Answers are grounded in this paper and cite their sources.
Near-infrared (NIR) spectroscopy enables nondestructive chemical analysis, yet quantitative measurements in aqueous environments are fundamentally hindered by water’s dominant, broadband absorption. This strong background obscures weak analyte signatures and can destabilize spectral inference in chip-scale spectrometers. Here, we demonstrate a chip-scale differential NIR spectrometer based on an integrated bidirectional photonic circuit that operates robustly in strongly water-absorbing environments. Counter-propagating light paths simultaneously encode solvent-dominated and analyte-sensitive spectral bands through a shared bidirectional photonic circuit, reducing internal encoding mismatch and enabling differential suppression of water-induced backgrounds and common-mode perturbations. We validate this approach using fully packaged silicon nitride spectrometers spanning two ultrabroad NIR windows. Quantitative measurements of water–ethanol mixtures yield a formal limit of detection of 0.05% based on the 3σ/slope criterion, whereas concentration differences down to 0.03% remain experimentally distinguishable. Beyond static analysis, the platform enables periodic, long-term tracking of ethanol generation from aliquots collected during a 10-day apple-juice fermentation. We establish a general and scalable strategy for robust chip-scale NIR spectroscopy in complex liquid environments, opening opportunities for robust chemical analysis on second-scale timescales and process monitoring in industrial, environmental, and biomedical applications.
No takes yet. Share an insight, caveat, or question.
Yao et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: