Doing the Splits: Acquisition of High-Resolution Raman Spectra Using Broadband Light Sources

• Universität Bremen

Introduction

Raman spectroscopy is a very versatile analytical tool that finds numerous applications in the pharmaceutical industry, for example, in the structural analysis of pharmaceuticals or as a means of process analytical technology. Besides the signal-to-background level, the spectral resolution is a key parameter determining the quality of a Raman spectrum. The spectral resolution is commonly referred to as the smallest difference in wavelength Δλ that can be distinguished at a given wavelength λ. In other words, if there are two signal peaks close to each other in a spectrum, the spectral resolution determines whether they can be identified as individual spectral features or appear as a single band.

 Figure 1. Observed spectral resolution as function of spectrometer resolution and bandwidth of light source.

In a Raman experiment, there are two main components, which determine the overall spectral resolution of the data: the light source and the spectrometer (including the detector). The spectrometer and detector define a limit of the resolution that is achievable as can be seen in Figure 1. In case of a narrowband light source such as a singlemode laser, the spectral resolution of any recorded spectrum is that of the spectrometer. However, when the bandwidth of the light source exceeds the resolution of the spectrometer, the spectrum will be governed by the light source. For this reason, broadband light sources such as light-emitting diodes (LEDs) and commercial laser pointers are usually deemed unsuitable for Raman spectroscopy. That is a pity as LEDs and laser pointers are cheap and widely available.

Experimental Approaches

There have been a number of attempts to develop experimental setups employing LEDs and laser pointers for Raman spectroscopy. The most straightforward approach is to spectrally narrow the radiation from the light source using a bandpass filter.¹ This was found suitable for obtaining sufficiently well resolved Raman spectra in the near-infrared, where a typical filter bandwidth of 1 nm translates into a resolution of ~20 cm^-1. At shorter wavelength, the bandpass filter would need to be even narrower in order to achieve the same spectral resolution, because the Raman shift is constant in frequency, while the dispersion in a spectrometer is linear with wavelength. Hence, recording Raman spectra with different excitation wavelengths in a spectrometer with a given resolution (in nm) will result in two spectra with different resolution in the wavenumber (cm^-1) domain. Another disadvantage of using a narrowband filter for spectrally shaping the excitation radiation is the significant loss in intensity and thus the reduced signal-to-noise ratio.

The concept proposed by Greer et al. was a game changer regarding the use of broadband light sources in Raman spectroscopy.² They collimated the radiation from an LED, sent it through an arrangement of two reflective diffraction gratings, and thereby spatially dispersed the excitation light in the focal plane inside the object of investigation. The signal emitted from the spatially dispersed radiation was imaged onto the entrance slit of an imaging spectrograph and eventually recorded as a 2D image. In this image, each pixel row represents a Raman spectrum recorded with a slightly different excitation wavelength. Even more important, the resolution of each of these spectra is determined by the spectrograph and the dispersion of the excitation radiation in the sample, but not by the bandwidth of the light source. Consequently, this approach enables high-resolution Raman spectroscopy with broadband light sources.

 Figure 2. Schematic iSERDS setup. LP = laser pointer, L1,L2 = lens, BD = beam dump, F = filter, CCD = charge coupled device detector.

The approach of Greer et al. was recently taken forward and extended to shifted-excitation Raman difference spectroscopy (SERDS). As the signal image contains Raman spectra with slightly different excitation wavelength, the new approach facilitates instantaneous SERDS spectroscopy without the need for recording two spectra sequentially.³ This iSERDS technique has been further modified by replacing the arrangement with two reflective gratings by a single transmission grating, and by replacing the LED by a commercial laser pointer.⁴ Thereby the setup, which is illustrated schematically in Figure 2, becomes much simpler, easier to align, and cheaper.

Conclusion

The use of broadband light sources in Raman spectroscopy has become a true alternative to the commonly employed narrowband lasers. Recent developments added experimental techniques to the menu that enable recording Raman spectra with a spectral resolution virtually independent of the bandwidth of the excitation radiation. Moreover, broadband light sources such as LEDs and laser pointers offer the advantage of being cheap and robust, and hence the light source is no longer a cost driver. A particularly promising area for LED- and laser pointer-based applications is the analysis of sensitive samples, where low intensities are desirable. This includes the analysis of high-value pharmaceutically active compounds as well as the microscopy of biological systems.

References

1. Schmidt MA, Kiefer J. Polarization-resolved high-resolution Raman spectroscopy with a light-emitting diode. Journal of Raman Spectroscopy 2013;44(11):1625-1627.
2. Greer JS, Petrov GI, Yakovlev VV. Raman Spectroscopy with LED Excitation Source. Journal of Raman Spectroscopy 2013;44(7):1058-1059.
3. Kiefer J. Instantaneous Shifted-Excitation Raman Difference Spectroscopy (iSERDS). Journal of Raman Spectroscopy 2014;45(10):980-983.
4. Grüber J, Kiefer J. Advanced Instantaneous Shifted-Excitation Raman Difference Spectroscopy (iSERDS) Using a Laser Pointer. Journal of Raman Spectroscopy 2016;47(9):1049-1055.

Author Biography

Prof. Dr. Johannes Kiefer is Chair Professor and Head of the division Technische Thermodynamik at the University of Bremen, Germany. In addition, he is an Honorary Professor at the University of Aberdeen, Scotland, and he holds a guest professorship of the Erlangen Graduate School in Advanced Optical Technologies (SAOT) at the University Erlangen-Nuremberg, Germany. His research interests are the areas of developing and applying spectroscopic techniques for the characterization of advanced materials and processes.