spectrum was recorded at about three hours after rradiation. Except for the 139 Ba and 51 Cr photopeaks, he most intense gamma ray is the 1369 keV peak of Na. To determine whether this was all due to the fast neutron reaction on the aluminum in the matrix, some rystals were irradiated in both RT-3 and RT-4, with a cooling period of about a month between ir adiations to be sure that all 24Na had decayed away. As figure 2 indicates, RT-3 terminates closer to the eactor core and thus has a substantially higher ratio of fast to thermal neutrons than RT-4. In both ir adiations the "concentration" of sodium, in μg/g, was calculated by comparison to a standard of pure sodium carbonate irradiated under the same condiions. If the 24Na were being produced by the reaction of thermal neutrons with trace amounts of sodium in he rubies, the concentration would be the same for both irradiations. However, if the sodium were being produced by fast neutrons from the Al in the matrix, ts "concentration" would be higher in RT-3 because of the higher proportion of fast neutrons. In this case he observed sodium level could be expected to be nversely proportional to the cadmium ratio. As table 2 shows, the sodium "concentration" does increase in T-3 relative to RT-4 and the increase is very similar to the decreases in the cadmium ratios for both gold and copper which have been determined for hese two facilities (6). Thus, it seems reasonable to conclude that in these crystals all the 24Na observed is enerated from the aluminum in the crystal itself. Again, if the crystals could be readily dissolved, the Na could be removed by passing the solution through column of hydrated antimony pentoxide (7); however, with such insoluble material it is more reasonable to choose the irradiation conditions to minimize its ormation. Concentration levels of the observed trace elements were determined by comparison with standards irradited under the same conditions as the samples. After radiation the standards were dissolved in a few drops f suitable solvent in a polyethylene vial having about he same cross-sectional area as the samples and sufcient solvent added to give a solution of the same mickness as the samples. This produced the same ounting geometry for both samples and standards and mus provided for optimum accuracy of the results. ure BaCO3 and SrCO; were used as Ba and Sr standrds and primary standard K2Cr2O7 was used as a Cr andard. For most of the other elements the standards ere either pure metals or oxides. 3. Results and Discussion To date a total of fifteen samples of eight different ruby boules have been analyzed by the method described above. Two samples of very pure synthetic sapphire crystal, prepared by a vapor-phase growth technique (8), have also been analyzed. Some of the results obtained are shown in table 3 which lists data on duplicate samples of four different crystals. For crystals "A" and "B" the duplicates are adjacent vertical slices while the duplicates for "C" and "D" are horizontally adjacent sections as shown in figure 1. This listing is fairly typical of the crystals analyzed. Some of the others contained additional trace elements such as Co, La, and Au. The synthetic sapphires had a much higher purity than the rubies. The only trace elements detected were Mn, 0.0005 μg/g; Cu, 0.03 and 0.04 μg/g; Sc, 0.010 μg/g; and Ir, 0.0005 and 0.002 μg/g. The presence of readily detectable amounts of Ir in all the Czochralski rubies was a surprise initially, but was quickly accounted for when it was learned that the crystals had been grown using Ir crucibles. With the system used in this work for gamma ray spectrometry the 317 keV peak of 192Ir and the 320 keV peak of 51Cr overlap, as shown in figure 4. The concentration of Ir can be determined from the 468 keV gamma ray, but since 51Cr has no other gamma rays, a correction must be made to obtain accurate results. This was accomplished by determining the area of the 296 keV peak of 192 Ir, calculating the area of the 317 keV peak from the published decay scheme (8) and subtracting this value from the total area of the 320 keV peak to obtain the 51Cr contribution. The spectrum shown in figure 4 is from crystal "A" which has 0.24 μg/g Ir. This is the highest Ir level observed in any of the crystals and thus requires the largest correction to the 51 Cr peak. For most crystals the Ir level is about 50 times lower and the correction is quite small. Except for 51Cr and 192Ir the radionuclides used for the determination of all the trace elements listed in table 3 have half-lives shorter or comparable to that of the 15-h 24Na generated from the matrix. As is shown in figure 3, many of the gamma ray peaks from the trace elements were quite small relative to that of 24Na. This illustrates the extreme importance of minimizing the formation of 24Na from the matrix during irradiation. In the case of copper the interference from 24Na is especially serious because the decay of 24Na results in the production of significant amounts of the same 511 keV annihilation radiation as that from 64Cu which is used to determine the copper concentration. The relative amount of this 511 keV radiation formed from 24Na ABLE 2. 