Hall voltage reversal and structural phase transition in VO 2 thin films
Cite as: Appl. Phys. Lett.116, 082106 (2020);doi: 10.1063/1.5143548 Submitted: 24 December 2019
.
Accepted: 16 February 2020.
Published Online: 28 February 2020
J. J.Feng,1,2C. F.Li,1C. L.Luo,2H.Yang,2A. H.Zhang,2Q.Li,2M.Guo,2D.Gao,2Z.Fan,2 D. Y.Chen,2M. H.Qin,2 M.Zeng,2 X. S.Gao,2 Y.Lin,3X. B.Lu,2,a)and J.–M.Liu1,2
AFFILIATIONS
1Laboratory of Solid State Microstructures, Nanjing University, Nanjing 210093, China
2Guangdong Provincial Laboratory of Quantum Engineering and Quantum Materials, and Institute for Advanced Materials, South China Academy of Advanced Optoelectronics, South China Normal University, Guangzhou 510006, China
3State Key Laboratory of Electronic Thin Films and Integrated Devices, University of Electronic Science and Technology of China, Chengdu, Sichuan 610054, China
a)Author to whom correspondence should be addressed:luxubing@m.scnu.edu.cn
ABSTRACT
In this work, we investigated the nanoscale conduction and charge transport characteristics of epitaxial VO2thin films around the metal- insulator transition (MIT) using the Hall transport measurement and conduction atomic force microscopy. Unlike the conventional oxides, the VO2thin films show unique transport characteristics. First, the dominant carrier type shows a critical change from electron to hole dur- ing the MIT sequence (cooling sequence) or from hole to electron during the reverse MIT sequence (heating sequence). Second, the carrier density measured during the MIT sequence is higher than that measured during the reverse MIT sequence, evidenced with a clear thermal hysteresis. Third, the volume fraction (area percentage) of the nanoscale high-conduction phase also shows a thermal hysteresis, evidenced with a larger volume fraction of the high-conduction region in the MIT sequence than the reverse MIT sequence. The first-principles calcula- tions indicate that the dominant carrier is the hole in the monoclinic phase, while it is the electron in the rutile phase, suggesting that the unique charge transport characteristics are attributed to the structural phase transition. Our work provides a deep insight into the nanoscale conduction and charge transports in VO2thin films.
Published under license by AIP Publishing.https://doi.org/10.1063/1.5143548
Mott insulators are the strongly correlated electronic materials receiving continuous attention in the past few decades.1–7The metal- insulator transition (MIT) as the core character of these materials has been highly concerned, while the associated mechanisms can be very fascinating, bringing us to challenges due to the strong electron corre- lation characteristics. Nevertheless, challenges also exist owing to the coupling of complicated structural ingredients, such as structural defects1–3and structural phase transition.6,7It is known that VO2is a typical Mott insulator that offers the d-band electrons with strong electron-electron correlations.8–10The MIT character of VO2has been particularly attractive not only because it occurs around room temper- ature and offers possible application potential but also due to its role as a core platform for studying strongly correlated electron physics and materials.
In the past few decades, two controversial mechanisms (following the Peierls scenario11,12and the Mott scenario13,14) have been pro- posed to explain the MIT effect of VO2. The Peierls mechanism indi- cates that a change in the electronic structure will be induced by the
structural phase transition (SPT) from the monoclinic phase (insulat- ing phase) to the rutile phase (metallic phase) near the temperature of T¼Tc341 K, whereTcis called the MIT point. The chemical bond- ing in VO2includes the hybridization of O 2porbitals with V 3dorbi- tals.15,16 The band structure changes with the SPT, leading to the variation of the Fermi level.16–19On the other hand, the Mott mecha- nism indicates that the MIT is of the first-order feature and can be attributed to a change in the Coulomb energy between the outermost electrons.20–22This means that the first-order MIT may be owing to the correlation effect instead of structural transition.23Recently, the Peierls-Mott synergistic phase transition mechanism was proposed and extensively discussed. It was believed that the SPT would occur earlier than the reverse MIT and the filling of thed-band would be later than the rearrangement of atoms in the high-Trutile phase, if photons are injected into VO2in the lowTmonoclinic phase.24It was also suggested that the SPT and MIT are basically simultaneous by probing the torsion angle variation of the nearest V-V chain coordi- nates.25Therefore, it seems that the first-order MIT in VO2emerges as
a consequence of both the electron-lattice interaction and electron- electron interaction,25exhibiting complexity to be further exploration.
