Objective For high-speed dynamic imaging applications such as automotive night vision and unmanned aerial vehicles, uncooled infrared focal plane arrays (UFPA) must simultaneously maintain high thermal sensitivity and fast response speed. The figure of merit (FOM = NETD (Noise Equivalent Temperature Difference) × τ (thermal time constant)) is a key metric for evaluating this comprehensive performance. While smaller pixel sizes offer advantages in size, weight, and power (SWaP), they suffer from reduced photosensitive area, resulting in a higher FOM and greater optimization difficulty. Achieving a balanced enhancement of sensitivity and response speed through structural design remains a core technical challenge.

Methods Through theoretical modeling, the key parameters influencing FOM—detector noise V_n, thermal capacitance C, and effective absorptivity A_eff (the product of fill factor and absorptivity)—were identified. The bridge area was determined to be the critical structural variable for modulating these parameters. Based on this insight, an optimization strategy centered on bridge area modulation was proposed. Multiple micro-bridge structures with varying dimensions were designed and fabricated to systematically investigate their impact on device performance.

Results and Discussions Experimental results reveal a nonlinear relationship between bridge area reduction and its effects on noise, thermal capacitance, and effective absorptivity. When the bridge area is large, 1/f noise contributes less to the total noise, resulting in a relatively stable noise level. However, as the area decreases below a critical threshold, 1/f noise rapidly increases and becomes dominant, leading to the degradation of NETD (Fig.3(b)). Thermal capacitance C decreases nearly proportionally with area, whereas effective absorptivity A_eff decays more slowly due to optical diffraction effects and edge absorption, causing the C/A_eff ratio to decrease with shrinking area (Fig.5). This trend favors a shorter thermal time constant per unit absorption. A comprehensive analysis of V_n, C, and A_eff shows that FOM initially decreases and then increases with further area reduction, indicating an optimal point (Fig.6). The minimum FOM is achieved at a bridge area of 30.7 μm². The non-monotonic dependence of FOM on bridge area is governed by the competing effects of noise and thermal time constant. At large areas, the 1/f noise is negligible, and the total noise is dominated by Johnson and readout circuit noise. As the area decreases, the 1/f noise, which scales inversely with the volume of the thermosensitive material, increases rapidly. This leads to a sharp rise in NETD, outweighing the benefit of reduced thermal time constant. Therefore, the optimal FOM is achieved at a bridge area of 30.7 μm², where the trade-off between noise and response speed is balanced.Based on this optimization, a 12 μm-pitch 640×512 detector was developed, supporting a maximum frame rate of 120 frame/s. The measured NETD is 32.4 mK (at 120 frame/s), the thermal time constant is 5.1 ms, and the FOM is as low as 165 mK·ms. This performance significantly surpasses that of current mainstream products, whose FOM typically ranges from 260 to 700 mK·ms with frame rates generally not exceeding 60 frame/s. In contrast, the developed detector achieves a shorter τ and supports 120 frame/s operation, enabling superior dynamic imaging capability. Imaging tests demonstrate that the detector effectively suppresses image smearing and the "jello effect" (rolling shutter distortion) when observing fast-moving targets, resulting in a substantial enhancement in image quality (Fig.8 and Fig.9).

Conclusions The bridge area is a critical structural parameter for optimizing the FOM of uncooled infrared detectors. Precise modulation of the bridge area enables a reduction in thermal capacitance while effectively controlling the increase in noise and the decrease in effective absorptivity, thereby achieving a lower thermal time constant without sacrificing sensitivity, leading to a significant improvement in FOM. The FOM optimization method based on bridge area modulation provides an effective technical solution for developing high-performance, high-frame-rate uncooled infrared detectors; the fabricated device demonstrates excellent dynamic imaging performance, validating the feasibility of the proposed approach.