# Fileset

[okabe_jpscp.38.011117.pdf](https://mdr.nims.go.jp/filesets/2167ae54-3dc6-413a-a1b8-05a08053f44c/download)

## Creator

Hirotaka Okabe, Masatoshi Hiraishi, Akihiro Koda, [Yoshitaka Matsushita](https://orcid.org/0000-0002-4968-8905), [Takeo Ohsawa](https://orcid.org/0000-0001-7528-8940), [Naoki Ohashi](https://orcid.org/0000-0002-4011-0031), Ryosuke Kadono

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Local Magnetism in the Spin-singlet State of VO<sub>2</sub>](https://mdr.nims.go.jp/datasets/b49acadf-1d18-4427-9b71-ab3b453db545)

## Fulltext

Local Magnetism in the Spin-singlet State of VO2  Hirotaka OKABE1,2*, Masatoshi HIRAISHI1,3, Akihiro KODA1, Yoshitaka MATSUSHITA4, Takeo OHSAWA4, Naoki OHASHI4 and Ryosuke KADONO1  1Institute of Materials and Structural Science, High Energy Accelerator Research Organization, 319-1106, Japan. 2Institute of Materials Research, Tohoku University, Aoba-ku, Sendai, 980-8577, Japan. 3Graduate School of Science and Engineering, Ibaraki University, Mito, 310-8512, Japan. 4National Institute for Materials Science, Tsukuba, 305-0047, Japan.  *E-mail: hirotaka.okabe.b4@tohoku.ac.jp (Received July 30, 2022)  Local magnetic field studies of vanadium dioxide (VO2), which has a spin-singlet phase under the metal-insulator transition at 340 K, were performed using SR method. By comparing the results of two samples with different qualities, it was discovered that a nonmagnetic state and microscopic magnetic ordered states coexist in the spin-singlet phase. This may be attributed to crystal lattice imperfections inherent in VO2. KEYWORDS: spin singlet, metal-insulator transition, SR   1.  Introduction  Vanadium dioxide (VO2) has several crystal polymorphs and exhibits a wide range of physical properties. The metal-insulator transition (MIT) between tetragonal rutile VO2(R) and monoclinic VO2(M1) has drawn the interest of numerous researchers since its discovery [1]. It has been actively investigated for applications in electronic devices such as memory, sensors, optical switches, and thermochromic dimming materials [2,3] because of the MIT occurring at around room temperature (340 K). Also in basic research, the transition mechanism has been the subject of much controversy (i.e. Mott-Hubbard [4], spin-Peierls [5], or molecular orbital crystals [6]).  Recently, Mengyan and coworkers reported the magnetism of VO2 via the muon spin research method [7]. Muon spin rotation/relaxation/resonance (μSR) method uses elementary muons as magnetic probes to detect small local magnetic fields in materials, such as nuclear magnetism, with high sensitivity. According to their research, some magnetic order is found even in the nonmagnetic singlet phase of VO2 below the MIT temperature. The origin of this order has been attributed to residual impurities and lattice defects. However there are no other reports of such magnetic order, and it is vital to re-examine this issue, including sample quality. To this end, we prepared two types of VO2 samples of different quality (high purity powder and deteriorated powder samples) and performed a comparative µSR study of local magnetic states in the presumedly nonmagnetic singlet phase. Our goals are to verify whether the aforementioned magnetic order is intrinsic and to gain insight into the MIT mechanism of VO2. JPS Conf. Proc. , 011117 (2023)©2023 The Author(s)https://doi.org/10.7566/JPSCP.38.01111738must maintain attribution to the author(s) and the title of the article, journal citation, and DOI.Proc. 29th Int. Conf. Low Temperature Physics (LT29)011117-1This article is published by the Physical Society of Japan under the terms of the Creative Commons Attribution 4.0 License. Any further distribution of this workProceedings of the 29th International Conference on Low Temperature Physics (LT29)Downloaded from journals.jps.jp by （研）物質・材料研究機構 on 04/11/25http://creativecommons.org/licenses/by/4.0/ 2.  