ALBERT

Notes: Abstract To investigate the point defect chemistry and the kinetic properties of manganese olivine Mn2SiO4, electrical conductivity (’) of single crystals was measured along either the [100] or the [010] direction. The experiments were carried out at temperatures T=850–1200 °C and oxygen fugacities $$f_{{\text{O}}_{\text{2}} } = 10^{ - 11} - 10^2 $$ atm under both Mn oxide (MO) buffered and MnSiO3 (MS) buffered conditions. Under the same thermodynamic conditions, charge transport along [100] is 2.5–3.0 times faster than along [010]. At high oxygen fugacities, the electrical conductivity of samples buffered against MS is ∼1.6 times larger than that of samples buffered against MO; while at low oxygen fugacities, the electrical conductivity is nearly identical for the two buffer cases. The dependencies of electrical conductivity on oxygen fugacity and temperature are essentially the same for conduction along the [100] and [010] directions, as well as for samples coexisting with a solid-state buffer of either MO or MS. Hence, it is proposed that the same conduction mechanisms operate for samples of either orientation in contact with either solid-state buffer. The electrical conductivity data lie on concave upward curves on a log-log plot of σ vs $$f_{{\text{O}}_{\text{2}} } $$ , giving rise to two $$f_{{\text{O}}_{\text{2}} } = 10^{ - 11} - 10^2 $$ regimes with different oxygen fugacity exponents. In the low- $$f_{{\text{O}}_{\text{2}} } = 10^{ - 11} - 10^2 $$ regime $$\left( {f_{{\text{O}}_{\text{2}} } 〈 10^{ - 7} {\text{atm}}} \right)$$ , the $$f_{{\text{O}}_{\text{2}} } = 10^{ - 11} - 10^2 $$ exponent, m, is 0, the MnSiO3-activity exponent, q, is ∼0, and the activation energy, Q, is 45 kJ/mol. In the high $$f_{{\text{O}}_{\text{2}} } = 10^{ - 11} - 10^2 $$ regime $$\left( {f_{{\text{O}}_{\text{2}} } 〉 10^{ - 7} {\text{atm}}} \right)$$ , m=1/6, q=1/4–1/3, and Q=45 and 200 kJ/mol for T〈1100 °C and T〉1100 °C, respectively. Based on a comparison of experimental data with results from point defect chemistry calculations, it is proposed that the change in m with $$f_{{\text{O}}_{\text{2}} } = 10^{ - 11} - 10^2 $$ is induced by a switch in charge neutrality condition. At low $$f_{{\text{O}}_{\text{2}} } = 10^{ - 11} - 10^2 $$ s, the charge neutrality condition is [e′]=[Mn Mn { ]; the hopping motion of electron holes h . is the dominant conduction mechanism. At high $$f_{{\text{O}}_{\text{2}} } = 10^{ - 11} - 10^2 $$ s, the charge neutrality condition is 2[V Mn · ]=[Mn Mn · ]; the hopping motion of electron holes h . and the migration of Mn ions associated with a counter flow of divalent Mn vacancies V Mn · control electrical conduction at T〈1100 °C and T〉1100 °C, respectively.

Notes: Abstract Grain growth kinetics in CaTiO3-perovskite + FeO-wüstite aggregates were studied at the conditions of T = 1223–1623 K, P = 0.1 MPa and P = 200 MPa. Starting samples were fabricated by hot-pressing mechanically mixed powders of CaTiO3 + FeO with FeO = 0%, 1%, 3%, 6%, 10%, 20% and 100% by weight in a gas-medium apparatus at 1323 K and 300 MPa for 5 h. The increase of grain size (G) of CaTiO3 with time (t) follows a growth law: G n −G n 0 = κ·t(κ=κ0exp(−(Q/RT)). Two grain growth regimes are observed at T 〈 1523 K and T ≥ 1523 K. For T 〈 1523 K, the best fits of the data to the growth law yield growth exponents of n = 2.2 ± 0.2, 3.0 ± 0.3 and 3.5 ± 0.3 for samples with FeO = 0%, 3% and 10% respectively. Under these conditions the rate constants, κ, obey an Arrhenius relation with Q = 206 ± 35 kJ/mol and 385 ± 65 kJ/mol for samples with FeO = 3% and 10%. Grain growth of CaTiO3 becomes sluggish when FeO content exceeds 6%. For T ≥ 1523 K, the best fits of the data to the growth law yield n = 2.5 ± 0.2 for both samples with FeO = 3% and 10%. The activation energies (Q ) were determined as 71 ± 30 kJ/mol and 229 ± 45 kJ/mol for samples with FeO = 3% and 10%, respectively. The TEM observations show a remarkable difference in the distribution and geometry of FeO below and above 1523 K: nanometer-sized particles of FeO were observed along CaTiO3 grain boundaries in samples annealed at T 〈 1523 K. No FeO particles were detected along CaTiO3 grain boundaries in samples annealed at T ≥ 1523 K, but large clusters of FeO particles are observed locally indicating a fast separation of FeO from CaTiO3. Thus we conclude that the slow growth rate of CaTiO3 at T 〈 1523 K is due to the pinning by FeO particles at grain boundary, and that the change of grain growth kinetics in CaTiO3 at T ≥ 1523 K may relate to the separation of FeO from CaTiO3, which we interpret as due to the phase transformation of CaTiO3 at around 1523 K.