Pure spin currents induced by asymmetric H-passivation in B₃C₂P₃ nanoribbons

Abstract

Inspired by the recently reported novel two-dimensional material B₃C₂P₃, we performed one-dimensional shearing along the zigzag direction to obtain four B₃C₂P₃ nanoribbons with various edge atom combinations. An asymmetric hydrogen passivation scheme was employed to modulate the electronic properties and successfully open the band gap, especially the 2H-1H passivation with dihydrogenation and monohydrogenation at the top and bottom edges, respectively, achieving bipolar magnetic semiconductors with edge P-atoms contributing to the main magnetism. Furthermore, three crucial spin-polarized transmission spectra yielded a significant spin-dependent Seebeck effect (SDSE), displaying superior thermoelectric conversion capabilities by generating pure spin currents. Our work shows that this asymmetric H-passivation effectively enables the enhancement of the spin caloritronic transport properties of the B₃C₂P₃, which is of great significance for the exploitation of novel materials and their applications in spintronics.

---

Introduction

Nowadays, the demand for powerful and low energy-consuming electronic devices has been growing increasingly, hence how to reduce the energy loss triggered by waste heat generated during the operation of electronic devices needs to be solved urgently.^1–5 Spin caloritronics adds a spin dimension to the conventional thermoelectric effect and generates a variety of spin-dependent thermoelectric effects through the interactions between spin, charge, and heat transfer, improving the thermoelectric conversion efficiency.⁶ Among them, the spin-dependent Seebeck effect (SDSE) stands out for being driven only by temperature differences to amplify spin currents while suppressing charge currents.^7,8 Ideally, the currents of the different spin components are opposite in direction and equal in magnitude, achieving perfect thermoelectric conversion with pure spin currents and zero charge currents.

However, in previous reports, spin caloritronic devices have been mostly designed in single-element regular materials such as carbon-based^9–12 and silicon-based ones.^13–16 Yet, due to the high plasticity of light atomic mass nonmetallic elements such as carbon, silicon, boron, and phosphorus,^17–19 it is more advisable to dig into new diversified hybrid materials consisting of them, such as phosphorus carbide, etc.^20–22 Diversified combinations of mainstream elements in low-dimensional materials into function-rich novel hybrid materials can provide a wider choice of materials in the thermoelectric field, thereby opening up ideas for designing better spin caloritronic devices.

Recently, based on the geometry of the 2D phosphorus carbide,²³ Kistanov²⁴et al. predicted a novel 2D graphene-like material, B₃C₂P₃, which was shown to be structurally and thermodynamically stable, exhibiting excellent stability in oxygen and water environments and have great potential to be prepared by carbon doping techniques^25,26 in boron phosphide,^27–29 or direct arc discharge and laser ablation vaporization.³⁰ Given the highly tunable direct bandgap under strain and substrate engineering action, as well as excellent electronic and mechanical properties of B₃C₂P₃, both Wu³¹et al. and the Al-Jayyousi group³² explored its promise as an anchoring material. Additionally, the linear photogalvanic effects (PGEs) in the B₃C₂P₃ photodetector have been recently investigated by Fu³³et al. All these works demonstrate the great potential of B₃C₂P₃ for spintronics applications and thus inspire us to explore the possibilities of this novel hybrid material applied in spin caloritronics.

In this paper, the B₃C₂P₃ monolayer is one-dimensionally tailored in the zigzag direction to obtain four zigzag B₃C₂P₃ nanoribbons with different edge-atom combinations, altering the eigen-electronic properties of the material by marginalising the atoms and simultaneously enhancing its magnetic properties. On this basis, we propose eight schemes for passivating the edges of these nanoribbons with different numbers of H atoms, bringing different electronic structures and magnetic characteristics to B₃C₂P₃. Surprisingly, it is found that, when saturated dihydrogen is at the top edge and monohydrogen is at the bottom edge, the B₃C₂P₃ nanoribbons behave as bipolar magnetic semiconductors (BMSs) in the ferromagnetic state, which exhibit exceptional spin caloritronics properties and achieve superior SDSE. Therefore, our scheme of this edge-asymmetric H-passivation makes B₃C₂P₃ a promising reserve player in spin caloritronics.

