ANALYTICAL CHEMISTRY
16 (53) S h t a n d e l , A. E.,and S h o s t e n k o , Y . V., Zhur. Priklad. Khim., 21, 173 (1948). (54) Sohn, A. W., Das Papier, 4,379 (1950). (55) S o p h i a n , L.H., and C o n n o l l y , V. J., J . Phys. & Colloid Chem., 55, 712 (1951). (56) Spencer, B.,and Williams, R. T., Biochem. J., 48, 537 (1951). (57) Stubbs, A. L.,S. African Ind. Chemist, 2, 133 (1948).
(58) T e l e p , G.,and Boltz, D. F., ABAL. CHEM.,23, 901 (1951). (59) Tousey, R., J o h n s o n , F. S., R i c h a r d s o n , J., and Toran, N., J. Optical SOC.Am., 41, 696 (1951). (60) V a r s a n y i , G., Magyar Folyoirat. 56, 47 (1950). (61)Yamabe, S., Japan. J . Pharm. Chem., 22,23 (1950). RECEIVED h-ox-ember 12, 1961.
X-RAY ABSORPTION HERMAN A. LIEBHAFSKlGeneral Electric Co.,Schenectady, ,V.Y .
M
OST of the present review is devot,ed to the determination of tetraethyllead fluid in gasoline by methods involving x-ray absorption. T E T R A E T H Y L L E A D F L U I D IN G A S O L I N E
,
The determination of tetraethyllead fluid in gasoline is among the most impartant routine analyses carried out in the petroleum industry. It exemplifies the advantages and disadvantages of all the common methods involving x-ray absorption-relative to each other, and relative t o the chemical and the polarographic. The determination was the subject of an American Society for Testing Materials symposium held in 1950 (I), most papers of which appeared too late for the previous review (SO). These papers are so extensive that it is possible t o review only the salient features of each for comparison with those of published investigations not included in the symposium, The earlier attempts t o use x-rays in this determination are described in various papers (8, 22, 41). The methods are described briefly in Figure l and Table I. The classification of Table I is retained in the subsequent discussion. Method I [Birks, Brooks, Friedman, and Roe (6)J. In principle, the x-ray fluorescence method is simple and specific. The actual situation is more complex, as is shown by the following comments: The polychromatic x-ray beam from a molybdenum target entered the leaded gasoline sample to excite the characteristic x-ray lines, of which the L a line of lead (1.17 A.) and the K , line of bromine (1.04 A.) were used to establish the amounts present of the two important constituents of one ethyl fluid-namely, tetraethyllead and ethylene dibromide. Intensity measurements (more precisely, measurements of radiant flux or power) were made by taking 1-minute counts on a Geiger counter. In welcome contrast with another application (4), the resolution here was good enough so that counts a t the peak intensities (once these were located) and of the background sufficed to establish the concentrations sought. A slight overlapping of the relatively weak ,584 line of lead with the Ka line of bromine could be regarded ae contributing to the background on one,side of the bromine line. Self-absorption, or the decrease in intensity of a characteristic line owing to absorption by the sample of both incident and emitted beams, had to be considered-for example, ethylene
dibromide a t the same concentration in gasoline will yield the K , line of bromine a t lower intensity if tetraethyllead is also present. The effect of self-absorption can be allowed for if the composition of the ethyl fluid is known or determined, and if suitable calibration curves are available. The contribution of sulfur to self-absorption should ordinarily be negligible.
I
I
I
! /
t H ~~
p 2
9
Figure 1. Relationship of Yarious Analytical Methods to Fundamental X-Ray Data for Lead (33) I n Method I, energy of wave lengths below, say l A , excites t h e L line, whose intensity is measured. I n Method 11, absorptiometric d a t a are taken a t wave lengths above and below a n absorption edge. Extrapolation t o the edge gives the change in absorbance from which the amount of the corresponding element is calculated, Several absorbance measurements are necessary. I n Method I11 a n absorbance measurement is made a t a single wave length preferabl; one somewhat shorter than t h a t of a n absorption edge. Everyklement free or combined, in a sample will absorb x-rays, which keepe the method frob being specific. I n Method IV, a n absorbance measurement is made b y use of a beam containing a considerable range of wave lengths. The polychromatic character of the beam can cause complications. The second sentence describing hlethod I11 applies here.
