Taking Raman Spectroscopy to Other Worlds

Hank Hogan

When future rovers go to Mars or other planets, equipping them with instruments capable of deep-UV Raman spectroscopy will give them a better chance of answering questions about their surroundings. That is the conclusion arrived at by a group of Germany-based investigators from Friedrich Schiller University in Jena and from the aerospace company Kayser-Threde GmbH in Munich, which has demonstrated the capabilities of UV Raman spectroscopy on several Martian meteorites.

A Raman image (left) and corresponding spectra (right) illustrate data acquired from a Martian meteorite. An excitation wavelength of 257 nm was used over an area ~250 μm². Reprinted with permission of Analytical Chemistry.

The researchers showed a 30 to 100 signal-to-noise ratio for UV Raman spectra versus a 2 to 15 ratio for longer wavelengths. Jürgen Popp, a physical chemistry professor at the university, noted that the results were clear. “Deep-UV Raman spectroscopy is superior to near-infrared or visible Raman spectroscopy.”

When light hits a substance, a small fraction of the photons undergoes an inelastic collision and emerges with a Raman shift that provides information about the chemical makeup of the material. Unfortunately, the Raman effect is weak unless the excitation takes place within an electronic resonance band of the material. Then the scattering cross section can be enhanced as much as 100 million times.

Most materials have an electronic resonance in the deep-UV, meaning an excitation wavelength of 250 nm or less. Moreover, the Raman signal is wavelength-dependent to the inverse fourth power, so the signal at 244 nm is 360 times stronger than that at 1064 nm, even without the increased cross section.

Also, the Raman shift is typically less than 30 nm above the excitation wavelength. Most fluorescence emission from material is above 280 nm for a deep-UV source. Therefore, the fluorescence is spectrally separated from the Raman signal, further improving the technique’s signal-to-noise ratio.

Those benefits, however, come at a technological cost that has kept the method from becoming widely used, which also explains why it has not heretofore been applied to meteorites from Mars. “The reason why UV Raman imaging has not been done before is mainly due to the lack of the right UV Raman excitation laser,” Popp said.

In a series of experiments, the investigators studied three Martian meteorite samples using five excitation wavelengths — 244, 257, 532, 633 and 830 nm — and a micro-Raman instrument from Horiba Jobin Yvon GmbH of Munich.

For the UV excitation source, they used a frequency-doubled argon-ion laser made by Coherent Inc. of Santa Clara, Calif. For visible and infrared excitation, they used a frequency-doubled Nd:YAG laser and a helium-neon laser, also from Coherent, as well as an external cavity semiconductor laser from Sacher Lasertechnik Group of Marburg, Germany.

Comparison of measurements taken on the same spots on the meteorites using different wavelengths showed that deep-UV yielded better signal-to-noise ratios. In addition, almost all spots exhibited useful Raman spectra when illuminated with deep-UV, but more than half did not with either visible or near-infrared excitation.

These results are promising, but spaceflight would require a miniaturized instrument, a version of which was developed recently at the Institut für Physikalische Hochtechnologie in Jena, where Popp is scientific director. There also is the problem of the laser, which must be compact and efficient. Popp reported that investigations of a new approach that could overcome this challenge are under way. “We are testing now a laser based on hollow cathode technology.”

Analytical Chemistry, Feb. 1, 2007, pp. 1101-1108.