One of the recent discoveries on VO2is the coexistence of metal and insulating regions evidenced by using near-field scanning micro- wave microscopy.13,26The area fraction (or volume fraction in the three-dimensional microstructure) of the nanoscale high-conduction and low-conduction phases shows a remarkable T-dependence.27 Nevertheless, it is still unclear what is responsible for suchT-dependen- ces. An investigation of this issue would be interesting for fundamental physics and nanoscale device applications. In the earliest times, Mott believed that the MIT is driven by the change in carrier density, the classical scenario for MIT.28,29When the carrier density is lower than a critical value, a semiconducting or insulating state is favored due to the on-site Coulomb interaction. Otherwise, a high-conductive or even metallic conduction phase ensues.28,29Given this scenario, more com- plexity would be expected if one further considers the microstructural and electronic inhomogeneities such as electronic phase separation that is intrinsically associated with transition metal oxides.5,30,31This is the reason why the transport behavior of VO2, which has been investi- gated for many years, remains to be a hot topic to date.
Besides, because the MIT in VO2is of first-order, one expects a remarkable hysteresis effect in terms of the transport behavior.
Therefore, the physics of electrical transport could be path-dependent if one checks the cooling and heating sequences passing through the MIT point separately. At least there are two issues deserving attention here. First, how does the transport behavior change if the low-T insulating phase and high-Tconductive phase have different dominant carriers? In fact, earlier calculations proposed some hint on this issue and a clarification would be needed.23Second, due to the SPT along with the MIT, one wonders if the two phase coexistence around the MIT would add more complexity to the transport behavior. Earlier investigations focused on the macroscopic transport of VO2, and these issues have been less touched.26,32In our previous work, the one-to-one correspondence between the macroscopic transport and the microscopic transport mapping using conductive atomic force microscopy (CAFM) was established.27 It provides a platform on which we can address the two issues mentioned above.
In this work, we investigate in detail the electrical transport behaviors of epitaxial VO2thin films upon the heating-cooling cycle across the MIT pointTc, including the longitudinal and transverse (Hall) transport behaviors as well as the CAFM mapping of the micro- structures. It was found that the structural phase transition (SPT) and MIT occur simultaneously over a broadT-range, and surprisingly, the dominant carriers do change from hole to electron or vice versa. The nanoscale two-phase coexistence during the heating-cooling cycle allows the unusual transport behaviors.
In our experiment, the VO2thin films were epitaxially deposited on (0001) sapphire (a-Al2O3) substrates using chemical solution depo- sition and the details of the fabrication process can be found in earlier work.33The micro-structures of the VO2thin films have been investi- gated by X-ray diffraction (XRD) and atomic force microscopy (AFM), and the results can be found in Fig. S1 in thesupplementary material. For the transport measurement, the samples were carefully fabricated and the longitudinal electrical resistivityqand Hall voltage VHwere measured using a standard Hall bar with the physical prop- erty measurement system (PPMS9, Quantum Design).Figures 1(a) and1(b)show a schematic configuration of the Hall measurement and
the actual photo of the fabricated Hall bar used in this work, respec- tively. The nanoscale conductivity imaging was carried out using the CAFM technique (CAFM apparatus, Asylum Research Cypher). The CAFM tips used in this work have a curvature radius less than 25 nm, which can guarantee the nanoscaledc-impedance measurement. The as-measured VO2 thin films are 40 nm in thickness determined using cross-sectional transmission electron microscopy.
