik, 4, 441-53 (1952). (38) EdlBn, B., J . Opt. Sac. Amer., 43,339-44 (1953). (39) Eshbach, F. E., and Fisher, R. il., Ibid., p. 331. (40) Estabrook, F. B., Astrophys. J., 115,571-3 (1952). (41) Fassel, V. A., Cook, H. D., Krotz, L. C., and Kehres, P. W., Spectrochim. Acta, 5 , 201-9 (1952). (42) Fisher, R.A., and Sowers,H. L., J . Opt. Sac. Amer., 43,330 (1953). (43) Foote, P. D., Ibid., 42,886-97 (1952). (44) Fred. RI.. and Tomkins. F. S.. Phws. Rev.. 89.318 (1953). (45) Fred, &I., Tomkins, F. S.,Brodyy J. K., and Hamermksh, AI., Ibid., 82,40G21 (1951). (46) Frederickson, L. D., Jr., and Smith, L., ANAL.CHEM.,23,742-4 (1951). (47) Fuwa, I. K., J . Chem. Sac. J a p a n , Pure Chem. Sect., 72,985-6 (1951). and Kling, C. E., Spectrochim. Acta, 4,439-45 (48) Gamble, L.W., (1952). (49) Garstang, R. H., Astrophys. J . , 115,506-8 (1952). (50) Gerber, W.O.,Jr., A p p l . Spectroscopy, 6, No. 3,5-13 (1952). (51) Gillieson, A.H., and Birks, Congr. groupe, avance. mhthod. anal. spectrog. produits mit., 14th Congr., 1951, 155-73. (52) Gillis, J., Hout, J. E., and Kemp, X., Rev. universelle mines, 8, 284-8 (1952). (53) Green, J. B., and Loring, R. A., Phys. Rev., 90,80-2 (1953). (54) Green, L. C., Weber, N. E., and Krawitz, E., Astrophys. J., 113,690-6 (1951). (55) Halperin, A., and Sambursky, S., J . Opt. Sac. Amer., 42,47581 (1952). (56) Ham, A. J., Koar, J., and Reynolds, J. R., Analyst, 77, 766-73 (1952). (57) Headlee, A. J. W., and Hunter, R. G., Ind. Eng. Chem., 45, 548-51 (1953). (58) Holmes, J . R., and Deloume, F., J . Opt. Sac. Amer., 42,77-9 (1952). (59) Hughes, H. K., et al., ANAL.CHEM.,24, 1349-54 (1952). (60) Hughes, R. C., Spectrochim. Acta, 5, 210-17 (1952). (61) Hughes, R.H., P h y s . Rev., 91, 457 (1953). (62) Huldt, L.,and Lagerqvist, .4.,Arkiv Fysik,5,91-5 (1952). (63) Humphreys, C. J., J . Research Natl. Bur. Standards, 50, 1-6 (1953). (64) Humphreys, C. J., andKostkowski, H. J.,Ibid., 49,73-84(1952). (65) Inagaki, RI., J . Chem. SOC.J a p a n , P u r e C h a . Sect., 74, 19-22 (1953). A p p l . Spectroscopy, 6,17-19 (1952). (66) Jaycox, E.K., (67) Jones, M. C. K., and Hardy, R. L., I n d . Eng. Chem., 44,2615-19 (1952). (68) Karchmer, J. H., and Gunn, E. L., ANAL.C H E M . ,24, 1733-41 (1952). (69) Keller, R. E.,and Smith, L., Ibid., 796-9 (1952). (70) Kelly, F. M., Proc. P h y s . SOC.( L o n d o n ) , A65, 250-4 (1952). (71) Kessler, K. G., and Trees, R. E., P h y s . Rev., 92, 303-7 (1953). (72) Kiess, C. C.,J . Opt. Soc. Amer., 43,1024-6 (1953). (73) Kiess, C . C.,J.Research Natl.Bur Standards,47,385-426(1951). (74) Ibid., 51, 247 (1953).
