Open Access
Add to BibSonomyAbstract
While vanadium dioxide (VO2) is one of the most extensively studied highly correlated materials, there are intriguing similarities and differences worth exploring in another highly correlated oxide, niobium dioxide (NbO2). Both materials exhibit a thermally-induced first-order insulator-metal transition at a material-dependent critical temperature, which is considerably higher in NbO2 than in VO2 – approximately 1080 K and 340 K in bulk, respectively. This transition, evidenced by up to 6 orders of magnitude change in DC and optical conductivities, can also be induced in VO2 via photo-doping on a sub-picosecond timescale. Here, we present the first ultrafast pump-probe studies on the optically-induced transition of NbO2 thin films and the comparison with similar VO2 films. It is observed that NbO2 films transition faster and exhibit significantly faster recovery time than VO2 films of similar thickness and microstructure, showcasing that NbO2 is a promising material for next generation high-speed optoelectronic devices.
© 2016 Optical Society of America
1. Introduction
There is a great deal of interest in light-material interactions for applications such as ultrafast optical switches for telecommunication [1], optical limiters [2,3], and other optoelectronic devices [4]. Because constrained geometries have profound effects on many physical properties in materials, thin films are a popular choice for these types of applications. In particular, vanadium dioxide (VO2) thin films used as optical switches have garnered recent attention, due in part to their large change in optical conductivity – 2 orders of magnitude [5,6] – and even larger change in DC conductivity – at least 4 to 5 orders of magnitude [7,8] increase in DC. Additionally, in 2004, Cavalleri et al. [9] demonstrated that the light-induced transition from the insulating to the metallic phase, initiated by hole photo-doping into the valence band with a visible laser, exhibited extremely fast switching speeds (<100 fs) [10] at very low switching energies (on the order of 1 pJ/μm2) [11]. However, it was also observed that such films typically exhibited longer recovery time back to the insulating state (>20 ns) [12], hindering applications requiring faster ON/OFF/ON transitions.
A related material, niobium dioxide (NbO2), may offer a solution. Both materials undergo an insulator-metal transition (IMT) at a material-dependent critical temperature, 340 K in VO2 and 1080 K in NbO2. This large difference in critical temperature may indicate advantageous differences in NbO22, but this material has not been studied as extensively. There is well-documented work on the electrically- and thermally- induced IMT in NbO2 [13–16], but, while some work has been done recently on the electronic and optical properties [17,18], the optically-induced IMT has never been demonstrated in NbO2. In what follows, we describe our ultrafast pump-probe studies on NbO2 thin films, which show that not only can the IMT in NbO2 be induced optically, but also that the fluence needed to drive the transition at room temperature is lower and the recovery time is much faster than for VO2, thus enabling additional possibilities for ultrafast optical switching applications.
2. Experimental details
2.1 Sample structure
The thermally-induced IMT in NbO2, like VO2, is accompanied by a structural change in the lattice from a monoclinic structure at room temperature to a high-temperature rutile structure [19,20]. The nature of this transition is the subject of significant study in both materials [21].
Both the NbO2 and VO2 films studied were grown on c-plane sapphire substrates [22]. The NbO2 sample studied was 212 nm thick and covered by a 5.65 nm capping layer of AlOx to prevent degradation [22]; the VO2 film considered was of comparable thickness (101 nm) with no capping layer, since it was previously demonstrated that this material is robust against degradation. High-angle 2θ-ω x-ray diffraction (XRD) scans on the samples, shown in Fig. 1, indicate good crystalline structure in the room-temperature monoclinic phase for both films [17,23].
Fig. 1 High-angle 2θ-ω scans of the NbO2 (top) and VO2 (bottom) samples studied. The dashed line indicates the position of the bulk substrate reflection, centered at 2θ = 41.685°. The NbO2(440) peak is centered at 2θ = 37.223°, while the VO2(020) peak is centered at 2θ = 40.069°.
