Документ взят из кэша поисковой машины. Адрес оригинального документа : http://geo.web.ru/conf/khitariada/1-2003/informbul-1_2003/magm-1e.pdf Дата изменения: Fri Oct 31 16:28:38 2003 Дата индексирования: Sat Dec 22 05:04:05 2007 Кодировка:

INFLUENCE OF COMPOSITION AND 2 ON CHROMIUM DISTRIBUTION BETWEEN FORSTERITE CRYSTAL AND MELT
Dudnikova V.B. (GEOCHI RAN), Gaister A.V. (IOFAN), Zharikov E.V. (IOFAN), Gulko N.I. (GEOCHI RAN), Senin V.G. (GEOCHI RAN), Urusov V.S. (GEOCHI RAN).
vdudnikova@mtu-net.ru Keywords: forsterite, distribution coefficient, oxygen fugacity.

The dependencies of distribution coefficient of chromium (Cr) between forsterite single crystal and its melt upon the initial chromium concentration in the melt, oxygen fugacity of growth atmosphere as well as upon additional doping by other elements (Li, Na, Al) were investigated. Forsterite single crystals were grown from the melt by the Czochralski method. Chromium concentration in the crystals was measured by microprobe analysis, lithium concentration ­ by atomic emission spectrometry with inductively 0,3 coupled plasma. Cr was calculated by extrapolation of the dependence S/CL on crystallized fraction of initial melt g to zero (S ­ chromium concentration in the crystal, 0,2 L ­ initial chromium concentration in the 1 2 melt). Fig. 1 shows the dependencies Cr on CL 0,1 and CSo (CSo was found by the extrapolation CS to g=0). One can see that Cr decreases K Cr = -0,18C Lo + 0,23 more than three times when the initial 0,0 chromium concentration in the melt increases from 0.07 to 0.97 wt.%. The extrapolation of 0,01 0,10 1,00 C S ; C L wt.% the dependence Cr upon CL to CL0 gives the limiting value Cr equal 0.23. Cr can be Fig.1 Dependence K Cr on C So (curve 1) considered as the constant only for and C L (curve 2) (lgf O2 = -1,7). CL 0.1-0.2 wt.%. The further increase of dopant concentration in the melt retards the growth of chromium content in crystal and CSo becomes saturated what leads to decrease of Cr. Probably, the reason for this effect is the formation of chromium clusters in the melt. This makes further insertion of chromium into the crystal difficult. To study an influence fO2 on Cr the series of single crystals grown in the high purity Ar atmosphere as well as in the mixtures of N2 and O2 with oxygen content from 0.85 to 12 vol.% was explored. The initial chromium concentration in the melt was 0,3 K Cr =C S /C Lo 0.12±0.02 wt.%. Our results our data together with data from literature Cr [1] 0,2 are shown on Fig. 2 and display 3+ [2] that Kr decreases sharply at Cr [8] -4 fO2>10 . Three different models [9] 0,1 were examined for description 4+ 4+ C Cr /C Lo such behavior of Kr making Cr allowance for EPR-data [1] for 0,0 [1] determination of Cr4+­content in -8 -6 -4 -2 0 lg f O2 the crystals. In all the models we assume that chromium is present Fig.2. Dependence of Cr on f O2 and relation of different valent in the crystals in range f forms of chromium in the crystal All the points -the experimental 4 i /C
Lo

K

Cr


2

The experimental data can be closely approximated only at the assumption that Cr6+ ions exist in the forsterite melts (models II and III). According to these models when fO2 increases, Cr6+ ions accumulate in the melt, and at the same time the concentration of Cr3+ ions in the melt (and, as consequence, also in the crystal) decreases. The concentration of Cr4+ions only slightly varies with fO2 change. As a result the total chromium content in the crystals and Kr decrease sharply when fO2 increase. Partial distribution coefficient for Cr3+­ions (Cr3+) evaluated within models II and III equals 0.2; Cr4+ for model III equals 0.1. Ions Cr2+could appear in crystals by fO2<10-4 [2]. A value of Cr2+ can be evaluated by means of dependence of distribution coefficients of divalent impurities between forsterite single crystal and its melt upon their ionic radii. We investigated this dependence earlier in [4]. The value of Cr2+ equals ~ 0.5. Thereby, with increasing Cr valence its distribution coefficient decreases. The r3+ ions have a main influence on total Kr because these ions predominate in the crystals. The solubility of the r3+ions with the charge more than that of replaceable ions of Mg2+ can be increased by additional doping by ions which charge is less than that of the replaceable cations of the host crystal. To select the most suitable charge compensator we performed the computer simulation of the forsterite crystal structure using minimization of static energy of the crystal lattice using the GULP program [5]. The parameters of the interatomic potentials were taken from [6,7]. The calculations were made within the framework of an ionic model taking into account covalent effects by means a three-body O-Si-O bond-bending potential and electron polarization of oxygen ions. We compared the solution energies of r3+ ions for different mechanisms of charge compensation using magnesium vacancies vMg, monovalent ions in magnesium sublattice (LiMg, NaMg) and trivalent ion in silicon sublattice 0,35 (AlSi). According to our 3+ calculations, the r solubility decreases in the following row of 0,30 charge compensators: LiMg > NaMg > vMg > AlSi. 0,25 Kr was measured for the crystals doped only by chromium as well as by both chromium and 0,20 lithium. The initial chromium concentration in the melt was 0,15 0.06±0.01 wt.%, the initial 0 1 2 3 4 lithium content in the melt was Li/C r , a to mic ra tio in the c rys ta l varied in the range 0.01 ­ 0.42 wt.%. All the crystals were Fig . 3. Dep end enc e o f Cr fro m Li/C r ato mic ratio in the c rys tals grown in the high purity argon atmosphere since the crystals grown in such condition (lg fO2 -4) contain mainly Cr3+ [1]. First KCr increases as the lithium content in the melt increases and then becomes saturated (fig. 3). The value of KCr in case of the conjugate isomorphism with lithium can be increased about 1.5 times in comparison with that in the case of the individual impurity. The saturation of KCr dependence on lithium content occurs when the atomic ratio Li/Cr in the crystal is close to 1. This work is supported by RFBR (Grants 02-02-16360,00-02-16103, 00-15-98582). 1. 2. 3. 4. 5. 6. 7. 8. 9. References: Mass J.L. et al. J. Cryst. Growth. 1996. V.165. P.250-257. Yamaguchi Y. et al. J. Cryst. Growth. 1993. V.128. P.996-1000. Schreiber H.D., Haskin L.A. Proc. Lunar. Sci. Conf. 7th 1976. P. 1221-1258. Dudnikova V.B. et al. Materials of Electron Technique. 2000. 2. C.11-14. (in Russian) Gale J.D. J. Chem. Soc. Faraday Trans. 1997. V. 93. P. 629-637. Freeman C.M. J. Solid State Chem.1990. V.85.P.65-75. Jaoul O. et al. Phys. Earth Planet. Inter. 1995. V. 89 P.199-218. Higuchi M. et al. J. Cryst. Growth. 1995. V.148. P.140-147. Sugimoto . et al. Phys. Chem. Minerals. 1997. V.24. P.333-339.
K
Cr


3
Electronic Scientific Information Journal "Herald of the Department of Earth Sciences RAS" 1(21)2003 Informational Bulletin of the Annual Seminar of Experimental Mineralogy, Petrology and Geochemistry ­ 2003 URL: http://www.scgis.ru/russian/cp1251/h_dgggms/1-2003/informbul-1_2003/magm-1e.pdf Published on July 15, 2003 © Department of the Earth Sciences RAS, 1997-2003 All rights reserved