Lijie Wang

Flexible organic crystals offer significant opportunities for organic electronics; however, their potential applications in fields like flexible sensors are at present limited by poor charge transport that result in failure under reversible conditions. In this report, we propose a universal interfacing design strategy for the fabrication of versatile conductive MXene coatings on the surface of flexible organic crystals that preserve the electric conductivity of the MXenes while they also remain mechanically flexible and robust across a wide temperature range of nearly 300°C both above and below ambient temperature. This new sub‐class of hybrid materials based on adaptive organic crystals undergoes reversible changes in the crystal structure upon deformation, while they also engender stable electrical conductivity and retain deformation sensitivity from about −196°C to 100°C. A wireless sensing system based on these materials was constructed to monitor bionic finger movements that demonstrate the potential for human‐machine interfacing and medical monitoring applications. The structural stability of this and similar devices that utilize crystal‐MXene hybrid materials under non‐ambient temperatures provide an innovative solution to the current challenges with occasional failure of low‐temperature mechanical sensing systems in extreme environments, and specifically such that are used in lunar and deep‐space explorations.

Highlights A layered hybrid crystal integrates fluorescence, mechanical flexibility, conductivity, and cryogenic durability via reduced graphene oxide and thermally responsive polyurethane encapsulation. The hybrid crystal enables real-time dual-mode sensing of humidity (1.65% RH⁻¹) and temperature (0.46% °C⁻¹) with high sensitivity and cycling stability. Infrared-induced photothermal actuation at − 150 °C allows reversible crawling and walking under cryogenic conditions.

Traditional volatile organic compounds (VOCs) detection models separate component identification and concentration prediction, leading to low feature utilization and limited learning in small-sample scenarios. Here, we realize a Residual Fusion Network based on multi-task learning (MTL-RCANet) to implement component identification and concentration prediction of VOCs. The model integrates channel attention mechanisms and cross-fusion modules to enhance feature extraction capabilities and task synergy. To further balance the tasks, a dynamic weighted loss function is incorporated to adjust weights dynamically according to the training progress of each task, thereby enhancing the overall performance of the model. The proposed network achieves an accuracy of 94.86% and an R² score of 0.95. Comparative experiments reveal that using only 35% of the total data length as input data yields excellent identification performance. Moreover, multi-task learning effectively integrates feature information across tasks, significantly improving model efficiency compared to single-task learning.