The background count is of unusual importance, owing to scattering of the incident x-ray beam by the light elements (carbon and hydrogen) that make up most of the sample. At 4 ml. of tetraethyllead fluid per gallon, the background count (5100 Der minute) was half the total, so that the intensity of the La line of'lead (to which the concentration of tetraethyllead Table I. Incomplete Characterization of X-Ray Methods (33) fluid is proportional) was also measured Extreme Adaptability by half the total count. At lower conMeasured Voltage t o Control Method Name Type Specific Intensity Stabilization in Plant centrations, the background will be relaI X-ray fluores- 8pectrophotometric Yea Low Desirable Possibly tively larger. Furthermore, the backcence ground will vary with the composition I1 AbsorptionSpectrophotometric Yes Low Desirable Possibly edge of the sample, especially with the carbonI11 Absorptiometry Spectrophotometric No Low Desirable Possibly (monochromatic) hydrogen ratio. This v a r i a t i o n will IV Absorptiometry Photometric NO C a n be Sometimes Definitely (polychromatic) high mandatory ordinarily be negligible but may have to be considered in special circumstances.
V O L U M E 24, NO. 1, J A N U A R Y 1 9 5 2
17
closer t o the direct than to the comparative method (43). With truly monochromatic beams, such commutation is not so important because Relative Cost the effect of voltage fluctuations is less serious (31, Speed Operp. 1011). Elapsed Operator Samples Equipator time time per d a y Bccuracy ment time Hughes and Hochgesang also demonstrated that Hours Min. TEL M1. j?uid/gal. tetraethyllead and sulfur could be simultaneously Chemical 24 40 12 0.04-0.06 1 5.7 determined by making absorbance measurements Polarograph 1-4 30 16 0.05 10 4.3 a t two wave lengths (0.76 and 0.96 A,)-an exX-ray 0.12 7 60 0. l o a 15 1.0 ample of differential absorptiometry with mono= With base stock constant, or sulfur content known, 0.10 becomes 0.03. chromatic beams. This differential method takes advantage of the decreases in the absorbance of lead lcross the L absorption edges near 0.8 The authors emphasize the importance of selecting the x-ray A., and to this extent resembles the method of Glocker and Frohnmayer (29; see also above). target so as to minimize scattering. The authors propose the use of an auxiliary counting system t o Hughes and Hochgesang’s way of solving the sulfur problem monitor the incident beam in order to increase precision. This inrequired the use of a gas-filled x-ray tube with a thorium target, and it will become better adapted to routine applications if a genious proposal is likely to be simpler and more effective than more stable thorium-target tube ever becomes commercially attempts to stabilize voltage and current during long counting available. periods. They also make various suggestions useful for placing Method IV [Calingaert, Lamb, Miller, and Noakes (8)]. These their method on a routine basis. authors have carried out the most thorough investigation pubIn the opinion of the reviewer, the outstanding advantages of lished on the suitability of the General Electric x-ray photometer the x-ray fluorescence method are two. First, the mass or volume of sample need not be measured, provided the thickness exceeds (35) for the determination of tetraethyllead fluid in gasoline under various conditions. Their article should he compared with the minimum required to give the maximum count a t the wave that of Vollmar, Petterson, and Petruzzelli ( d l ) , who in 1949 relengths of interest. (The density of the sample may be needed for ported results for this determination on such an instrument used the calculation of results.) Second, lead and bromine can be determined simultaneously because the method is specific. Birks, without supplementary voltage regulation in a routine petroleum Brooks, Friedman, and Roe have pointed out the advantages and inspection laboratory. Some of these results were presented in greater detail by Vollmar a t an ASTM symposium (40). disadvantages of this method for the determination of tetraethyllead fluid in gasoline in a valuable exploratory investigation, The General Electric x-ray photometer is a split-beam nulltype instrument employing the comparative method. It is clear and an extensive study of the method (modified as they