# Fileset

[jpsj.94.053702.pdf](https://mdr.nims.go.jp/filesets/18b861dc-eab7-40be-8094-8c9ccb910b32/download)

## Creator

[Takako Konoike](https://orcid.org/0000-0002-6037-5782), [Shinya Uji](https://orcid.org/0000-0001-9351-6388), [Yuya Hattori](https://orcid.org/0000-0002-3805-4659), [Taichi Terashima](https://orcid.org/0000-0001-9239-0621)

## Rights

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

## Other metadata

[Control of the Electronic States of the Organic Conductor <i>α</i>-(BEDT-TTF)<sub>2</sub>I<sub>3</sub> by Uniaxial Tensile and Compressive Strain](https://mdr.nims.go.jp/datasets/05810b6d-21f0-4cfb-92c3-5c5af4572246)

## Fulltext

Control of the Electronic States of the Organic Conductor α-(BEDT-TTF)2I3 by Uniaxial Tensile and Compressive StrainControl of the Electronic States of the Organic Conductor α-(BEDT-TTF)2I3by Uniaxial Tensile and Compressive StrainTakako Konoike+ , Shinya Uji , Yuya Hattori , and Taichi TerashimaResearch Center for Materials Nanoarchitectonics (MANA), National Institute for Materials Science (NIMS),Tsukuba, Ibaraki 305-0003, Japan(Received March 11, 2025; accepted April 9, 2025; published online April 25, 2025)We developed a sample mounting method that enables the use of a piezoelectric-based strain cell, allowing us tomeasure the electrical resistance of the organic conductor α-(BEDT-TTF)2I3 under uniaxial tensile and compressivestrain. The metal–insulator (M–I) transition temperature changes with the application of uniaxial strain, exhibiting largeanisotropy with respect to the strain direction. We successfully controlled the electronic states from the metallic to theinsulating state by the continuous application of uniaxial strain at low temperatures. The strain dependence of theresistance and elastoresistance exhibits characteristic behavior around the M–I transition temperature. Based on previousX-ray diffraction measurements, we consider that the libration of BEDT-TTF molecules plays a crucial role in thecharacteristic behavior around the phase transition.Pressure has long been employed as an effective tool forinvestigating the physical properties of strongly correlatedelectron systems. Most of these studies have been conductedunder hydrostatic conditions, where isotropic pressure isapplied to the sample. For experiments up to about 3GPa,clamp-type pressure cells using liquids such as oil aspressure-transmitting medium are commonly used. However,when varying the applied pressure using this method, thepressure cell must be removed from the cryostat, making itimpossible to vary the pressure continuously. Therefore, it isdifficult to precisely investigate the pressure dependence ofphase transitions. In addition, experiments under hydrostaticpressure cannot determine which direction of pressure is mosteffective in altering electronic properties. Recently, Hickset al. developed a uniaxial strain device using a piezoelectricactuator.1) This apparatus enables continuous strain control atlow temperatures and can apply both uniaxial compressionand tension. Using this device, the symmetry of theunconventional superconducting gap2) and the electronicnematic state, where the rotational symmetry of the system isbroken, have been intensively studied.3–5) This apparatus hasbeen used for inorganic materials after processing them intolong, thin plates. On the other hand, applying this techniqueto organic materials is challenging because single crystals arefragile and small.As a method for applying uniaxial pressure to organicconductors, the epoxy-crystal method, which combines aclamp-type cell with a sample embedded in cylindrical epoxyresin, has been developed.6) However, as mentioned above,continuous pressure adjustment is not possible with a clamp-type cell. Another reported method applies effective uniaxialstrain continuously by bending a flexible substrate with anorganic conductor on it,7,8) but it requires transferring anultrathin sample (a few hundred nanometers thick) onto amicrofabricated electrode. In this study, we developed asample mounting method that enables the use of a piezo-electric-based strain cells for single crystals of organicconductors