Experiments  In this study, a commercially available powder of VO2 (99.9% purity, Kojundo Chemical Lab. Co., Ltd.) was used as a “pristine” sample (in the as-received condition without sintering or other heat treatment), and compared with another sample which was seemingly deteriorated as had been kept in air for more than 10 years. The sample qualities were checked using the powder X-ray diffractometer (SmartLab, Rigaku Co.) under ambient conditions. The magnetic susceptibility was measured using the SQUID magnetometer (MPMS, Quantum Design, Inc.) from 5 to 400 K. μSR experiments were performed using the ARTEMIS spectrometer installed in the S1 area at Muon Science Establishment (MUSE), Japan Proton Accelerator Research Complex (J-PARC).  3.  Results and Discussion  Figure 1 shows the powder X-ray diffraction patterns of pristine and deteriorated samples. The insets are photographs of each sample; the colors and textures are very different. While all diffraction peaks in the pristine sample can be indexed to monoclinic VO2 (M1) [8], there are several peaks that appear to be impurities in the deteriorated sample. Although the peaks of these impurities are too broad to be accurately identified, they are assumed to be VO2 (B) [9], a polymorph of VO2 (M1), and a hydrated VO2*xH2O (JCPDS data card 00-018-1445) [10], based on their approximate peak positions.  Figure 2 shows the temperature dependence of the magnetic susceptibility χ in these samples. A magnetic transition with hysteresis (MIT) was observed around 340 K in both samples. Below the MIT temperature, the pristine sample shows almost constant small values of χ, while the deteriorated sample shows large Curie paramagnetic behavior, which may be due to the impurities. As a reference, when pristine samples were annealed at 1373 K in Ar atmosphere, the magnitude of Curie component increased about twice as much as before treatment (data not shown).   Fig. 1.  Powder X-ray diffraction patterns of (a) pristine and (b) deteriorated VO2 samples. The vertical bars indicate the positions of the Bragg reflections of VO2 (M1). The insets are photographs of each sample. 9080706050403020102  (deg.)Intensity (arb.unit)(a)(b) Fig. 2.  Temperature dependence of magnetic susceptibility χ of (a) pristine and (b) deteriorated VO2 samples. Filled (Open) circles are data acquired during temperature rising (falling). 1.0x10-50.80.60.40.20.0 (emu g-1)1.0x10-50.80.60.40.20.0 (emu g-1)4003002001000T (K)(a)(b)011117-2JPS Conf. Proc. , 011117 (2023)38Proceedings of the 29th International Conference on Low Temperature Physics (LT29)Downloaded from journals.jps.jp by （研）物質・材料研究機構 on 04/11/25Figures 3 show the μSR time spectra in respective VO2 samples measured at 5, 100, 200, and 300 K. These spectra were measured under zero magnetic field (ZF) or a weak longitudinal magnetic field (LF) of 3~50 G. By comparing data from different LFs, μSR can estimate the magnitude of the local magnetic field generated inside a material and its fluctuations. Since these spectra show decoupling (recovery of muon spin polarization) under LFs of a few tens of gauss, relaxation is mainly due to the nuclear magnetic fields of V. However, in some of these spectra, there is a contribution of an exponential component suggesting relaxation due to electron spins. A disappearance of the initial signal intensity at t = 0 (a missing asymmetry) is a manifestation of a large local field resulting from a magnetic ordering. We analyzed these spectra using the following equation,  𝐴𝑃𝑧(𝑡) = 𝐴1𝐺KT(Δ, 𝜈)exp(−𝜆1𝑡) + 𝐴2exp(−𝜆2𝑡) + 𝐴const.   Figs. 3.  Zero field and longitudinal field μSR time spectra of (a)~(d) pristine and (e~h) deteriorated VO2 samples measured at selected temperatures (5, 100, 200, and 300 K). 0.250.200.150.100.050.00Asymmetry 0 G 5 G 10 G300 K0.250.200.150.100.050.00Asymmetry 0 G 3 G 5 G 10 G 20 G200 K0.250.200.150.100.050.00Asymmetry 0 G 5 G 10 G100 K0.250.200.150.100.050.00Asymmetry1614121086420t (s) 0 G 10 G 50 G5 K0.250.200.150.100.050.00Asymmetry 0 G 5 G 10 G 30 G300 K0.250.200.150.100.050.00Asymmetry200 K 0 G 5 G 10 G 30 G0.250.200.150.100.050.00Asymmetry 0 G 5 G 10 G 30 G100 K0.250.200.150.100.050.00Asymmetry1614121086420t (s) 0 G 5 G 10 G 30 G5 K(a)(b)(c)(d)(e)(f)(g)(h)011117-3JPS Conf. Proc. , 011117 (2023)38Proceedings of the 29th International Conference on Low Temperature Physics (LT29)Downloaded from journals.jps.jp by （研）物質・材料研究機構 on 04/11/25A is the signal amplitude, called asymmetry, which is proportional to the volume fraction of each component (A1, A2, Aconst). The first term on the right-hand side, GKT, is the Kubo-Toyabe function that represents a muon-spin relaxation due to randomly oriented nuclear spins [11].  