WorksPhysics and Astronomy
Generating Two-dimensional Ferromagnetic Charge Density Waves via External Fields
Abstract
Two-dimensional (2D) ferromagnetic charge density wave (CDW), an exotic quantum state for exploring the intertwining effect between correlated charge and spin orders in 2D limit, has not been discovered in the experiments yet. Here, we propose a feasible strategy to realize 2D ferromagnetic CDWs under external fields, which is demonstrated in monolayer VSe using first-principles calculations. Under external tensile strain, two novel ferromagnetic CDWs ( and 2 CDWs) can be generated, accompanied by distinguishable lattice reconstructions of magnetic V atoms. Remarkably, because the driving forces for generating these two ferromagnetic CDWs are strongly spin-dependent, fundamentally different from that in conventional CDWs, the and 2 CDWs can exhibit two dramatically different half-metallic phases under a large strain range, along with either a flat band or a Dirac cone around Fermi level. Our proposed strategy and material demonstration may open a door to generate and manipulate correlation effect between collective charge and spin orders via external fields.
The reduced dimensionality in two-dimensional (2D) materials leads to enhanced correlation effects, which is beneficial for generating multiple symmetry-breaking orders. Of particular interests are the coexistence or competition between charge density wave (CDW) with superconductivity and magnetism in 2D limit [1, 2, 3, 4]. While the coexistence of CDWs and superconductivity has been observed in some 2D transition-metal dichalcogenides (TMDs) [1, 2] and consequently stimulated intensive attentions in many other correlated systems [5, 6], the coexistence of CDWs and ferromagnetism in 2D systems, forming 2D ferromagnetic CDWs, has not been experimentally observed yet. Fundamentally, these unique time-reversal symmetry-breaking CDWs themselves or in combination with superconductivity and topology may contribute to a series of long-sought physics phenomena, e.g., unconventional anomalous Hall effect, chiral charge order, and Kondo lattice [5, 6, 7, 8]. Practically, the 100% spin-polarized ferromagnetic CDW, i.e., half-metallic CDW, could be an ideal platform to realize novel CDW-controlled metal-insulator transition in single spin channel for exotic spintronics and information storages [9], beyond the conventional nonmagnetic CDWs [10].
The realization of 2D ferromagnetic CDWs is tacitly accepted to be unlikely, mainly due to the mutually exclusive energy gain accounting for the formation of CDW and ferromagnetism: (i) forming CDWs usually reduces the density of states (DOS) at Fermi level (), accordingly decreasing possibility for forming Stoner itinerant ferromagnetism; (ii) forming ferromagnetism causes splitting of energy bands around , reshaping topology of Fermi surface (FS) and weakening the condition for FS nesting (FSN), one possible origin of CDWs [11, 12]. As a typical example, monolayer (ML) VSe is able to exhibit multiple CDW phases [13, 14, 15, 16, 17, 18], room-temperature magnet [19, 20], and even possible superconductivity [21]. However, after more than three decades of studies [3, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23], ferromagnetic CDWs have never been observed in VSe. A similar situation is also found in other 2D TMD systems [1, 2, 3]. Until now, a feasible strategy to realize 2D ferromagnetic CDW is still lacking, preventing the understanding of new correlation effects between collective charge and spin orders induced novel physics phenomena in monolayer limit.
A Strategy for Generating Ferromagnetic CDW. Here, we propose a feasible strategy to realize ferromagnetic CDW under external field, following by two possible criteria: (i) the system should exhibit ferromagnetic configuration and its competing nonmagnetic CDW as groundstate and metastable state, respectively, making it a possible platform for realizing new complex ground states (left panel, Figure 1); (ii) an external field could induce dynamic instability of this ferromagnetic configuration but maintains its total energy lower than that of nonmagnetic CDW; the dynamic instability increases as the external field increases. Once criteria (i)-(ii) are satisfied, upon certain external field a sufficiently strong dynamical instability may trigger a new CDW coexisting with ferromagnetic order (middle panel, Figure 1), as long as ferromagnetic order can survive the lattice reconstruction. Importantly, during the formation of ferromagnetic CDW, an additional criterion [criterion (iii)] may be required to form half-metallic CDW, that is, the major driving force for ferromagnetic CDW formation should be strongly spin-dependent. As a result, significantly different band renormalization could occur in different spin channels, providing a key ingredient for forming half-metallicity (right panel, Figure 1). In practice, the 2D systems are ideal platforms to realize this strategy due to their convenience to be applied with various external fields, e.g., strain [24, 25], electric [26, 27], and irradiation [28, 29] fields.

Figure 1. A strategy for generating ferromagnetic CDW. Criterion (i) requires the system holding ferromagnetic (FM) phase and nonmagnetic (NM) CDW as the groundstate and metastable state, respectively. Criterion (ii) represents an external field may trigger a new FM CDW formation. Criterion (iii) is an additional requirement for half-metallic (HM) CDW formation. See text for more explanation.
In this study, using first-principles calculations (see Methods [30]), we propose that ML-VSe is such an ideal platform to generate ferromagnetic CDWs via our strategy. First, ML-VSe holds ferromagnetic 1T phase (nonmagnetic CDWs) as groundstate (metastable states), satisfying criterion (i). Second, under in-plane tensile strain (), the total energy of ferromagnetic 1T phase maintains to be lower than that of nonmagnetic CDWs, but its dynamic instability increases as increases, satisfying criterion (ii). Therefore, two ferromagnetic CDW orders, and CDWs, can be generated in ML-VSe under certain critical . Importantly, the momentum-dependent electron-phonon coupling (MEPC) and FSN, major driving forces for forming these two ferromagnetic CDWs, are strongly spin-dependent, eventually satisfying criterion (iii). Consequently, the and CDWs exhibit fascinating half-metallic phases under a large range of . In particular, the CDW possesses A-type half-metallic state with a flat band around , whereas CDW holds B-type half-metallic state with a clean Dirac cone at .
Generating Ferromagnetic CDWs in ML-VSe
The structure of ML 1T-VSe has a space group of , where one V layer is sandwiched by two Se layers. The calculated lattice constant of ML 1T-VSe (3.34 Å) agrees well with the experimentally measured one (∼3.35 Å) [20]. Without external field, ferromagnetic 1T phase is the groundstate, whose total energy is ∼24 meV/f.u. lower than that of nonmagnetic one. This ferromagnetic 1T phase has been observed by some experiments at room-temperature [19, 20]. On the other hand, the phonon spectrum of nonmagnetic 1T phase shows strong instability with imaginary phonon modes near 1/2 Γ-M and 3/5 Γ-K (Fig. S1 [30]), indicating the existence of and CDWs (Table S1 [30]), respectively. Indeed, these nonmagnetic CDWs have also been observed in experiments [13, 14, 15, 16, 17, 18]. Importantly, the calculated total energy of ferromagnetic 1T phase is ∼13.8 and ∼11.7 meV/f.u. lower than that of nonmagnetic and CDWs, respectively, which indicates that ML-VSe meets the criterion (i) for forming ferromagnetic CDWs.
To check whether criterion (ii) can be simultaneously satisfied, we have further calculated the energy and dynamic instability of ferromagnetic 1T phase under external strain field. As shown in Figure 2(a), the energy difference () between ferromagnetic 1T phase and these two nonmagnetic CDWs increases (decreases) as a function of tensile (compressive) , indicating that the energetic stability of corresponding ferromagnetic 1T phase can be enhanced under tensile . Surprisingly, when tensile is applied, although the total energy of ferromagnetic 1T phase can be further lowered with respect to the nonmagnetic CDWs, its dynamic instability increases, as indicated by the calculated phonon spectra in Figure 2(b). The larger the , the softer the phonon modes. Therefore, ML-VSe can also meet the criterion (ii) for forming ferromagnetic CDWs. Particularly, when =5%, two imaginary phonon modes appear, i.e., one is located along M-K-Γ and the other is located at K. The maximum at M (K) suggests a () CDW, while maximum near 3/4 Γ-K suggests a CDW (Table S1 [30]).
Figure 2. Strain-tunable ferromagnetic CDWs in ML-VSe. (a) Calculated total energies of FM 1T phase and two NM CDWs in ML-VSe as a function of strain. (b) Lowest three branches of phonon spectra of FM 1T phase under three typical tensile strain. (c) Top views of () and () configurations. Primitive cells are marked as dashed-lines. Variable magnetic moments in V atoms are illustrated with different colors (in unit of ). (d) Total energies of FM 1T and various CDW phases as a function of tensile strain. (e) Energy diagram of strain-tunable CDW phase transitions.
