Thermochromic VO₂ Thin Films Mapped at the Nanoscale with Tip-Enhanced Raman Spectroscopy
Researchers at the University of Stavanger and ETH Zurich have used hyperspectral tip-enhanced Raman spectroscopy (TERS) to map cross-sections of thermochromic vanadium dioxide films, reporting the work in ACS Applied Materials & Interfaces.
They studied a single VO₂ film and a VO₂/TiO₂ bilayer, both grown on glass by pulsed laser deposition.
In the VO₂ they found small crystals facing in different directions; in the TiO₂ layer, an unexpected brookite phase alongside anatase; and between the two, an interface where the oxides intermix rather than meeting cleanly. The authors believe it is the first TERS mapping of polycrystalline VO₂ films.
Science made simple
Vanadium dioxide acts a bit like a thermostat built into the material. Below its switching temperature it lets infrared light, which carries heat, pass through. Above it, the material turns metallic and reflects much of that heat away. Coat a window with it and, in principle, the glass lets the sun’s warmth in on cold days and keeps it out on hot ones.
To see what’s going on inside these films, the researchers used Raman spectroscopy. A laser is shone at the material and a tiny fraction of the light comes back at a slightly different colour, shifted by the way the atoms vibrate. Each crystal structure gives its own fingerprint. The catch is that an ordinary laser spot is a few hundred nanometres across, so it averages over many grains at once. TERS gets round this by placing a sharp, silver-coated tip in the laser spot. The silver gathers the light at its point, much as a lightning rod gathers electric charge, so the signal comes mainly from the tiny patch directly beneath it. Moving the tip across the sample builds up a map point by point.
Why nanoscale detail matters
VO₂ switches at around 68 °C, too high for practical coatings. One common way to lower it is to grow the VO₂ on titanium dioxide (TiO₂), which puts the film under strain. Whether that works depends on features far smaller than a micrometre: grain size and orientation, unwanted phases, and what happens where the layers meet. X-ray diffraction struggles with VO₂ phases whose signals overlap, electron microscopy needs thinned samples and can damage the film, and conventional Raman can’t resolve detail below the diffraction limit.
Using silver-coated probes
The team made their own probes by coating silicon atomic force microscope cantilevers with a 120 nm layer of 99.99% pure silver, deposited by thermal evaporation. The finished tips had apex diameters of about 86 nm. Measured against conventional Raman spectra taken at the same spots, they gave signals up to five times stronger, an enhancement the authors describe as critical for analysis at this scale.
The samples were cut and polished so the probe could scan across each cross-section, with laser power kept low enough that the VO₂ didn’t switch during mapping.
What the researchers found
The VO₂ contains many small crystals facing in different directions. An orientation-sensitive Raman signal varied between neighbouring points, which the authors read as many small grains and a high density of grain boundaries. Earlier studies have shown that VO₂ films with larger grains, and so fewer boundaries, switch at lower temperatures. On that basis the authors expect this structure to raise the switching temperature. They did not measure it in these films.
The TiO₂ layer contains two crystal forms. It was mostly anatase, but brookite, a less common form, was spread through the layer. Shifts in its Raman bands point to tensile strain, though the authors note that turning those shifts into strain values is difficult for brookite. The map changed from anatase to brookite within a single 10 nm step, which puts the technique’s resolution at 10 nm or better.
The boundary between the layers is not sharp. TiO₂ signal reached up to about 50 nm into the VO₂, evidence that the two oxides intermix. The authors call this a qualitative estimate, noting that TiO₂’s stronger Raman signal may exaggerate its reach. They suggest the mixing likely comes from differences in growth conditions: the VO₂ was deposited at a lower oxygen pressure than the TiO₂, leaving more oxygen vacancies for titanium to move into.
For anyone tuning a thermochromic coating, these are the features to watch. Grain size has a known link to switching temperature, and the TiO₂ is there specifically to strain the VO₂, so an unexpected phase in that layer or a blurred interface could make a film behave differently from its design.
Region | What TERS showed | Authors' caveat |
VO₂ film | M1 phase throughout; orientation-sensitive peak ratio varying from 0.5 to 1 between nearby points | Grain-boundary effect on switching is expected, not measured |
TiO₂ layer | Mostly anatase, with brookite distributed through the layer; brookite bands red-shifted by 10 and 12 cm⁻¹ | Converting Raman shifts into strain values is difficult for brookite |
VO₂/TiO₂ interface | TiO₂ signal extending up to about 50 nm into the VO₂ | A qualitative estimate; far-field contributions cannot be fully excluded |
The silver on the tip
Films that look uniform at the micrometre scale turned out to vary in grain orientation, phase and strain at the nanoscale, and TERS mapped this on a polished cross-section without the sample thinning electron microscopy needs. The paper didn’t measure these films’ switching temperatures, so the effect of those features remains an expectation from earlier work.
The mapping relied on the silver at the tip. A 120 nm coating of 99.99% silver turned a standard AFM cantilever into a probe that could resolve detail down to 10 nm and give signals up to five times stronger than conventional Raman.
Advent supplies 99.99% silver in a range of forms, including wire, rod, foil and tube, under an ISO 9001:2015 certified quality system.
Conclusion
Films that look uniform at the micrometre scale turned out to vary in grain orientation, phase and strain at the nanoscale, and TERS mapped this on a polished cross-section without the sample thinning electron microscopy needs. The paper didn’t measure these films’ switching temperatures, so the effect of those features remains an expectation from earlier work.
The mapping relied on the silver at the tip. A 120 nm coating of 99.99% silver turned a standard AFM cantilever into a probe that could resolve detail down to 10 nm and give signals up to five times stronger than conventional Raman. Advent supplies 99.99% silver in wire and rod forms under an ISO 9001:2015 certified quality system.
SOURCE:
Title: Nanoscale Structural Insights into Thermochromic VO₂ Thin Films Using Tip-Enhanced Raman Spectroscopy
Journal: ACS Applied Materials & Interfaces, Vol. 17, Issue 22, pp. 32625–32634
Authors: Ayushi Rai, Siiri Bienz, Vidar F. Hansen, Renato Zenobi, Naresh Kumar
Published: 22 May 2025
DOI: 10.1021/acsami.5c04827
Abstract: Thermochromic vanadium dioxide (VO2) thin films, known for their reversible metal–insulator transition (MIT) near 68 °C, are promising candidates for energy-efficient applications such as smart window coatings. However, optimizing their structural and interfacial properties to enhance…
