Breaking SWIR detection limits with optically addressed LC-SLM–based single-pixel imaging

An optically addressed spatial light modulator (OASLM) combining an InGaAs photosensor with a liquid crystal (LC) layer is demonstrated for infrared-to-visible image up-conversion in the short-wave infrared (SWIR) spectral range with ultrahigh performance. When SWIR radiation is incident on the InGaAs layer, the resulting photocurrent establishes a spatially varying electric field that reorients the LC molecules and, in turn, modulates the polarization state of a co-propagating visible readout beam. A conventional Si CMOS camera then captures the resulting visible image, providing a direct reconstruction of the original infrared scene. The improvements include back-surface polishing of the InP substrate, which reduces scattering of the SWIR beam, thus improving sensitivity and resolution; an LC material with high dielectric anisotropy (Δε = 60) to minimize the operation voltage; and reading at a shorter optimised wavelength to enhance reflection at the LC-InP interface and reduce the writing effect. Under an illumination of 6.36 μW/cm 2 , the device achieves a contrast of approximately 70 % and is most stable at modulation frequencies between 70 and 90 Hz at 1.3 V. Spatial resolution estimated from the line spread function to be ∼470 lp/mm, and transient characterization gives rise and decay times of 44 ms and 76 ms, respectively. The minimum measured SWIR intensity required to switch ON the device was 2.8 nW/cm 2 at 17 % contrast, and the estimated switching energy per bit is 5.53 attoJoul/bit (equivalent to 40 photons), and 9.56 attoJoul/bit (70 photons) for switching OFF, thus highlighting the importance of this device for optical computing, not only as a highly sensitive SWIR camera. When the green readout LED is replaced with a broadband white LED, the voltage-controlled birefringence of the LC layer introduces wavelength-dependent retardation, producing spectral shifts and perceptible colour variations in the reconstructed images. The distinction between glycerol and water is demonstrated using this camera as a way of example. Taken together, these findings confirm that the proposed OASLM architecture offers a compact, low-cost route to real-time SWIR imaging that can be integrated with any Si-based camera vision system to convert into a SWIR imaging system.