- Email: [email protected]
Technical notes
TECHNICAL
of each experimental run the temperature was quickly lowered to below IOO’C, and the pressure then reduced. The samples were removed from the apparatus and examined by X-ray diffraction, using a Debye-Scherrer camera of 114 mm. dia. In each case, complete transformation to a spine1 structure was observed, and even in the case of FeMnGeO, no .additional lines could be observed in the powder patterns after long exposures. The spine1 lattice parameters given in Table 1 were obtained after applying film shrinkage corrections and extrapolating the unit cell values to 28 = 180”, making use of back angle reff ections and appropriate extrapolation functions [ 1 I]. These new compounds provide a substantial extension of known olivine-spine1 transformations, Table 2. The density increases are similar to those previously observed, as expected for the transformation of one structure type to another[l2].
A. E. RINGWOOD A. F. REID
Department of Geophysics and Ge~hemist~, Australian National University and Division of Mineral Chemistry, Commonwealth Scientific Industrial Research Organization, Melbourne, Australia. RRl?ERFBCEs
1. BERNAL J. D., Observatory, 59,268 (1936). 2. JEFFREYS H., Mon. Not. R. astrn. Sot. geophys.
Suppi., 4,50 (1937). 3. GOLDSCHMIDT V., Nachr. Ges. Wiss. Gottingen, Math-Physik. KI ., 184 (193 1). 4. RINGWOOD A. E., In “Advances in Earth Science” (Edited by P. Hurley), p. 357, M.I.T. Press (1966). 5. RINGWOOD A. E. and MAJOR A., Earth &planet. Sci. Lett. 1,241 (1966). 6. RINGWOOD A. E. and MAJOR A., Phys. Earth Planer. Interiors, in press, (I 970). 7. BLASSE G., .J. inora. nuci. Chem.. 25. 230 (19631. 8. RINGWOOD A. E. &d REID A.F: (i~~~p~atiu~j. 9. WADSLEY A. D., REID A. F. and RINGWOOD A. E., Acta crystaiiogr., B&$,740 (1970). 10. GREEN D. H. and RINGWOOD A. E., Geochim.
cosmochim. Acta, 31,767 (1966). L. I., “Handbook of X-ray Analysis of
11. MIRKIN
2793
NOTES
Poiyc~staiiine Materials”, p. 509, English transiation., Consultants Bureau, New York ( 1964). 12. REID A. F. and RINGWOOD A. E., J. Solid State Chem., Wadsiey Memorial Issue, 1, 557 (1970).
J. Phys. Chem. Soiids
Vol. 3 I, pp. 2793-2795.
Acoustic relaxation in manganese ferrites (Mn,Fq_,O,) at temperatures between 90 and 400 Kforl.O
loss measurements were performed on manganese ferrites (Mn,Fe,_,OJ with a manganese content of 1.0 < x < 1.6 at frequencies between 50 kHz and 50 MHz and temperatures between 80 and 400” K. The activation energies calculated from these measurements are in good agreement with those calculated from magnetic relaxation and electrical conductivity me~urements. The maximum loss was frequency independent and seems to be proportional to (1 -x). Gibbons [ I] reported some results of acoustical loss measurements on manganese ferrite single crystals with a manganese content of x = 1.18 and x = 1.37, where x is given in on the MnZFe3_s04. The measurements sample with x = l-18 were performed’ at a frequency of I50 kHz. The loss peak was found at a temperature of 225 K. The acoustic activation energy calculated from the measurements for x = 1.37 is between 0*270.34 eV. In our laboratory these measurements were extented to other manganese contents and higher frequencies. Both poIycrystalline and single-c~stalline samples were used. The single crystals were prepared by a floating zone technique using an f arc image furnace. By analysis the accuracy of the given manganese content was found to be within 1 mol.
