Purely electronic insulator-metal transition in rutile VO2 | Nature Communications

Nature Communications volume 16, Article number: 5444 (2025) Cite this article Volatile resistive switching in neuromorphic computing can be tuned by external stimuli such as temperature or

Nature Communications volume 16, Article number: 5444 (2025) Cite this article

Volatile resistive switching in neuromorphic computing can be tuned by external stimuli such as temperature or electric-field. However, this type of switching is generally coupled to structural changes, resulting in slower reaction speed and higher energy consumption when incorporated into an electronic device. The vanadium dioxide (VO2), which has near room temperature metal-insulator transition (MIT), is an archetypical volatile resistive switching system. Here, we demonstrate an isostructural MIT in an ultrathin VO2 film capped with a photoconductive cadmium sulfide (CdS) layer. Transmission electron microscopy, resistivity experiments, and first-principles calculations show that the hole carriers induced by CdS photovoltaic effect are driving the MIT in rutile VO2. The insulating-rutile VO2 phase has been proved and can remain stable for hours. Our finding provides a new approach to produce purely electronically driven MIT in VO2, and widens its applications in fast-response, low-energy neuromorphic devices.

In the cognitive era, volatile and nonvolatile resistive switching has been proposed as the building block to engineer the next generation energy efficient neuromorphic computing technologies1,2. Among the volatile switching, metal-insulator transition (MIT) in Mott systems is particularly promising because of its low energy consumption, fast reaction speed, controllable, and repeatable behavior3. Most of the current methods for modulating resistive switching in MIT materials are mainly limited by heating and/or biasing, which largely restricts their applications. Furthermore, the MIT, which manifest itself in the electronic transport, is also intertwined with the structural phase transition, further limiting the switching speed4,5,6,7. Since metal-to-insulator and structural transitions on the same material are difficult to separate, additional functionality can be brought to a hybrid system by interfacing the MIT material with another functional material that may have high sensitivity to different external stimuli8,9,10,11,12. Therefore, it is exciting and urgent to understand and manipulate the resistive switching process in hybrid systems.

Vanadium dioxide (VO2) is a model MIT system, with a near room-temperature transition, making it suitable for electronic device applications13,14. VO2 changes from insulating monoclinic to conductive rutile phases from low temperature to high temperature. The electronic MIT is accompanied by structural transition, and monoclinic phase has dimerization of V ions along the c-axis. However, ultrafast experiments have shown that the electronic transition could be decoupled from the structural phase transition15. Moreover, some experiments have shown that the nonequilibrium metallic monoclinic phase can be obtained by photo-excitation, showing that isostructural MIT is possible under nonequilibrium conditions15,16. Recently, Lee et al. successfully realized isostructural MIT in monoclinic VO2 by electron doping17. Moreover, Mondal et al. used modulation doping (electron doping) to control the electronic structure of VO218. It is therefore interesting to consider the potential isostructural MIT in rutile VO2 system, especially by doping effect.

In this work, we use a photoactive CdS layer to experimentally and theoretically explore hole doping effects on VO2. Our results surprisingly establish a doping driven transition in VO2, i.e., creating a (meta)stable insulating rutile VO2. This finding completes the VO2 electronic-crystal structure phase diagram (as shown in Supplementary fig. 1). The MIT can occur entirely within the monoclinic phase and now entirely within the rutile phase. The full elucidation of the pure electronic transition would represent a major step forward in the understanding of the MIT in this important material.

An ultrathin 8 nm VO2 film was epitaxially grown on a TiO2 substrate to maximize the effect on the interfacial coupling. A CdS layer was subsequently deposited (about 80 nm thick), forming a CdS/VO2/TiO2 heterostructure. Illumination of CdS (band gap of 2.4 eV) with visible light creates photoinduced carriers19,20,21,22,23,24 which can then drift into the VO2 and modify its electronic properties. It is worthwhile to mention that the 80 nm CdS was chosen because this is the minimum thickness for which we can conveniently measure the electrical transport, as thinner films have too high a resistance (see Supplementary fig. 2). Figure 1A shows the schematic of experimental geometry. The samples and the Au/Pt layers were prepared using a standard focused ion beam (FIB) lift-out process25. The physical and structural properties were measured before and after exposure to a solar light generator. Figure 1B shows the resistance vs. temperature measurement (R-T) obtained from the cooling branch with the light on and off states. The red line is the R-T result measured before shining solar light (off state), and the black line is the result acquired after shining light (on state). The whole R-T loops and time related relaxation study results are shown in the Supplementary figs. 3–6. In the blue region, the VO2 is insulating, while in the pink region, the VO2 is conductive, no matter the light is on or off. However, in the yellow region (near room temperature), the VO2 changes from insulating to metallic state when the light is turned on. Figure 1C shows a high-angle annular dark field scanning transmission electron microscopy (HAADF-STEM) image of the cross-section of the FIB sample. The energy dispersive spectrometer mapping results of the sample are shown in Supplementary fig. 7.