Figure 1(c)depicts theq-Tcurves of the VO2thin film in one heating-cooling cycle. During the heating process, the resistivity decreases with increasing T and it begins to drop sharply around T333 K. Up toT340 K, the resistivity decreases by three orders of magnitude, marking the reverse MIT sequence. BeyondT340 K, a further decrease inqcan be seen with increasingT. In the cooling sequence, the resistivity shows an increase with decreasingT. The MIT occurs in-between 330 and 340 K, evidencing a marked hysteresis with aT-window of10 K. These results are well consistent with earlier reports.14,26
The Hall measurements were carried out to check the carrier type and carrier density with varyingTvalues across the MIT point.
The geometry for the Hall measurement is illustrated schematically in Figs. 2(a) and 2(b)with the vertically applied magnetic field of B 3.0 T. Upon heating-up from the low-Tside where the sample is in the monoclinic phase, as shown inFig. 2(c), the observed Hall voltage is negative in the low T end and the voltage magnitude decreases gradually with increasing T before a rapid decrease back close to zero atT332 K. A zoom-in of the data unveils surprisingly the sign reversal of the Hall voltage from the negative sign to the posi- tive one, occurring atT¼Trh334 K, as shown clearly in the inset of Fig. 2(c). Here, the subscript “rh” withTrhrefers to the temperature for the sign reversal during the heating sequence. Further increasingT beyondTrh, the Hall voltage remains positive although its magnitude is much smaller than the Hall voltage far belowTrh. This is reasonable since the low-Trange belowTrh334 K is dominated with the mono- clinic insulating phase and the Hall voltage must be much larger than that in the high-Trange in which the dominant phase is the metallic rutile phase.
FIG. 1.(a) A schematic diagram of the sample structure for the Hall effect mea- surement, (b) actual photo of the Hall bar used in the Hall effect measurement, and (c) resistivity (q)—temperature (T) characteristics of the VO2thin film.
The sign reversal atTrhsuggests the change of dominant carrier type from hole to electron. As a complementary evidence, we also check the variation of Hall voltage in response to the cooling sequence from the high-Tend (>360 K) where the sample is occupied with the metallic rutile phase. The results are plotted inFig. 2(d). Again, one sees the sign reversal of the Hall voltage from the positive sign to the negative one atTrc324 K, indicating the change of dominant car- riers from electron to hole. It is seen that the Hall voltage vsTcurves for the heating and cooling sequences are very similar althoughTrh
andTrcare different, suggesting that the measured data are reliable and the MIT and reverse MIT are nicely reproducible. It should be noticed that kinks were observed in the Hall voltage curves at tempera- tures aroundTrhandTrc, which are assumed to be due to an inhomo- geneous composite of VO2with mixed monoclinic and rutile phases.
These kink phenomena resulting from phase coexistence have also been found by magnetic measurements in VO2 based synthetic multilayers.34,35
To understand the sign reversal of the Hall voltage in both the heating and cooling sequences, a reasonable argument goes obviously to the difference in the dominant carrier type between the low-Tinsu- lating and high-T conductive phase. This issue can be qualitatively addressed by looking at the band structures of the two phases. As a reasonable approach, the first-principles calculations of the band struc- tures for the insulating monoclinic phase and conductive rutile phase become the first priority to see the possible carrier type each. The cal- culations were performed based on the projected augmented wave (PAW) pseudo-potentials using the Vienna ab Initio Simulation Package (VASP).36–38The electronic interactions are described using Perdew–Burke–Ernzerhof (PBE) parametrization of generalized gradi- ent approximation (GGA).39The plane wave cutoff was set to 550 eV.
A 666 mesh for the monoclinic phase and an 888 mesh for the rutile structure are used for Brillouin-zone sampling.40Dudarev implementation41was adopted to add an on-site Coulomb interaction Ueff(¼U–J)¼3.0 eV to the 3dorbitals of vanadium. The lattice con- stants and atomic positions were fully optimized iteratively until the Hellman–Feynman forces converged to less than 5.0 meV/A˚ . The cor- responding lattice parameters used for the calculation can be referred to Table S1 in thesupplementary material.