ANALYTICAL CHEMISTRY
58 (75) Kingsbury, G.W. J., and Temple, R. B., J . A p p l . Chem. (London),1,406-11(1951). (76) Klinkenberg, P. F. A., Revs. Mod. Phys., 24,63-73 (1952). (77) Kopfermann, H., and Wessel, G . , 2.P h y s i k , 130,100-8 (1951). (78) Kosters, W., and Engelhard, M. E., Proc. verb. comm. int. poids mesures, 22,137-42 (1950). (79) Kulp, J. L., Turekian, K., and Boyd, D. W., B u l l . GeoZ. SOC. Amer., 63,701-16 (1952). (80) Landolt-Bornstein, “Zahlenwerte und Funktionen,” 6 Auflage, I Band, 1 Teil, pp. 260-71,Berlin, Springer-Verlag, 1950. (81) IbM., 5 Teil, pp. 149,1952. (82) Littlefield, T.A., and Turnbull, D. T., Proc. Roy. SOC.(London), 218A,577-86 (1953). (83) McNally, J. R.,Jr., Am. J . Phys., 20,15240 (1952). (84) Mansfield, W. O., Jr., Fuhrmeister, J. C., and Fry, D. L., .J. Opt. SOC.Amer., 41,412-16 (1951). (85) Mayer, F.X., Spectrochim. Acta, 5,63-72 (1952). (86) Meeker, R. F., and Pomatti, R. C., ANAL.CHEM.,25,151-4 (1953). (87) Meggers, W. F., Ibid., 21,29-31 (1949). (88) Ibid., 22,18-23 (1950). (89)Ibid., 24,23-27 (1952). (90) Meggers, W. F.,J . Research Natl. B u r . Standards, 49,187-216 (1952). (91) Meggers, W. F., Spectrochim. Acta, 4,317-26(1951). (92) Meggers, W. F., and Bosman, W. R., unpublished work, 1952. (93) Meggers, W.F.,Catalin, M. A., and Velasco, R., J . Opt. SOC. Amer., 42,288 (1952). (94) illeggers, W. F., and Harrison, G. R., Ibid., 43,816(1953). (95) Meggers, W. F., and Murphy, R. J., J . Research Natl. Bur. Standards, 48,33444(1952). (96) Meggers, W. F., and Scribner, B. F., “Index t o the Literature on Spectrochemical Analysis,” Part I, 1920-1939,Part 11, 1940-1945,Part 111, 1946-1950,Am. SOC.Testing SIaterials. Philadelphia, Pa. (Si) Merrill, P. W., Science, 115,484 (1952). (98) Monfds, A., Ottelet, I., and Rosen, B., Industrie china. belge, 16, 675-6 (1951). (99) Moore, C. E.,“An Ultraviolet Multiplet Table,” Natl. Bur. Standards, Circ. 488,Part 1 (1949),Part 2 (1952). (100)Moore, C. E.,“Atomic Energy Levels,” Vol. 11, Katl. Bur. Standards, Circ. 467 (1952). (101)Moore, C. E., Science, 114,59 (1951). (102) Morowitz, H. J., and Broida, H. P., ANAL.CHEM.,24,1657-8 (1952). (103) Mortimore, D.M., and Noble, L. A., Ibid., 25,296-8 (1953). (104) hlurakawa. K.. J . Phvs. SOC.J a v a n . 8.213-14 (19533. . , Ibid., pp. 215-18. Ibid., pp. 382-7. Murakawa, K., and Kamei, T., Phys. Rec., 92,325-7 (1953). >furakana, K., and Suwa, S., Ibid., 87,1048-9(1952). Ibid., 85,6834 (1952). hIurakawa, K., and Suwa, S., Rept. Inst. Sci. and Technol., Unia. T O ~ U 6O. .209-15 (1952). Xelkowski, H., and Krebs, K.; hraturwissensclr*Llten, 40, 268 (1953).
