Specific heat and electrical resistivity of FeAl alloys

The specific heat Cp and electrical resistivity p of four iron-aluminum alloys (30, 38, 43, and 48 at.% Al) were measured as a function of temperature and composition. The alloys were step-cooled from 1000°C to minimize vacancy concentration and induce maximum order. At room temperature the 30 at.% Al alloy was in the FesAl phase region (D0&sub3; structure) while the other three were in the FeAl phase region (B2 structure). The resistivity at 25°C increased with increasing disorder (increasing Al content for Fe&sub3;Al and increasing Fe content for FeAl), reaching a maximum value of ∼165 μΩ-cm at the Fe&sub3;Al-FeAl phase boundary.

A pulse-heating calorimeter was used to measure simultaneously C&subp; and ρ from 25 to approximately 1000°C. The samples were initially heated at 73°C/s (except for the 43 at.% Al which was heated at 90°C/s), then cooled back to 25°C at 1.5°C/s. They were then repulsed to examine the effect of the pulse treatment and of heating rate. For the 30 at.% Al alloy, ρ increased to a maximum at 485°C, at which a slight decrease occurred. This was taken as due to the D0&sub3 to B2 phase transformation. Then in the FeAl region, ρ decreased slightly to the upper temperature of measurement (∼1200°C). The slope of the ρ-T curves decreased with decreasing Al content, becoming negative for the 30 at.% Al alloy in the FeAl region. The relatively high resistivities and the tendency for the ρ-T curves to approximate a common value at high temperatures is »milar to the saturation behavior of some amorphous and disorder transition alloys discovered by Mooij.

For the 30, 38 and 43 at.% A1 alloys, the ρ-T curves were identical for the initial slowly cooled condition and after cooling from high temperature at 1.5°C/s. The curve for the 48 at.% A1 alloy was about 30% higher after cooling at 1.5°C/s from high temperature. The ρ-T curves were the same for the heating rates that were used in this study.

The specific heat was determined for the alloys, and it showed a dramatic increase at high temperatures. The C&subp;-T curves were the same for the initial slowly cooled condition and for pulse treatment. For the 43 and 48 at.% A1 alloys the C&subp;-T curves wCTe independent of heating rate for the rates tested, but for the 38 at.% A1 alloy, the temperature corresponding to the dramatic rise in C&subp; shifted to higher temperature with higher heating rate.

The C&subp;-T curves in the lower temperature linear region were extrapolated to higher temperature, and subtracted from the experimental C&subp; curve to determine the enthalpy of formation of defects responsible for the dramatic increase in C&subp;. The enthalpy of formation decreased from 135 kJ/mol at 30 at.% A1 to 95 kJ/mol at 48 at.% Al. The vacancy concentrations nv were calculated from these values by assuming triple defect formation. At 900°C n&subv; was constant (<0.2 at.%) from 30 to 38 at.% Al but increased to ∼1.2 at.% for 48 at.% Al. This is in agreement with the results from other experimental studies and the Chang-Neumann model.

For the 38 at.% Al alloy, C⊂p; was affected by the heating rate but ρ was not. Therefore, it was determined that the β sublattice dominates the resistivity since vacancies form primarily on the α sublattice through the formation of triple defects.