A Schematic experimental geometry. A cross-sectioned FIB (focused ion beam) sample illuminated by white light. B Change in the resistance across the metal-insulator phase transition measured on the cooling branch with the light ON and OFF. C Low magnification high angle annular dark field (HAADF) image showing the layered configuration of the sample. D, E Selected area electron diffraction patterns for the “ON” and “OFF” light states. The dashed rectangular areas are enlarged in the upper left corner of the micrograph. The TEM experiments (imaging, electron diffraction, and spectroscopy) were conducted at room temperature. Temperature dependent XRD results of (F) VO2/TiO2 system and (G) CdS/VO2/TiO2 system. The thickness of VO2 film for measuring the XRD is 8 nm.

Since the system will relax back in 8 h after shining light (Supplementary fig. 3), systematic transmission electron microscopy (TEM) studies can be carried out to understand the mechanism of this photo-assisted MIT. For consistency, the TEM experiments (including electron diffraction and spectroscopy) were conducted within 3 h after the illumination. Figure 1D shows the selected area electron diffraction pattern acquired at room temperature at the VO2/TiO2 interface in CdS/VO2/TiO2 system before exposing light. Both VO2 and TiO2 exhibit a rutile structure, since the ultrathin VO2 is under in-plane tensile strain from TiO2 substrate, which is consistent with the published results25,26. After light exposure, yellow region in Fig. 1B, the conductivity changes dramatically (from insulating state to metallic state), while the VO2 maintains the rutile structure as seen in Fig. 1E. This shows that the light-driven transition from insulator to metallic state is purely electronic and not driven by a structural phase transition. During the TEM observation, less than 1 nA beam current was used, which will cause less than 1 °C temperature increase27.

Numerous studies have examined the metal–insulator transition in ultrathin VO2 films17,18,28,29,30. The reported behavior varies significantly depending on parameters such as film thickness, substrate choice, and the presence of capping layers. While some investigations observe behavior consistent with that of bulk VO218,28, others suggest that ultrathin films may undergo an isostructural MIT17,29,30. In this work, we investigate VO2 films capped with CdS—a configuration that has rarely been studied. To probe possible temperature-induced structural changes, we performed X-ray diffraction (XRD) measurements above and below the transition temperature (360 K and 260 K, respectively). Clear structural changes are observed in our samples consisting of uncapped VO2 on TiO2 substrates (see Fig. 1F), consistent with the well-known rutile-to-monoclinic transition upon cooling18. In contrast, XRD data for our CdS-capped VO2 on TiO2 (Fig. 1G) show no detectable structural changes across the same temperature range, suggesting the absence of a structural transition. Given that electron diffraction confirms the rutile structure at room temperature (see Fig. 1D), we attribute the observed MIT in the capped films to a rutile-to-rutile transition, consistent with an isostructural mechanism.

Electron energy loss spectroscopy (EELS) in TEM has been widely applied to study the electronic structure of materials. The signal of valence EELS (low-loss part of the spectrum) is two orders of magnitude stronger than the core-loss part. The dominant feature of valence EELS is from collective excitations (plasmons). The position of the plasma peak (the highest peak) is related to the density of valence electrons (and their effective mass)31,32,33. Thus, the peak position change reflects the evolution of valence electrons (the details can be found in the Supplementary Information).