2.2 Experimental setup
The ultrafast pump-probe setup used for this experiment is shown in Fig. 2. The Ti:sapphire laser system emitted pulses of ~120 fs at a center wavelength of 800 nm and a repetition rate of 1 kHz. The laser output was attenuated and split into a pump beam and a weaker probe beam using an 80/20 beam splitter. The pump beam passed through an optical chopper with a frequency of 500 Hz and was further attenuated using a variable neutral-density filter in order to control the fluence before focusing to an ~80 μm diameter spot on the sample. The probe beam reflects off a rooftop mirror mounted on a variable delay stage to control the relative delay between the pump and probe beams by up to 4 ns. Following the delay stage, the probe beam was then heavily attenuated to a fluence far below that of the pump beam (3-5 orders of magnitude weaker). Also following the delay stage, the polarization of the probe beam was rotated 90°, thus perpendicular to the polarization of the pump beam; this allowed for rejection of scattered pump beam from the detector. The probe beam was focused onto the sample on the same spot as the pump beam but to a smaller diameter (~40 μm) using a shorter focal length lens to ensure probing of only the central region of the pumped region where the effects of the optical excitation can be considered one-dimensional in the direction perpendicular to the surface.
The probe beam reflected off the sample, passed through a polarizer set to pass only the probe polarization, and was focused onto a silicon photodetector. The photodetector signal was input to a boxcar integrator triggered by a 1 kHz reference from the pulse controller for the Legend. Using the boxcar integrator, the signal of the photodetector from each probe pulse was integrated over a window of 230 ns. The “last sample” output of the boxcar, which is a scaled DC voltage of the integrated signal, is updated at the 1 kHz repetition rate of the laser and provides the input to a lock-in amplifier. This measurement scheme eliminates the dead time between laser pulses and produces a very low noise floor and thus an improved signal-to-noise ratio. All data is reported as ΔR/R, computed by dividing the change in the reflected probe beam (magnitude of the lock-in signal) by the baseline reflectivity of the sample (lock-in signal from the probe beam only).
2.3 Measurements
A LabView® program was used to control the delay stage position and acquire data from both the boxcar integrator and the lock-in amplifier for each position of the stage. The program moves the stage in steps as small as 1.25 μm, corresponding to a relative delay of 8.34 fs, well below the time resolution of the measurement scheme. The smallest steps used here were 2 μm (13.3 fs of delay).
Pump-probe measurements were made on both the NbO2 and VO2 samples for a relative delay of 0-2.87 ns with a fixed probe fluence of 3.2 μJ/cm2. The pump fluence was varied between scans over a range from 2.2 mJ/cm2 to 422 mJ/cm2 for the NbO2 sample and 17.5 mJ/cm2 to 422 mJ/cm2 for the VO2. Higher fluences were attempted but the scans showed indications of sample damage and thus are not included here. At fluences lower than 2.2 mJ/cm2, no transition was observed in the NbO2 film; in the VO2 film, no transition was observed at fluences lower than 17.5 mJ/cm2, showing that a higher minimum fluence is needed to drive the transition in VO2 than in NbO2.
3. Results and discussion
3.1 Experimental results
The purely electronic component of the IMT that is photoinduced by an ultrashort optical excitation has been detected in VO2 using time-resolved terahertz spectroscopy [24], IR transmittance measurements [25], and photoelectron spectroscopy (TR-ARPES) [26], as well as in the thermally-induced IMT using electron and photoelectron spectroscopy and microscopy [27]. The detection of this metal-like monoclinic phase, indicative of an electronic IMT, enables isolation of the purely electronic response (i.e. Mott-Hubbard) from the response due to the structural IMT (i.e. Peierls) in VO2. This same electronic component of the photoinduced IMT has been detected for the first time in NbO2, as shown in Fig. 3.
Fig. 3 Pump-probe measurements and fits for the lowest fluence at which the IMT was seen in each material. (a) Fitted data showing the initial response and recovery of the film. Data are plotted as disconnected points, while fits are solid lines. (b) Full-track data.
It is important to note that the threshold for detecting the optically-induced electronic IMT, i.e. the fluence below which no change in reflectivity was seen, in NbO2 is lower (2.2 mJ/cm2) than in VO2 (17.5 mJ/cm2). Additionally, the initial fast transient response of the NbO2 is larger than that of the VO2 and remains so until a pump fluence of 70.0 mJ/cm2 (Fig. 4), at which point the thermal barrier to the structural transition in VO2 is overcome and the change in reflectance is dominated by the development of the metallic rutile phase. This is consistent with previous studies of the ultrafast dynamics of VO2 thin films that show the response of the film to varying fluences [24,28,29], as well as reports of chaotic behavior at the onset of the structural IMT [30], which can be seen in the non-uniform change in the VO2 signal at times greater than 5 ps.
Fig. 4 Pump-probe measurements at a pump fluence of 70.0 mJ/cm2, when the structural transition begins in the VO2 film, causing the signal to become larger than that of the NbO2 film.