While it is found that CDW is dynamically unstable with a spontaneous transformation to CDW within a doubled supercell, both and CDWs are dynamically stable under tensile strain (Fig. S2 [30]). Importantly, the ferromagnetic order of both CDWs can survive the formation of CDW. As shown in Figure 2(c), the CDW (space group: ) is formed by anti-trimerization of V atoms in 1T phase while the CDW (space group: ) is formed by the reconstruction of V atoms into a zigzag stripe. Besides, the local magnetic moments of V atoms are uniformly distributed in a CDW but exhibit a novel spin oscillation along axis in CDW (Figure 2(c)).
In Figure 2(d), we have calculated the total energies of these ferromagnetic CDWs as a function of tensile , compared with ferromagnetic 1T phase and nonmagnetic CDW. Interestingly, -tunable multiple phase transitions between ferromagnetic 1T and ferromagnetic CDWs are observed. When , the ferromagnetic 1T is groundstate, while ferromagnetic @bibliography CDWs cannot exist [I in Figure 2(e)]. When , CDW appears as a metastable state [II in Figure 2(e)]. When , CDW emerges as another metastable state [III in Figure 2(e)]. Importantly, when , the first groundstate phase transition occurs from 1T to CDW [IV in Figure 2(e)], i.e., the ferromagnetic CDW is now stabilized as the groundstate. When , the second groundstate phase transition occurs from CDW to CDW; meanwhile, the 1T phase is no longer a metastable state [V in Figure 2(e)]. On the other hand, the between nonmagnetic CDW and these two ferromagnetic CDWs increase as increases, i.e., the larger the , the stronger the instability of nonmagnetic CDW.
We have further confirmed the energetic stability of ferromagnetic orders in and CDWs by comparing them with other typical antiferromagnetic orders (Fig. S3 and Table S2 [30]). It turns out that ferromagnetic orders are always energetically favorable for both and CDWs under tensile strain. The larger results in more stable ferromagnetic orders. Interestingly, calculated magnetocrystalline anisotropy energy (MAE) shows that the easy axis of ferromagnetic CDWs is along in-plane direction, similar to CrCl [31]; and the larger the , the larger the MAE. These results indicate that a larger can induce a higher in these ferromagnetic CDWs.
Electronic Structures of Ferromagnetic CDWs
It is important to further understand the electronic structures of these two ferromagnetic CDWs. Here spin-polarization ratio is defined as [32], where and are the DOS values at in spin (majority) and spin (minority) channels, respectively. The calculated spin-dependent for groundstates is summarized in Figure 3(a). When , the of ferromagnetic 1T phase is very small () and insensitive to . Remarkably, when , the of CDW keeps to be , denoted as A-type half-metallicity [majority (minority) spin channel is insulating (metallic)]. When , the bandgap in spin () channel of CDW gradually decreases (increases) (Fig. S4 [30]), along with gradually changing from to . When , the of CDW keeps to be , denoted as B-type half-metallicity [majority (minority) spin channel is conducting (insulating)]. In practice, these two different types of half-metallic phases can be indistinguishable using the spin-resolved STM measurements (see Fig. S5 [30] and related discussion).

Figure 3. Electronic structures of ferromagnetic CDWs. (a) Spin-polarization ratio as a function of strain for different groundstates. Unfolded band structures and DOS for CDW () in (b) spin and (c) spin channels. (d)-(e): same as (b)-(c) but for CDW (). Band structures of FM 1T VSe are also plotted here for comparison. Fermi level is set to zero.
In the following, the () and () CDWs are selected to explore their unusual half-metallic states. For CDW, in spin channel [Figure 3(b)], the hole pocket centered at K in FM-1T phase is largely suppressed and pushed down to a lower energy position during the CDW formation, forming a Mexican-Hat-shape band dispersion with a large DOS peak at VBM. Importantly, the strong band renormalization results in a unusual flat band, composed with twofold V orbitals, in the bottom of conduction band along M-K-, reflected by the very sharp DOS peak at CBM. The formation of this flat-band might be due to the anti-trimerization of V atoms in a triangular-like frustrated lattice [33]. Furthermore, this flat band can gradually shift down to under larger tensile and electron doping (Fig. S6 [30]). In spin channel [Figure 3(c)], the energy bands are much less changed during the CDW formation, i.e., the electron pocket centered at M is slightly pushed up. Eventually, an A-type half-metallic phase is formed. Interestingly, DOS at in spin () channel is fully suppressed (remarkably increased) during the CDW formation, invaliding the common expectation that CDW formation is accompanied by DOS reduction at [34, 35, 36]. This highly asymmetrical band renormalization in spin and channels result in a large half-metallic bandgap eV.