Figure 2A is a typical HAADF-STEM image showing the CdS/VO2/TiO2 structure. EELS line scan was taken across CdS/VO2 interface (the white arrow area). Figure 2B, C show the reference valence EELS spectra of CdS and VO2 acquired from the regions far from the interface, respectively. To retrieve plasma components from these valence EELS, multiple Lorentzian functions have been used to fit every valence EELS spectrum33,34,35. The position of the plasma peak is determined from the fitting (as shown in Fig. 2A, B). Two peaks were identified for CdS, and three peaks were fitted for VO2. Figure 2D shows typical extracted valence EELS spectra from different positions before light exposure, i.e., the black line is the valence EELS acquired in CdS (3.24 nm from the CdS/VO2 interface). The extracted valence EELS spectra after light exposure are shown in Fig. 2E.

A High-resolution HAADF-STEM image of the CdS/VO2/TiO2 interface. The white arrow indicates the EELS line scan position. Horizontal dashed lines have marked the interfaces. B Typical experimental EELS spectrum for CdS. Two Lorentzian functions were used to fit the spectrum. C Experimental EELS spectrum for VO2. Three peaks were fitted. EELS spectra were measured at different positions from the CdS/VO2 interface with light OFF (D) and ON (E). (F, G) are the calculated EELS spectra using the experimentally extracted dielectric function from (D, E), respectively. The black arrows show the blue shift of the spectra across the interface from CdS to VO2.

The valence EELS in an inhomogeneous medium reflects the electron state at the measured point and the dielectric coupling between the measured point and nearby areas. The dielectric coupling in valence EELS has been considered based on the dielectric function theory36,37,38. Figure 2F, G are calculated valence EELS spectra by using the dielectric functions retrieved from the experiments (see Supplementary Information), which share similarities to the experimental spectra in Fig. 2D, E, respectively36. The current dielectric function theory36,37,38 assumes that the materials at the two sides of the interface are homogeneous, and their dielectric functions have step changes across the interface. However, this ideal case may not represent experimentally obtained interfaces and the homogeneity near the interface may be destroyed by interfacial strain39,40,41,42, atomic rearrangement, charge transfer, etc.43,44. So, the deviation between calculated and experimental EELS mainly comes from aforementioned inhomogeneous structure near the interface.

The plasma peak dominates the valence EELS, and the position of the plasma peak relates to the density of valence electrons (the details can be found in Supplementary Information):

where E is the energy loss of plasma peak (i.e. the highest plasma peak position in EELS), Nv is the density of the valence electrons, ωP is the plasmon frequency, e and m are electron charge and mass33. The shift of plasma peak is related to the change of the density of the valence electrons.

Figure 3A, B show the energy of the plasma peak as a function of distance x from the interface. x < 0 and x > 0 are the CdS and VO2 areas, respectively. The figures show that the plasmon peaks shift to lower energy after the light is turned on (Fig. 3B). Figure 3C, D show the plasmon peak positions (i.e., the highest peak positions illustrated in Fig. 2B, C) at different coordinates and the corresponding calculation results (i.e., the highest peak positions acquired from the calculated EELS) before and after light exposure. A comparison of Fig. 3C, D shows that both CdS and VO2 lost electrons (gained holes) upon turning the light on, implied by their plasma peaks shifting to lower energy. The shift of plasma peaks caused only by the inhomogeneous electron density near the CdS/VO2 interface can be obtained by subtracting the shift caused by the dielectric coupling. This is accomplished by subtracting the red data points from the black data points in Fig. 3C, D. Using Eq. 1, the changes in electron density can be estimated, and the results before and after illumination are shown in blue and orange in Fig. 3E. Figure 3E shows that the change in the electron density near the interface is slightly different for the two data sets, indicating the effect of the illumination on the electron density. This difference obtained by subtracting the blue data points from the orange ones in Fig. 3E is a measure of the changes in electron density caused by illumination as shown in Fig. 3F.

A, B Contour maps of the EELS line scan intensity across the VO2/CdS interface before and after illumination. The intensity of EELS in the figures decreases from red to blue. Comparing (A, B), the peak positions shift to lower energy (redshift) in each location after the light exposure. The vertical dashed lines are located at 20 eV and guided to the eyes. C, D The energy of the highest plasmon peak vs. position coordinate before and after light exposure. The calculation corresponds to the maximum of the peaks computed in Fig. 2E, F. E Uneven electron density distributions caused by interfacial inhomogeneity as a function of position. The blue and orange dots are the electron density changes measured before and after light exposure. F Differences in electron density changes caused by beam shining (the orange dots (after light exposure) subtract the blue dots (before light exposure) in (E), since the interfacial inhomogeneity effects have been removed.