2794
TECHNICAL
per cent and the number of cation vacancies was less than 2 pro mille. The measurements of the mechanical loss at frequencies of 10 to 100 MHz were performed with an apparatus analogous to that of Chick, Anderson and Truell[2]. At frequencies below 500 kHz the electromagnetic drive method described by Fine and Ellis[3] was used. The polycrystalline samples were used at frequencies below 500 kHz. At frequencies above 10MHz the measurements were performed on single crystals in the (100) and ( 111) direction with and without a magnetic field. The field strength was greater than the saturation value. Below 500 kHz or when the measurements were performed with a magnetic field at constant frequency the losses were found to have a maximum at a temperature T,,. The relation between frequency and temperature fits the following formula
NOTES
EleY10.6-
0.2
oMagnetlc l
Actlvatlon
Energy
accordmg
to ref 4
Acoustic
Actlvatlon
Energy
Actlvatton to ref 4
Energy
0.1 t 01 M
*Electric accordmg 1.1
1.2
1.3
l.4
15
16
-x Fig. 1. Acoustic and magnetic activation energy of Mn,F%_,O, plotted against the manganese content x.
wherefis the frequency; r the relaxation time; k the Boltzman constant and E the activationf energy; the results are given in Table 1. In Fig. 1 the acoustic activation energy E is Table 1. composition
E.,,, in eV
- 14,,
x= 1.1
0~31~0~02 o-35 f 0.02 0*39*0.01 0.42 k 0.02 0.48 -c 0.02
12.8 13.1 13.6 13.6 14.2
x= x= x=
1.2 1.3 1.4 _x= 1-6
tan 6,, (max) 7x 14 x 20 x 27 x 42 x
10-S 10-Z 10-J
1O-3 1O-3
plotted as a function of the manganese content in the range between 1.0 and 16 The result of conductivity and magnetic measurements according to Brabers et af.[4] are also shown. The maximum losses in samples with a ( 100) direction and in samples with a ( 111) direction occur at the same temperature. Between 50 and 500 kHz the maximum losses
Fig. 2. The maximum loss in the temperature range 80400” K of Mn,Fe,_,O, plotted against the manganese content X.
were not dependent on the frequency. The results are given in Fig. 2. They were proportional to (x- 1) and suggest that the relaxation process is related to the Mn3+ ions on the octahedral sites [5]. The agreement between the activation energies of the acoustical and magnetic relaxation, and that of the electrical conductivity at manganese content between x = 1*O
TECHNICAL
and x = 1.6 gives rise to the supposition that they are caused by the same process namely “electron transfer’. The proportionality of the maximum losses to the supposed Mn3+ content suggests that the acoustic relaxation process is caused by a Jahn-Teller induced electron transfer. Acknowledgement-The authors wish to thank ir. V.A.M. Brabers for preparing the single crystalline and polycrystalline mangenese ferrites. J. H. HENDRIKS F. J. DlJKSTRA Laboratorium voor Materiaalkunde, T. H. Eindhoven, Netherlands REFERENCES F.,J.appl.Phys.28,810(1957). CHICK D., ANDERSON G. and TRUELL R., J. Acoust Sot. Am. 32, 186(1960). FINE M. E. and ELLIS W. C., Trans. Am. Inst. BRABERS V. A. M. and HENDRIKS J. H. Solid State Commun. 6,795 (1968). BRABERS V. A. M. and DEKKER P., Phys. Status Sofidi 29, K73 ( 1968).
1. GIBBONSD.
2. 3. 4. 5.
J. Phys. Chem. Solids
Cation
Vol. 3 1, pp.
2795-2797.
diffusion and electrical conduction in KC1 crystals
(Received 8 May 1969; in revisedform 27 January 1970)
CATION self diffusion and electrical conduction in alkali halide crystals occur by the migration of cation vacancies from one lattice site to another and are given by NernstEinstein’s relation [ I] D = o-fKT/Ne2 where D is the cation self diffusion coefficient, m is the electrical conductivity, f is the correlation factor, N is the number of anion cation pairs per unit volume and the other symbols have their usual significance. The validity of the Nernst-Einstein’s relation in a
JPCSVd.3l.No.
