1147
V O L U M E 2 7 , NO. 7, J U L Y 1 9 5 5 When 200 to 400 mg. of silica are present, attack on the glassware does not amount to more than 1 or 2 mg. As the amount of silica added is increased, the recovery of fluorine in the distillate decreases. K h e n the sides of the apparatus were washed down halfn-ay through the distillation, the recovery of fluorine was better but still not complete. J t may be assumed that part of the material which gets thrown onto the walls entraps fluorine and does not take part in the reaction. However, the results for silica were not substantially affected b y the recovery of fluorine The explanation offered is that the fluorine remaining attached to the silica is liberated as hydrogen fluoride by a hydrolysis reaction during the ignition of the silica. From a consideration of the equilibrium SiF4
+ 4H20 % Si(OH), t 4HF
which has been investigated b y Lenfesty and coworkers (7‘) it follows that a s the concentration of fluoride is reduced, the ratio of silicon tetrafluoride to hydrogen fluoride will also drop. This suggests that, in the above analyses, if the fluorine remaining in the silica residue is liberated slowly and with free access of moist air, it escapes mainly as hydrogen fluoride. This would not introduce any error into the results. T h a t the fluorine in fact escapes during the ignition was confirnied b y fusing such an ignited residue with alkali and distilling from perchloric acid. Practically no fluorine was recovered.
Table 111. Determination of Silicon and Fluorinenin Tennessee Standard Phosphate Rockb Weight of sample, grams Silica added, gram 1 s t distillation Silica in distillate, grain Fluorine in distillate, gram 2nd distillation Silica in distillate, gram Silica in flask, gram Fluorine in distillate, gram Silica found, % Fluorine found, % a Accurate procedure. b Sational Bureau of Standards. S o . j 6 b . silica, 3.4% fluorine.
1
2
2.5046
0.2735
2.5069 0.2735
0.0345 0.0798
0.0338 0.0850
Sone
0 4939 0.0058 10.18 3 42
None
0.4926 0,0029 10.09 3 31
Certified analj-sis.
10.1%
I n the experiments reported in Tables I and I1 the fluorine-tosilicon atomic ratio in the distillate varied between 4 and 10, the higher values, in general, correqponding with lower concen-
trations of fluorine in the distilling flask. I n the second distillation reported in Table 111,where the fluorine concentration was much lower, silicon could not be detected in the distillate. This effect, is in agreement with the results reported by Shell and Craig ( 1 1 ) . This means that the fluorine, when present in solution is high concentration, distills largely as silicon tetrafluoride, and when present in low concentration it dist,ills mainly as hydrogen fluoride. I n Table 11, sample S P R 1 gave practically the same results whether or not sodium silicate was added, and so did the standard Tennessee rock phosphate; on the other hand, sample XPR 2 gave a very high result in absence of added silicate. This indicates the presence of sufficient freely available silica in the first two cases and of a deficiency in the other case. Total silica content is no indication of the amount freely available to hydrofluoric acid, as shov-n by quartz being almost entirely immune to attack under the experimental conditions. ACKNOWLEDGRIENT
The authors wish to t,hank D. Kellerman for valuable euggestions and criticisms and R. Rosenblatt for assistance with the experiments. LITERATURE CITED (1) Brabson, J. A,, Mattraw, H. C., Maxwell, G. E., Darrow, dnita, and Needham, M.F., Ax.4~.CHEM.,20, 504 (1948). (2) Furman, N. H., ed., “Scott’s Standard Methodj of Chemical Analysis,” 5th ed., Vol. I, p. 803, Van Sostrand, New York, 1939. (3) Groves, A. W., “Silicate -halysis,” 2nd ed., pp. 212-13, George, Allen & Unwin, London, 1951. (4) Hillebrand, W. F., Lundell, G. E. F., Bright, H. A , , and Hoffman, J. I., “Applied Inorganic Analysis,” 2nd ed., pp. 943-5, Wiley, New York, 1929. (5) Hoffman, J. I., and Lundell, G. E. F., J . Research Natl. Bur. Standards, 20, 607 (1938). (6) Jarobson, C. A , , J . P h y s . Chem., 27, 577, 761 (1923); 28, 506 (1924). (7) Lenfesty, F. A,, Farr, T. D., and Brosheer, T. C., I n d . Eng. Chem., 44, 1448 (1952).
(8) Lundell, G. E. F., and Hoffman, J. I., Bur. Standards J . Rcsearch, 3 , 581 (1930). (9) Pietska. G., and Ehrlich, P., Angezo. Chem., 65, 131 (1953). (10) Sawaya, T., Technol. Repts. Tohoku Univ., 16, 17 (1951). (11) Shell, H. R., and Craig, R. L., ANAL.CHEY.,26, 996 (1954). (12) Travers. JI.,Chmpt. rend., 173, 714 (1921). (13) Willard, H. H., and Winter, 0. B., IKD. Esc. CHEM. SAL. ED..5, 7 (1933). R E C E I Y Efor D reriew S e y t ~ i n b e r14. 1954.
Accepted .\larch 12. 1955
Miniature Fluorescent X-Ray Spedrograph L. S. BIRKS and E. U. S. N a v a l
J. BROOKS
Research Laboratory, Washington
25, D. C.