have from the nature of the comparative method (43) that the use of suggested) under actual operating conditions ought to be underpolychromatic beams introduces complications not present when taken. Method I1 [Calingaert, Lamb, Miller, and Noakes (S)]. In the beams are monochromatic, and that these complications decooperation with L. K. Frevel, these authors attempted to apply crease in importance as standard and unknown approach each other in ultimate composition and in mass, provided commutathe method of Glocker and Frohnmayer (19), which has been distion between the two is sufficiently rapid. cussed in earlier review (28-30), to the tetraethyllead problem. Unfortunately, preliminary measurements indicated that the Dow Calingaert, Lamb, Miller, and Noakes report improving the perautomatic x-ray absorption specpometer could not resolve satisformance of the General Electric x-ray photometer by making factorily the bromine and the lead LIII absorption edges (8). several changes, the most important of \%hichwere the installation of a Sorensen voltage regulator and the elimination of the Preliminary tests indicated that the General Electric XRD-3 spectrometer could accomplish this resolution, and further work copper oxide rectifiers in the phase-sensing detector bridge rirwith this instrument is in progress. cuit. Method I11 [Hughes and Hochgesang ( Z Z ) ] . The authors The performance of the photometer as modified is described in have developed a monochromatic absorptiometric method, which the folloping quotation: has been employed extensively and successfully for routine deThe input voltage could be set a t a desired voltage nith an terminations of tetraethyllead fluid. accuracy of f0.05 volt and it would remain constant to kO.10 In a modified Xorelco x-ray photometer, the absorption cell volt for 0.5 to 3 hours. In fact, it mTas found that after a warm-up was placed between a monochromator (rock salt crystal) and the period of 15 to 30 minutes with the x-ray tube operating, the input Geiger counter housing. The K , line of molybdenum (0.71 A.) voltage would register within 0.10 volt of the voltage used the previous day. . . , , , , The absorbance readings of aluminum was used because targets of that metal are available in commerversus aluminum were reproducible to AO.1 mil and those of cial tubes, and because this line is strongly absorbed by lead, hydrocarbons versus aluminum to rt0.4 mil for periods of several which has its LI absorption edge a t 0.781 A. The time to was deweeks (8). terlhined for 10,000 counts with a nickel standard in the beam. The corresponding time, t , with the sample in the beam was next The authors investigated very carefully the desirability of measured and followed by a subsequent determination of to, the having standard and unknown approach each other in ultimate latter being done to establish an average t o for both this sample composition. They introduced the useful device of employing a and the next. From the ratio t/%, tetraethyllead contents were solid hydrocarbon block (polystyrene) as a reference standard obtained with the aid of a nomograph. Table I1 summarizes the [for use of solid standards by another laboratory (26, 27), see authors’ evaluation of their method. below]. With this standard, the sensitivity to voltage fluctuaHughes and Hochgesang tested their method on polychrotions of absorbance measurements on hydrocarbons was reduced matic beams with a hydrocarbon standard and reached the tenfold (presumably from what it had been with aluminum as thoroughly logical conclusion that monochromatic beams are to reference standard, cf. 8, Table 11). When an x-ray beam passes through matter, the intensity be preferred under their conditions. The polychromatic beams (strictly, the power) of the beam is reduced owing to true (photoas generated were of course too powerful for the Geiger counter and had to be stopped down. electric) absorption and to scattering. Although the two processes together determine the value of the mass absorption coHughes and Hochgesang did not provide for rapid commutation between standard and unknown, so that their method lies efficient (fl, pp. 515, 535), the latter process is often ignored beTable 11. Comparison of Chemical, Polarographic, and X-Ray Methods for Determination of Tetraethyllead in Gasoline (22)
. .