12-N
NOTES
2795
particular crystal depends upon the concentration of defects other than the free cation vacancies which may contribute to the diffusion coefficient without contributing to the electrical conduction or the other way round. In actual practice, these crystals contain other defects like the impurity ions, anion vacancies, impurity-vacancy pairs and vacancy-vacancy pairs as a result of which, the NemstEinstein’s relation may or may not be valid at different temperatures. Simultaneous measurement of cation self diffusion coefficient and electrical conduction in alkali halide crystals was first made by Mapother, Crooks and Maurer[2]. It was shown that in NaCl and NaBr crystals in the impurity region the diffusion coefficient measured by tracer technique (Dtr) is greater than dithrsion coefficient calculated from the electrical conductivity (D,,J and that in the intrinsic region Dt, = Deal except at temperatures close to the melting point where Dt, is less than D,,,. These results were satisfactorily explained by these authors [2] and the deviation from the NemstEinstein’s relation was attributed to cation diffusion by the migration of pairs in the impurity region and to the electrical conduction by the anion vacancies at high temperatures in the intrinsic region. A check on the validity of the Nemst-Einstein’s relation in KC1 crystal by Witt[3] and Aschner[4] gave different results in the intrinsic region. The author has made simultaneous measurements of potassium ion diffusion and electrical conduction in KC1 crystals in the temperature range 400-700°C and the results are reported here. Potassium chloride crystals were grown in air from Analar grade material obtained from M/s British Drug Houses Private Ltd., Bombay, India, by Kyropoulos technique and cooled to room temperature at a rate of about l”C/min. The electrical conductivity was measured using Veb Funkwerk Erfiut Type 1001 Tera Ohmmeter[5] when the resistance of the crystal was greater than lo6 fi and by the pulse method[6] when the resistance was
of each experimental run the temperature was quickly lowered to below IOO’C, and the pressure then reduced. The samples were removed from the apparatus and examined by X-ray diffraction, using a Debye-Scherrer camera of 114 mm. dia. In each case, complete transformation to a spine1 structure was observed, and even in the case of FeMnGeO, no .additional lines could be observed in the powder patterns after long exposures. The spine1 lattice parameters given in Table 1 were obtained after applying film shrinkage corrections and extrapolating the unit cell values to 28 = 180”, making use of back angle reff ections and appropriate extrapolation functions [ 1 I]. These new compounds provide a substantial extension of known olivine-spine1 transformations, Table 2. The density increases are similar to those previously observed, as expected for the transformation of one structure type to another[l2].
A. E. RINGWOOD A. F. REID
Department of Geophysics and Ge~hemist~, Australian National University and Division of Mineral Chemistry, Commonwealth Scientific Industrial Research Organization, Melbourne, Australia. RRl?ERFBCEs
1. BERNAL J. D., Observatory, 59,268 (1936). 2. JEFFREYS H., Mon. Not. R. astrn. Sot. geophys.
Suppi., 4,50 (1937). 3. GOLDSCHMIDT V., Nachr. Ges. Wiss. Gottingen, Math-Physik. KI ., 184 (193 1). 4. RINGWOOD A. E., In “Advances in Earth Science” (Edited by P. Hurley), p. 357, M.I.T. Press (1966). 5. RINGWOOD A. E. and MAJOR A., Earth &planet. Sci. Lett. 1,241 (1966). 6. RINGWOOD A. E. and MAJOR A., Phys. Earth Planer. Interiors, in press, (I 970). 7. BLASSE G., .J. inora. nuci. Chem.. 25. 230 (19631. 8. RINGWOOD A. E. &d REID A.F: (i~~~p~atiu~j. 9. WADSLEY A. D., REID A. F. and RINGWOOD A. E., Acta crystaiiogr., B&$,740 (1970). 10. GREEN D. H. and RINGWOOD A. E., Geochim.
cosmochim. Acta, 31,767 (1966). L. I., “Handbook of X-ray Analysis of
11. MIRKIN
2793
NOTES
Poiyc~staiiine Materials”, p. 509, English transiation., Consultants Bureau, New York ( 1964). 12. REID A. F. and RINGWOOD A. E., J. Solid State Chem., Wadsiey Memorial Issue, 1, 557 (1970).