Models and methods

The two-dimensional B₃C₂P₃ material (see Fig. S1, ESI†) is a planar honeycomb structure that looks like BP nanoribbons connected by C atoms. For graphene-like structures, zigzag direction cutting is more likely to introduce magnetism compared to armchair direction cutting. Therefore, we tailor the B₃C₂P₃ monolayer along the zigzag direction, exposing the outermost edge atoms. Taking nanoribbons with widths of 4–6 atomic chain widths as examples, four bare zigzag B₃C₂P₃ nanoribbons with different edge configurations can be obtained based on the combinations of the outermost atoms at the top and bottom edges, as shown in Fig. 1(a). Furthermore, we have designed eight H-passivation schemes for each edge configuration. According to different passivation schemes, the 1s orbitals of H atoms hybridize with the 2p orbitals of B and C atoms as well as the 3p orbitals of P atoms at the edges in different degrees, resulting in variations of the electronic states and total magnetic moments of the B₃C₂P₃ nanoribbons, thereby modulating the B₃C₂P₃ nanoribbons with different electronic and magnetic characteristics. Taking the zCB_CP edge configuration as an example, the xy-plane side views of the H-passivated structures are presented in Fig. 1(b), which are named nH-zCB_CP_mH, with n/m referring to the number of passivated H-atoms at the top/bottom edges.

Fig. 1 (a) Bare zigzag nanoribbons in four edge configurations. (b) Eight H-passivation schemes exemplified by zCB_CP. (c) and (d) Front and top views of a spin caloritronic device exemplified by 2H-zCB_CP-1H. All vacuum layers are not drawn here for display convenience.

There are a total of 36 bare and hydrogen-passivated nanoribbons, and their total energies E_total in NFM, FM, and AFM states are calculated in Table S1 (ESI†), where it can be found that the ground state of individual structures is the NM state, most of them are in the FM state, and the rest are in the AFM ground state, but the difference in energies between those in the AFM and the FM is not very large. Since the FM states can be stabilized by external magnetic fields in experiments and applications,^34,35 all the work in this paper focuses only on the performance modulation in the FM state. For further calculations of the spin transport properties, spin caloritronic devices are designed, exemplified by the 2H-zCB_CP-1H configuration, as shown in Fig. 1(c) and (d), which consists of a central scattering region and two left and right semi-infinite electrodes with the same structure as it, without any other chemical interface between them in gapless connection.

These structures are fully geometrically relaxed by the DS-PAW (device studio-projector augmented wave)³⁶ procedure based on a plane wave basis set, reaching the most structurally stable state with forces less than 0.01 eV Å⁻¹ between each atom. The electronic and transport properties are calculated by the first principles calculation software Nanodcal³⁷ package, based on the double-zeta polarized (DZP) atomic orbital basis set³⁸ combining density-functional theory (DFT)^39,40 with the non-equilibrium Green's function (NEGF).^41,42 The exchange correlation energy is described by the Perdew–Burke–Ernzerhof (PBE) functional within the generalized gradient approximation (GGA).⁴³ The norm-conserving pseudopotential is used to define the atomic cores.⁴⁴ The mesh grids in k-space for calculating the transport properties are set to 1 × 1 × 100. The vacuum layers in the non-periodic direction of these structures are all larger than 15 Å excluding interactions between atomic layers.

Different temperatures are applied at the left and right terminals of the device without bias voltage, and the spin-resolved current driven by temperature difference is defined according to the Landauer–Büttiker formula:⁴⁵

where a negative value e is the electron charge, h denotes the Planck constant, T_L/R denotes the temperature of the left/right electrodes, and E is the energy. f_L/R denotes the Fermi Dirac distribution of the left and right electrodes in the non-equilibrium state, which is determined by eqn (2):⁴⁶where k_B is the Boltzmann constant. τ_↑/↓ is the spin-resolved transmission coefficient of the bipolar system, which is calculated as:⁴⁷