27Al (n, a) 24Na background in ruby crystals depends on the particular detector used and for the 47-cm3 Ge(Li) detector employed in these analyses was 3.9 percent of the 1369 keV photopeak. The peak areas for 64 Cu were corrected accordingly. The results in table 3 show that Cr, Ga, and Ir are distributed quite homogeneously within a given crystal, while Ba and Sr exhibit extreme fluctuations. Since these fluctuations occur between adjacent horizontal sections as well as adjacent vertical slices, some sort of precipitation or other segregation of Ba and Sr Ba concentration was as high as 25 percent. One cluster of these inclusions is shown in figure 5. The bright areas in the photograph are formed by the Ba x rays excited by the electron beam. 4. Detection Limits for Other Elements Although only the elements noted above were a tually observed in the crystals analyzed, it is possible to determine upper concentration limits for many other elements which could have been detected present. Table 4 lists these detection limits as well a those for the elements observed in one or more sam ples. Like any set of detection limits, these are com pletely arbitrary and are valid only for the particula conditions used in this work, including the 47-cm Ge(Li) detector used for counting. These conditions an summarized at the end of table 4. Obviously, in the case of some element of special interest the irradiation time or counting time could be increased or a detecto of higher efficiency could be used to improve the sens tivity. Many of these detection limits are experimen tally measured values; the others were calculated from nuclear constants, relating the analytical gamma ray to one of similar energy which had been measured. The limits in table 4 assume no interference from other gamma rays except, as noted, for 51Cr and 2 Ir. 192 Na and Mg are not listed in table 4 because they are subject to interference from the matrix as shown in table 1 and thus are not detectable except at relatively high levels. Sn is not listed because its detection limit is so high (2000 μg/g) that the presence of trace amounts could not be detected by this method. Other elements not listed in table 4 produce radionuclides which either have very short half-lives or do not emit gamma rays and so would not be detected under the conditions used in this work, even if present. a These elements would be observed if present in ruby at or above the levels indicated, assumming the following conditions: 1 g sample, 1 h irradiation at 1013n cm-2s-1, 1 h cooling time for decay of 28Al, 10 min counting time for short-lived activities (T < 2 h), 100 min counting time for short-lived activities (T2>2 h), all counting done with 47-cm3 Ge(Li) detector. For elements marked the limit will be about a factor of ten poorer in the presence of Ir and Cr at the levels normally observed to date. RE 5. Scanning x-ray image of Ba-rich inclusions in ruby crystal, Ba La1, 4.6μ/cm. he authors thank K. F. J. Heinrich and C. E. Fiori, of the Analytical Chemistry Division, NBS, who ducted the electron probe microanalysis investiga1. Special thanks are due to G. W. Cleek and J. L. gesen of the Inorganic Materials Division and the titute for Materials Research, NBS, who provided mples and many helpful discussions. This work was pported in part by ARPA. This paper was presented part at the 159th National Meeting, American emical Society, Houston, Texas, February 1970. 6. References Albert, P., Caron, M., and Chaudron, G., Compt. rend. 233, 1108 (1951); Albert, P., Ann. Chim. 13, 827-96 (1956). [2] For a detailed listing see Activation Analysis: A Bibliography, Lutz, G. J., Boreni, R. J., Maddock, R. S., Meinke, W. W., Eds., Nat. Bur. Stand. (U.S.), Tech. Note 467, 264 pages (Dec. 1969). Weiner, J. R., O'Connor, J. J., and Rubin, B., J. Electrochem. Soc. 110, 1160 (1963). [3] [4] Ortega, R. F., Modern Trends in Activation Analysis, Nat. Bur. Stand. (U.S.), Spec. Publ. 312, Vol. 1, 691 pages (1969) (see pp. 536-540). [5] ARPA-NBS Program of Research on High Temperature Materials and Laser Materials, Franklin, A. D., and Bennett, H. S., Eds., Nat. Bur. Stand. (U.S.), Tech. Note 531, 75 pages (June 1970). [6] Becker, D. A., and LaFleur, P. D., Activation Analysis Section: Summary of