as well as inorganic materials. Using this device,we investigated α-(BEDT-TTF)2I3 and successfully con-trolled its electronic state. This compound is known to forma Dirac electron system when the metal–insulator (M–I)transition is suppressed under high hydrostatic pressure above12 kbar.9,10) We discuss the relationship between the M–Itransition and the libration of BEDT-TTF molecules based onresistance behavior near the phase transition temperature.Uniaxial strain experiments were performed using apiezoelectric-based strain cell [Razorbill InstrumentsFC100, Fig. 1(a)]. Single crystals of α-(BEDT-TTF)2I3 weregrown by electrochemical oxidation.11) We selected thincrystals with a typical size of 0:8 � 0:3 � 0:02mm3. Goldwires (10 µm in diameter) were attached to the sample withcarbon paste for electrical resistance measurements. First,a plate of epoxy resin (Stycast 1266), which has elasticitysimilar to that of organic conductors, was prepared. To enablethe application of stronger strain, the plate was thinned to athickness of 0.2mm, leaving only the screw-fastening areasfor the device. The sample and the strain gauge withelectrodes were placed on the plate [Fig. 1(b)] and embeddedin the epoxy resin by coating them with epoxy, as shown inpiezoelectricstackssamplestrain gaugeouterinnerouterepoxy plate(stycast 1266)(b)thickness 0.2 mmStraingauge1.5mm16 mmAu wire4 mmTop view Side view(c)(a)5 mm25.4 mm9 mmFig. 1. (Color online) (a) Schematic illustration of a piezoelectric-basedstrain cell. The arrows on the outer and inner piezoelectric stacks indicatethe direction of stack movement when compressive strain is applied to thesample. (b, c) Top view and side view of the epoxy plate, respectively.Journal of the Physical Society of Japan 94, 053702 (2025)https://doi.org/10.7566/JPSJ.94.053702Letters053702-1 ©2025 The Physical Society of Japanmaintain attribution to the author(s) and the title of the article, journal citation, and DOI.©2025 The Author(s)This 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 work mustJ. Phys. Soc. Jpn.Downloaded from journals.jps.jp by （研）物質・材料研究機構 on 05/07/25https://orcid.org/0000-0002-6037-5782https://orcid.org/0000-0001-9351-6388https://orcid.org/0000-0002-3805-4659https://orcid.org/0000-0001-9239-0621https://doi.org/10.7566/JPSJ.94.053702http://creativecommons.org/licenses/by/4.0/http://crossmark.crossref.org/dialog/?doi=10.7566%2FJPSJ.94.053702&domain=pdf&date_stamp=2025-04-25Fig. 1(c). To prevent damage to the sample from the uncuredepoxy, the surface of the sample was pre-coated with enamel.Finally, the epoxy plate was mounted on the strain cell withscrews, as shown in Fig. 1(a). Uniaxial compressive strainwas generated by applying positive and negative voltages tothe outer and inner piezoelectric stacks, respectively, whileuniaxial tensile strain was generated by applying the oppositevoltages. In this apparatus, the three stacks are arrangedso that the thermal contraction effect of the piezoelectricelements is canceled out. Electrical resistivity was measuredusing a standard four-terminal method, and the strain wasmonitored by a strain gauge placed next to the sample. Thestrain cell was attached to the bottom of the probe andinserted into the 4He cryostat. Uniaxial compression andtension could be applied repeatedly as long as the sampleremained properly fixed to the epoxy substrate.In α-(BEDT-TTF)2I3, the M–I transition occurs at 135Kunder ambient pressure,12) and it has been shown that theinsulating ground state is charge ordered (CO) state inducedby the nearest neighbor Coulomb interaction.13–16) In thiscompound, an anisotropic triangular lattice is formed, asshown by the dotted lines in Fig. 2(a). At room temperature,the unit cell (thin lines) contains three crystallographicallyindependent molecules: A, B, and C. Molecule AA isequivalent to A owing to the inversion symmetry. In COstate below TCO ¼ 135K, molecules A and AA becomeinequivalent, and a horizontal type CO state stabilizes, wheremolecules A and B are charge rich, while molecules AA and Care charge poor [Fig. 2(b)].Figures 2(c) and 2(d) show the temperature dependenceof the resistance of