τ_↑/↓(E) = Tr[Γ_L(E)G^R_↑/↓(E)Γ_R(E)G^A_↑/↓] (3)

where Γ_L/R and G^R/A_↑/↓ denote the broadening function of the coupling between the central scattering region and the left/right electrodes and the retarded/advanced Green's function of the central scattering region, respectively.^48,49 The intermediate function L_n,↑/↓ is introduced and defined as follows:³⁴where n = 0, 1, then the spin-dependent Seebeck coefficient in the linear response state is calculated as:

The charge Seebeck coefficient S_c and spin Seebeck coefficient S_s are expressed as and S_s = S_↑ − S_↓, respectively. When different temperatures T_L and T_R are applied to the left and right ends of the device, the spin current formed by the spin voltage generated by the temperature gradient ΔT = T_L − T_R is denoted as I_s = I_↑ − I_↓, and the charge current formed by the charge voltage is denoted as I_c = I_↑ + I_↓. S_s(S_c) indicates the ability to convert the temperature gradient into spin (charge) voltage.⁵⁰

Results and discussion

Spintronic

The spin-resolved bands determine the spintronic properties. The first edge configuration, zCB_CP, was initially investigated by calculating the band structures of its bare and eight hydrogenated ones in the FM state, as shown in Fig. 2(a)–(i). Apparently, after self-consistent iteration with the FM state as the initial spin state, all other structures converge to the FM state, except for 2H-zCB_CP-2H in Fig. 2(i) which converges to the nonmagnetic state, showing nonmagnetism. Among them, bare zCB_CP exhibits magnetic metallic properties, and when its lower edge is saturated with single hydrogen (i.e., 0H-zCB_CP-1H), the C and P atoms at the lower edge shift from sp-hybridization to sp²-hybridization, but the band morphology near E_f does not undergo a significant change. Similarly, from 1H-zCB_CP-0H to 1H-zCB_CP-1H, the bottom edge turns from bare to monohydrogenated, and this change still does not have much effect on the spin-splitting band near E_f. Thus, for zCB_CP, saturation with monohydrogen on the bare CP bottom edge essentially does not change the electronic properties of the structure. This conclusion also holds true in the case of the dihydrogenated top edge, from 2H-zCB_CP-0H to 2H-zCB_CP-1H, both of which are semiconductors. However, by comparing these three pairs (from zCB_CP to 1H-zCB_CP-0H, from 0H-zCB_CP-1H to 1H-zCB_CP-1H, and from 0H-zCB_CP-2H to 1H-zCB_CP-2H, whose bare top edges are monohydrogenated), it can be seen that, regardless of whether the bottom edges are bare or mono- or di-hydrogenated, as long as the CB top edge is monohydrogenated, the structure is metallic. Furthermore, no matter whether the top or bottom edges are dihydrogenated, such as 2H-zCB_CP-0H and 0H-zCB_CP-2H, or both the top and bottom edges are dihydrogenated, i.e., 2H-zCB_CP-2H, as long as one of the edges is dihydrogenated, the configuration tends to be semiconducting. When this semiconducting dihydrogen saturation coexists with the metallic top-edge monohydrogenation (i.e., 1H-zCB_CP-2H), the bands traversing E_f are still present and are contained within the two semiconducting subbands, see Fig. 2(g), and thus 1H-zCB_CP-2H combines both top and bottom edge properties. For the second edge configuration zCB_P, its bare ribbon is also metallic (see Fig. 3(a)), and since its top edge is the same as that of zCB_CP, the dominant effect of hydrogen saturation at the top edge on the electronic properties remains unchanged, i.e., when the top edge is monohydrogenated, 1H-zCB_P-0H, 1H-zCB_P-1H, and 1H-zCB_P-2H show metallic properties, and when the top edge is dihydrogenated, the band gap at E_f opens and shows semiconducting properties, such as 2H-zCB_P-0H, 2H-zCB_P-1H, and 2H-zCB_P-2H. However, unlike zCB_CP, bare zCB_P has only P atoms on the bottom edge, and when these P atoms are passivated by a single H atom, a transition to semiconductors occurs (i.e., 0H-zCB_P-1H). This mechanism combined with the metallicity mechanism driven by the top monohydrogenated CB edge results in 1H-zCB_P-1H behaving as a metal.