18
ANALYTICAL CHEMISTRY apparent anomalies encountered in absorbance measurements are accounted for by scattering and filtering effects.” It is the virtue of the X 100 Aluminum comparative method that all these anomalies 9.1 become negligible if standard and unknown are 5.6 properly chosen. 2.5 1.3 Method IV [Levine and Okamoto (27)]. This 0.8 investigation shows that . polychromatic x-ray beams can be used successfullv with a Geiger counter as detector in the routine determination of tetraethyllead in gasoline, provided the comparative method is employed. A modified Korth American Philips 90” Geiger-counter x-ray spectrometer was used, the most important modification being the addition of a sliding dual cradle (26, Figure 4) to permit the rapid and reproducible insertion into the polychromatic x-ray beam of either standard or unknown. (This cradle performs the function of the stand in 43, Figure 2.) Special care was taken to work with a beam of effective wave length 0.53 A,, for which carbon and hydrogen have virtually identical mass absorption coefficients. Extensive use wae made of polystyrene rods as standards to replace liquid hydrocarbons (see 8 and 26, p. 701 and Figure 6). The authors give a careful discussion of the various corrections (including that involving the resolving time of the Geiger counter) that must be made, and show how this can be accomplished in straightforward fashion. Method IV [Liebhafsky and Winslow (33)]. After introducing the ASThI symposium, these authors presented published results (@) on the determination of tetraethyllead in gasoline by absorptiometry with polychromatic beams according to the comparative method on the laboratory photometer. Furthermore, they demonstrated that differential absorptiometry with such beams could be used to determine both tetraethyllead and sulfur in a hydrocarbon medium. The guiding principle here was that also applied by Hughes and Hochgesang (22; see also above). It needed to be demonstrated, however, that the complications possible with polychromatic beams could be reduced so as to make the method practical with a satisfactory detecting system. Blending. Various authors (notably 26, 41) have pointed out that blending operations meet the conditions under which the x-ray absorbance measurements vnder discussion have proved satisfactory-namely, analysis of leaded gasoline tyith base stock available for comparison. Liebhafsky and Winslow (33, Figure 2) give a schematic diagram of a system proposed by J. P. Smith for controlling the automatic blending of tetraethyllead fluid and base
Table 111. Per Cent Absorption Due to Scattering (8) A, A.
0.5 0.6 0.8 1.0 1.2
Mass Absorption Coefficient, p m Carbon Aluminum 0.336 1.950 0,449 3.226 0.812 7.318 1.402 13.95 2.280 23.68
Mass Scattering Coefficient, um Carbon Aluminum 0.1835 0.1769 0.1870 0.1803 0.1910 0.1841 0.1920 0.1851 0.1948 0.1870
um/pm
’Carbon 54.6 41.6 23.5 13.7 8.5
I
cause it is unimportant except a t low atomic numbers and small wave lengths (28, p. 18). Birks, Brooks, Friedman, and Roe (6; see also above) have emphasized the importance of scattering in analysis by x-ray fluorescence, where it enhances the bmkground. Calingaert, Lamb, Miller, and Noakes have shown its importance in the absorptionietric determination of tetraethyllead fluid in gasoline. As these authors point out, apparent anomalies in absorptiometry that have their root in the scattering of polychromatic beams are difficult to disentangle from those caused by changes in the effective wave length of such beams, owing to filtFation by the sample. Severtheless, they succeeded in accomplishing the disentanglement by measuring the absorbance of a 2.7-cm. polyethylene block placed a t successively higher levels in the absorption cell of the General Electric photometer. It was shown that the measured absorbance, A , is given approximately by
A =a
- c(l/D2)
(1)
where a and c are constants and D is the distance from the bottom of the block to the x-ray detector. The intercept, a, is the maximum absorbance, which will be measured when the scattered x-rays reaching the receiver are a minimum ( D very large). The 1/02 term results from geometrical considerations that govern the intensity of scattered x-rays reaching the receiver. The observed departure from the linear relationship (Equation 1) a t decreasing D no doubt results mainly because the intensity of scattered x-rays is highest for small angles (11, p. 124). The calculated per cent absorbance due to scattering as given by Calingaert, Lamb, Miller, and Xoakes is shown in Table 111. These data are especially significant because it is desirable to carry out absorbance measurements on hydrocarbons a t effective wave lengths near 0.5 A,, where carbon and hydrogen have closely similar mass absorption coefficients. Calingaert, Lamb, Miller, and Noakes say in summary: “All
~
Table IV. Reference
X-Ray Method0
Detector
~~
~
~~~~~
~
~~
~
Relative Reliability of AbsorptiometricMethods Reference Method
No. of Comparisons
Average Deviation
Maximum Deviationb
Comments
M1.T E L Ruid/gal. MI. TELfluid/gal.