While the bandgap value if any may be slightly underestimated in the present computation scheme, the qualitative characteristics remain similar. The calculated band structures for the two phases are given inFigs. 3(a)and3(b)where the total and partial density of states (DOS) spectra are plotted. As shown inFig. 3(a)for the monoclinic phase, the valence band near the Fermi level (EF) is mainly composed of V-3dstates, while the V-3d state is dominant at the top of the valence band, implying that the dominant carriers would be holes. The bandgap, as modified by applying Ueff, is0.65 eV, consistent with FIG. 2.Schematic diagrams of the Hall effect measured at low temperature (a) and high temperature (b); a downward magnetic fieldB¼ 3 T and sense currentI¼100lA are adopted for the measurements. Temperature dependent Hall voltage (VH) during heating (c) and cooling (d) processes. The insets of (c) and (d) show the critical tempera- ture regions, at which the Hall voltage sign changes. (e) Temperature dependent Hall voltages for both the heating and cooling processes, also showing a clear hysteresis effect. It should be noted that because the error values in the high temperature area, especially in the area above MIT temperature, are very small, the error bars in these tem- perature regions cannot be displayed on the graph basically.
FIG. 3.Band structure, total density of states (DOS), and partial density of states of VO2obtained by first-principles theory calculations. (a) VO2with themonoclinic phase; (b) VO2with the rutile phase.
earlier predicted results42and other reported estimated/measured val- ues.43,44Considering the fact that the low-Tregion far below the MIT pointTcwould be occupied with the insulating monoclinic phase, one can assume that the low-Tphase prefers holes as the dominant carriers.
On the other hand, one looks at the rutile phase, and its band structure plotted inFig. 3(b)shows again the dominant V-3dstate in the conduc- tion band. The Fermi levelEFenters the conduction band, implying the metallic-like electronic structure. Nevertheless, it is clear that the domi- nant carrier in this phase is the electron. Similarly, one can assume that the dominant carriers in the high-Tregion far aboveTcshould be elec- trons since the rutile phase is dominant in the high-Tregion.
Based on the calculated results inFig. 3, one can explain at least qualitatively the observed sign reversal of the Hall voltage upon heating-cooling cycling. For a qualitative discussion, the MIT sequence is accompanied by the SPT sequence. Therefore, the low-Tregion, mainly occupied with the monoclinic phase, must have holes as the dominant carriers. The high-Tregion, mainly occupied with the rutile phase, must have electrons as the dominant carriers. In other words, the SPT sequence not only participates in the MIT sequence but also controls the electric transport behaviors more. It is worth noting that in the work by Luoet al.,44the carriers of VO2thin films are electrons no matter in the high temperature or in the low temperature region.
The structural phase transition induced carrier type change has not been observed. The reason is that in their work, the VO2thin film is grown by magnetron sputtering and contains a lot of oxygen vacan- cies. In that case, the majority carrier type in VO2thin films is domi- nated by defects like oxygen vacancies within the whole temperature range.45–47
Now, there remain two more issues to be discussed. The first issue deals with the dramatic difference in the Hall voltage difference between the low-Tregion and the high-Tregion. The second issue is related to the microstructural characteristics during the MIT sequence.
For the first issue, it is seen fromFigs. 2(c)and2(d)that the Hall vol- tages at, e.g.,T 300 K (low-Tregion end) andT350 K (high-T region) differ by more than two orders of magnitude, implying the dramatic difference in the carrier density between the two regions.
Such a big difference in the carrier density, no matter holes or elec- trons, is the reason for the difference in the Hall voltage. The Hall volt- age also shows a difference at the sameT during the heating and cooling process, which is also attributed to the difference in carrier density. To confirm this argument, one can estimate the carrier density hysteresis from the measured Hall data, and the results are plotted in Fig. 4, where the carrier densitynon the log-scale is plotted againstT in the heating and cooling cycle. Clearly, one sees the big difference in the carrier density between the two phases, as expected.