J. Phys. Chem. Soiids
Vol. 3 I, pp. 2793-2795.
Acoustic relaxation in manganese ferrites (Mn,Fq_,O,) at temperatures between 90 and 400 Kforl.O
loss measurements were performed on manganese ferrites (Mn,Fe,_,OJ with a manganese content of 1.0 < x < 1.6 at frequencies between 50 kHz and 50 MHz and temperatures between 80 and 400” K. The activation energies calculated from these measurements are in good agreement with those calculated from magnetic relaxation and electrical conductivity me~urements. The maximum loss was frequency independent and seems to be proportional to (1 -x). Gibbons [ I] reported some results of acoustical loss measurements on manganese ferrite single crystals with a manganese content of x = 1.18 and x = 1.37, where x is given in on the MnZFe3_s04. The measurements sample with x = l-18 were performed’ at a frequency of I50 kHz. The loss peak was found at a temperature of 225 K. The acoustic activation energy calculated from the measurements for x = 1.37 is between 0*270.34 eV. In our laboratory these measurements were extented to other manganese contents and higher frequencies. Both poIycrystalline and single-c~stalline samples were used. The single crystals were prepared by a floating zone technique using an f arc image furnace. By analysis the accuracy of the given manganese content was found to be within 1 mol.
2794
TECHNICAL
per cent and the number of cation vacancies was less than 2 pro mille. The measurements of the mechanical loss at frequencies of 10 to 100 MHz were performed with an apparatus analogous to that of Chick, Anderson and Truell[2]. At frequencies below 500 kHz the electromagnetic drive method described by Fine and Ellis[3] was used. The polycrystalline samples were used at frequencies below 500 kHz. At frequencies above 10MHz the measurements were performed on single crystals in the (100) and ( 111) direction with and without a magnetic field. The field strength was greater than the saturation value. Below 500 kHz or when the measurements were performed with a magnetic field at constant frequency the losses were found to have a maximum at a temperature T,,. The relation between frequency and temperature fits the following formula
NOTES
EleY10.6-
0.2
oMagnetlc l
Actlvatlon
Energy
accordmg
to ref 4
Acoustic
Actlvatlon
Energy
Actlvatton to ref 4
Energy
0.1 t 01 M
*Electric accordmg 1.1
1.2
1.3
l.4
15
16
-x Fig. 1. Acoustic and magnetic activation energy of Mn,F%_,O, plotted against the manganese content x.
wherefis the frequency; r the relaxation time; k the Boltzman constant and E the activationf energy; the results are given in Table 1. In Fig. 1 the acoustic activation energy E is Table 1. composition
E.,,, in eV
- 14,,
x= 1.1
0~31~0~02 o-35 f 0.02 0*39*0.01 0.42 k 0.02 0.48 -c 0.02
12.8 13.1 13.6 13.6 14.2
x= x= x=
1.2 1.3 1.4 _x= 1-6
tan 6,, (max) 7x 14 x 20 x 27 x 42 x
10-S 10-Z 10-J
1O-3 1O-3
plotted as a function of the manganese content in the range between 1.0 and 16 The result of conductivity and magnetic measurements according to Brabers et af.[4] are also shown. The maximum losses in samples with a ( 100) direction and in samples with a ( 111) direction occur at the same temperature. Between 50 and 500 kHz the maximum losses
Fig. 2. The maximum loss in the temperature range 80400” K of Mn,Fe,_,O, plotted against the manganese content X.