Four NbO2 scans covering a range of pump fluences are presented in Fig. 5 with the corresponding fit parameters listed in Table 1 and discussed below. In Fig. 6, a scan at a fluence that fully transitions the VO2 is presented; the associated fit parameters are also listed in Table 1. It is important to note that in VO2, the structural IMT completely dominates the response of the film such that it is not possible to detect either purely electronic processes or e-p scattering processes (t1 and t2 respectively, described below). Figure 7 shows normalized scans of both the NbO2 and VO2 films at 422 mJ/cm2, just below the damage threshold for both films. It is clear that the NbO2 has a faster initial transient response (tp), or “turn on” of the film’s response, than VO2. Further, the NbO2 film exhibits a clear purely electronic response (t1) even at the highest fluence, while the response of the VO2 is dominated by the optically-driven structural IMT, rendering the purely electronic IMT practically undetectable. This electronic response in the NbO2 film shows a recovery of roughly 70% within ~5 ps at all fluences before the optically-driven structural IMT becomes evident.
Fig. 5 Scans of NbO2 with pump fluences ranging from 8.8 mJ/cm2 to 422 mJ/cm2. (a) Fitted data showing the initial response and recovery of the film. Data are plotted as disconnected points, while fits are solid lines. (b) Full-track data.
Table 1. Fit parameters for the scans shown in Figs. 3, 5, and 6. The two highlighted columns correspond to the plots shown in Fig. 7.
Fig. 6 Scan of fully-transitioned VO2. (a) Fitted data showing the initial response of the film. Data are plotted as disconnected points, while fits are solid lines. (b) Full-track data.
Fig. 7 Normalized scans of NbO2 and VO2 at 422 mJ/cm2, the highest fluence achieved without damaging the samples. (a) Fitted data showing the initial response of the films. Data are plotted as disconnected points, while fits are solid lines. (b) Full-track data.
While the electronic IMT has been detected at very low fluences in the optical response of VO2, the results presented here show that a purely electronic IMT is more readily isolated from the structural IMT in NbO2 and can be detected optically. This is likely due to the higher thermal barrier to the structural IMT in NbO2, which requires twice the fluence required in VO2 to initiate optically (140 mJ/cm2 in NbO2, as discussed below, compared to 70 mJ/cm2 in VO2, as stated above).
3.2 Data fitting and analysis
Using an analysis typical of pump-probe studies [31], experimental time-resolved optical reflectivity data were analyzed using a three-time-constant fit given by:
where Erf is the standard error function, used here to model the initial fast transient response of the film. The time designated by toff represents the delay from the start of the scan to the initial transient response, while the time constant tp is related to both the laser pulse-width and the rate of the initial fast transient response of the film to the ultrashort laser pulse. While all of the parameters are initially free, we find that both of these values are constant for a given sample; the value of toff can vary due to small differences in the position of the pump and probe beam from sample to sample, and while the laser pulse-width is held constant, the initial fast transient response can vary between samples, producing different fits for tp. The time constants t1, t2, and t3 are related to the scattering and relaxation processes following absorption of the ultrashort laser pulse [32,33] and are described in more detail below. The constant A0 sets the overall scaling of the fit to the experimental data and increases with increasing fluence, while A1, A2, and A3 allow for scaling of the effects from the different processes described below.At 800 nm (1.55 eV), the laser pulse energy is absorbed by the electrons and produces a nonequilibrium between the effective temperature of the electrons and the lattice (phonons) of the material. The time constant t1 describes scattering processes only within the electronic system (i.e. e-e scattering) and results in a redistribution of energy within the electron system on a timescale of hundreds of femtoseconds. As is the case for toff and tp, the time constant t1 and its scaling factor A1 are constant for a given sample, since the relevant processes do not depend, or are weakly dependent, on the amount of energy deposited by the pump pulse. The exception to this is for the measurements on VO2 at the highest fluence (Fig. 6) where the structural IMT completely dominates the response. On a slightly longer timescale, the time constant t2 describes scattering between the electrons and phonons (e-p scattering) and the resulting transfer of energy to the lattice. As the fluence increases, both t2 and the associated scaling factor A2 decrease. These effects are likely caused by the new phonon spectrum due to the change of the lattice structure as the energy density becomes high enough to begin driving the structural IMT. The third time constant, t3, describes slower processes that, at lower fluences, are dominated by the recovery of the excited film, with most of the NbO2 remaining in the monoclinic structure and the hot carriers relaxing back to the nominal distribution of the room temperature insulating phase. It is likely that the incomplete recovery of the transient reflectance response at lower fluences is indicative of microscopic regions of the metallic rutile phase of NbO2. The formation and growth of microscopic regions of rutile phase within the insulating matrix has been studied previously in VO2 by nearfield [34,35] and farfield [36] optical measurements, and is expected to also occur in the structural IMT of NbO2. At fluences of 140 mJ/cm2 and above, A2 goes to zero and t3 becomes negative. While mathematically this would indicate that the A3 term diverges with increased delay time, physically this indicates the beginning of growth of the metallic rutile phase, i.e. that sufficient energy is transferred to the phonon system to drive the more complete structural transformation associated with the thermally-driven structural IMT. Note that each of the free parameters (t2, A2, t3, A3) was varied programmatically in an iterative process to minimize the RMS of the least square fit as listed in Table 1. To test the sensitivity of the fit, the RMS was computed for a ± 10% change from the optimal for each free parameter. The results of this sensitivity study are presented in Table 2, and indicate the reliability of the fit.