Interestingly, the magnitude of the carrier density in the heating- cooling cycle also showsT-dependent hysteresis, similar to theq-T curves shown inFig. 1(b). Starting from the low-Tend, the carrier density is1019cm3on the order of magnitude, and it increases gradually with increasingT, reaches up to1021cm3upon the insu- lating to metal transition around 330 K, and becomes saturated at 61021cm3. It is worth noting that the carrier density as a function ofTshows a kink atT332 K, the point at which the dominant car- rier type changes from hole to electron, according to the Hall data. In the cooling sequence from the high-T end, the carrier density decreases gradually with decreasingT, reaching the point at which a kink-like feature appears, corresponding to the switching of the carrier
type from electron to hole. Beyond this point, the carrier density fur- ther falls and reaches down to 21019cm3atT300 K.
Considering that VO2 is a strongly correlated electron system, the observed phenomena described above can be understood without surprising. It is well known that transition metal oxides often show inhomogeneities in electronic states and transport behaviors. In our previous work,27the nanoscale high-conduction and low-conduction regions coexist in a broadT-window covering the nominal MIT point, suggesting the coexistence of the monoclinic phase and rutile phase. In addition, the area (volume) fraction of the high-conduction region changes gradually with varyingTin the heating-cooling cycle, consis- tent with the claim that the conductive rutile phase and insulating monoclinic phase coexist over a broadT-range.
In order to further confirm this claim, we performed a set of CAFM imaging of the sample used for the Hall voltage measurement, by addressing the one-to-one correspondence between the CAFM imaged microstructure and the carrier density/type. CAFM imaging was performed onto a 2lm2lm square area upon the heating and cooling cycle. The details of the measuring setup and procedure for the CAFM imaging were given in our previous work.27Here, the bright contrast represents the highly conductive region and the black contrast marks the weakly conductive region.
Apart from the carrier density data plotted in Fig. 4 for the heating-cooling cycle, we insert a set of probed CAFM images for illus- trating the evolution of the conductive phase and insulating phase on the nanoscale during the cycle. The area (volume) fraction of the con- ductive regions as a function ofTin the cycle shows the one-to-one correspondence with the carrier density loop.
This correspondence between the CAFM imaging and carrier density against temperature in the heating-cooling cycle suggests that the two phase coexistence has twofold scenarios. On one hand, the electronic phase coexistence over the broad T-range is identified, where the low-Tinsulating phase with the carrier density as low as 1019cm3and the high-Tmetallic phase with the carrier density as FIG. 4.Changes in carrier density (n) as a function of measurement temperature during the heating and cooling processes. The red and blue dotted boxes corre- spond to the temperature at which the dominant carriers change during heating (Trh) and cooling (Trc), respectively. In this figure, changes in nanoscale conduction with temperature measured by CAFM are also shown for both the heating and cool- ing processes. The area percentage of the nanoscale high-conduction shows a one-to-one correspondence with the carrier density loop.
high as1022cm3were confirmed to coexist by the CAFM imaging and transport measurement. On the other hand, the structural phase coexistence over theT-range is also identified, where the hole-dominant monoclinic insulating phase and electron-dominant rutile conductive phase were confirmed to coexist by the Hall voltage measurement.
In summary, the Hall effect and CAFM measurements were carried out to investigate the nanoscale conduction and charge transport characteristics of VO2thin films. The dominant carrier type shows a critical change from electron to hole during the MIT sequence (cooling sequence) or from hole to electron during the reverse MIT sequence (heating sequence). The first-principle cal- culations demonstrate that the monoclinic and rutile phases of VO2 films show the p-type and n-type conduction, respectively.
This means that the carrier type change during the MIT or reverse MIT sequences is due to the SPT sequence. Both the carrier density and the area fraction of the conductive regions show a clear tem- perature hysteresis effect during the lowT!highT!lowTmea- surement cycle. The present results prove that the structure phase transition plays a critical role in their nanoscale conduction and unique charge transport behaviors of VO2thin films.
See thesupplementary materialfor XRD and AFM measurement results and relevant VO2lattice parameters used in the first-principles calculations.
This work was supported by the National Key Research Project of China (Grant No. 2016YFA0300101). X.B.L acknowledges the support of the Project for Guangdong Province Universities and the Colleges Pearl River Scholar Funded Scheme (2016).