To develop an understanding of the mechanism behind the experimentally observed phenomena in rutile VO2, density functional theory (DFT) calculations were carried out. Figure 4A illustrates the original band structures of Au/Pt, CdS, and VO2. The band gap, valence band (EV), conduction band (EC), and work function (W) are listed in the figure45,46,47. The work function of rutile VO2 (4.7 eV) is slightly lower than that of CdS (4.8 eV), thereby facilitating hole doping into the VO2. The strained rutile VO2, that results from anchoring to the TiO2 substrate, is insulator at room temperature, as indicated in Fig. 4A.

A Schematic diagrams of the Au/Pt, CdS, and R VO2 band structures. The Fermi energy in each material is indicated by dashed lines. B Total (green), O (red), and V (blue) DOS for the VO2 systems, (C) 0.25 holes per formula unit of rutile VO2. The system with no holes is insulating, whereas the system with 0.25 holes exhibits a metallic behavior.

We have performed DFT + U calculations to investigate the effect of photodoping on the electronic structure of the strained rutile system by introducing holes in the system. Because the change of electron density is about −4 × 1027 electron/m3 (Fig. 3F) at the interfacial area and changes to 0 away from the interface (within 5 nm), we assume an average doping density of −2 × 1027 electron/m3, i.e., 2 × 1027 hole/m3. The unit cell parameters for rutile VO2 are a = b = 4.6 Å, c = 2.9 Å, which means each VO2 has 2 × 2 x 1027 hole/m3 x 4.6 × 10−10 × 4.6 × 10−10 × 2.9 × 10−10 m3 / cell ≈ 0.25 hole/cell. Thus, we adopted 0.25 hole per cell for an order of magnitude calculation. These calculations show that introducing 0.25 holes per VO2 unit closes the insulating gap, as shown in Fig. 4B, C. This finding mimics our experiment shown in Fig. 1B, in which illumination of CdS switches the rutile VO2 from an insulating to a conductive state.

In the unstrained state, rutile VO2 is a metal, which can be observed from the band structure obtained from the Materials Project (mp-19094). However, when under tensile strain from TiO2 substrate, our calculations using the same parameters as the Materials Project find that a band gap of 0.6 eV opens up in rutile VO2. This gap is closed with hole doping. While the DFT + U calculation of hole-doped VO2 shows metallic behavior, the experimental temperature-resistance curve in Fig. 1B shows semiconducting behavior. The physical scenario that is suggested by our calculations is as follows: At low temperature, no matter the light is on or off, VO2 is rutile with a relatively big gap and hence holes are uneasy to be injected. The presence of the gap at low temperatures suggests the emergence of an ordered state with our DFT + U calculations, indicating that AFM ordering is the most likely candidate. Moreover, our calculations also suggest that, as the temperature increases reaching the proximity of the ordinary TMIT, the gap may weaken facilitating the hole-injection by the CdS. This process, eventually drives down the transition by 15 K.

Though CdS could dope carriers inside of VO2, its thickness matters. We measured the resistance of several films with different CdS thicknesses as a function of temperature with and without applied light (see Figure S2). We see a strong trend towards higher resistivity in thinner films and a strong reduction in resistivity in the light on state with respect to the light off. The resistivity of the 80 nm film was so high that it could not be conveniently measured without illumination.

The insulating rutile VO2 has a nominal filling of d1, thus it should be a Mott insulator. Our experiments and calculations suggest that its MIT through photo doping is a filling-controlled Mott transition. It should be mentioned that, in the current device configuration, we couldn’t directly measure the electronic structure of the equilibrium for the hole-doped rutile VO2. A direct angle-resolved photoemission spectroscopy of a related system, V2O3, concluded a different mechanism for the MIT48. The MIT in V2O3 is caused by a progressive decrease in spectral weight of an itinerant, dispersive quasi particle conduction band without noticeable changes in its dispersion and effective mass. This is accompanied by an energy shift and increase in spectral weight, of a quasi-localized state, which goes from an energy close to the Fermi energy in the metallic state to an energy close to the bottom of the vanishing dispersive band in the insulating state. In our case, the doping is the driving force of the transition.