Figs. 3(d) and 3(e) shows the calculated spin and band structures of CDW, respectively. A larger band renormalization also occurs in spin channel than in spin channel during CDW formation, along with largely-reduced (fully-suppressed) DOS at in spin () channel. Accordingly, a B-type half-metallic phase with insulating state located in spin channel is formed in CDW. Unexpectedly, as shown in Figure 3(d), a clean Dirac cone located at M around appears in the unfolded band structure of spin channel, accompanied by the suppression of local bandgap during CDW formation, which may be visible under spin-resolved ARPES measurements [37]. In spin channel [Figure 3(e)], the hole pocket centered at is strongly pushed down by eV, which not only plays a key role in forming a half-metallic bandgap eV, but also introduces a new valley state around .
Origins of Ferromagnetic CDWs under Strain
The driving force behind the CDW transition is an important subject with ongoing controversy [4, 38, 39]. The strong spin-dependent band renormalization in and CDWs indicate that the driving force for these CDW formation may have a strong spin dependence feature. Generally, FSN and/or MEPC, common origins of the observed CDWs in many TMD systems [3, 12], are spin-independent. The real (Re) and imaginary (Im) parts of electron susceptibility reflect the electron instability and FSN of a system [38], respectively. When both Re and Im peak at the same wave vector , a FSN-induced CDW with may be triggered. Furthermore, MEPC can be evaluated by calculating phonon linewidth [40].
Figure 4 shows the calculated spin-resolved Im and of ferromagnetic 1T-VSe under two typical , which indeed exhibit strong spin dependence, satisfying criterion (iii) for forming half-metallic CDWs. For =3%, as shown in Figure 4(a) and Fig. S7 [30], noticeable peaks appear around K in both Re and Im of spin channel, but not of spin channel. Contributed by hole pocket round K [Figure 3(b)], FSN of spin channel rather than spin channel could contribute to the formation of half-metallic CDW (Table S1 [30]). In addition, it may also provide a valid understanding for the much larger band renormalization in spin Figure 3(b) than spin Figure 3(c) channel, a key factor for forming half-metallicity. Meanwhile, as shown in Figure 4(b), high intensity peaks of appear around K in both spin channels, suggesting that MEPC may be another possible reason to form half-metallic CDW. Therefore, the joint MEPC and (spin-dependent) FSN may contribute to the half-metallic CDW formation.

Figure 4. Origins of strain-dependent ferromagnetic CDWs. (a) Spin-resolved Im and (b) (contributed by the lowest phonon mode in the phonon spectrum) for FM 1T-VSe under =3%. (c)-(d): same as (a)-(b) but under =5%. Unit of is THz.
The situation is different for . A hole pocket around K Figure 3(b) of spin channel in ferromagnetic 1T phase under is now converted to electron pocket at M at FS Figure 3(d), changing the FSN condition dramatically. Indeed, as shown in Figure 4(c), the peak in shifts from K to M in spin channel, which, unfortunately, cannot contribute to the CDW formation (Table S1 [30]). In the spectra of Figure 4(d), we can observe three distinctive peaks around M, -K and K in spin channel. The peak at -K in spin channel indicates that the key driving force to form half-metallic CDW may be MEPC. Meanwhile, no noticeable peaks are observed in both and in spin channel. This may account for the observed larger band renormalization in spin Figure 3(d) than spin Figure 3(e) channel. Hence, the spin-selective MEPC, rather than FSN, may play a major role in forming half-metallic CDW. Figure 4 reveals the unusual relationship between CDW transition and half-metallicity formation.
Outlook. In the current experiments, some popular substrates with smaller lattice-constants than ML-VSe are frequently applied for growing ML-VSe [15, 17, 18, 19]. It might induce some compressive to ML-VSe, effectively suppressing the energetic stability of ferromagnetic order and even stabilizing the nonmagnetic CDWs as groundstates Figure 2(a). To meet the criteria to form ferromagnetic CDWs, we suggest that the ML-VSe should be prepared on the substrates with larger lattice-constant, or directly apply tensile strain, e.g., via bending [41]. In general, our strategy for generating ferromagnetic CDWs are valid in many other 2D systems under various external fields, not limited to strain field.
Acknowledgement. The authors thank Prof. Z. Liu for helpful discussions. This work is supported by the NSFC (Grant No. 12088101) and NSAF (Grant No. U1930402). Computations are done at Tianhe-JK supercomputer at CSRC.
References
References
- [1]T. Kiss, T. Yokoya, A. Chainani, S. Shin, T. Hanaguri, M. Nohara, and H. Takagi, Nat. Phys. 3, 720 (2007).