Table 2. Results of the sensitivity study on Eq. (1), showing the percent change in the RMS value of the fit after changing the given parameter by ± 10%.
4. Conclusion
The ultrafast pump-probe studies presented here show for the first time that, as in VO2, the IMT of NbO2 can be induced optically. More important is the existence of a clear electronic response in the NbO2 film that can be optically detected and that recovers within picoseconds at all fluences, showing a strong similarity to the recently-discovered monoclinic metallic phase in VO2 [24–27]. NbO2 films, however, show a significantly faster electronic recovery than VO2 films, indicating that thin films of NbO2 may be better suited for all-optical ultrafast switching applications. It is also important to note that the IMT of the NbO2 films can be optically excited with a lower fluence (2.2 mJ/cm2) than VO2 (17.5 mJ/cm2). This lower minimum fluence further indicates that the optical response of NbO2 can be driven more efficiently in ultrafast switching applications where the higher fluence needed to transition the VO2 could be damaging to other components. Our results show that the ultrafast electronic response in NbO2 was evident all the way up to the damage threshold, indicating that this material could provide robust ultrafast switching over a large dynamic range of operation. In addition, the slightly faster initial response of the NbO2 films could provide an advantage over VO2 in devices requiring somewhat faster OFF/ON switching times. Of further interest is the combination of NbO2 and VO2 thin films in optoelectronic devices – the NbO2 will switch off in less than 10 ps, while the VO2 will still be metallic, not switching off for ~20 ps, offering the possibility of combining them in double-function optical switches or memory devices.
Funding
National Science Foundation (NSF) (DMR-1006013); Virginia Microelectronics Consortium (VMEC).
Acknowledgments
The authors would like to thank Dr. Irina Novikova for her assistance with the ultrafast pump-probe setup, as well as Dr. Elizabeth Radue and Scott Madaras for their assistance on technical aspects of the measurements.
References and links
1. W. Yoshiki and T. Tanabe, “All-optical switching using Kerr effect in a silica toroid microcavity,” Opt. Express 22(20), 24332–24341 (2014). [CrossRef] [PubMed]
2. R. C. Hollins, “Materials for optical limiters,” Curr. Opin. Solid State Mater. Sci. 4(2), 189–196 (1999). [CrossRef]
3. S. Perumbilavil, P. Sankar, T. Priya Rose, and R. Philip, “White light Z-scan measurements of ultrafast optical nonlinearity in reduced graphene oxide nanosheets in the 400–700 nm region,” Appl. Phys. Lett. 107(5), 051104 (2015). [CrossRef]
4. G. Steinmeyer, “A review of ultrafast optics and optoelectronics,” J. Opt. A, Pure Appl. Opt. 5(1), R1–R15 (2003). [CrossRef]
5. H. W. Verleur, A. S. Barker Jr, and C. N. Berglund, “Optical properties of VO2 between 0.25 and 5 eV,” Phys. Rev. 172(3), 788–798 (1968). [CrossRef]
6. D. I. Khomskii, Transtion Metal Compounds (Cambridge University 2014).
7. L. Wang, I. Novikova, J. M. Klopf, S. Madaras, G. P. Williams, E. Madaras, J. Lu, S. A. Wolf, and R. A. Lukaszew, “Distinct length scales in the VO2 metal–insulator transition revealed by bi-chromatic optical probing,” Adv. Opt. Mat. 2(1), 30–33 (2014). [CrossRef]
8. A. Zylbersztejn and N. F. Mott, “Metal-insulator transition in vanadium dioxide,” Phys. Rev. B 11(11), 4383–4395 (1975). [CrossRef]
9. A. Cavalleri, Th. Dekorsy, H. H. W. Chong, J. C. Kieffer, and R. W. Schoenlein, “Evidence for a structurally-driven insulator-to-metal transition in VO2: A view from the ultrafast timescale,” Phys. Rev. B 70, 161102 (2004).
