
Congratulations to Samuel White in the Haglund lab! Dr. White’s article “Solid-State Dewetting of Tungsten-Doped Vanadium Dioxide Nanoparticles: Implications for Thermochromic Coatings” has been selected as a VINSE spotlight publication.
Sam White received his Ph.D. in Physics from 91Թ in 2022 under the supervision of Professor Richard Haglund. Dr. White has completed an NRC Postdoctoral Research Associateship at the Naval Research Laboratory under the supervision of Dr. Rachael Myers-Ward and is now working as a Physicist for Resilience Engineering and Operations at IonQ.
Vanadium dioxide (VO2) undergoes a solid-solid phase transition at a critical temperature (Tc) of about 70°C, or when excited electrically or optically. A small change in crystal structure is accompanied by a large change in both electrical and optical properties, making it useful for a wide variety of devices, such as passive thermal control films and active photonics. Most applications, however, have requirements which the properties of native VO2 cannot satisfy, such as Tc near or below room-temperature, or certain transmission/absorption levels in a specific spectral window. Chemical changes (such as adding dopant atoms) and morphological changes (achieved by different growth/processing methods) can have significant impacts on these properties, and much ongoing research seeks new and better ways to tailor VO2 to specific use cases. Doped nanoparticles in particular have achieved very low Tc without loss of the all-important optical contrast. Our group already excels at producing high-quality VO2 thin films, both doped and undoped; patterning these films into nanoparticles by lithographic means achieves unparalleled control, but is slow and expensive.
In this paper, we study the formation of doped and undoped VO2 nanoparticles via solid-state dewetting of thin films, achieved simply by adjusting the conditions under which films are annealed, a necessary step making VO2 films. When annealed at ~500°C for ~10 minutes, VO2 thin films on silicon (with native oxide) spontaneously dewet to form nanoparticles. As the anneal time is increased from 10-20 minutes, these nanoparticles grow. We model this particle coarsening by Smoluchowski aggregation, in which randomly-moving particles meet and coalesce, and successfully reproduce the measured particles size distributions. At still longer times or higher temperatures, the VO2 becomes oxidized into the more stable V2O5. We show that the same process also produces nanoparticles from W-doped films, though with different nanoparticle morphology, and that the W-dopants concentrate at the nanoparticle surface. These results yield new insights into substrate-VO2 interaction and the behavior of dopants in VO2, expanding our ability to fine tune this versatile material for its broad range of applications.
Authors: Samuel T. White, James R. Taylor, Ivan Chukhryaev, Silas M. Bailey, Joshua M. Queen, James R. McBride, Richard F. Haglund Jr.
Abstract: Doped vanadium dioxide (VO2) nanoparticles (NPs) have significant potential for applications requiring temperature-dependent emissivity, reflectivity, or transmission. Thermochromic coatings in particular enable energy-saving smart windows and passive thermal radiators but are subject to tight performance constraints. A major challenge is preparing uniform layers of NPs, over large areas, with controllable size distributions and transition temperatures (Tc). We describe the growth and transition characteristics of randomly distributed undoped and W-doped VO2 NPs formed by solid-state dewetting. Sizes and size distributions are controlled by anneal time, as particles grow via Smoluchowski aggregation before oxidizing into V2O5; shapes are determined by the interfacial energies between VO2(2O5) and the silicon substrate. Tungsten dopants concentrate at the NP surface, increasing the energy barrier for and slowing the rate of dewetting, aggregation, and oxidization. Surprisingly, the doped NPs exhibit lower Tc and sharper hysteresis than comparably doped thin films. These results advance our capacity to engineer doped VO2 NPs, yield valuable insights into VO2–substrate interactions, and highlight the distribution of W-dopants in VO2ʲ.