The new device, described in an early view article in Analytical Chemistry , sequentially illuminates bacteria with different ultraviolet (UV) wavelengths and measures a spectrum at each wavelength allowing a 2 dimensional (2D) spectral fingerprint to be collected that can differentiate between different species of bacteria.

Such a device could find application in many areas, such as in the identification of drug-resistant bacteria for drug screening purposes, identification of bacteria in hospitals, food and pharmaceutical manufacturing

Researchers from the Naval Research Laboratory in Washington DC, US and Research Support Instruments, Maryland US developed the swept wavelength optical resonant-Raman device (SWOrRD) to rapidly identify bacteria in a variety of settings without the need for reagents or lengthy sample culturing.

The SWOrRD consists of a variable-wavelength laser that illuminates the sample, optics that collect the light scattered from the sample and a two-stage spectrometer.

A series of 30 different spectra were collected for five bacterial species, Staphylococcus epidermidis (S. epidermidis), Bacillus cereus (B. cereus), Bacillus thuringiensis (B. thuringiensis), Escherichia coli (E. coli) and Yersinia rohdei (Y. rohdei) using the SWOrRD.

These spectra were processed and aligned on a grid to give 2D spectral maps for the bacteria that can be easily distinguished from each other.

Small spectral differences that are difficult to distinguish in single spectra are more readily identified in the 2D spectra as the signals evolve as the wavelength changes.

Differences in the 16S rDNA of bacteria has been shown to be a very sensitive method of determining how related bacteria are, and even B. cereus and B. thuringiensis that are 98 per cent related can be distinguished using the technique.

The spectral signature of one type of bacteria can be distinguished from that of another using the Optical Real-time Adaptive Spectral Identification System (ORASIS) that was developed for the analysis of satellite and hyper-spectral imagery.