Fig. 2 Band structures of zCB_CP for (a) bare and (b)–(i) eight H-passivation schemes in the initial FM state. The black dashed lines represent Fermi level E_f.

Fig. 3 Band structures of (a) zCB_P, (b) zB_CP and (c) zB_P for bare and 8 edge H-passivation schemes in the initial FM state. The black dashed lines represent Fermi level E_f.

For the other two edge configurations depicted in Fig. 3(b) and (c), zB_CP shows a transition from a semimetal to a magnetic metal to a bipolar magnetic semiconductor, whereas for zB_P, the system undergoes spin-splitting only when the top edge is exposed, while it is spin-degenerated in all other cases.

From the above discussion, it can be concluded that among these 36 structures, different edge hydrogen passivation schemes can effectively regulate the spintronic properties of the B₃C₂P₃ nanoribbons, especially the 2H-1H passivation schemes, i.e., 2H-zCB_CP-1H, 2H-zCB_P-1H, and 2H-zB_CP-1H. These three structures show remarkable bipolar magnetic semiconductor properties with the potential for excellent spin-thermoelectric effects, and the thermodynamic stability of the structures was verified by ab initio molecular dynamics (AIMD) simulations (see Fig. S2, ESI†). Thus, the following discussion will focus on these three structures. Fig. 4(a) displays the spin density distribution of 2H-zCB_CP-1H, 2H-zCB_P-1H, and 2H-zB_CP-1H, which represent the spatial distribution of the charge difference between the spin up and down electrons.

Fig. 4 (a) Spin density distribution of 2H-zCB_CP-1H, 2H-zCB_P-1H, and 2H-zB_CP-1H with an isosurface value of 0.02 Å⁻³, with the purple and blue isosurfaces denoting higher spin-up and spin-down charge densities, respectively. (b) Atomic magnetic moment and (c)–(e) projected density of states of 2H-zCB_CP-1H, 2H-zCB_P-1H, and 2H-zB_CP-1H.

Fig. 4(a) displays the spin density distribution of 2H-zCB_CP-1H, 2H-zCB_P-1H, and 2H-zB_CP-1H, which represents the spatial distribution of the charge difference between the spin up and down electrons.⁵¹ Apparently, the spin density of 2H-zCB_CP-1H features a bulk distribution, whereas those of 2H-zCB_P-1H and 2H-zB_CP-1H are localized at the top and bottom edges, respectively, and all of them exhibit significant magnetism. This is also supported by the calculated magnetic moments in Fig. 4(b), where 2H-zCB_CP-1H has the maximum moment of 3.27μ_B, and 2H-zCB_P-1H and 2H-zB_CP-1H have moments of 1.60μ_B and 1.70μ_B, respectively, and it is evident that P atoms in all the three structures account for the largest proportion of the total magnetic moment. Furthermore, in Fig. 4(c)–(e), the projected density of states (PDOS) near E_f are also mainly contributed by P atoms, especially their 3p orbitals. Thus, in all three B₃C₂P₃ nanoribbons, the P atoms are the main source of magnetism and energy states around E_f.

Spin caloritronic transport properties

A particular transmission spectrum is necessary to produce excellent thermoelectric properties. As shown in Fig. 5, in agreement with the previous band structures and DOS, the transmissions of all three structures are suppressed at E_f, creating a transmission gap with spin-up and spin-down transmission channels located on either side of E_f, which is highly favorable for the emergence of SDSE. Specifically, 2H-zCB_CP-1H has the largest transmission gap at E_f and the narrowest transmission peaks on both sides, while 2H-zCB_P-1H has an intermediate transmission gap; noticeably, it has extremely symmetric spin-up and spin-down transmission peaks near E_f, which is more conducive to producing perfect SDSEs. 2H-zB_CP-1H, on the other hand, has the narrowest transmission gap and the widest transmission peaks with the highest transmission coefficient, which is favorable for generating large spin thermal currents. The spin-dependent transmission spectra of the 2H-zCB_CP-1H, 2H-zCB_P-1H, and 2H-zB_CP-1H show the commonality of bipolar magnetic semiconductors, as well as the presence of unique transport properties that differ from each other, which has intrigued us about their behavior in spin-thermoelectric transport.

