Ultrasensitive imaging-based sensor unlocked by differential guided-mode resonance

Imaging-based sensors convert physicochemical parameters of analytes into visible patterns, yet a high sensitivity remains constrained. Here, we introduce the concept of differential guided-mode resonance with thickness modulation at a tens-nanometer scale to greatly enhance the sensitivity, alleviating the sensitivity-dynamic range tradeoff. Experimental results reveal a sensitivity of up to a million-level pixels per refractive index unit (RIU), surpassing existing technologies by nearly three orders of magnitude, with a large dynamic range reconfigured by the incident angle. With the present method, a moderate value (about 100) of the Q factor suffices to make a record high sensitivity and the Figure of Merit (FOM) can reach 10⁴ RIU^-1 level. We also demonstrate a portable device, highlighting its potential for practical applications, including 2D distribution sensing. This method unlocks the potential of imaging-based sensors with both record high sensitivity and tremendous dynamic range for accurate medical diagnosis, biochemical analysis, dynamic pollution monitoring, etc.