Activities, July 1969 to June 1970, Nat. Bur. Stand. (U.S.), Tech. Note 548, 165 pages (1970) (see p. 14). [7] Gills, T. E., Marlow, W. F., and Thompson, B. A., Anal. Chem. 42, 1831 (1970). [8] Parker, H. S., and Harding, C. A., J. Am. Ceramic Soc. 53, 583 (1970). [9] Lederer, C. M., Hollander, J. M., and Perlman, I., Table of Isotopes, Sixth Edition (John Wiley and Sons, Inc., New York, 1967). (Paper 75A5-676) 429-270 O 713 JOURNAL OF RESEARCH of the National Bureau of Standards - A. Physics and Chemistry Crystallography of Some Double Sulfates H. F. McMurdie,* M. C. Morris,* J. deGroot,* and H. E. Swanson** Institute for Basic Standards, National Bureau of Standards, Washington, D.C. 20234 (June 2, 1971) New information is given on cell parameters, density and methods of preparation of 50 compounds of the langbeinite and Tutton salt groups. The langbeinites have the general formula (A+)2(B2+)2(XO4)3 and the Tutton salts the general formula (A+)2(B2+)(XO4)2 ·6H2O, where A is K, Rb, (NH4), Tl or Cs; B is Mg, Ni, Cu, Co, or Zn; and X is S or Cr. A comprehensive list of references on the crystallography of the compounds is included. Key words: Chromates; langbeinites; lattice constants; sulfates; Tutton salts. 1. Introduction The crystallographic data on langbeinite A+B+(XO4)3 id Tutton salt A+B2+(XO4)2 · 6H2O type compounds ported here were obtained as part of a project at the ational Bureau of Standards1 in which x-ray powder tterns of pure phases are prepared to extend and prove the Powder Diffraction File (PDF).2 Sulfates d chromates belonging to these groups of comunds, which are not represented in the PDF by good ay powder patterns, and which could be prepared th satisfactory purity, were studied. A literature arch on other pertinent material was made and is mmarized. Completely indexed x-ray patterns, optiproperties, and a description of our methods are ported in the series NBS Monographs 253 [41, 42, ] for each phase studied. Briefly, the x-ray work s done on powders using a diffractometer, with internal standard (Ag or W) for calibration of the d acing measurements. The two groups of compounds are treated separately. each group, the methods of preparation of the sames are outlined briefly, and tables summarizing our dings and those from the literature are given. Research Associates of the Joint Committee on Powder Diffraction Standards. This project is sponsored in part by the Joint Committee on Powder Diffraction National Bureau of Standards Monograph 25, Standard X-ray Diffraction Patterns, Sec. and Circ. 539, Sec. 1-10, have been published. They are available from the U.S. Gov. ment Printing Office, Washington, D.C. 20402. Figures in brackets indicate the literature references at the end of this paper. As is customary in crystallographic studies, the cell parameters are given in terms of strom (Á) equal to 10-10 m. structure of this compound has been determined by Zemann and Zemann [64]. The basic structure is cubic with a tetramolecular unit cell and the space group P21/3 (No. 198). The SO4 tetrahedron is regular, and the B2+ ions are octahedrally coordinated. The A+ ions are of two kinds, one of which is surrounded by four oxygens at varying distances, and the other surrounded by three oxygens. Selenates or chromates of this structure have not previously been reported. The conditions for preparation of the langbeinite type compounds are given in table 1. In this study four isostructural chromates were prepared, all with B2+ = Mg2+. No attempt was made to prepare selenates. Two compounds, K2Ca2(SO4)3 and K2Cd2(SO4)3, were found to have orthorhombic distortions of the langbeinite structure. b. Results and Discussion 5 Table 2 gives unit cell size, density, PDF card number if in the File and other relevant references. These data are given for the various langbeinite compounds from both our studies and from the literature. Two compounds were distorted from cubic and have been indexed as orthorhombic; data for these are given in table 3. It is of interest to note that Jona and Pepinsky [20] reported a transition in (NH1);Cd(SO1); at about 87 K, doubtless to a distorted form. It is assumed that solid solutions will occur between most or all of these phases. The only studied solid solution here was between KM (SO); and K Ca (SO); the latter of which was orthorhombic. The compound midway between. these end members was found to be cubic, confirming the findings of Morey et al. [36]. No langbeinites could be prepared with Na+ as A and only a limited number with Cs. Also none were prepared with cations larger than Ca2+ as B2+. It is of interest to note that the order of the cell size with constant A is not in the one generally accepted as that of increasing radius of the B2+ |