- [2]M. M. Ugeda, A. J. Bradley, Y. Zhang, S. Onishi, Y. Chen, W. Ruan, C. Ojeda-Aristizabal, H. Ryu, M. T. Edmonds, H.-Z. Tsai, A. Riss, S.-K. Mo, D. Lee, A. Zettl, Z. Hussain, Z.-X. Shen, and M. F. Crommie, Nat. Phys. 12, 92 (2015).
- [3]S. Manzeli, D. Ovchinnikov, D. Pasquier, O. V. Yazyev, and A. Kis, Nat. Rev. Mater. 2, 17033 (2017).
- [4]B.-X. Zheng, C.-M. Chung, P. Corboz, G. Ehlers, M.-P. Qin, R. M. Noack, H. Shi, S. R. White, S. Zhang, and G. K.-L. Chan, Science 358, 1155 (2017).
- [5]T. Neupert, M. M. Denner, J.-X. Yin, R. Thomale, and M. Z. Hasan, Nat. Phys. 18, 137 (2021).
- [6]C. Mielke, D. Das, J.-X. Yin, H. Liu, R. Gupta, Y.-X. Jiang, M. Medarde, X. Wu, H. C. Lei, J. Chang, P. Dai, Q. Si, H. Miao, R. Thomale, T. Neupert, Y. Shi, R. Khasanov, M. Z. Hasan, H. Luetkens, and Z. Guguchia, Nature 602, 245 (2022).
- [7]S.-Y. Yang, Y. Wang, B. R. Ortiz, D. Liu, J. Gayles, E. Derunova, R. Gonzalez-Hernandez, L. Šmejkal, Y. Chen, S. S. P. Parkin, S. D. Wilson, E. S. Toberer, T. McQueen, and M. N. Ali, Sci. Adv. 6, eabb6003 (2020).
- [8]V. Vaño, M. Amini, S. C. Ganguli, G. Chen, J. L. Lado, S. Kezilebieke, and P. Liljeroth, Nature 599, 582 (2021).
- [9]I. Žutić, J. Fabian, and S. D. Sarma, Rev. Mod. Phys. 76, 323 (2004).
- [10]M. Yoshida, R. Suzuki, Y. Zhang, M. Nakano, and Y. Iwasa, Sci. Adv. 1, e1500606 (2015).
- [11]S.-K. Chan and V. Heine, J. Phys. F: Met. Phys. 3, 795 (1973).
- [12]K. Rossnagel, J. Phys.: Condens. Matter 23, 213001 (2011).
- [13]D. Zhang, J. Ha, H. Baek, Y.-H. Chan, F. D. Natterer, A. F. Myers, J. D. Schumacher, W. G. Cullen, A. V. Davydov, Y. Kuk, M. Y. Chou, N. B. Zhitenov, and J. A. Stroscio, Phys. Rev. Mater. 1, 024005 (2017).
- [14]P. Chen, W. W. Pai, Y.-H. Chan, V. Madhavan, M. Chou, S.-K. Mo, A.-V. Fedorov, and T.-C. Chiang, Phys. Rev. Lett. 121, 196402 (2018).
- [15]J. Feng, D. Biswas, A. Rajan, M. D. Watson, F. Mazzola, O. J. Clark, K. Underwood, I. Marković, M. McLaren, A. Hunter, D. M. Burn, L. B. Duffy, S. Barua, G. Balakrishnan, F. Bertran, P. L. Fèvre, T. K. Kim, G. van der Laan, T. Hesjedal, P. Wahl, and P. D. C. King, Nano Lett. 18, 4493 (2018).
- [16]G. Duvjir, B. K. Choi, I. Jang, S. Ulstrup, S. Kang, T. T. Ly, S. Kim, Y. H. Choi, C. Jozwiak, A. Bostwick, E. Rotenberg, J.-G. Park, R. Sankar, K.-S. Kim, J. Kim, and Y. J. Chang, Nano Lett. 18, 5432 (2018).
- [17]A. O. Fumega, M. Gobbi, P. Dreher, W. Wan, C. González-Orellana, M. Peña-Díaz, C. Rogero, J. Herrero-Martín, P. Gargiani, M. Ilyn, M. M. Ugeda, V. Pardo, and S. Blanco-Canosa, J. Phys. Chem. C 123, 27802 (2019).
- [18]P. M. Coelho, K. N. Cong, M. Bonilla, S. Kolekar, M.-H. Phan, J. Avila, M. C. Asensio, I. I. Oleynik, and M. Batizll, J. Phys. Chem. C 123, 14089 (2019).