α-(BEDT-TTF)2I3 under uniaxial strainalong the a- and the b-axes, with a constant voltage appliedto the piezoelectric stacks, respectively. Here, the straingenerated by the applied voltage varies slightly withtemperature; the strain applied to the sample decreases byapproximately 10% for every 10K drop in temperature.Therefore, each curve in Figs. 2(c) and 2(d) does not strictlycorrespond to the result under the same strain intensity. Theintensity of strain in each curve was determined at 130K.Under the a-axis strain, the transition temperature TCO issuppressed by compressive strain, while it increases withtensile strain, as shown in Fig. 2(c). The results indicate thatapproximately 0.3% tensile or compressive strain canincrease or decrease TCO by about 10K, respectively. Thisbehavior resembles the results observed under hydrostaticpressure,17,18) suggesting that this compound is highlysensitive to pressure along the a-axis. The effect of 0.3%compression along the a-axis is nearly equivalent to that of1.4 kbar of hydrostatic pressure. Using the epoxy-crystalmethod,6) a decrease in resistance suggesting superconduc-tivity has been observed at 7.2K under a-axis uniaxialpressure of 2 kbar.19) Additionally, a transition to a Diracelectron state has been observed under an effective a-axisstrain induced by bending the substrate.8) In our measure-ments down to 1.5K, neither phenomenon was observed,suggesting that a larger compressive strain is required torealize these states. On the other hand, the strain dependenceof TCO qualitatively agrees with the results obtained by bothexperimental methods. The stabilization of the CO stateobserved under the a-axis tensile strain cannot be achievedby using a clamp-type cell, which can only applycompressive stress. This highlights the importance of usingthis device to elucidate the electronic properties of thematerial, and it is even expected that a new electronic phase,which is not observed under compression, could be realized.Figure 2(d) shows the results when strain is applied alongthe b-axis. In the high-temperature region, the resistancedecreases under compressive strain and increases undertensile strain, similar to the behavior observed under thea-axis strain. However, a significant change in resistanceassociated with the CO transition was not observed undertensile strain. Additionally, at low temperatures belowapproximately 120K, the strain dependence of the resistanceexhibits the opposite tendency: the resistance decreases undertensile strain along the b-axis. This result appears to becaused by the Poisson effect, which refers to the phenomenon103104105R (Ω)170160150140130120110100T (K) +0.24 % +0.12 %      0 %  -0.12 %  -0.24 %T > T CO(a) T < T CO(b)(c)(d)compressiontensioncompressiontensionstrain // a-axisstrain // b-axis103104105R (Ω)170160150140130120110100T (K)  +0.31%  +0.21%          0%   -0.19%   -0.28%  at 130Kat 130KFig. 2. (Color online) (a, b) Arrangement of BEDT-TTF molecules in theconducting layer of α-(BEDT-TTF)2I3 above and below TCO, respectively.Above TCO, molecules A and AA in the unit cell (thin lines) arecrystallographically equivalent due to inversion symmetry. (c, d) Temper-ature dependence of resistance under the a- and the b-axes strain,respectively. The strain intensity in each curve was determined at 130K.J. Phys. Soc. Jpn. 94, 053702 (2025) Letters T. Konoike et al.053702-2 ©2025 The Physical Society of Japan©2025 The Author(s)J. Phys. Soc. Jpn.Downloaded from journals.jps.jp by （研）物質・材料研究機構 on 05/07/25where compression (expansion) induces slight expansion(compression) in the perpendicular direction. As previouslymentioned, this compound is sensitive to the a-axis strain.The compressive effect along the a-axis, induced by thePoisson effect during the application of the b-axis tensilestrain, shifts TCO to lower temperatures and results in lowerresistance. Thus, the influence of the Poisson effect should betaken into consideration when discussing the experimentalresults under the b-axis strain, especially at low temperatures.These results indicate a large anisotropy in the effect ofuniaxial strain direction. The fact that the CO