=.
+
-
($6)
I11
Geiger counter
Polarographic
25.
0,052
0.10
0.16
(8)
IV
Multiplier phototube
Chemical
68
0.05
0.12
0.11
(97)
IV
Geiger counter
ChemicalC
0
0.021
0.05
0.02
(97)
IV
Geiger counter
ChemicalC
54
0.02
(40)
IV
Multiplier phototube
ChemicalC
40
0.026d
0.10
0.05
(33)
IV
Multiplier phototube
Chemical
14
0.011a
0.01
O.1lf
Sen Table I
Not given
Assumed S content, 0.07%. Corresponding correction applied. Absorbance of unleaded base stock corrected for Sulfur determined and corrected for Sulfur determined and corrected for Combined experience of two laboratories. Sulfur corrected for Laboratory photometer used. Absorbance of unleaded base stock corrected for
V O L U M E 2 4 , NO. 1, J A N U A R Y 1 9 5 2 stock by means of a General Electric x-ray photometer. Other important blending operations could be carried out similarly. Reliability of X-Ray Methods. Were sufficient data published, it would be highly desirable to calculate the standard deviation for each of the methods of Table I under the best operating conditions. One could then take the position that the results obtained under these conditions can be “guaranteed” t o within three standard deviations (32). For Method I, it is of course possible to estimate the minimum standard deviation, s, from the statistical considerations governing the standard deviations of the counting rates (18). At 4 ml. of tetraethyllead fluid per gallon, s = 0.097 ml. per gallon:
8
=
4d10,200 f 5100 = 0.097 nil. per gallon 5100
(2)
where 10,200 is the counts per minute for the sample and 5100 the counts per minute for the background. This minimum s decreases very slowly with the tetraethyllead fluid content so long as the background remains unchanged-e.g., s = 0.083 ml. per gallon a t 0.1 ml. of tetraethyllead fluid per gallon. Obviously, the precision of the fluorescence method will become prohibitively low as the fluid content decreases. For the other methods, the problem of evaluating the best precision attainable is more complex. For Method I11 as an example, this point is made evident by Levine and Okamoto (l7, Table VII). All investigators of absorptiometric methods agree that the highest precision will be attained when the unleaded base stock is available for comparison, and that uncertainty in the sulfur content of the base stock is the most serious difficulty in other cases. With aviation gasolines, this difficulty is always tolerable because the sulfur contents are low. Owing t o the situation just described, it seems advisable a t this time to attempt no more than a summary of what has been accomplished by absorptiometric methods relative to others (see Table IV). I n practically all cases, the average deviation (Table IT, column 6) is comparable with the standard deviation of the reference method. Even though no more precise conclusion can he drawn now, it is clear that the x-ray methods in Table IV have proved successful and should ultimately displace the others except in special cases.
