Phase competition underlies many remarkable and technologically important phenomena in transition metal oxides. Vanadium dioxide (VO2) exhibits a first-order metal–insulator transition (MIT) near room temperature, where conductivity is suppressed and the lattice changes from tetragonal to monoclinic on cooling. Ongoing attempts to explain this coupled structural and electronic transition begin with two alternative starting points: a Peierls MIT driven by instabilities in electron–lattice dynamics and a Mott MIT where strong electron–electron correlations drive charge localization1,2,3,4,5,6,7,8,9,10. A key missing piece of the VO2 puzzle is the role of lattice vibrations. Moreover, a comprehensive thermodynamic treatment must integrate both entropic and energetic aspects of the transition. Here we report that the entropy driving the MIT in VO2 is dominated by strongly anharmonic phonons rather than electronic contributions, and provide a direct determination of phonon dispersions. Our ab initio calculations identify softer bonding in the tetragonal phase, relative to the monoclinic phase, as the origin of the large vibrational entropy stabilizing the metallic rutile phase. They further reveal how a balance between higher entropy in the metal and orbital-driven lower energy in the insulator fully describes the thermodynamic forces controlling the MIT. Our study illustrates the critical role of anharmonic lattice dynamics in metal oxide phase competition, and provides guidance for the predictive design of new materials.
Access optionsAccess options
Subscribe to Journal
Get full journal access for 1 year
only $3.90 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Research by J.D.B., O.D., M.E.M., E.D.S., L.A.B. and R.J.M. was supported by the US Department of Energy (DOE), Basic Energy Sciences (BES), Materials Sciences and Engineering Division (MSED). Research by J.H. was supported by the Center for Accelerating Materials Modeling, funded by the US DOE, BES, MSED. Experimental work by C.W.L. was sponsored by the Laboratory Directed Research and Development Program of ORNL (Principal Investigator, O.D.). Research by D.L.A. at the Spallation Neutron Source and J.Z.T., A.H.S. and B.M.L. at the Advanced Photon Source (APS), Argonne National Laboratory (ANL), was supported by the US DOE, BES, Scientific User Facilities Division. We thank A. Tselev, S. Nagler, A. Banerjee, H. Krakauer and V. Cooper for interesting discussions on VO2. Inelastic neutron scattering measurements were performed using the ARCS facility at the ORNL Spallation Neutron Source, which is sponsored by the Scientific User Facilities Division, Office of Basic Energy Sciences, US Department of Energy. We thank J. Niedziela for help with the sample environment at ARCS. IXS measurements were performed using the X-ray Operations and Research (XOR) beamline 30-ID (HERIX) at the APS. Diffuse X-ray scattering measurements were performed using the XOR beamline 33-BM-C at the APS. We thank J. Karapetrova and C. Schleputz for assistance in setting up experiments at UNICAT. Use of the APS, an Office of Science User Facility operated for the US DOE Office of Science by ANL, was supported by the US DOE under contract no. DE-AC02-06CH11357. Theoretical calculations were performed using resources of the National Energy Research Scientific Computing Center, a DOE Office of Science User Facility supported by the Office of Science of the US Department of Energy under contract no. DE-AC02-05CH11231. We thank O. Hellman for providing the temperature-dependent effective potential software and assistance.
Extended data figures
Extended data tables
Temperature-dependent x-ray diffuse scattering measurements from VO2 single crystals were measured using an area detector as the sample was rotated. The reciprocal lattice vector was calculated for each pixel of the detector at each angle setting. Two-dimensional and three dimensional diffuse scattering maps were calculated by averaging the counts of all pixels contained in each voxel of a regular 0.05 x 0.05 x 0.025 rutile reciprocal-lattice grid. The Supplementary Video illustrates that, as discussed in the article, the 3-D thermal diffuse H+K+L=2n, n≠0) in reciprocal space.
About this article
Scientific Reports (2017)