insulating stateis more easily controlled by the a-axis strain than by the b-axis strain is consistent with previous studies using clamp-type cell and substrate bending.8,19)Next, we investigated the resistance as a function ofuniaxial strain. The sample was first compressed uniaxially ata constant temperature and then expanded along the samedirection. Our experimental method enables continuous strainapplication at low temperatures and in-situ resistancemeasurements. This is the first time such measurements havebeen performed on this compound, and the results allow usto investigate the details of the M–I transition. Figure 3(a)shows the a-axis strain dependence of the resistance of α-(BEDT-TTF)2I3. At high temperatures, the resistance varieslinearly with strain. As the temperature decreases, theresistance becomes nonlinear with strain and changesdrastically near TCO. This behavior indicates a transitionfrom a low-resistance metallic state to a high-resistance COstate. Thus, we successfully induced the CO state throughuniaxial strain, demonstrating the effectiveness of theuniaxial strain in controlling the electronic state of organicconductors. Around TCO, we note that the electrical resistanceinitially increases slightly with tensile strain, then remainsnearly constant, and subsequently rises significantly. Thisbehavior suggests the existence of a region where theresistance is less affected by strain.Figure 3(b) shows the resistance of α-(BEDT-TTF)2I3 asa function of the b-axis strain at various temperatures.Compared to the results under the a-axis strain, the change inresistance is smaller. At temperatures near TCO, a stepwiseincrease in resistance is observed. Since the resistance valuesare reproducible, this behavior is not caused by cracks in thesample. The resistance step appears under weaker tensilestrain at lower temperatures. This structure is likely relatedto the structural phase transition accompanying the COtransition, as discussed later.20) The reason why the resistancedoes not increase significantly after the stepwise increase isprobably due to the a-axis compressive strain resulting fromthe Poisson effect.Next, we focus on the low-strain region. Figures 4(a) and4(b) show the a- and the b-axes strain dependence of thesample resistance normalized by its value without strain,respectively. Here, the strain ϵ is estimated as � ¼ð1=KÞ�Rg=Rg using the strain gauge resistance Rg, where�Rg is the change in resistance from its value without strain,and K is 2.1 for our gauge. The normalized sample resistanceis linear with strain in the low-strain region, and its slope,dð�R=RÞ=d�, is defined as the elastoresistance, which reflectsthe sensitivity of the electronic system to strain. Figure 4(c)shows the elastoresistance as a function of temperature forstrain along the a-axis (red filled circles) and the b-axis (blueopen circles), respectively. For reference, the temperaturedependence of resistance (black line) is also plotted on theright axis. As the temperature decreases, the elastoresistancedecreases monotonically under the b-axis strain. On the otherhand, under the a-axis strain, the elastoresistance initiallyincreases with decreasing temperature but decreases slightlyaround 160K before increasing divergently from 140K,very close to TCO. This result indicates the existence ofa temperature region where the electronic system becomesinsensitive to strain before the CO transition. The elastore-sistance also exhibits anisotropy, although detailed discus-sion is difficult because the Poisson effect becomes non-negligible below TCO under the b-axis strain.Organic conductors generally have highly anisotropiccrystal structures. In α-(BEDT-TTF)2I3, anisotropy exists inthe orientation of BEDT-TTF molecules within the con-duction plane, as shown in Fig. 2(a). The stacking directionof the molecules corresponds to the a-axis, while the inter-columnar direction corresponds to the b-axis. Electricalconduction is carried by π electrons that extend perpendic-ularly from the nearly flat BEDT-TTF molecules. Con-sequently, the electronic state is highly sensitive to molecularorientation, resulting in significant