10. S. Lysenko, A. Rúa, V. Vikhnin, F. Fernández, and H. Liu, “Insulator-to-metal phase transition and recovery processes in VO2 thin films after femtosecond laser excitation,” Phys. Rev. B 76(3), 035104 (2007). [CrossRef]
11. M. Rini, Z. Hao, R. W. Schoenlein, C. Giannetti, F. Parmigiani, S. Fourmaux, J. C. Kieffer, A. Fujimori, M. Onoda, S. Wall, and A. Cavalleri, “Optical switching in VO2 films by below-gap excitation,” Appl. Phys. Lett. 92(18), 181904 (2008). [CrossRef]
12. M. Rodriguez-Vega, M. T. Simons, E. Radue, S. Kittiwatanakul, J. Lu, S. A. Wolf, R. A. Lukaszew, I. Novikova, and E. Rossi, “Effect of inhomogeneties and substrate on the dynamics of the metal-insulator transition in VO2 thin films,” Phys. Rev. B 92(11), 115420 (2015). [CrossRef]
13. G. C. Vezzoli, “Recovery curve for threshold‐switching NbO2,” J. Appl. Phys. 50(10), 6390 (1979). [CrossRef]
14. G. C. Vezzoli, “On‐state decay in NbO2: Relationship to recombination and to nonlinear I‐V,” J. Appl. Phys. 51(5), 2693 (1980). [CrossRef]
15. S. K. Nandi, X. Liu, D. K. Venkatachalam, and R. G. Elliman, “Threshold current reduction for the metal–insulator transition in NbO2−x-selector devices: the effect of ReRAM integration,” J. Phys. D Appl. Phys. 48(19), 195105 (2015). [CrossRef]
16. T. Joshi, T. R. Senty, P. Borisov, A. D. Bristow, and D. Lederman, “Preparation, characterization, and electrical properties of epitaxial NbO2 thin film lateral devices,” J. Phys. D Appl. Phys. 48(33), 335308 (2015). [CrossRef]
17. A. B. Posadas, A. O’Hara, S. Rangan, R. A. Bartynski, and A. A. Demkov, “Band gap of epitaxial in-plane-dimerized single-phase NbO2 films,” Appl. Phys. Lett. 104(9), 092901 (2014). [CrossRef]
18. A. O’Hara, T. N. Nunley, A. B. Posadas, S. Zollner, and A. A. Demkov, “Electronic and optical properties of NbO2,” J. Appl. Phys. 116(21), 213705 (2014). [CrossRef]
19. V. Eyert, “The metal-insulator transition of NbO2: An embedded Peierls instability,” Europhys. Lett. 58(6), 851–856 (2002). [CrossRef]
20. A. A. Bolzan, C. Fong, B. J. Kennedy, and C. J. Howard, “A powder neutron diffraction study of semiconducting and metallic niobium dioxide,” J. Solid State Chem. 113(1), 9–14 (1994). [CrossRef]
21. N. B. Aetukuri, A. X. Gray, M. Drouard, M. Cossale, L. Gao, A. H. Reid, R. Kukreja, H. Ohldag, C. A. Jenkins, E. Arenholz, K. P. Roche, H. A. Dürr, M. G. Samant, and S. S. P. Parkin, “Control of the metal–insulator transition in vanadium dioxide by modifying orbital occupancy,” Nat. Phys. 9(10), 661–666 (2013). [CrossRef]
22. Y. Wang, R. B. Comes, S. Kittiwatanakul, S. A. Wolf, and J. Lu, “Epitaxial niobium dioxide thin films by reactive-biased target ion beam deposition,” J. Vac. Sci. Technol. A 33(2), 021516 (2015). [CrossRef]
23. L. Wang, E. Radue, S. Kittiwatanakul, C. Clavero, J. Lu, S. A. Wolf, I. Novikova, and R. A. Lukaszew, “Surface plasmon polaritons in VO2 thin films for tunable low-loss plasmonic applications,” Opt. Lett. 37(20), 4335–4337 (2012). [CrossRef] [PubMed]
24. T. L. Cocker, L. V. Titova, S. Fourmaux, G. Holloway, H.-C. Bandulet, D. Brassard, J.-C. Kieffer, M. A. El Khakani, and F. A. Hegmann, “Phase diagram of the ultrafast photoinduced insulator-metal transition in vanadium dioxide,” Phys. Rev. B 85(15), 155120 (2012). [CrossRef]