A more comprehensive measurement of the detailed structure of CdS-capped VO2 on TiO2 would be a valuable future study. This would ideally include both electron diffraction, which has excellent sensitivity to small sample volumes, and X-ray diffraction measurements on off-specular Bragg peaks, which has the advantage of macroscopic sampling of the film, along with more straightforward possibilities to model the structure due to the kinematic interaction of X-rays.

In conclusion, a photo-assisted, isostructural doping-driven MIT was found in ultrathin rutile VO2 film capped with photoactive CdS. The TEM and transport studies together with DFT calculations show that besides the previously reported insulating monoclinic phase, conductive rutile phase, and (meta)stable conductive monoclinic phase, an insulating rutile phase can be stabilized by anchoring between rutile TiO2 substrate and CdS capping layer. Thus, another missing piece of the puzzle in the VO2 electronic-crystal structural phase diagram has been found. The tensile strain induced by the substrate puts VO2 in a new type of stable state, which is the whole idea of hybrid multilayer materials. In contrast, the gradual decay over time of the “on-state” reflects a slow relaxation process rather than a true metastability phenomenon. In addition to the commonly investigated thermal and electric field induced transitions, our results show that heterostructuring a Mott insulator with a photoactive material adds a new degree of freedom. This provides a new way to modulate the MIT, thus widening the applications of VO2 to photo-active electronic devices. In particular, the isostructural transition reported here shows that VO2 is a potential material platform for implementing (quasi)nonvolatile, low-energy, fast-response switching devices.

A FEI Helios G4 FIB was used for the TEM sample preparation. A JEOL 200CF TEM at Brookhaven National Laboratory was used for the TEM studies. Dual EELS was used, and all EELS spectra were corrected by the zero-loss peaks. For HAADF images, the convergent angle is 21.2 mrad, and the collection angle is between 67 and 275 mrad.

An 8 nm thick epitaxial VO2 film was grown by reactive sputtering on the TiO2 substrate (oriented along the (001) plane). A 4-mtorr argon/oxygen mixture (8% O2) was used during deposition, and the substrate was kept at 350 °C during the growth and later cooled down at a rate of 12 °C/min. The CdS, 80 nm thick film was grown on top of the VO2 using radio frequency magnetron sputtering from a CdS target in a 2 mTorr pure argon atmosphere at 150 °C. Two Au (20 nm) electrodes were patterned on the CdS/VO2 heterostructure films. Room temperature XRD measurements were done in a Rigaku SmartLab system. Single-phase growth was confirmed by XRD, textured along 〈002〉 for VO2 and H〈002〉 for CdS. Transport measurements were carried out on a Montana C2 S50 Cryocooler and TTPX Lakeshore cryogenic probe station, using a Keithley 6221 current source and a Keithley 2182 A nanovoltmeter.

The simulated solar light is provided by a 150 W solar simulator equipped with an air mass 1.5 global (AM 1.5 G) filter (Newport), and the light power was calibrated to 1 Sun (100 mW/cm2) using a quartz-windowed Si solar cell (Newport).

The authors declare that the data supporting the findings of this study are available within the paper and its supplementary information file. Source data are provided with this paper.

Marković, D., Mizrahi, A., Querlioz, D. & Grollier, J. Physics for neuromorphic computing. Nat. Rev. Phys. 2, 499–510 (2020).

Article Google Scholar

Valle, J. D. et al. Spatiotemporal characterization of the field-induced insulator-to-metal transition. Science 373, 907–911 (2021).

Article ADS PubMed Google Scholar

Wang, Z. et al. Resistive switching materials for information processing. Nat. Rev. Mater. 5, 173–195 (2020).

Article ADS CAS Google Scholar

Rocco, R. et al. Exponential Escape Rate of Filamentary Incubation in Mott Spiking Neurons. Phys. Rev. Appl. 17, 024028 (2022).

Article ADS CAS Google Scholar

Valle, J. D., Ramírez, J. G., Rozenberg, M. J. & Schuller, I. K. Challenges in materials and devices for resistive-switching-based neuromorphic computing. J. Appl. Phys. 124, 211101 (2018).

