Magnetic materials are composed of atom-scale polarisations known as local moments, which act as little magnets with orientations that can form ordered patterns called magnetic phases/states. Their dynamics and fluctuations, as well as phase transitions between them, underline their functionality for caloric refrigeration. First-order phase transitions, which are discontinuous transformations between different magnetic states, present the largest cooling responses thanks to the fact that small external stimuli suffice to drive an abrupt magnetic change and consequent giant caloric effect. One of the most successful computational approach and workhorse of materials science modelling from first principles is the Density Functional Theory (DFT). In this project we combine DFT with statistical mechanics methods for the local moments to describe magnetic phase transitions and caloric cooling by means of the so-called Disordered Local Moment (DLM) picture, one of the few approaches for finite temperature magnetism from first principles that exists.
The project strongly focused on Mn-based antiperovskite systems, which are magnetic materials with a prominent first-order transition to a triangular magnetic state given rise by multisite interactions. Most importantly, the application of mechanical stress to antiperovskites is not only accompanied by giant caloric cooling, but also by the stabilization of another collinear magnetic phase that should be responsive to a magnetic field. Hence, antiperovskite materials are good multicaloric candidates. We have carefully analysed the electronic properties of antiperovskite materials as well as performed multicaloric experiments on an optimal antiperovskite sample created and characterised by collaborators in the United Kingdom. To this end, we have designed and constructed a novel experimental device that can measure multicaloric effects driven by the simultaneous application of a magnetic field and uniaxial stress. However, an enhanced cooling response driven by the multicaloric effect in this sample has not been observed yet.
Central theoretical work has been centred on the extension of DLM-DFT theory to include the coupling of magnetism with the atom vibrations, a challenging task that posed important fundamental and computational questions. These developments have shown that the size and consequent caloric boost of multisite interactions is highly dependent on the atom vibrations in antiperovskite materials, which has explained previous contradictions between experiment and theory. We plan to apply our new computational tool to predict new antiperovskite materials with optimal properties, as well as other multicaloric materials.
The project also focused on multicaloric effects in La(FexSi1-x)13, one of the most famous magnetic refrigerants, and CrGeTe3, an important material in spintronics. DLM-DFT calculations have successfully unrevealed the origin of cooling responses in La(FexSi1-x)13 in excellent agreement with experiments, explaining the small role of multisite interactions somewhat challenging the current understanding of this material. Moreover, experimental work on CrGeTe3 has shown that its cooling performance can be indeed enhanced via the multicaloric effect. DLM-DFT calculations have demonstrated that multisite interactions are substantial in CrGeTe3 and that a first-order phase transition can be induced by uniaxial stress, promoting its extraordinary multicaloric potential. Our work along these lines has been presented in several scientific events and has produced two publications, another two being in preparation.