25. V. R. Morrison, R. P. Chatelain, K. L. Tiwari, A. Hendaoui, A. Bruhács, M. Chaker, and B. J. Siwick, “A photoinduced metal-like phase of monoclinic VO2 revealed by ultrafast electron diffraction,” Science 346(6208), 445–448 (2014). [CrossRef] [PubMed]
26. D. Wegkamp, M. Herzog, L. Xian, M. Gatti, P. Cudazzo, C. L. McGahan, R. E. Marvel, R. F. Haglund Jr, A. Rubio, M. Wolf, and J. Stähler, “Instantaneous band gap collapse in photoexcited monoclinic VO2 due to photocarrier doping,” Phys. Rev. Lett. 113(21), 216401 (2014). [CrossRef] [PubMed]
27. J. Laverock, S. Kittiwatanakul, A. A. Zakharov, Y. R. Niu, B. Chen, S. A. Wolf, J. W. Lu, and K. E. Smith, “Direct observation of decoupled structural and electronic transitions and an ambient pressure monocliniclike metallic phase of VO2.,” Phys. Rev. Lett. 113(21), 216402 (2014). [CrossRef] [PubMed]
28. E. Radue, L. Wang, S. Kittiwatanakul, J. Lu, S. A. Wolf, E. Rossi, R. A. Lukaszew, and I. Novikova, “Substrate-induced microstructure effects on the dynamics of the photo-induced metal–insulator transition in VO2 thin films,” J. Opt. 17(2), 025503 (2015). [CrossRef]
29. D. J. Hilton, R. P. Prasankumar, S. Fourmaux, A. Cavalleri, D. Brassard, M. A. El Khakani, J. C. Kieffer, A. J. Taylor, and R. D. Averitt, “Enhanced photosusceptibility near Tc for the light-induced insulator-to-metal phase transition in vanadium dioxide,” Phys. Rev. Lett. 99(22), 226401 (2007). [CrossRef] [PubMed]
30. L. A. L. de Almeida, G. S. Deep, A. M. N. Lima, and H. Neff, “Thermal dynamics of VO2 films within the metal–insulator transition: Evidence for chaos near percolation threshold,” Appl. Phys. Lett. 77(26), 4365–4367 (2000). [CrossRef]
31. A. Cavalleri, C. Tóth, C. W. Siders, J. A. Squier, F. Ráksi, P. Forget, and J. C. Kieffer, “Femtosecond structural dynamics in VO2 during an ultrafast solid-solid phase transition,” Phys. Rev. Lett. 87(23), 237401 (2001). [CrossRef] [PubMed]
32. J. Shah, Hot Carriers in Semiconductor Nanostructures, Physics and Applications (Academic 1992).
33. J. M. Klopf and P. M. Norris, “Subpicosecond observation of photoexcited carrier thermalization and relaxation in InP-based films,” Int. J. Thermophys. 26(1), 127–140 (2005). [CrossRef]
34. M. M. Qazilbash, Z. Q. Li, V. Podzorov, M. Brehm, F. Keilmann, B. G. Chae, H. T. Kim, and D. N. Basov, “Electrostatic modification of infrared response in gated structures based on VO2,” Appl. Phys. Lett. 92(24), 241906 (2008). [CrossRef]
35. M. M. Qazilbash, M. Brehm, B.-G. Chae, P.-C. Ho, G. O. Andreev, B.-J. Kim, S. J. Yun, A. V. Balatsky, M. B. Maple, F. Keilmann, H.-T. Kim, and D. N. Basov, “Mott transition in VO2 revealed by infrared spectroscopy and nano-imaging,” Science 318(5857), 1750–1753 (2007). [CrossRef] [PubMed]
36. L. Wang, I. Novikova, J. M. Klopf, S. Madaras, G. P. Williams, E. Madaras, J. Lu, S. A. Wolf, and R. A. Lukaszew, “Distinct length scales in the VO2 metal–insulator transition revealed by bi-chromatic optical probing,” Adv. Opt. Mater. 2(1), 30–33 (2014). [CrossRef]
