Magnetic cooling is a highly efficient refrigeration technique with the potential to replace the traditional vapor compression cycle. It is based on the magnetocaloric effect, which is associated with the temperature change of a material when placed in a magnetic field. We present experimental evidence for the origin of the giant entropy change found in the most promising materials, in the form of an electronic reconstruction caused by the competition between magnetism and bonding. The effect manifests itself as a redistribution of the electron density, which was measured by X-ray absorption and diffraction on MnFe(P,Si,B). The electronic redistribution is consistent with the formation of a covalent bond, resulting in a large drop in the Fe magnetic moments. The simultaneous change in bond length and strength, magnetism, and electron density provides the basis of the giant magnetocaloric effect. This new understanding of the mechanism of first order magneto-elastic phase transitions provides an essential step for new and improved magnetic refrigerants.

Magnetic refrigeration became one of the most popular fields of research among the magnetic materials. After the discovery of the Giant magnetocaloric effect (GMCE) in Gd₅Si₂Ge₂, many new materials (viz. LaFeSi₁₃, Heusler alloys, FeRh etc.) have been proposed in the past decade. Mn_x-Fe_1-xSi_yP_1-y series of materials are among the best performing materials, in terms of efficiency, availability, cost, environmental friendliness etc [1]. MnFeSiP materials show a simultaneous transiton of the magnetic state accompanied by an elastic transition, which is considered to be the reason behind its high magnetocaloric efficiency. Recently the underlying mechanism of this 'magneto-elastic' transition was explained by employing density functional theory. The coexistance of strongly and weakly magnetic atoms give rise to the effect called 'mixed magnetism'. [2] The efficiency of these materials are determined by the adiabatic temperature change and magnetic entropy change at the Curie temperature. These two quantities are directly proportional to the latent heat of magnetization (L). So the determination of L is essential for determining the usefulness of such system. Experimentally L can be determined using differential scanning calorimetry. Determination of the latent heat can also be done using ab-initio calculation, which is useful in characterizing such systems based on their efficiency. The accurate values of the free energy above and below the transition temperature are needed to obtain the magnetic latent heat and the transition temperature.