is the nuclear magnetic distribution width,  is the muon hopping rate, and 1,2 are the relaxation rates caused by the electron spins of V. The second term on the right-hand side represents the rapid-relaxation component that corresponds to a missing asymmetry. Unfortunately, the second term cannot be fitted correctly due to the time resolution of the instrument. Aconst is the constant term which mainly represents the background component such as the sample holder. The solid lines in Figure 3 show the fitting curves obtained by using the above equation. In the following sections, we will discuss the fitting parameters only obtained from ZF spectra, considering the fitting results of these LF spectra (not shown here). Figures 4 show the temperature dependence of the corrected asymmetry A1 [A1/(A-Aconst)], 1, , and  of each sample. As the temperature decreased, A1 tends to decrease stepwise for both samples. The decreases in A1 suggest the emergence of magnetic orders. However, the decreases in A1 are not steep enough to determine the magnetic transition temperatures. The similar behaviors are observed in the earlier μSR results [7]. However, it does not necessarily correspond to the magnetic order they observed. The value of A1 at the lowest temperature is larger for the deteriorated sample (~0.6) than for the pristine sample (~0.4), indicating that more missing asymmetry (1-A1) exists in the pristine sample. This contradicts the prediction that more impurities induce the increase in the volume fraction of the microscopic magnetic order in VO2 by  Figs. 4.  Temperature dependence of μSR fitting parameters under zero magnetic field for pristine (blue circle) and deteriorated (green triangle) VO2 samples. (a) corrected asymmetry A1, (b) relaxation rate 1, (c) nuclear magnetic distribution width , and (d) muon hopping rate . Note that Δ is restricted to the parameter (Δ < 0.3) based on the fitting results in Figure 3 (not shown). 1.00.80.60.40.20.0corrected A1300250200150100500T (K)0.40.30.20.10.0s-1)300250200150100500T (K)1.00.80.60.40.20.0s-1)300250200150100500T (K)0.350.300.250.200.150.100.050.00s-1)300250200150100500T (K)(a) (b)(c) (d)011117-4JPS Conf. Proc. , 011117 (2023)38Proceedings of the 29th International Conference on Low Temperature Physics (LT29)Downloaded from journals.jps.jp by （研）物質・材料研究機構 on 04/11/25destroying more V-V dimers. However, this is presumably due to the excessive sample deterioration, which has further transformed the microscopic magnetic order into paramagnetic, where the spin fluctuation rate is out of the μSR time range (>1011 Hz). This is consistent with the behavior of  that the impurity effect almost appears exclusively as an increase of Curie paramagnetic component; the clear MIT is still observed in the deteriorated sample (see Fig. 2). In other words, the effect of deterioration (increased impurity content due to moisture, etc.) on the magnetic state of VO2 is localized, and it is not reasonable to assume that the microscopic magnetic orders with large volume fractions are due solely to impurities. The longitudinal relaxation rates 1 gradually increase as A1 begin to decrease [Fig.4(b)]. The increase of 1 is more pronounced in the deteriorated sample, suggesting the enhancement of muon spin relaxation from the V electron spins. In particular, the peak structures below 200 K is reminiscent of magnetic transitions. Because of the correlation of 1 with Δ and ν, it is not possible to clearly distinguish whether these peaks are phase transitions or not. However, considering the stepwise decrease of A1, several different microscopic magnetic orders (clusters?) appear to coexist in the singlet phase. A significant number of V spins do not form dimers regardless of the amount of impurities, because the both samples follow the similar trend. Therefore, it seems unlikely that the microscopic magnetic order was only caused by external factors, such as the type and amount of impurities. Similar to titanium oxides, vanadium oxides have a number of compounds with slight different oxygen ratios, such as the Magnéli and Wadsley phases [12]. In these