High thermal conductivity electronic materials are critical components for high-performance electronic and photonic devices as both active functional materials and thermal management materials. We report an isotropic high thermal conductivity exceeding 500 W m−1K−1 at room temperature in high-quality wafer-scale cubic silicon carbide (3C-SiC) crystals, which is the second highest among large crystals (only surpassed by diamond). Furthermore, the corresponding 3C-SiC thin films are found to have record-high in-plane and cross-plane thermal conductivity, even higher than diamond thin films with equivalent thicknesses. Our results resolve a long-standing puzzle that the literature values of thermal conductivity for 3C-SiC are lower than the structurally more complex 6H-SiC. We show that the observed high thermal conductivity in this work arises from the high purity and high crystal quality of 3C-SiC crystals which avoids the exceptionally strong defect-phonon scatterings. Moreover, 3C-SiC is a SiC polytype which can be epitaxially grown on Si. We show that the measured 3C-SiC-Si thermal boundary conductance is among the highest for semiconductor interfaces. These findings provide insights for fundamental phonon transport mechanisms, and suggest that 3C-SiC is an excellent wide-bandgap semiconductor for applications of next-generation power electronics as both active components and substrates.

The emergence of wide and ultrawide bandgap semiconductors has revolutionized the advancement of next-generation power, radio frequency, and opto- electronics, paving the way for chargers, renewable energy inverters, 5G base stations, satellite communications, radars, and light-emitting diodes. However, the thermal boundary resistance at semiconductor interfaces accounts for a large portion of the near-junction thermal resistance, impeding heat dissipation and becoming a bottleneck in the devices’ development. Over the past two decades, many new ultrahigh thermal conductivity materials have emerged as potential substrates, and numerous novel growth, integration, and characterization techniques have emerged to improve the TBC, holding great promise for efficient cooling. At the same time, numerous simulation methods have been developed to advance the understanding and prediction of TBC. Despite these advancements, the existing literature reports are widely dispersed, presenting varying TBC results even on the same heterostructure, and there is a large gap between experiments and simulations. Herein, we comprehensively review the various experimental and simulation works that reported TBCs of wide and ultrawide bandgap semiconductor heterostructures, aiming to build a structure–property relationship between TBCs and interfacial nanostructures and to further boost the TBCs. The advantages and disadvantages of various experimental and theoretical methods are summarized. Future directions for experimental and theoretical research are proposed.

Silicon nitride (Si3N4) is a promising substrate for high-power electronics due to its superior mechanical properties and potential outstanding thermal conductivity (κ). As experiments keep pushing the upper limit of κ of Si3N4, it is believed that it can reach 450 W/mK, similar to SiC, based on classical models and molecular dynamics simulations. In this work, we reveal from first principles that the theoretical κ upper limits of β-Si3N4 are only 169 and 57 W/mK along the c and a axes at room temperature, respectively. Those of α-Si3N4 are about 116 and 87 W/mK, respectively. The predicted temperature-dependent κ matches well with the highest available experimental data, which supports the accuracy of our calculations, and suggests that the κ upper limit of Si3N4 has already been reached in the experiment. Compared to other promising semiconductors (e.g., SiC, AlN, and GaN), Si3N4 has a much lower κ than expected even though the chemical bonding and mechanical strengths are close or even stronger. We find the underlying reason is that Si3N4 has much lower phonon lifetimes and mean free paths (<0.5 μm) due to the larger three-phonon scattering phase space and stronger anharmonicity. Interestingly, we find that the larger unit cell (with more basis atoms) that leads to a smaller fraction of acoustic phonons is not the reason for lower κ. Grain size-dependent κ indicates that the grain boundary scattering plays a negligible role in most experimental samples. This work clarifies the theoretical κ upper limits of Si3N4 and can guide experimental research.

Ligand-assisted wet chemical synthesis is a versatile methodology to produce controllable nanocrystals (NCs). The post-treatment of ligands is significant for the performance of functional devices. Herein, a method that retains ligands of colloidal-synthesized nanomaterials to produce thermoelectric nanomaterials is proposed, compared to the conventional methods that strip ligands using multi-step cumbersome processes. The retained ligands method can control over the size and dispersity of nanocrystals during the consolidation of the NCs into dense pellets, in which retained ligands are transformed into organic carbon within the matrices, and intimate contacts between carbon and nanoparticles establish clear organic-inorganic interfaces. Characterizations of non-stripped and stripped samples confirm that this strategy can affect electric transport slightly but reduce the thermal conductivity largely. As a result, the materials (e.g., SnSe, Cu2-xS, AgBiSe2, and Cu2ZnSnSe4) with ligands retained achieve higher peak zT and better mechanical properties. This method not only can be applied to other colloidal thermoelectric NCs but also other functional materials.