A simple, inexpensive x-ray spectrograph with no moving parts and with provision for recording the spectra from two specimens simultaneously, side by side, on photographic film is described. The instrument is about the size of a small x-ray powder diffraction camera and is used on ordinary x-ray diffractiontype equipment. Its main use is in qualitative and semiquantitative analysis, and it covers the same range of elements as the usual Geiger-counter, fluorescent, x-ray spectrometers. With exposure times of 0.5 to 1 hour, concentrations of 270 manganese in an iron matrix and of less than 0.570 manganese in an aluminum matrix may be detected. Resolution is good enough to separate manganeseKa radiation at 2.10 A. from chromium-Kp radiation at 2.08 A.
F
LUORESCEXT x-ray spectroscopy has become widely recognized in the past five years as a very pon-erful analgtical tool. This recognition is based primarily on the development of Geiger-counter, x-ray spectrometers of great sensitivity, high resolution, and ease of automation. There remains, hoTever, a large area of qualitative and semiquantitative chemical analysis viherein x-ray spectroscopy is not applied because the problems do not warrant the rather elaborate and expensive Geiger-counter equipment, and unfortunately, no simple, inespensive spectrographs have been available. I n a n attempt t o satisfy the requirements of this large middle ground of x-ray spectroscopy, the present simple spectrograph has been constructed. It records the x-ray spectra on photographic3 film and is used on ordinary x-ray diffraction apparatus in much
A N A L Y T I C A L CHEMISTRY
1148
the same manner 8 s powder difIraction cameras and with comparable exposure times of 0.5 to 1 hour. INSTRUMENTATION
The principle of the present spectrograph is shown schematically in the plan view of Figure 1. Primary x-ray8 strike the specimen in the u m d fashion and excite fluorescent x-radiation. It is the crystal arrangement which distinguishes this instrument and permits the whole spectrum to be recorded simultaneously without m y moving parts. In this figure, diffraotion is by plane8 which are parallel to the narrow edge of the crystal so that only a narrow bundle of radiation of each wave length is passed on to the photographic a m . Thus, the diffracting region limits resolution and eliminates the need for any collimating device. This narrow diffracting region is obtained either by using a very thin crystal slab (thicknesses of 0.004 inch are easily obtained with the alkali halides) or hy moving the limiting edge in close to act as a half slit 80 that only radiation diffraoted from a m o w region of B thicker crystal is allowed t o pass on to the photographic film. The shield next to the crystal prevents stray radiation from being transmitted through the crystal.
graph. The arc, G, for holding the film has a radius of 57.3 mm., so that 1 mm. on the film represents 1 degree 8. The separator, H , is used to keep the two spectra from overlapping; a similar separator is used hetween the two specimens but is not shown in the figure. Thus with both crystals alike, a stsndard and unknown specimen may be compared, or two crystals of different spacing may be used with a single, large specimen to cover a greater range in wave length. EXPERIMENTAL RESUL'
Both qualitative and semiquantitative results have been obtained for B range of elements, and two examples are shown in Figures 3 and 4. Figure 3 shows typical spectra for a series of salts containing a range of elements from chromium (24) to lead (82). A standard. tunesten-tartreet. x-rav diffrsction tube Table I.
Comoosition of Nickel-Chromium Steels
3
0.57
9.1
5.5 2.95
4 5
6.8
25.7
2.07
1.57 0.23
was operated a t 50 kv. and 20 ma. with exposures between 0.5 and 1 hour. I n making satisfactory prints for publication, the contrast on the original films wa8 increased snd therefore has broadened the darker lines considerably. The left edge of each spectra represents 0 degree 8; the high 8 angle ends on the right have been removed above ahout 50 degrees 8. These speotra were taken two a t a time with a lithium fluoride analyzing crystal for each specimen. I n spectra 1 and 2, the adjacent elements of copper and nickel are easily distinguished. In spec-
Figure 1.
Principle of simple x-ray speetrograph with no moving parts
I n the figure fluorescent radiation of wave length, XI, emerging in a parallel bundle from only one particular small area on the specimen is diffracted by the crystal while radiation of wave must emerge from a different area; this mean8 the length, the specimens must be homogeneous if semiquantitative analyses me to be performed. With a fixed crystal, the angular measure along the film is in terms of 0 rather than 2 8, so that a 90' arc covers the complete diffraction region. Actually, the instrument as shown in Figure 2 is two spectrographs side by side-that is, two units such shown in Figure 1 with two specimens and two crystals yielding two speotra simultaneously. I n Figure 2, the housing, A, has window, B, for the primary beam (may be covered with Mylar) and a pumping tube, C, for evacuation or admitting helium; D is a light-tight shutter for the window, B, so that the film may be loaded away from the x-ray tube. The specimens are mounted a t E at about 45' with respect to the primary beam; this angle is not critical and might be changed to a smaller or greater angle to favor either the longer or shorter wave lengths. Specimen size and distance from the crystal are not important as long as the specimen covers the desired angular range with respect to the crystal. The crystal is shown a t F but only the top edge is visible in the photo-
A . Housing B. Window fi C. Tube fore D. Light-tigt E.
Specimen
F. Cryatal
(0
G. Arc for holding photographia tilm
Hr.
s.gerators
trum l5, chromium shows up even in the presence of lead. Sp