REFERENCES
1J. Jeong, N. Aetukuri, T. Graf, T. D. Schladt, M. G. Samant, and S. P. Parkin, Science339, 1402 (2013).
2Z. Yang, C. Ko, and S. Ramanathan,Annu. Rev. Mater. Res.41, 337 (2011).
3Z. H. Tian, G. Lu, S. Greg, Z. Lei, Z. Y. Xia, F. Eugene, W. K. Derek, C. Subhasis, P. Hanjong, A. M. Jarrett, B. Michael, G. S. Darrell, V. B. John, D. Suman, G.
Venkatraman, and E. H. Roman,Adv. Funct. Mater.26, 6612 (2016).
4M. Kratochvilova, A. D. Hillier, A. R. Wildes, L. Wang, S. –W. Cheong, and J.
–G. Park,npj Quantum Mater.2, 42 (2017).
5R. F. Need, P. B. Marshall, E. Kenney, A. Suter, T. Prokscha, Z. Salman, B. J.
Kirby, S. Stemmer, M. J. Graf, and S. D. Wilson,npj Quantum Mater.3, 7 (2018).
6B. J. Kim, Y. Lee, S. Choi, J. W. Lim, S. Yun, H. T. Kim, T. J. Shin, and H. S.
Yun,Phys. Rev. B77, 235401 (2008).
7M. Imada, A. Fujimori, and Y. Tokura,Rev. Mod. Phys.70, 1039 (1998).
8S. Jin, T. H. Tiefel, M. McCormack, R. A. Fastnacht, R. Ramesh, and L. H.
Chen,Science264, 413 (1994).
9Y. Tokura and N. Nagaosa,Science288, 462 (2000).
10R. J. Zeches, M. D. Rossell, J. X. Zhang, A. J. Hatt, Q. He, C. H. Yang, A.
Kumar, C. H. Wang, A. Melville, C. Adamo, G. Sheng, Y. H. Chu, J. F. Ihlefeld, R. Erni, C. Ederer, V. Gopalan, L. Q. Chen, D. G. Schlom, N. A. Spaldin, L. W.
Martin, and R. Ramesh,Science326, 977 (2009).
11T. M. Rice, H. Launojs, and J. P. Pouget,Phys. Rev. Lett.73, 3042 (1994).
12R. H. Wentzcovitch, W. W. Schulz, and P. B. Allen,Rev. Lett.72, 3389 (1994).
13M. 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, Science318, 1750–1753 (2007).
14H. T. Kim, B. G. Chae, D. H. Youn, S. L. Maeng, G. Kim, K. Y. Kang, and Y. S.
Lim,New J. Phys.6, 52 (2004).
15N. B. Aetukuri, A. X. Gray, M. Drouard, M. Cossale, L. Gao, A. H. Reid, R.
Kukreja, H. Ohldag, C. A. Jenkins, E. Arenholz, and K. P. Roche,Nat. Phys.9, 661 (2013).
16J. B. Goodenough,J. Solid State. Chem.3, 490 (1971).
17T. C. Koethe, Z. Hu, M. W. Haverkort, C. S. Langeheine, F. Venturini, N. B.
Brookes, O. Tjernberg, W. Reichelt, H. H. Hsieh, H. J. Lin, C. Y. Chen, and L.
H. Tjeng,Phys. Rev. Lett.97, 116402 (2006).
18V. Eyert,Ann. Phys.11, 650 (2002).
19L. Whittaker, C. Jaye, Z. Fu, D. A. Fischer, and S. Banerjee,J. Am. Chem. Soc.
131, 8884 (2009).
20F. Brinkman and T. M. Rice,Phys. Rev. B2, 4302 (1970).
21H. T. Kim, B. J. Kim, Y. W. Lee, B. G. Chae, S. J. Yun, and K. Y. Kang,Physica C460, 1076 (2007).
22H. T. Kim, B. J. Kim, Y. W. Lee, B. G. Chae, S. J. Yun, and K. Y. Kang,Physica B403, 1434 (2008).
