Developing raman excitation spectroscopy for high-speed material characterization

Abstract

Spontaneous Raman Spectroscopy has greatly aided material characterization due to its high specificity, label-free operation, and minimal sample preparation requirement. These advantages have made it an indispensable tool for a range of applications such as automated inspection of industrial products, identification of illicit substances, in-vivo biomedical diagnoses, and more. However, spontaneous Raman spectroscopy lags far behind enhanced Raman techniques when it comes to the speed with which samples can be analyzed.

Enhanced techniques such as stimulated Raman scattering or surface-enhanced Raman scattering overcome Raman spectroscopy’s biggest challenge: weak Raman signals compared to more prominent phenomena such as fluorescence. Doing so allows them to operate at much higher throughputs albeit with certain trade-offs such as labelled detection, non-linear generation, or highly specialized instruments. These drawbacks motivate the development of faster spontaneous systems that can help bridge the speed gap while retaining wide applicability and flexibility of operation.

The difference between excitation spectroscopy and conventional methods is the use of a swept excitation wavelength and a fixed detection wavelength instead of the opposite configuration to measure a Raman spectrum. This allows the detection of a Raman spectrum to be completed using a single-pixel detector with high sensitivity and high gain compared to a conventional multi-pixel camera sensor. Acquiring Raman excitation spectra this way is only possible due to Raman scattering occurring at fixed shifts beyond the excitation wavelength regardless of the actual wavelength itself.

Through a preliminary study, it was empirically proven that Raman spectra acquired through both methods are comparable under non-resonant conditions. With these results, a Raman excitation spectrometer was designed and several key factors were identified that ultimately dictate the spectrometer’s performance such as the detector’s transimpedance. These factors were used to construct a guideline to achieve the maximum throughput a system like this is capable of.

These guidelines were then applied towards the construction of a standalone system that was assessed using three main criteria: high-speed measurements, signal-to-noise ratio/sensitivity, and fluorescence minimization. A range of samples including biological tissues, pigments, chemicals, and microparticles were chosen to evaluate the system’s performance. This system acquired equivalent or better spectra when compared to a commercial system at similar acquisition times. When compared to such systems in current literature, acquisition speed improvement upwards of 300x and beyond 1000x was achieved. It was also shown that simply increasing acquisition time in this system was not equivalent to a longer integration time in conventional systems. Rather, averaging acquisitions for an equal amount of time produced more comparable results.

The system was then extended towards point-scanned imaging and dynamic sample measurements. A Raman sample with high spectral contrast was constructed and measured with the excitation wavelength specifically targeting the sample’s characteristic Raman peak. A maximum speed of 1.5 ms per pixel was demonstrated. In the final test, a polystyrene microbead rapidly moving across the focal spot at a speed of roughly 1 mm/s was measured. The system was able to capture 5 complete Raman spectra during the movement, indicating that even faster particle movement could be accommodated. A complete characterization of the time-based response of such systems could potentially yield significant improvements.

The Raman excitation spectroscopy system developed in this work significantly accelerates the acquisition of spontaneous Raman spectra, acquiring Raman spectra at up to 1 ms per 100 1/cm of Raman shift with 6 – 18 1/cm resolution, without sacrificing its label-free, specific, and widely applicable nature. This research contributes to establishing the budding field of swept-source or Raman excitation spectroscopy as a viable high-throughput modality for demanding characterization applications.