|
Probe-based Raman systems are widely used for in-situ measurements. However, the intrinsically weak Raman signals limit its applications. A common approach to collect more Raman photons is by including more collection fibers with taller detectors. However, it is costly and requires modification of the spectrograph.
Most Raman spectra have broad silent spectral regions. Here, we increase the throughput by introducing horizontally shifted collection fibers rather than vertically. Our machine learning technique successfully deconvoluted the original spectra and improved the limit of detection. Our approach is a simple, cost-effective, and universal method to increase the throughput without modifying existing Raman spectrometers.
|