23M. Kang, S. W. Kim, and J. W. Ryu,J. Appl. Phys.118, 035105 (2015).
24A. Cavalleri, T. Dekorsy, H. H. W. Chong, J. C. Kieffer, and R. W. Schoenlein, Phys. Rev. B70, 161102 (2004).
25T. Yao, X. Zhang, Z. Sun, S. Liu, Y. Huang, Y. Xie, C. Wu, X. Yuan, W. Zhang, Z. Wu, G. Pan, F. Hu, L. Wu, Q. Liu, and S. Wei,Phys. Rev. Lett.105, 226405 (2010).
26H. Madan, M. Jerry, A. Pogrebnyakov, Y. Mayer, and S. Datta,ACS Nano9, 2009 (2015).
27J. J. Feng, C. Yang, A. H. Zhang, Q. Li, Z. Fan, M. H. Qin, M. Zeng, X. S. Gao, Y. Lin, G. F. Zhou, X. B. Lu, and J.–M. Liu,Appl. Phys. Lett.113, 173104 (2018).
28N. F. Mott,Rev. Mod. Phys.40, 677 (1968).
29N. F. Mott,Proc. Phys. Soc. LondonA62, 416 (1949).
30S. Samanta and A. K. Raychaudhuri,Phys. Rev. B92, 195125 (2015).
31C. McGahan, S. Gamage, J. Liang, B. Cross, R. E. Marvel, R. F. Haglund, and Y.
Abate,Nanotechnology28, 085701 (2017).
32J. Choi, B. J. Kim, G. Seo, H. T. Kim, S. Cho, and Y. W. Lee,Curr. Appl. Phys.
16, 335 (2016).
33W. Liang, M. Gao, C. Lu, Z. Zhang, C. H. Chan, L. Zhuge, J. Dai, H. Yang, C.
Chen, B. H. Park, Q. Jia, and Y. Lin,ACS Appl. Mater. Interfaces10, 8341 (2018).
34G. Wei, X. Lin, Z. Si, N. Lei, Y. Chen, S. Eimer, and W. Zhao,Appl. Phys. Lett.
114, 012407 (2019).
35G. Wei, X. Lin, Z. Si, D. Wang, X. Wang, X. Fan, K. Deng, K. Liu, K. Jiang, N.
Lei, Y. Chen, S. Mangin, E. Fullerton, and W. Zhao, Adv. Quantum Technol.3, 1900104 (2019).
36G. Kresse and J. Furthm€uller,Phys. Rev. B54, 11169 (1996).
37G. Kresse and J. Furthm€uller,Comput. Mater. Sci.6, 15 (1996).
38G. Kresse and D. Joubert,Phys. Rev. B59, 1758 (1999).
39J. P. Perdew, K. Burke, and M. Ernzerhof,Phys. Rev. Lett.77, 3865 (1996).
40R. Resta,Rev. Mod. Phys.66, 899 (1994).
41S. L. Dudarev, G. A. Botton, S. Y. Savrasov, C. J. Humphreys, and A. P. Suton, Phys. Rev. B57, 1505 (1998).
42C. N. Berglund and H. J. Guggenheim,Phys. Rev.185, 1022 (1969).
43J. Wan, Q. Ren, N. Wu, and Y. Gao,J. Alloys Compd.662, 621 (2016).
44J. B. Goodenough and H. Y. P. Hong,Phys. Rev. B8, 1323 (1973).
45H. Luo, B. Wang, E. Wang, X. Wang, Y. Sun, Q. Li, S. Fan, C. Cheng, and K.
Liu,Appl. Phys. Rev.6, 041407 (2019).
46C. Chen, Y. Zhao, X. Pan, V. Kuryatkov, A. Bernussi, M. Holtz, and Z. Fan, J. Appl. Phys.110, 023707 (2011).
47M. Nazari, C. Chen, A. A. Bernussi, Z. Y. Fan, and M. Holtz,Appl. Phys. Lett 99, 071902 (2011).