Article ADS Google Scholar

Qiu, E. et al. Stochasticity in the synchronization of strongly coupled spiking oscillators. Appl. Phys. Lett. 122, 094105 (2023).

Article ADS CAS Google Scholar

Qiu, E. et al. Stochastic transition in synchronized spiking nanooscillators. Proc. Natl Acad. Sci. 120, e2303765120 (2023).

Article CAS PubMed PubMed Central Google Scholar

Shabalin, A. G. et al. Nanoscale Imaging and Control of Volatile and Non-Volatile Resistive Switching in VO2. Small 16, 2005439 (2020).

Article CAS Google Scholar

Wang, S. et al. Ultrafast photo-induced dynamics across the metal-insulator transition of VO2. Europhys. Lett. 118, 27005 (2017).

Article ADS Google Scholar

Butakov, N. A. et al. Broadband Electrically Tunable Dielectric Resonators Using Metal–Insulator Transitions. ACS Photonics 5, 4056–4060 (2018).

Article CAS Google Scholar

Dietze, S. H. et al. X-ray-induced persistent photoconductivity in vanadium dioxide. Phys. Rev. B 90, 165109 (2014).

Article ADS Google Scholar

Li, J. et al. Laser-induced selective local patterning of vanadium oxide phases. Adv. Compos. Hybrid. Mater. 8, 157 (2025).

Article CAS Google Scholar

Morin, F. J. Oxides Which Show a Metal-to-Insulator Transition at the Neel Temperature. Phys. Rev. Lett. 3, 34–36 (1959).

Article ADS CAS Google Scholar

Luo, R. et al. Spin Seebeck effect at low temperatures in the nominally paramagnetic insulating state of vanadium dioxide. Appl. Phys. Lett. 121, 102404 (2022).

Article ADS CAS Google Scholar

Morrison, V. R. et al. A photoinduced metal-like phase of monoclinic VO2 revealed by ultrafast electron diffraction. Science 346, 445–448 (2014).

Article ADS CAS PubMed Google Scholar

Wegkamp, D. et al. Instantaneous Band Gap Collapse in Photoexcited Monoclinic VO2 due to Photocarrier Doping. Phys. Rev. Lett. 113, 216401 (2014).

Article ADS PubMed Google Scholar

Lee, D. et al. Isostructural metal-insulator transition in VO2. Science 362, 1037–1040 (2018).

Article ADS CAS PubMed Google Scholar

Mondal, D. et al. Modulation-doping a correlated electron insulator. Nat. Commun. 14, 6210 (2023).

Article ADS CAS PubMed PubMed Central Google Scholar

Kalsi, T., Mitra, H., Roy, T. K., Godara, S. K. & Kumar, P. Comprehensive Analysis of Band Gap and Nanotwinning in Cd1–xMgxS QDs. Cryst. Growth Des. 20, 6699–6706 (2020).

Article CAS Google Scholar

Navarro, H. et al. A hybrid optoelectronic Mott insulator. Appl. Phys. Lett. 118, 141901 (2021).

Article ADS CAS Google Scholar

Adda, C. et al. An optoelectronic heterostructure for neuromorphic computing: CdS/V3O5. Appl. Phys. Lett. 121, 041901 (2022).

Article ADS CAS Google Scholar

Cheng, L., Xiang, Q., Liao, Y. & Zhang, H. CdS-Based photocatalysts. Energy Environ. Sci. 11, 1362–1391 (2018).

Article CAS Google Scholar

Navarro, H. et al. Light-Induced Decoupling of Electronic and Magnetic Properties in Manganites. Phys. Rev. Appl. 19, 044077 (2023).

Article ADS CAS Google Scholar

Navarro, H. et al. Disentangling transport mechanisms in a correlated oxide by photoinduced charge injection. Phys. Rev. Mater. 7, L123201 (2023).

Article ADS CAS Google Scholar

Cheng, S. et al. Operando characterization of conductive filaments during resistive switching in Mott VO2. Proc. Natl Acad. Sci. 118, e2013676118 (2021).

Article CAS PubMed PubMed Central Google Scholar

Cheng, S. et al. Inherent stochasticity during insulator-metal transition in VO2. Proc. Natl Acad. Sci. 118, e2105895118 (2021).

