The RAMANTIS project aims to develop a unique system capable of ‘seeing’ the chemical composition of materials and biological tissues. Raman spectroscopy is a ubiquitous technique for material inspection, offering accurate and non-destructive identification of the composition of samples. Raman imaging (RI) applies this technique to extended two-dimensional samples. Current RI systems require scanning times of seconds to even days to produce a single Raman image. This project will build the first proof-of-concept real-time Raman imaging (RTRI) system capable of producing, and displaying live, 2D maps of the composition of materials.

The importance of Raman spectroscopy can be gauged from its global market size of ~$1.5 billion and its annual growth rate of ~10%. It is used on biomedical samples, novel materials, pharmaceuticals, semiconductors, and more. The availability of RTRI would open up a large new domain of applications and research opportunities, as it would not only drastically shorten scanning times, but also enable the characterisation of dynamical processes as they happen. Some examples of potential applications are in vivo identification of malignant tissue, monitoring of the growth of carbon nanotubes and graphene, characterisation of the (de)hydration of salt crystals used in seasonal heat storage, and cheaper quality control during the production of computer chips.

Raman scattering is the process by which photons lose or gain a bit of energy, and thus shift in wavelength, as they scatter off a material. The energy lost or gained depends on intra- and intermolecular vibrations of the material, and can therefore have only very specific values. Consequently, one can illuminate a material with light of a single colour and decompose the reflected light into its constituent colours to identify it. To perform RTRI, we need a detector capable of simultaneously resolving the spatial information and wavelength, as well as a data acquisition system capable of analysing the data in real-time. Current state-of-the-art Raman imagers take thousands of spectra as a laser is raster-scanned across the surface, or use wavelength-tuneable filters to take sets of single-colour images sequentially. Since these systems rely on scanning, they are not suitable for real-time applications. Furthermore, if full spectra are recorded at all pixel locations (as is common), the amount of data generated is intractable for real-time analysis.

We will take away these limitations by repurposing the multispectral imaging system MANTIS, recently developed for fusion research, as a vastly improved detector capable of synchronously imaging light from 10 Raman peaks. Combined with an innovative data acquisition system and advanced image-processing algorithms, it will enable the capture of 2D images with both high chemical specificity and high temporal resolution. As a first proof-of-concept, we will determine the number of graphene layers on a silicon wafer in real-time.