fluorescence in about one twentieth of the time required by the usual chemical procedures ( 7 ) . X-ray fluorescence analysis really began when ?*loseleyshowed in 1913 that the characteristic lines emitted by a brass anticathode were stronger for copper than for zinc. Electrons were similarly used to excite characteristic lines in a series of important subsequent researches (U), but were gradually replaced by x-ray beams because these were more convenient to use and were less likely to disturb the sample. Recent work by Castaing (9) shows that an electron beam is to be preferred in certain cases when an electron microscope is available. Castaing (9) modified an electron microscope by introducing a second electrostatic lens to reduce the cathode image further, and by attaching an x-ray spectrometer to resolve the characteristic lines produced when the electron beam struck the sample placed in the microscope. Outstanding features of the technique are the small size of sample (one micron cube, or thereabouts) that suffices, and the absence of pronounced self-absorption effects. The latter advantage is realized because of the nature of the exciting beam and because this beam does not penetrate appreciably into the sample. Remarkable quantitative results are cited for copper alloys; the results in parentheses are the quotients I.4/Icu ( 1 4 = intensity of K , for copper in the alloy; IC“= corresponding intensity for pure copper) for alloys containing the following weight fractions of copper: 0.01 (0.0099); 0.04 (0.0373); 0.53 (0.504); 0.88 (0.867). This excellent agreement can be further improved by applying suitable corrections. There probably is no better way of analyzing metallic samples for which a point-to-point exploration is required. MISCELLANEOUS
References are given here to recent investigations of topica treated in previous reviews or reserved for future discussion. X-ray absorption in the life sciences (14-17) 6-rap absorption (10, 20, 37, 38) Uranium in solution by methods involving x-ray absorption (2, 5) Sulfur in hydrocarbons by x-ray absorp(ion (33-26) y-rays from cobalt 60 (3, 12, 13, 36, 39) (’omplications possible with polychromatic beams (34) Density measurements in supersonic flow (42) Sondestructive testing methods based on x-ray absorption (I) ACKNOWLEDGMENT
X-RAY FLUORESCENCE
The attractiveness of analytical methods employing x-ray fluorescence is greatly enhanced if high resolution a t sufficient intensity can be achieved for the characteristic lines (cf. Method I). T h e n this situation exists, measurements need be made only a t the goniometer setting corresponding to maximum intensity and a t a nearby setting to fix the background. Otherwise, additional measures must be taken, as by Birks and Brooks (4) in their investigation of the zirconium-hafnium and niobium-tantalum systems. Brissey ( 7 ) has shown that the General Electric XRD3unit with fluorescence analyzer, which diffracts the x-rays from the 337 planes of a bent mica crystal, is capable of resolving the La,lines of both tungsten and tantalum from the second-order K , line of niobium without appreciable loss of intensity. On this spectrometer, therefore, the above-mentioned situation exists not only with respect to niobium and tantalum alone, but with respect to these elements in the presence of tungsten. Furthermore, a t the Thomson Laboratory, X-ray fluorescence analysis is being used for routine determinations of molybdenum, nickel, cobalt, chromium, and iron in high temperature alloys. Chemically analyzed samples possessing x-ray absorption properties similar to the unknowns are used for standards. Analyses for these elements are made by x-ray
The author wishes to thank his various correspondents and hia colleagues, H. G. Pfeiffer and E. H. Winslow, for help with this review. LITERATURE CITED
(1) Am. SOC.Testing Materials, Symposium on Role of h’ondestructive Testing in Economics of Production, 53rd Annual Meeting, Atlantic City, N. J., June 27, 1950, Spec. Tec. Publ. 112, especially pp. 43-55 (1951). (2) Bartlett, T. W., ANAL.CHEM.,23, 705 (1951). (3) Birchenall, C. E., and Philbrook, W.O., Iron Age, 164, No. 19, 77, 174 (1949). (4) Birks, L. S.,and Brooks, E. J., ANAL.CHEX.22, 1017 (1950). (5) Ibid.. 23, 707 (1951). (6) Birks, L. S.,Brooks, E. J., Friedman, H., and Roe, R. M., Ibid., 22.1258 (1950). . . (7) Brissey, R. M., General Electric Co., Lynn, Mass., letter to H. A. L.. Oct. 17. 1951. (8) Calingaeri,’G., Lamb, F. W., Miller, H. L., and Noakes, G. E., ANAL.CHEM.,22, 1238 (1950). (9) Castaing, R., Recherche abronaut., No. 23, 41 (1951). (10) Clapp, C. W., and Bernstein, S., Gen. EZec. Rev.,53, 31 (March 1950). (11) Compton, -4. H., and Allison, S. K., “X-Rays in Theory and Experiment,” New York, D. Van Nostrand Co., 1935. (12) Czygan, W., ITon Age, 166,68 (Aug. 24, 1950). (13) Dutli, J. W., Tenney, G. H., and Withrow, H. E., U. S. Atomia Energy Comm., AECD-2719 (LADC-712) (Sept. 23, 1949). (14) Engstrom, A., Rev. Sci. Instruments, 18, 681 (1947).
