Magnetocaloric refrigeration is one of the most promising next-generation solid-state caloric techniques to revolutionize the traditional air-compression technique. The La(Fe,Si)13-based materials are recognized as candidates with potential for practical applications. However, flexible strategies to improve the Curie temperature (TC) and further achieve the tunable giant magnetocaloric effect (GMCE) still need to be developed. Here, the systematic experimental investigation on a series of light elements (C, F, S) modified LaFe11.6Si1.4 compounds are presented. It is found that all modified samples exhibit a higher TC, with a negligible impact on the thermal hysteresis. The GMCE performance in C- and S-modified samples is significantly degraded, but the maximum magnetic entropy change |Δ sm| for the optimally doped F sample can be well maintained at 19.2 J kg−1 K−1 for a field change of 2 T. The preferential site occupancy of dopants is determined, and the microstructural observation and metastable atomic changes have also been analyzed. It is concluded that interstitial doping is more efficient to shift TC. The first-order transition can however not be maintained upon doping due to changes in the hybridization. These findings highlight the importance of the interplay between the lattice pressure effect and the covalent hybridization for this material family.

Compared with traditional techniques, solid-state magnetocaloric phase transition materials (MPTMs), based on the giant magnetocaloric effect (GMCE), can achieve a higher energy conversion efficiency for caloric applications. As one of the most promising MPTMs, the hexagonal (Mn,Fe)₂(P,Si)-based compounds host some advantages, but the existing hysteresis and relatively unstable GMCE properties need to be properly tackled. In this study, it is found that substitutions with Ni, Pd, and Pt can maintain and even enhance the GMCE (≈8.7% maximum improvement of |Δs_m|). For a magnetic field change of Δμ₀H = 2 T, all samples obtain a |Δs_m| in the range of 20–25 J kg⁻¹ K⁻¹ with a low thermal hysteresis ΔT_hys (≤5.6 K). The performance surpasses almost all other (Mn,Fe)₂(P,Si)-based materials with ΔT_hys (<10 K) reported until now. The occupancy of substitutional Ni/Pd/Pt atoms is determined by X-ray diffraction, neutron diffraction, and density functional theory calculations. The difference in GMCE properties upon doping is understood from the competition between a weakening of the magnetic exchange interactions and the different degrees of orbital hybridization among 3d-4d-5d elements. The studies elaborate on the responsible mechanism and provide a general strategy through d-block doping to further optimize the GMCE of this materials family.