Calorimetric study of the giant magnetocaloric effect in (MnNiSi)0.56(FeNiGe)0.44

(MnNiSi)0.56(FeNiGe)0.44 belongs to a new family of alloys, similar to MnAs, showing a magnetostructural first-order transition near room temperature with large latent heat and magnetocaloric effect (MCE). From isothermal magnetization, remarkable values of the entropy change have been reported, such as |ΔST| = 11.5 J/kg K at 290 K for the small field change of 1 T, or |ΔST| = 70.1 J/kg K for 5 T on a slightly different composition. Strangely, this last value almost doubles that obtained from the Clausius-Clapeyron equation. Therefore, a very detailed set of calorimetric determinations have been made including heat capacity by adiabatic and differential scanning calorimetry (DSC), and direct measurement of the heat absorbed on isothermal demagnetization for several magnetothermal histories of the sample. We found high values of |ΔST| = 45.7 J/kg K at 289.4 K for a field change of 7 T, for a sample previously cooled to a low temperature and then heated under magnetic field to the target temperature. This value is high but very far from previously reported data, which happened to be nonphysical but obtained from a very frequently used, but incorrect, experimental protocol to determine ΔST from magnetization, via the Maxwell relation. The spurious contribution has been analyzed and computed, explaining the nonphysical reported values. The strong thermal and magnetic hysteresis of 8.4 K and 8.3 T, respectively, make this alloy useless for magnetic refrigeration, but the results encourage searching for other derivatives with lower hysteresis which could have even higher |ΔST |.