Huasong Qin

As thermal management approaches the low-dimensional limit in post-Moore electronics, two-dimensional (2D) materials offer a versatile platform for engineering lattice thermal conductivity (κL). However, a unified framework capable of discovering both ultrahigh- and ultralow-κL materials remains absent because these extremes originate from fundamentally different phonon-transport physics. Here, we develop a unified physics-guided framework integrating a modified 2D Slack-type model with hierarchical random forest learning for accurate full-range κL prediction. The framework establishes a physically interpretable descriptor space, achieves high predictive accuracy (R2 = 0.962), and reveals distinct descriptor spaces governing ultrahigh- and ultralow-κL transport. Screening 16,488 2D materials identifies 48 ultrahigh-κL candidates exceeding most conventional metals and 12 ultralow-κL candidates below liquid-like systems, establishing a unified paradigm for discovering and engineering thermal conductivity extremes in 2D materials.

As chip technologies move towards higher power density, greater integration and more extreme operating environments, thermal management has become inseparable from mechanical reliability. In such systems, heat transport, structural evolution and damage accumulation are strongly coupled from atomistic lattice dynamics to system-level heat-flow interaction. Here we review recent progress in understanding multiscale thermo-mechanical coupling in chip systems. We outline the roles of atom-level defects, bonding and crystal symmetry in intrinsic thermal transport, the effects of microstructural heterogeneity and interfaces on heat-flow continuity and stress localization, and the impact of device- and system-level architectural complexity on hot-spot formation, thermal mismatch, and reliability degradation. We then examine the coupled degradation induced by high-temperature operation, including creep, delamination, fatigue and warpage, and summarize regulation strategies based on defect engineering, interface optimization, package design, advanced cooling and adaptive control. Finally, we highlight emerging AI-assisted approaches for multiscale correlated materials screening, thermal prediction and inverse design, and identify the key challenges that must be addressed to enable predictive thermal management in next-generation chip systems.

Precise control of interfacial thermal resistance (ITR) is essential for thermal management in van der Waals (vdW) devices. Here, we report a counterintuitive temperature-dependent reversal of ITR in twisted PtX2 (X = S, Se, Te) interfaces. At near-commensurate small twist angles, elevated temperatures enhance anharmonic scattering, suppressing direct phonon transmission and dramatically increasing ITR. In contrast, at incommensurate large twist angles, heat transport paradoxically improves because direct transmission is inherently limited by structural mismatch, allowing thermally activated inelastic vibrational pathways to dominate and reduce ITR. Spectral energy density and spectral heat current analyses confirm that this reversal arises from a dynamic competition between the degradation of direct channels and the activation of inelastic channels. The phenomenon extends universally to other vdW architectures, including graphene and MoS2, establishing twist-temperature co-regulation as an effective route for tailoring interfacial thermal transport in vdW materials.

Programmable surface morphing is reshaping geometrical matter from static, prespecified forms into systems capable of adaptive, on-demand control, yet precise and extensive reprogrammable morphing of complex surfaces remains a central challenge through external stimulated material response. Here, we present a reprogrammable digital-intelligent (digintel) metasurface (DMS) built from digitally addressable bistable unit cells whose stable configuration switching generates local eigenstrain, as well as the digitally programmable surface morphing. By digitally encoding the stable state of unit cells, DMS exhibits extensive reprogrammable morphing with the complexity and type of surface morphology exponentially increasing with the number of unit cells. To precisely design the target surface, we further develop a machine learning-assisted inverse-design framework that maps the target surface to digital state codes, enabling deterministic reconstruction of complex surfaces. Experiments and finite element analyses validate the mechanical response of the unit cell and DMS, manifested as state-dependent stiffness, stability margins, load-bearing capacity, and morphing-enabled flow-field modulation across reconfigured geometries. This work establishes a generalizable platform for digintel engineering of reconfigurable functional surfaces, with potential impact on robotics, medical assistance, and aerospace morphing structures.

In this work, we systematically investigate the cross-plane thermal conductivity in stacked diamane via molecular dynamics simulations, focusing on the effects of stacking orientation, external pressure, interlayer coupling strength, and twist angle. By combining spectral heat current decomposition and spectral energy density analysis, we identify high-frequency optical phonons within 10–15 THz as the dominant heat carriers across interfaces—a selectivity arising from interfacial hydrogen termination that strengthens the coupling of these modes. Their lifetimes and coupling strengths are highly sensitive to structural configuration and mechanical modulation. Our results reveal that parallel stacking, increased pressure, and stronger interlayer coupling significantly enhance thermal conductivity by stabilizing optical modes and suppressing phonon scattering. In contrast, increasing the twist angle introduces structural incommensurability, which reduces phonon lifetimes and leads to substantial suppression of thermal transport. These findings underscore the pivotal role of optical phonons in governing interfacial thermal transport as well as establish an optimal strategy for engineering thermal conductivity in 2D vdW materials.

Aerogels are known for their high porosity and very low density and can be made from a range of materials, but are limited by structural instability under extreme thermomechanical conditions. We report on 194 types of dome-celled ultralight aerogels that maintain superior elasticity spanning from 4.2 kelvin (K) to 2273 K, realized by a two-dimensional channel–confined chemistry method. Such aerogels exhibit superelasticity under 99% strain for 20,000 cycles and thermal shock resistance at 2273 K over 100 cycles. The high-entropy carbide aerogel achieves a thermal conductivity of 53.4 mW·m ⁻¹ ·K ⁻¹ at 1273 K and 171.1 mW·m ⁻¹ ·K ⁻¹ at 2273 K. The combination of temperature-invariant elasticity and chemical diversity makes such aerogels highly promising for extreme thermomechanics, from heat-insulated industries to deep space exploration.

Janus transition metal dichalcogenides (TMDs) exhibit exceptional electronic, optical, and catalytic properties due to their unique asymmetric structures. The article systematically investigates the stability of Janus MoSSe with typical vacancy defects using first‐principles calculations. The results reveal that the Gibbs free energy for the hydrogen evolution reaction (HER) is significantly reduced to ≈0.5 eV, lower than that of pristine MoSSe and conventional MoS ₂ monolayers. Notably, the application of external strain further enhances the HER performance of defect‐engineered Janus MoSSe. This improvement is attributed to the adaptive release of concentrated strain by dangling bonds at the defect region, resulting in distinct tunable patterns. The findings elucidate the underlying mechanism behind the enhanced HER performance of MoSSe through strain engineering, providing theoretical support for the optimal design of efficient HER catalysts based on defective Janus TMDs.