- [19]M. Bonilla, S. Kolekar, Y. Ma, H. C. Diaz, V. Kalappattil, R. Das, T. Eggers, H. R. Gutierrez, M.-H. Phan, and M. Batizll, Nat. Nanotechnol. 13, 289 (2018).
- [20]W. Yu, J. Li, T. S. Herng, Z. Wang, X. Zhao, X. Chi, W. Fu, I. Abdelwahab, J. Zhou, J. Dan, Z. Chen, Z. Chen, Z. Li, J. Lu, S. J. Pennycook, Y. P. Feng, J. Ding, and K. P. Loh, Adv. Mater. 31, 1903779 (2019).
- [21]T. Yilmaz, E. Vescovo, J. T. Sadowski, and B. Sinkovic, Spectroscopic evidence of high temperature superconductivity in VSe₂, ArXiv:2110.05652 (2021).DOI
- [22]C. van Bruggen and C. Haas, Solid State Commun. 20, 251 (1976).
- [23]R. Chua, J. Henke, S. Saha, Y. Huang, J. Gou, X. He, T. Das, J. van Wezel, A. Soumyanarayanan, and A. T. S. Wee, ACS Nano 16, 783 (2021).
- [24]C. Lee, X. Wei, J. W. Kysar, and J. Hone, Science 321, 385 (2008).
- [25]F. Guinea, M. I. Katsnelson, and A. K. Geim, Nat. Phys. 6, 30 (2009).
- [26]K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva, and A. A. Firsov, Science 306, 666 (2004).
- [27]Y. Zhang, T.-T. Tang, C. Girit, Z. Hao, M. C. Martin, A. Zettl, M. F. Crommie, Y. R. Shen, and F. Wang, Nature 459, 820 (2009).
- [28]S. Cho, S. Kim, J. H. Kim, J. Zhao, J. Seok, D. H. Keum, J. Baik, D.-H. Choe, K. J. Chang, K. Suenaga, S. W. Kim, Y. H. Lee, and H. Yang, Science 349, 625 (2015).
- [29]Z. Sun, A. Martinez, and F. Wang, Nat. Photonics 10, 227 (2016).
- [30]See Supplemental Material for more information about computational method, relation between CDW supercell and primitive cell, phonon spectrum of NM 1T phase and FM CDWs, AFM configurations, electronic structure of FM CDWs, real part of electron susceptibility and full STM simulation images, which includes Refs. [42-51].
- [31]A. Bedoya-Pinto, J.-R. Ji, A. K. Pandeya, P. Gargiani, M. Valvidares, P. Sessi, J. M. Taylor, F. Radu, K. Chang, and S. S. P. Parkin, Science 374, 616 (2021).
- [32]Y. Sun, Z. Zhuo, X. Wu, and J. Yang, Nano Lett. 17, 2771 (2017).
- [33]D. Călugăru, A. Chew, L. Elcoro, Y. Xu, N. Regnault, Z.-D. Song, and B. A. Bernevig, Nat. Phys. 18, 185 (2021).
- [34]C.-S. Lian, C. Si, and W. Duan, Nano Lett. 18, 2924 (2018).
- [35]H. Tan, Y. Liu, Z. Wang, and B. Yan, Phys. Rev. Lett. 127, 046401 (2021).
- [36]S. Seong, H. Kim, K. Kim, B. I. Min, Y. S. Kwon, S. W. Han, B.-G. Park, R. Stania, Y. Seo, and J.-S. Kang, Phys. Rev. B 104, 195153 (2021).
- [37]B. Lv, T. Qian, and H. Ding, Nat. Rev. Phys. 1, 609 (2019).
- [38]M. D. Johannes and I. I. Mazin, Phys. Rev. B 77, 165135 (2008).
- [39]X. Zhu, Y. Cao, J. Zhang, E. W. Plummer, and J. Guo, Proc. Natl. Acad. Sci. USA 112, 2367 (2015).
- [40]J. Diego, A. H. Said, S. K. Mahatha, R. Bianco, L. Monacelli, M. Calandra, F. Mauri, K. Rossnagel, I. Errea, and S. Blanco-Canosa, Nat. Commun. 12, 598 (2021).
- [41]E. Han, J. Yu, E. Annevelink, J. Son, D. A. Kang, K. Watanabe, T. Taniguchi, E. Ertekin, P. Y. Huang, and A. M. van der Zande, Nat. Mater. 19, 305 (2019).
