Nano-Raman Images on the Nanoscale with a Light Touch

Hank Hogan

When dealing with something delicate, you must be gentle. Researchers at Vanderbilt and Fisk universities, both in Nashville, Tenn., and at Istituto di Struttura della Materia del Consiglio Nazionale delle Ricerche in Rome thus turned to scanning near-field optical microscopy in the collection mode to perform nano-Raman mapping of porous glass ceramics. They needed the subdiffraction-limited resolution of scanning near-field optical microscopy but did not want the act of imaging to change what they were trying to observe.

Using an optical fiber aperture to collect the signal in a scanning near-field optical microscope, researchers created this phase ratio map of TiO₂ to PO₄ and cross-section profile for calcium-titanium-phosphorus (CTP) glass ceramics in a nano-Raman scan that covered 0.85 x 0.85 μm² and revealed chemical composition. The inset shows the ratio variation along the line from A-A'. The white represents high ratio points, and the black, lower ones. The map showed no correlation with a simultaneously acquired topography map, demonstrating significant lateral variations in ceramic crystalline skeleton composition.

“This approach, collection mode, is less disturbing of the local fields on a sample compared to the apertureless surface-enhancing methods,” said Andrey Zavalin, a research professor of physics and astronomy at Vanderbilt.

In scanning near-field optical microscopy, a probe held very near the surface of a sample is moved across it in a raster pattern. The probe enables resolution below the classical diffraction limit, and such near-field capabilities are crucial to achieving resolutions of less than 100 nm using optical sources.

The nano-Raman scanning near-field optical microscope setup was based on an Olympus IX71 inverted microscope. The scanner head is on the top, protected with acoustic/thermal shielding.

Nano-Raman, the combination of scanning near-field optical microscopy and Raman imaging, enables the chemical composition of a sample to be mapped on the nanoscale. One method is to use a metal tip illuminated by an excitation source as an apertureless probe. Unfortunately, the enhanced field created by such probes can induce surface-enhanced Raman scattering — a particular problem for porous glass ceramics, which are used as a surface-enhanced Raman scattering substrate — or otherwise disturb the sample.

The other approach involves an optical fiber aperture, which does not excite the substrate. The research team used this technique, constructing an instrument out of an Olympus Corp. inverted microscope, a homemade scanning near-field optical microscope scanner, a Princeton Instruments/Acton Research Corp. spectrometer and a 488-nm Coherent Inc. optically pumped semiconductor laser.

The laser illuminated the sample while the scanner collected the Raman signal. The scanner consisted of a small aperture, a less-than-100-nm-diameter end of an optical fiber coated with gold. The investigators moved this across the surface, using a feedback mechanism to keep it close. They mapped the TiO₂/PO₄ phase distribution in ceramics, demonstrating a lateral resolution of better than 100 nm, while looking at the intensity of Ti-O and P-O bands.

Zavalin noted that the spectrometer and light-coupling optics had to be extremely efficient, as the signal was not strong. The entire setup had to be mechanically stable to enable long scans without mechanical disturbances.

Plans call for improving the stability of the instrument. Possible applications include studying hot spots on surface-enhanced Raman scattering active substrates — regions where the scattering effect is particularly pronounced. There also may be uses in mapping the composition and structure of chips on the micro- and nanoscale and biological cells, Zavalin said.

Applied Physics Letters, March 27, 2006, 133126.