Article CAS PubMed PubMed Central Google Scholar

Park, J. et al. Direct Quantification of Heat Generation Due to Inelastic Scattering of Electrons Using a Nanocalorimeter. Adv. Sci. 8, 2002876 (2021).

Article CAS Google Scholar

Shiga, D. et al. Thickness dependence of electronic structures in VO2 ultrathin films: Suppression of the cooperative Mott-Peierls transition. Phys. Rev. B 102, 115114 (2020).

Article ADS CAS Google Scholar

Yang, M. et al. Suppression of Structural Phase Transition in VO2 by Epitaxial Strain in Vicinity of Metal-insulator Transition. Sci. Rep. 6, 23119 (2016).

Article ADS CAS PubMed PubMed Central Google Scholar

Fan, L. L. et al. Strain Dynamics of Ultrathin VO2 Film Grown on TiO2 (001) and the Associated Phase Transition Modulation. Nano Lett. 14, 4036–4043 (2014).

Article ADS CAS PubMed Google Scholar

Mecklenburg, M. et al. Nanoscale temperature mapping in operating microelectronic devices. Science 347, 629–632 (2015).

Article ADS CAS PubMed Google Scholar

Rez, P. & Muller, D. A. The Theory and Interpretation of Electron Energy Loss Near-Edge Fine Structure. Annu. Rev. Mater. Res. 38, 535–558 (2008).

Article ADS CAS Google Scholar

Pines, D. Elementary Excitations in Solids. (W.A. Benjamin, New York; 1963).

Reed, B. W., Chen, J. M., MacDonald, N. C., Silcox, J. & Bertsch, G. F. Fabrication and STEM/EELS measurements of nanometer-scale silicon tips and filaments. Phys. Rev. B 60, 5641–5652 (1999).

Article ADS CAS Google Scholar

Meng, Q. et al. Quantification of Charge Transfer at the Interfaces of Oxide Thin Films. J. Phys. Chem. A 123, 4632–4637 (2019).

Article CAS PubMed Google Scholar

Moreau, P., Brun, N., Walsh, C. A., Colliex, C. & Howie, A. Relativistic effects in electron-energy-loss-spectroscopy observations of the Si/SiO2 interface plasmon peak. Phys. Rev. B 56, 6774–6781 (1997).

Article ADS CAS Google Scholar

Garcia-Molina, R., Gras-Marti, A., Howie, A. & Ritchie, R. H. Retardation effects in the interaction of charged particle beams with bounded condensed media. J. Phys. C: Solid State Phys. 18, 5335–5345 (1985).

Article ADS CAS Google Scholar

Bolton, J. P. R. & Chen, M. Electron energy loss in multilayered slabs. Ultramicroscopy 60, 247–263 (1995).

Article CAS Google Scholar

Aruta, C. et al. Orbital occupation, atomic moments, and magnetic ordering at interfaces of manganite thin films. Phys. Rev. B 80, 014431 (2009).

Article ADS Google Scholar

Pinta, C. et al. Suppression of spin-state transition in epitaxially strained LaCoO3. Phys. Rev. B 78, 174402 (2008).

Article ADS Google Scholar

Merwe, J. H. v.d. On the Stresses and Energies associated with Inter-Crystalline Boundaries. Proc. Phys. Soc. Sect. A 63, 616–637 (1950).

Article ADS Google Scholar

Meng, Q., Wu, L. & Zhu, Y. Phonon scattering of interfacial strain field between dissimilar lattices. Phys. Rev. B 87, 064102 (2013).

Article ADS Google Scholar

Nakagawa, N., Hwang, H. Y. & Muller, D. A. Why some interfaces cannot be sharp. Nat. Mater. 5, 204–209 (2006).

Article ADS CAS Google Scholar

Vaz, C. A. F. et al. Origin of the Magnetoelectric Coupling Effect in Pb(Zr0.2Ti0.8)O3 / La0.8Sr0.2MnO3 Multiferroic Heterostructures. Phys. Rev. Lett. 104, 127202 (2010).

Article ADS CAS PubMed Google Scholar

Ngo, T. D. et al. Fermi-Level Pinning Free High-Performance 2D CMOS Inverter Fabricated with Van Der Waals Bottom Contacts. Adv. Electron. Mater. 7, 2001212 (2021).

