Spectral-engineered upconversion heterostructure for high performance self-powered dual-mode photodetectors.

The rapid advancement of intelligent driving, LiDAR, and object recognition creates a pressing need for self-powered photodetectors (PDs) combining broadband response with wavelength-specific detection. However, a central challenge lies in achieving this dual functionality with high sensitivity in the near-infrared (NIR). The heterogeneous integration of two-dimensional materials with upconversion nanoparticles (UCNPs) presents a viable strategy, yet its application's potential is hampered by precise spectral matching and inefficient carrier separation. Herein, we developed an innovative heterostructure based on GaAs/MoS2/NaYF4: 20%Yb3+, 15%Ho3+, x mol%Ce3+ (x = 0, 2, 5, 10, 20) UCNPs for high performance self-powered photodetection. With an optimal Ce3+ doping concentration of 5 mol%, we tailored the UCNPs through precise luminescence engineering to exhibit a dominant and enhanced emission peak at 650 nm under 980 nm excitation, achieving optimal spectral overlap with the high-absorption bands of both multilayer MoS2 and GaAs. The heterostructure capitalizes on the built-in electric field at the GaAs/MoS2 interface to enable efficient carrier separation and zero bias operation. The resulting device demonstrates a remarkable dual-mode photoresponse, a broadband detection range spanning from violet to NIR (400-870 nm), and a sharp, selective narrowband response at 980 nm. At zero bias, it achieves a responsivity of 0.12 A/W and a specific detectivity of 2.4×1010 Jones at 980 nm, with a fast response time of ∼80/98 ms. Furthermore, the PDs exhibit excellent stability and anti-interference capability in the self-powered mode. This work successfully addresses the critical challenge in the collaborative optimization of broadband and narrowband detection through precise spectral engineering and heterojunction band design, and provides a new perspective for future compact and multifunctional optoelectronic systems.