Fig. 5 Spin-resolved transmission spectra of (a) 2H-zCB_CP-1H, (b) 2H-zCB_P-1H and (c) 2H-zB_CP-1H.

The spin-dependent Seebeck coefficient S_σ, σ = ↑, ↓ characterizes the thermoelectric conversion ability, and eqn (4) and (5) shows that S_σ is related to the transmission coefficient and is also a function of the chemical potential μ and temperature. Fig. 6(a)–(c) plots the S_σversus chemical potential at different temperatures of 2H-zCB_CP-1H, 2H-zCB_P-1H and 2H-zB_CP-1H. As the temperature increases, the S_↑ and S_↓ peaks closest to μ = 0 eV shift to the right and left, respectively, but both peaks gradually decrease. In addition, since the transmission spectra of all three structures analyzed earlier show spin-bipolar magnetic semiconductivity, S_↑ and S_↓ near μ = 0 eV maintain the opposite sign, which leads to the corresponding S_s being much larger than S_c, as can be observed in Fig. 6(d)–(f). At 200 K, the S_s of 2H-zCB_CP-1H reaches a huge peak of 2 mV K⁻¹ with a peak plateau at the interval of [−0.11, 0.12] eV, and the subsequent increase in temperature leads to smaller peaks and wider plateaus. At 100 K, the S_s peak of 2H-zCB_P-1H at μ = 0 eV is 1.5 mV K⁻¹ at, and that of S_s of 2H-zB_CP-1H is the smallest, 730 μV K⁻¹. After S_s reaches the peak, the peak S_s plateau broadens but then decreases as the temperature continues to increase, while the S_c at μ = 0 eV is always zero, and the S_s is much larger than the S_c. All three structures show obvious SDSE.

Fig. 6 S _σ of (a) 2H-zCB_CP-1H, (b) 2H-zCB_P-1H and (c) 2H-zB_CP-1H. S_s/c of (d) 2H-zCB_CP-1H, (e) 2H-zCB_P-1H and (f) 2H-zB_CP-1H.

Applying different temperatures to the left and right electrodes of the device under zero bias, the spin-resolved currents I_σ, σ = ↑, ↓ can be generated driven by the temperature gradient ΔT. As shown in Fig. 7(a)–(c), the temperature of the left electrode is set higher than that of the right electrode, and when the left electrode reaches a certain temperature (i.e., the threshold temperature T_th), the current starts to be produced, and then increases with the increase of temperature. Apparently, the signs of I_↑ and I_↓ are opposite for these three structures, with negative I_↑ and positive I_↓, indicating that significant SDSE is induced in all three B₃C₂P₃ nanoribbon devices. In contrast, 2H-zCB_CP-1H has the highest T_th of 250 K and the smallest current value of ∼10 nA, besides, at ΔT = 40 K, T_L = 450 K and ΔT = 80 K, T_L= 475 K, I_↓ begins to decrease with increasing T_L, resulting in a local negative differential thermoelectric resistance effect (NDTR). The T_th of 2H-zCB_P-1H is 225 K; in particular, I_↓ and I_↑ have the most comparable amplitude with the greatest potential to produce a perfect SDSE. 2H-zB_CP-1H has the smallest T_th of 100 K, indicating that this structure can be available for low-temperature driven devices, and furthermore, it can generate the largest I_σ of 200 nA, and produces a local NDTR at T_L = 450 K, ΔT = 60, 80 K. Comparing the T_th of these three structures, 2H-zCB_CP-1H > 2H-zCB_P-1H > 2H-zB_CP-1H, which is consistent with the trend of the transmission gap, suggesting a positive correlation between T_th and the size of the transmission gap.

Fig. 7 Spin-resolved current I_σ of (a) 2H-zCB_CP-1H, (b) 2H-zCB_P-1H and (c) 2H-zB_CP-1H. Spin/charge currents I_s/c of (d) 2H-zCB_CP-1H, (e) 2H-zCB_P-1H and (f) 2H-zB_CP-1H.