anisotropy in response toapplied strain.The crystal structure of α-(BEDT-TTF)2I3 before and afterthe CO transition was precisely elucidated by synchrotron X-ray diffraction measurements.20) These results revealed that45678910002345R (Ω)0.40.30.20.10.0-0.1-0.2Strain (%)(a)145 K(b)stepwise strain // a-axisstrain // b-axis150 K130 K125 K170 K160 K46810024681000246-0.4 -0.2 0.0 0.2 0.4Strain (%)R (Ω)Fig. 3. (Color online) Resistance as a function of uniaxial strain along(a) the a-axis and (b) the b-axis. The green dotted lines in (a) and (b) indicatethe regions where plateau-like structures and stepwise increases in resistancewere observed, respectively.J. Phys. Soc. Jpn. 94, 053702 (2025) Letters T. Konoike et al.053702-3 ©2025 The Physical Society of Japan©2025 The Author(s)J. Phys. Soc. Jpn.Downloaded from journals.jps.jp by （研）物質・材料研究機構 on 05/07/25the CO transition is accompanied by a structural phasetransition, and the main structural change is not thetranslational shift of the BEDT-TTF molecules but a changein the dihedral angle θ, as defined in Fig. 2(b). The change inthe angle θ affects the spatial overlap of the wave functions,leading to a drastic change in the transfer integrals during theCO transition, which ultimately results in a M–I transition.Raman spectroscopy measurements show the presence ofseveral low-energy phonons in this system, including thelibration of BEDT-TTF molecules.21,22) Here, the libration ofmolecules refers to the restricted rotational motion of themolecules around an equilibrium position. The libration ofthe BEDT-TTF molecules induces vibrations of the dihedralangle θ. As indicated by the X-ray diffraction measure-ments,20) changes in the dihedral angle θ affect the transferintegral. Since specific heat measurements have shown thatthe M–I transition is first-order,23) there is a region where themetallic and CO phases coexist. The double peak observed inX-ray diffraction measurements and the scanning near-fieldoptical microscopy image24) also suggest the coexistence ofthe two phases in the vicinity of TCO. In this region, librationaround the two equilibrium positions is likely to occur,leading to a broader spatial overlap of the wave function dueto the increased angular range of rotational motion. As aresult, the bandwidth is expected to be effectively broadened,reducing its sensitivity to tensile strain. Consequently, theresistance shows little variation with strain in the regionwhere the two phases coexist, which may explain the plateaustructure observed in Fig. 3(a). Similarly, the behaviorobserved in Fig. 4(c), where the elastoresistance becomesinsensitive to the a-axis strain near TCO, could be attributed tothe same mechanism.We speculate that by inducing a first-order phase transitionfrom the metallic to the CO state through the application ofuniaxial strain, we could observe these phenomena becausethe electronic state of this compound is sensitive to thelibration of BEDT-TTF molecules, the importance of whichhas already been pointed out.25,26) Further improvement inthe sample setting, where a portion of the sample surfaceis exposed, would allows X-ray and Raman spectroscopymeasurements, leading to a deeper understanding of thestrain-induced phase transitions.Acknowledgment This work was supported by JSPS KAKENHI Grants(No. 22K22444). MANA is supported by World Premier International ResearchCenter Initiative (WPI), MEXT, Japan.+KONOIKE.Takako@nims.go.jp1) C. W. Hicks, M. E. Barber, S. D. Edkins, D. O. Brodsky, and A. P.Mackenzie, Rev. Sci. Instrum. 85, 065003 (2014).2) C. W. Hicks, D. O. Brodsky, E. A. Yelland, A. S. Gibbs, J. A. N. Bruin,M. E. Barber, S. D. Edkins, K. Nishimura, S. Yonezawa, Y. Maeno,and A. P. Mackenzie, Science 344, 283 (2014).3) I. Kostylev, S. Yonezawa, and Y. Maeno, J. Appl. Phys. 125, 082535(2019).4) I. Kostylev, S. Yonezawa, Z. Wang, Y. Ando, and Y. Maeno, Nat.Commun. 11, 4152 (2020).5) J. M. Bartlett, A. Steppke, S. Hosoi, H. Noad, J. Park, C. Timm, T.Shibauchi, A. P. Mackenzie, and C. W. Hicks, Phys. Rev. X 11,021038 (2021).6) M. Maesato, Y. Kaga, R. Kondo, and S. Kagoshima, Rev. Sci. Instrum.71, 176 (2000).7) M. Suda, Y. Kawasugi, T. Minari, K. Tsukagoshi, R. Kato, and H. M.Yamamoto, Adv. Mater. 