No abstract

No abstract

The electrical conductivity and the Seebeck coefficient of single-crystal and polycrystalline NiO were measured as a function of temperature and of oxygen partial pressure over a large part of the phase field of stable NiO. In the temperature range 1000°–1600°C the electrical conductivity was found to be proportional to the fourth root of the oxygen partial pressure. This pressure dependence is that which one would expect for pure, nonstoichiometric NiO containing singly ionized metal vacancies as the predominant point defects. The activation energy for conduction, ΔHc, computed from σ = σ0exp (−ΔHc / kT), was found to be 0.92 ± 0.02 eV. The Seebeck coefficient of NiO was found to be p type over the entire range of temperatures and oxygen partial pressures measured. From the temperature dependence of the Seebeck coefficient, the enthalpy of formation of defects in NiO was calculated to be 0.66 ± 0.03 eV. From the measured temperature dependence of the electrical conductivity and the Seebeck coefficient it was found that the charge-carrier mobility increases with the temperature. Considering the drift mobility to be of the form μp∝T−1exp(−ΔHm / kT) a value of 0.37 eV could be attributed to the activation energy for motion of the charge carrier, ΔHm. The existence of such an activation energy supports the suggestion of a hopping process for the charge-carrier transport mechanism in NiO above about 1000°C.

No abstract

IiIi,rt itivcstigatiti~ the high-temperature defcct structure' anti electrical properties of wustite, an accurate kuowledgc o f tlic deviatioii from stoichiometry as a function of oxygen partial pressurr ant1 teniperature was desirable. This note presents tlic rrwilts of it tlieriiiogravirnetric study on wustite. Varying oxygen partial pressures were obtained by mixing COI and CO in tlie ran.qe (),% < /1(C02)/P(CO) < 7.9 (10-'5 attn < P ( 0 z ) < 1 0 -s :itin). Tlie wustite phase was studied in tlie range 1000" < 1' < 1300°C. Tlie iron-rich part of the wustite field, 1.05 < O/Fc < 1.10, was exitmitied in great detail. J(iliiisoii aiicl Mattlicy spectroscopically pure iron sheet (0.1 umtn tliick) was used as startirig material. A strip about 15 by 90 inin (totxl weiglit --.I g ) wa? rolled to a spiral, contact between the rolls being carefully avoided. To obtain well-grained samples, :iii :tniiealiiig, oxidation, aiid aging procedure described by Himmel ct The typical grain size ~v a s :i 3 i n i n widtli aiid half tlie specimen thickness. Xceiir:ttely cotitrolled inixturcs of CO1 and CO gases werc 1):isseti tlirougI1 the furnace from'below with a linear velocity of 0.8 ciii//sec, wliieli was ciiougli to prcvent thernmal segregation of '1'11~: oxygrii ptrtial pressure resulting froin the dis-,' 2 co:> Ft. 2 co + 0 2 is givcii hy 11ie rclatioii P (0,) -= f< ( 7') [ P ( C 0 2 ) / P (CO) J 2 (1)wlierc Ii(7') is tIir equilibrium constant of the reaction given by C~ugliliii.~ ,411 electrobalnrice was mounted on a vertical furnace in wliicli tlic sailiplcs were suspended by a platinum wire int.0 the Iiot zone of the furnace. 'I'he balance was used as a differential u~i g l~i i i g dcviw, the zero point being the apparent weight of tiia.qiic(ite. ('Hie a p p~e i i t weight includes hanging wire and lictitious weight due to the flowing gas; see Ref. 6.) 'I'lic deviatioii froiii stoicliiometry x in FeOL+, was calculated Ironi (lie rvlatioii:where M(FeaOa) and M(FeOl+,) are the apparent weights of the respective oxides, and M(Fe) is the absolute weight of the iron content. Because the weight of the iron is constant during oxidation and reduction, it is convenient to use in Eq.(2) the oxygenexcess notation of wustite, FeOl+,. The cation-deficiency notation Fel-,O is obtained from the relation y = x/l + x.I t is assumed in Eq. (2) that the magnetite phase is essentially stoichiometric, as Darken and Gurry had found.4 The apparent weight of the magnetite, M(FesOn), was determined separately a t each isotherm to cancel errors due t o differences in lifting forces. The errors introduced by using mixtures with different molecular weights are greater than the balance sensitivity, and, in terms ofx, amount to Ax = f0.0002. Errors in P(02) due to misreading the mixtures ratio correspond in x to twice this amount, the total experimental inaccuracy being Ax = 0.0006. The results were not affected by oxidation to magnetite and further reduction to wustite. Results obtaiued on different saniples were reproducible within the experiment...