were not dependent on the frequency. The results are given in Fig. 2. They were proportional to (x- 1) and suggest that the relaxation process is related to the Mn3+ ions on the octahedral sites [5]. The agreement between the activation energies of the acoustical and magnetic relaxation, and that of the electrical conductivity at manganese content between x = 1*O
TECHNICAL
and x = 1.6 gives rise to the supposition that they are caused by the same process namely “electron transfer’. The proportionality of the maximum losses to the supposed Mn3+ content suggests that the acoustic relaxation process is caused by a Jahn-Teller induced electron transfer. Acknowledgement-The authors wish to thank ir. V.A.M. Brabers for preparing the single crystalline and polycrystalline mangenese ferrites. J. H. HENDRIKS F. J. DlJKSTRA Laboratorium voor Materiaalkunde, T. H. Eindhoven, Netherlands REFERENCES F.,J.appl.Phys.28,810(1957). CHICK D., ANDERSON G. and TRUELL R., J. Acoust Sot. Am. 32, 186(1960). FINE M. E. and ELLIS W. C., Trans. Am. Inst. BRABERS V. A. M. and HENDRIKS J. H. Solid State Commun. 6,795 (1968). BRABERS V. A. M. and DEKKER P., Phys. Status Sofidi 29, K73 ( 1968).
1. GIBBONSD.
2. 3. 4. 5.
J. Phys. Chem. Solids
Cation
Vol. 3 1, pp.
2795-2797.
diffusion and electrical conduction in KC1 crystals
(Received 8 May 1969; in revisedform 27 January 1970)
CATION self diffusion and electrical conduction in alkali halide crystals occur by the migration of cation vacancies from one lattice site to another and are given by NernstEinstein’s relation [ I] D = o-fKT/Ne2 where D is the cation self diffusion coefficient, m is the electrical conductivity, f is the correlation factor, N is the number of anion cation pairs per unit volume and the other symbols have their usual significance. The validity of the Nernst-Einstein’s relation in a
JPCSVd.3l.No.
12-N
NOTES
2795
particular crystal depends upon the concentration of defects other than the free cation vacancies which may contribute to the diffusion coefficient without contributing to the electrical conduction or the other way round. In actual practice, these crystals contain other defects like the impurity ions, anion vacancies, impurity-vacancy pairs and vacancy-vacancy pairs as a result of which, the NemstEinstein’s relation may or may not be valid at different temperatures. Simultaneous measurement of cation self diffusion coefficient and electrical conduction in alkali halide crystals was first made by Mapother, Crooks and Maurer[2]. It was shown that in NaCl and NaBr crystals in the impurity region the diffusion coefficient measured by tracer technique (Dtr) is greater than dithrsion coefficient calculated from the electrical conductivity (D,,J and that in the intrinsic region Dt, = Deal except at temperatures close to the melting point where Dt, is less than D,,,. These results were satisfactorily explained by these authors [2] and the deviation from the NemstEinstein’s relation was attributed to cation diffusion by the migration of pairs in the impurity region and to the electrical conduction by the anion vacancies at high temperatures in the intrinsic region. A check on the validity of the Nemst-Einstein’s relation in KC1 crystal by Witt[3] and Aschner[4] gave different results in the intrinsic region. The author has made simultaneous measurements of potassium ion diffusion and electrical conduction in KC1 crystals in the temperature range 400-700°C and the results are reported here. Potassium chloride crystals were grown in air from Analar grade material obtained from M/s British Drug Houses Private Ltd., Bombay, India, by Kyropoulos technique and cooled to room temperature at a rate of about l”C/min. The electrical conductivity was measured using Veb Funkwerk Erfiut Type 1001 Tera Ohmmeter[5] when the resistance of the crystal was greater than lo6 fi and by the pulse method[6] when the resistance was