20
ANALYTICAL CHEMISTRY
(15) Engstrom, A., and Amprino, R., Esperientia, VI/7, 267 (1950). (16) Engstrom, A., and Click, A., Science, 111, 37%(1950). (17) Engstrom, A., and Wegstedt, L., Acta Chem. Scand., 3, 1442 (1949). (18) Friedlander, G., and Kennedy, J. W.,“Introduction to Radiochemistry,” Chap. IX, S e w York, John Wiley 8: Sons, 1949. (19) Glocker, R., and Frohnmayer, W.,Ann. Physik, 76, 369 (1925). (20) Heller, R. B., Sturcken, E. F., and Weber, A. H., Rev. Sci. Instruments, 21, No. 11, 898 (1950). (21) Hevesy, G. von, “Chemical Analysis by X-Rays and Its Applications,” Chap. VI and VIII, Kew York, McGraw-Hill Book Co., 1932. (22) Hughes, H. K., and Hochgesang, F. P., ANAL.CHEM., 22, 1248 (1950). (23) Hughes, H. K., and Wilcsewski, J. W., Proc. Mid-Year Meeting, Am. Petroleum Inst., 30M (III), 11 (1950). (24) Kehl, W’. L., and Hart, J. C., Proc. Am. PetroleumInst., 111, 28, 9 (1948). (25) Levine, S. W., and Okamoto, A. H., ANAL. CHEM.,23, 682 (1951). (26) Ibid., p. 699. (27) Ibid., p. 1293. (28) Liebhafsky, H. 9.,Ibid., 21, 17 (1949).
(29) Ibid., 22, 15 (1950). (30) Ibid.,23, 14 (1951). (31) Liebhafsky, H. A, Ann. N . Y . Acad. Sci., 53, 997 (1951). (32) Liebhafsky, H. A,, Pfeiffer, H. G., and Balis, E. W.,ANAL. CHEM., 23, 1531 (1951). (33) Liebhafsky, H. A., and Winslow, E. H., .48i”hf Bull., KO. 167, 67 (1950). (34) Liebhafsky, H. A., and Zemany, P. D., A s . 4 ~ . CHEM.,23, 970 (1951). (35) Michel, T. C., and Rich, T. A., Gen. Elec. Re$., 50, No. 2, 45 (1947). (36) Morrison, A., .Vucleonics, 5 , No. 6, 19 (1949). (37) Muller, R. H., and Wise, E. S . ,ANAL.CHEM.,23, 207 (1951). (38) Persico, E., and Goeffrion, C., Rev. Sci. Instruments, 21, 945 (1950). (39) Schwinn, TV. L., Welding Engr., 35, 24 (December 1950). (40) Vollmar, R. C., ASTM Symposium on Instrumental Methods of A4nalysis,San Francisco, Oct. 11, 1949. (41) Vollmar, R. C., Petterson, E. E., and Petrusselli, P. A , ANAL. CHEM.,21,1491 (1949). (42) Winkler, E. M., J . Applied Phys., 22, 201 (1951). (43) Zemany, P. D., Winslow,E. H., Poellmitz, G. S.,and Liebhafsky, H. A,, ANAL. CHEM., 21, 493 (1949). RECEIVED Nov. 17,1951.