(112) Newbound, K. B., and Fish, F. H., Can. J . Phys., 29,357-61 (1951). (113) Norris, J. A., and Pepper, C. E., ANAL.CHEX., 24, 1399-403 (1952). (114) Parker, L. W., and Holmes, J. R., J . Opt. SOC.Amer., 43,103-9 (1953). (115) Parker, L. W., and Holmes, J. R., P h y s . Rev., 90,142-3 (1953). (116) Poehlman, W.J., and Sarnowski, R. E., J . Opt. Soc. Amer., 42,489-92 (1952). (117) Rosen, B., Rev. unicerselle mines, 9,445-54(1953). (118) Ross, J. S.,and Murakawa, K., Phys. Rev., 85,559-63 (1952). (119) Saito, K., Shimogawara, T., and Morinaga, K., J . Chem. SOC. J a p a n , Pure Chem. Sect., 72,269-71(1951). (120) Sales, M.,A d . rea2 soc. espafi. f6s. y qufm. ( M a d r i d ) , 49, 15-30 (1953). (121) Salpeter, E.E., Ricerche spettroscop. lab. astrofs. s p k c o h cat& cana, 2,1-63 (1952). (122) Scalise, M., Metallurgia ital., 44,568-9 (1952). (123) Schrenk, W.G., and Clements, H. E., ANAL.CHEM.,23,1467-9 (1951). (124) Schuler, H.,and hlichel, A,, Spectrochim. Acta, 5, 322-6 (1953). (125) Scribner, B. F., and Corliss, C. H., ANAL.CHEX.,23,1548-52 (1951). (126) Shenstone, A. G., Proc. R o y . SOC. (London), 219A, 419-25 (1953). Stukenbroeker, G. L., and McKally, J. R.. ,Jr., (127’)Smith, D.D., Phys. Rev., 84,383 (1951). (128) Smith, D.M., Spectrochim. Acta, 5,1-4 (1952). (129) Sprague, G., and Tomboulian, D. H., Phus. Ret., 92, 105-8 (1953). (130) Steudel, -4., 2. P h y s i k , 132,42945(1952). (131)Ibtd.. 133,438-42 (1952). (132) Stukenbroeker, G.L.,and McNally, J K.,J r . , J . Opt. SOC. Amer., 43,36-41 (1953). (133) Stukenbroeker, G . L., Smith, D. D., Werner, G. K., and LZcNally, J. R., Jr., Ibid., 42,383-6 (1952). (134) Suwa, S.,J . Phys. SOC.J a p a n , 8,377-81 (1953). (135) Sun-a, S., Phys. Rea., 83,1258-9 (1951). (136)Ibid., 86,247-8(1952). (137’)Transactions of the Joint Commission for Spectroscopy, J . Opt. SOC. Amer., 43,410-30 (1953). (138) Trees, R. E.,and Harvey, M. M.,J. Research Natl. B u r . Standards, 49,397-408 (1952). (139) Trefftz, E., and Biermann, L., Z . Astrophys., 30,275-81(1952). (140) Twyman, F.,“Metal Spectroscopy,” London, Charles Griffin and Co., 1951. (141) Villars, D.S.,J . Opt. SOC.Amer., 42,652-8 (1952). (142) Yoigt, G.E., Deut. Z . ges. gerichtl. Med., 41,151-3 (1952). (143)Katanabe, K., and Inn, E. C. Y . ,J . O p t . SOC.Amer., 43,32-5 (1953). (144) Werner, O., Angew. Chem., 65,69-78 (1953). (145) Wilets, L., and Bradley, L. C., P h y s . Rez., 84,1055 (1951) (146) Yeager, E., Yeager, J., and K o l f , W., Ihid., 81,164 (1951). (147) Zelikoff, M., Wyckoff, P. H., Aschenbrand, L. hI., and Loomis, R. S., J . Opt. SOC.Amer., 42,818-19 (1952)
Mass Spectrometry VERNON H. DIBELER National Bureau o f Standards, Washington,
T
D. C.
HE broad survey of the literature of mass spectrometry
attempted in the preceding review (61) is continued in the present one on the premise of expanded mutual interests of chemists and physicists in the many fields of science t o which mass spectrometry contributes. An effort has been made to include papers, especially of foreign origin, published too late t o be included in the previous review. Activity in mass spectrometry continues to increase in scope and intensity. Aside from the magnitude of the present review, comprehensive accounts of the theory, instrumentation, and applicatjons of mass spectrometry (16, 84, 160, 208, 209, 230) attest to this fact. Other reviews (57, 66,149,167,181, 182, 191, 2565,257)consider recent developments and applications in varying detail, stressing for the most part chemical and isotopic analysis.
INSTRUMEYTATION, METHODS AND TECHNIQUES
Wattauch (903, 204) gives an extensive description of the various types of electrical and magnetic field arrangements used in conventional instruments for precise mass measurements. He also reports a method for determining dispersion in mass spectrography (207). The construction of a Mattauch-type mass spectrograph with a resolution of about 1/12,000 is reported by Delfosse and NBve de Mevergnies ( 5 9 ) . Herzog (140) describes a new instrument with anastigmatic image which utilizes a combination of a spherical condenser and a homogeneous magnetic field. A new instrument utilizing no magnetic field is described b y Paul and Steinwedel (246). Miller and coworkers (212) report the development and certain clinical applications of a port-