26, 3490 (2014).8) Y. Kawasugi, H. Suzuki, H. M. Yamamoto, R. Kato, and N. Tajima,Appl. Phys. Lett. 122, 123102 (2023).9) S. Katayama, A. Kobayashi, and Y. Suzumura, J. Phys. Soc. Jpn. 75,054705 (2006).10) N. Tajima, S. Sugawara, M. Tamura, Y. Nishio, and K. Kajita, J. Phys.Soc. Jpn. 75, 051010 (2006).11) K. Bender, I. Hennig, D. Schweitzer, K. Dietz, H. Endres, and H. J.Keller, Mol. Cryst. Liq. Cryst. 108, 359 (1984).12) K. Bender, K. Dietz, H. Endres, H. W. Helberg, I. Henning, H. J.Keller, H. W. Schafer, and D. Schweitzer, Mol. Cryst. Liq. Cryst. 107,45 (1984).13) H. Kino and H. Fukuyama, J. Phys. Soc. Jpn. 64, 1877 (1995).14) H. Seo, J. Phys. Soc. Jpn. 69, 805 (2000).15) Y. Takano, K. Hiraki, H. M. Yamamoto, T. Nakamura, and T.Takahashi, J. Phys. Chem. Solids 62, 393 (2001).16) T. Takahashi, Synth. Met. 133–134, 261 (2003).17) T. Ojiro, K. Kajita, Y. Nishio, H. Kobayashi, A. Kobayashi, R. Kato,and Y. Iye, Synth. Met. 56, 2268 (1993).18) N. Tajima, M. Tamura, Y. Nishio, K. Kajita, and Y. Iye, J. Phys. Soc.Jpn. 69, 543 (2000).19) N. Tajima, A. Ebina-Tajima, M. Tamura, Y. Nishio, and K. Kajita,J. Phys. Soc. Jpn. 71, 1832 (2002).20) T. Kakiuchi, Y. Wakabayashi, H. Sawa, T. Takahashi, and T.Nakamura, J. Phys. Soc. Jpn. 76, 113702 (2007).21) K. I. Pokhodnia, A. Graja, M. Weger, and D. Schweitzer, Z. Phys. B90, 127 (1993).22) A. Graja, Synth. Met. 56, 2477 (1993).23) N. A. Fortune, K. Murata, M. Ishibashi, M. Tokumoto, N. Kinoshita,and H. Anzai, Solid State Commun. 79, 265 (1991).24) A. Pustogow, A. S. McLeod, Y. Saito, D. N. Basov, and M. Dressel,Sci. Adv. 4, eaau9123 (2018).25) S. Iwai, K. Yamamoto, A. Kashiwazaki, F. Hiramatsu, H. Nakaya, Y.Kawakami, K. Yakushi, H. Okamoto, H. Mori, and Y. Nishio, Phys.Rev. Lett. 98, 097402 (2007).26) S. Iwai, K. Yamamoto, F. Hiramatsu, H. Nakaya, Y. Kawakami, and K.Yakushi, Phys. Rev. B 77, 125131 (2008).-0.2-0.10.00.1ΔR / R-100 x10-6 0 100ε  147 K  150 K  155 K  160 K  170 K(a)(c)145 K136 K -0.04-0.020.000.020.04ΔR / R-100 x10-6 0 100ε 160 K 150 K 140 K 135 K(b)strain // a-axis strain // b-axis700600500400300200100d(ΔR/R) / dε240220200180160140120T (K)46810024681000246R (Ω) strain // a strain // b ResistanceFig. 4. (Color online) (a, b) Strain dependence of the normalized resis-tance along (a) the a-axis and (b) the b-axis, respectively. (c) elastoresistanceas a function of temperature under the a-axis (red) and the b-axis (blue)strain. For reference, the temperature dependence of resistance (black) isplotted on the right axis.J. Phys. Soc. Jpn. 94, 053702 (2025) Letters T. Konoike et al.053702-4 ©2025 The Physical Society of Japan©2025 The Author(s)J. Phys. Soc. Jpn.Downloaded from journals.jps.jp by （研）物質・材料研究機構 on 05/07/25https://doi.org/10.1063/1.4881611https://doi.org/10.1126/science.1248292https://doi.org/10.1063/1.5063729https://doi.org/10.1063/1.5063729https://doi.org/10.1038/s41467-020-17913-yhttps://doi.org/10.1038/s41467-020-17913-yhttps://doi.org/10.1103/PhysRevX.11.021038https://doi.org/10.1103/PhysRevX.11.021038https://doi.org/10.1063/1.1150180https://doi.org/10.1063/1.1150180https://doi.org/10.1002/adma.201305797https://doi.org/10.1063/5.0141023https://doi.org/10.1143/JPSJ.75.054705https://doi.org/10.1143/JPSJ.75.054705https://doi.org/10.1143/JPSJ.75.051010https://doi.org/10.1143/JPSJ.75.051010https://doi.org/10.1080/00268948408078687https://doi.org/10.1080/00268948408072071https://doi.org/10.1080/00268948408072071https://doi.org/10.1143/JPSJ.64.1877https://doi.org/10.1143/JPSJ.69.805https://doi.org/10.1016/S0022-3697(00)00173-6https://doi.org/10.1016/S0379-6779(02)00404-6https://doi.org/10.1016/0379-6779(93)90409-Phttps://doi.org/10.1143/JPSJ.69.543https://doi.org/10.1143/JPSJ.69.543https://doi.org/10.1143/JPSJ.71.1832https://doi.org/10.1143/JPSJ.76.113702https://doi.org/10.1007/BF02198145https://doi.org/10.1007/BF02198145https://doi.org/10.1016/0379-6779(93)90444-2https://doi.org/10.1016/0038-1098(91)90647-Ehttps://doi.org/10.1126/sciadv.aau9123https://doi.org/10.1103/PhysRevLett.98.097402https://doi.org/10.1103/PhysRevLett.98.097402https://doi.org/10.1103/PhysRevB.77.125131