phases, the stacking patterns of the transition metal octahedra are different, and MIT and magnetic transitions occur at various temperatures. There should be slight compositional deviations with partially deviated octahedral stacking patterns in VO2, because VO2 is also one of the Magnéli-phase compounds. The aforementioned behaviors of A1 and 1 can be explained by the existence of multiple local structures in the VO2 singlet phase that exhibit different magnetic transition temperatures. Such a mixed state can be regarded as a type of lattice imperfection identical to a disordered dimer crystal structure. Another possibility for the emergence of local magnetic fields in the singlet phase is muon-induced effects; positive muon (+) behaves similarly to hydrogen ions in matter. If implanted + binds to the oxygen formed VO6 octahedron, the V-V dimer can be destroyed by changing the valence state of the adjacent V ions. However, single + in VO2 cannot develop a magnetic order resulting in the missing asymmetry. It is also unlikely that µ+ forms polaronic state (Mu0) with an excess electron in such a narrow-gap (~0.6 eV) semiconductor [13]. Furthermore, the concentration of injected muons at a time is extremely small, so that muons cannot work together to develop a magnetic order. From the above reasons, we conclude that the singlet state and microscopic magnetic orders coexist in VO2 at low temperatures. This also suggests that the range within which the magnetic state of VO2 is affected by a single impurity or defect is narrow; probably a certain amount of V-V dimers could survive even if heavily doped. This tendency is also consistent with the local character of V-V dimers reported in TiO2-VO2 mixed crystal system that exhibit a spinodal decomposition [14]. In any case, SR and other local probes would be useful to elucidate the MIT mechanism of VO2.  011117-5JPS Conf. Proc. , 011117 (2023)38Proceedings of the 29th International Conference on Low Temperature Physics (LT29)Downloaded from journals.jps.jp by （研）物質・材料研究機構 on 04/11/25 4.  Summary  The local magnetic field in the singlet phase VO2 was investigated using the SR method for two samples with different impurity contents. In both samples, the coexistence of a singlet state and microscopic magnetic orders was observed. The coexisting state may be originated from structural imperfections inherent in VO2 rather than external impurity effects.  Acknowledgment  This work was supported by the MEXT Elements Strategy Initiative to Form Core Research Centers, from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (MEXT) under Grant No. JPMXP0112101001 and partially supported by JSPS KAKENHI (Grant No. 20K05312). The SR experiments were performed under user programs (Proposal No. 2019MS02) at the Materials and Life Science Experimental Facility of the J-PARC. We also acknowledge the Neutron Science and Technology Center, CROSS for the use of PPMS in their user laboratories.  References  [1] F. J. Morin: Phys. Rev. Lett. 3, 34 (1959). [2] H. Jerominek, F. Picard and D. Vincent: Opt. Engineering 32, 2092 (1993). [3] A. Cavalleri, C. Toth, C. W. Siders and J. A. Squier: Phys. Rev. Lett. 87, 237401 (2001). [4] A. Zylbersztejn and N. F. Mott: Phys. Rev. B 11, 4383 (1975). [5] D. Paquet and P. Leroux-Hugon: Phys. Rev. B 22, 5284 (1980). [6] Z. Hiroi: Prog. Sol. Sta. Chem. 43, 47 (2015). [7] P. W. Mengyan, R. L. Lichti, B. B. Baker and G. Jayarathna: J. Phys.: Conf. Ser. 551, 012017 (2014). [8] G. Andersson: Acta Chem. Scand. 8, 1599 (1954). [9] F. Theobald, R. Cabala and J. Bernard: J. Sol. Sta. Chem. 438, 431 (1976). [10] A. Deschanvres: C. R. Hebd. Seances Acad. Sci. 261, 31 (1965). [11] R. S. Hayano, Y. J. Uemura, J. Imazato, N. Nishida, T. Yamazaki and R. Kubo: Phys. Rev. B 20, 850 (1979). [12] N. Bahlawane and D. Lenoble: Chem. Vap. Deposition 20, 299 (2014). [13] M. Hiraishi, H. Okabe, A. Koda, R. Kadono, and H. Hosono, to be published in J. Appl. Phys.  [14] Z. Hiroi, H. Hayamizu, T. Yoshida, Y. Muraoka, Y. Okamoto, J. Yamaura and Y. Ueda: Chem. Mater. 25, 2202 (2013). 011117-6JPS Conf. Proc. , 011117 (2023)38Proceedings of the 29th International Conference on Low Temperature Physics (LT29)Downloaded from journals.jps.jp by （研）物質・材料研究機構 on 04/11/25