Single-transistor spectrometer via Burstein–Moss absorption edge modulation in compound semiconductors
Abstract
As on-chip photonic systems evolve toward miniaturization and integration, achieving high-performance spectral analysis within a minimalist architecture remains a critical challenge, typically constrained by the trade-off between spatial resolution and spectral channels. Here, we propose a universal design strategy for a single-transistor computational spectrometer based on the gate-controlled Burstein–Moss effect, which enables the dynamic electrical modulation of the optical absorption cutoff, thereby allowing the incident spectrum to be reconstructed solely from a single current–voltage (I–V) curve. Through self-consistent simulations, systematic selection criteria for the active layer are established, prioritizing a small electron effective mass (me*) and a high band edge extinction coefficient (κ). Simultaneously, the gate dielectric design is optimized based on the trade-off between electrostatic control capability and tunneling leakage suppression. As a proof-of-concept, a miniaturized single-transistor spectrometer model integrating a GaAs-on-insulator active layer with HfO2 is numerically demonstrated. Coupled with an adaptive regularization algorithm, the spectrometer exhibits exceptional spectral reconstruction capabilities, achieving a narrowband spectral resolution of ∼4.2 nm and an ultra-high wavelength positioning accuracy of ∼0.04 nm, along with the high-fidelity recovery of dense doublets (∼6 nm separation) and broadband continuous spectra. Furthermore, it demonstrates powerful spatial-spectral decoupling capabilities in array-level imaging and exceptional algorithmic resilience against physical degradation. This work provides a versatile device design blueprint for developing low-cost, high-performance, and process-compatible on-chip integrated spectral analysis systems.
Article Details
Journal Info
Applied Physics Letters
American Institute of Physics
Authors (10)
Xi Chen
Haoyang Shi
Lan Ma
Chenyan Wang
Ze Shang
School of Microelectronics, Fudan University 1 , Shanghai 200433,
Linlong Yang
School of Microelectronics, Fudan University 1 , Shanghai 200433,
Conglin Zhang
School of Microelectronics, Fudan University 1 , Shanghai 200433,
Ming Yang
Shi-Jin Ding
School of Microelectronics, Fudan University 1 , Shanghai 200433,
Xiaohan Wu