X-RAY DIFFRACTION H. S. KAUFMAN, M . W . Kellogg Co., New York, N. Y . , AND ISIDOR FANKUCHEN, Polytechnic Institute of Brooklyn, Brooklyn, N. Y .
I
S THE course of the past two years, since the appearance of
the last review on x-ray diffraction in this journal, there have been continued developments of apparatus and techniques which have further increased the utility of this tool for analytical purposes. These new developments, together with some of the most recent applications, are reviewed in this paper. APPARATUS AND TECHNIQUES
The redesigned x-ray diffraction units, manufactured by the General Electric Co. and North ilmerican Philips, are now in use in many laboratories. Both of these units are characterized by their highly stabilized x-ray outputs and the sensitive, high resolution Geiger counter spectrometers nith which they are equipped. Ohio X-ray, Bedford, Ohio, has also developed a recording x-ray Geiger counter spectrometer. Units such as these are responsible for the increased use of Geiger counter methods compared to the‘ previously favored photographic method for the detection and measurement of x-ray diffraction phenomena. Hilger and Watts (U. S.distributor, Jarrell-Ash Co., Boston, Mass.) have redesigned their demountable tube x-ray unit and are equipping it with a Geiger counter spectrometer with monitored x-ray beam to compensate for fluctuations. They have also announced a microfocus x-ray unit in Thich the loading per unit area is about 100 times that of ordinary tubes. This unit promises to be very useful for the analysis of micro specimens. For those who desire to build their own x-ray units, Garrod (38) has discussed the considerations in the design of demountable x-raj tubes. Lees and Armitage (60) have presented a circuit for automatic stabilization of continuously pumped tubes. A method has been described for increasing the safe power input into x-ray tubes by improving the water cooling efficiency (18). A great deal of attention has been given to camera modifications and design. Parsons (78)has described a method of adapting powder cameras for reflection patterns. An x-ray scanning camera (96) has been developed which eliminates diffraction line graininess due t o large crystallites. The realization that many liquid and gaseous compounds may be characterized by low temperature x-ray diffraction has provided a stimulating, rapidly growing interest in this field. -4
number of new or improved low temperature cameras and techniques have been described (1, 28, 7 4 ) . High temperature cameras have also received their share of attention. Goldschmidt and Cunningham (39) and Rilliams (105) have described such cameras. Gordon (41) has developed a precision camera and technique suitable for the high temperature measurement of the coefficient of expansion of beryllium. A novel high temperature camera utilizing a thin nickel foil as a combination x-ray filter and furnace has been described by Steward (86). The centering and mounting of high temperature camera specimens (87) and the calibration and manipulation (19) of such cameras have been discussed. Special cameras have been described by Lawson and coworkers (58, 69) for the examination of powder specimens a t high pressures. Alexander (3)has presented an analysis of the intensities from slits and pinholes in powder techniques. Considerations in the design of slits have been discussed by Garrod (38). Slits are effectively employed as a means of increasing the intensity, particularly where monochromatic radiation is employed. Wrightson and Fankuchen (111)have described a simple monochromator designed for powder work. A bent crystal monochromator has been described by Warren (103). Dumond (34) has developed point focus x-ray monochromators particularly adaptable to small-angle scattering. A collimator producing a beam of small divergence and high intensity (61) and an x-ray micro beam for examination of plastically deformed metals have been described (56). I n the field of small-angle scatter, Bolduan and Bear ( 8 1 ) have discussed the effective use of collimating apertures, while Yudowitch (21s) has evaluated the collimation errors encountered, Double crystal and slit methods for use in small-angle scattering have been compared (77). An improved double crystal spectrometer has been described by Broussard (86). Crystals suitable for this application have been discussed (85). A new x-ray method involving the use of a double crystal spectrometer for the study of imperfections has been described (76). Schulz (80, 81) has developed a method for determining preferred orientation in flat transmission and reflection samples using Geiger counter spectrometers.