Article CAS Google Scholar

Heo, J. et al. Clean thermal decomposition of tertiary-alkyl metal thiolates to metal sulfides: environmentally-benign, non-polar inks for solution-processed chalcopyrite solar cells. Sci. Rep. 6, 36608 (2016).

Article ADS CAS PubMed PubMed Central Google Scholar

Sohn, A. et al. Evolution of local work function in epitaxial VO2 thin films spanning the metal-insulator transition. Appl. Phys. Lett. 101, 191605 (2012).

Article ADS Google Scholar

Thees, M. et al. Imaging the itinerant-to-localized transmutation of electrons across the metal-to-insulator transition in V2O3. Sci. Adv. 7, eabj1164 (2021).

Article ADS CAS PubMed PubMed Central Google Scholar

Download references

We would like to acknowledge Tianxing D. Wang and Juan A. Hofer from University of California San Diego (UCSD) for the help with XRD measurements. The structural characterization, transport, TEM and collaboration between Brookhaven National Laboratory (BNL), CNRS and the UCSD were supported through an Energy Frontier Research Center program funded by the US Department of Energy (DOE), Office of Science, Basic Energy Sciences, under Grant DE-SC0019273. Electron microscopy work at BNL and the use of BNL’s Center for Functional Nanomaterials were supported by DOE-BES, the Division of Materials Science and Engineering, and the Division of Science User Facility, respectively, under Contract no. DE-SC0012704.

These authors contributed equally: Shaobo Cheng, Henry Navarro.

Department of Condensed Matter Physics and Materials Science, Brookhaven National Laboratory, Upton, NY, USA

Shaobo Cheng, Xing Li, Alexandre Pofelski, Qingping Meng, Mark P. M. Dean & Yimei Zhu

School of Physics and Microelectronics, Zhengzhou University, Zhengzhou, China

Shaobo Cheng & Xing Li

Department of Physics, Center for Advanced Nanoscience, University of California San Diego, La Jolla, CA, USA

Henry Navarro, Ali C. Basaran & Ivan K. Schuller

Department of Physics, Andrews University, Berrien Springs, MI, USA

Henry Navarro

Department of NanoEngineering, University of California San Diego, 9500 Gilman Dr, Mail Code 0448, La Jolla, CA, USA

Zishen Wang, Jasleen Kaur, Chi Chen & Shyue Ping Ong

Department of Materials Science & Engineering, University of California San Diego, 9500 Gilman Dr, Mail Code 0448, La Jolla, CA, USA

Jasleen Kaur

Center for Functional Nanomaterials, Brookhaven National Laboratory, Upton, NY, USA

Chenyu Zhou & Mingzhao Liu

Laboratoire de Physique des Solides, CNRS, Université Paris Saclay, Orsay Cedex, France

Marcelo Rozenberg

Search author on:PubMed Google Scholar

T.E.M. studies: S.B.C., X.L., A.P.; Thin film growth: H.N.; Theoretical calculations: Z.W., J.K., Q.M., C.C., M.R., S.P.O.; Methodology: S.B.C., H.N., C.Z., M.L., A.C.B., M.P.M.D.; Supervision: I.K.S., Y.Z.; Writing—Original draft: S.B.C., H.N., Z.W.; Writing—review &editing: M.R., I.K.S., Y.Z.

Correspondence to Yimei Zhu.

The authors declare no competing interests.

Nature Communications thanks Teruo Kanki, and the other, anonymous, reviewers for their contribution to the peer review of this work. A peer review file is available.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

Cheng, S., Navarro, H., Wang, Z. et al. Purely electronic insulator-metal transition in rutile VO2. Nat Commun 16, 5444 (2025). https://doi.org/10.1038/s41467-025-60243-0

Download citation

Received: 29 April 2024

Accepted: 19 May 2025

Published: 01 July 2025

DOI: https://doi.org/10.1038/s41467-025-60243-0

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Florida Grand Opera Renews Assistant Conductor Felix Jarrar's Contract for 2025-26 Season - OperaWire OperaWire

BUMAX Announces Strategic Partnership with Leading Turkish Fastener Distributor | impeller.net - The Online Pump Magazine

News

Purely electronic insulator-metal transition in rutile VO2 | Nature Communications