Additionally, the corresponding spin and charge currents of these three structures are calculated in Fig. 7(d)–(f). It is easy to find that after the threshold temperature is reached at the left electrode, I_s of all three structures rise speedily with the increase of T_L, whereas I_c only fluctuates slightly, and such discrepancy becomes more pronounced with the growth of ΔT, which results in the spin currents being much larger than the charge currents and all of them manifest the typical SDSE. Particularly noticeably, the 2H-zCB_P-1H structure exhibits near-perfect SDSE, due to its most spin-symmetric transmission spectrum. When I_s of 2H-zCB_P-1H increases continuously with the left electrode and the temperature difference, the opposite is true for I_c, which is persistently suppressed extremely close to zero, enabling the generation of pure spin currents in 2H-zCB_P-1H. The excellent performance over a wide temperature range ensures that 2H-zCB_P-1H is suitable for a wide range of temperatures, making 2H-zCB_P-1H an ideal spin caloritronic device. On the other hand, 2H-zB_CP-1H generates the largest I_s at the same temperature configuration compared to the other two structures, and attains a huge I_s of 344.6 nA at T_L = 500 K and ΔT = 80 K. Although the resulting I_c is non-negligible, I_s is still much larger than I_c, and does not have to be driven by excessively high temperatures, hence the 2H-zB_CP-1H also performs a remarkable SDSE and is an excellent spin caloritronic device.

The formation mechanism of SDSE is analyzed by taking 2H-zCB_P-1H as an example. According to eqn (1), I_σ is obviously determined by the difference of the Fermi distributions together with the transmission spectrum. As can be seen from Fig. 8(a), above E_f, the Fermi Dirac distribution in the left high-temperature region is larger than that in the right low-temperature region, i.e., the electron concentration in the high-temperature region is higher than that in the low-temperature region, and that what exists above E_f is the spin-down transmission channel (see Fig. 5(c)), so that the electrons diffuse from the high-temperature region to the low-temperature region by means of the spin-down transmission channel, i.e., they flow from left to right, and form a right-to-left I_↓. Similarly, below E_f, the concentration of holes on the left lead is higher than that on the right lead, so holes move from left to right through the spin-up transmission channel, forming a left-to-right I_↑, and thus I_↑ and I_↓ are in opposite directions, resulting in SDSE. In order to further shed light on the current variations, the current spectrum J_σ(E) = τ_σ(E)[f_L(E, T_L) − f_R(E, T_R)], σ = ↑, ↓⁵² is plotted in Fig. 8(b) for different ΔT at T_L = 500 K. Clearly, J_↑ is always opposite in sign to J_↓, and the values of J_σ increase gradually as the ΔT increases. One can also find that there is only one J_σ in the energy range [−0.3, 0.3] eV near E_f owing to the Fermi distribution difference, which implies that I_σ is determined by the transmission coefficients and Fermi distribution around E_f. When the spin up and spin down transmission coefficients on the left and right sides of E_f are exactly equal in a bipolar magnetic semiconductor, I_↑ and I_↓ can be generated in the opposite direction and with exactly equal magnitude, resulting in a large I_s while a zero I_c, thereby yielding an optimal SDSE.

Fig. 8 (a) Left is the Fermi Dirac distribution of T_L = 500 K, right is the Fermi Dirac distribution of T_R = 420 K, and the middle is the difference between the Fermi distributions of the left and right electrodes. (b) Current spectrum of 2H-zCB_P-1H.

Conclusions

In this paper, based on the recently reported two-dimensional B₃C₂P₃, four edge configurations of B₃C₂P₃ nanoribbons are obtained by cutting it along the zigzag direction, including zCB_CP, zCB_P, zB_CP and zB_P. Various edge H-passivation schemes have been proposed to modulate their transport properties. Among them, the asymmetric passivation scheme 2H-1H has resulted in all three structures exhibiting bipolar magnetic semiconductor properties, except for 2H-zB_P-1H, which has spin-degenerated bands. The analytical results show that the magnetism in these three structures originates mainly from edge P-atoms. In addition, the S_s of these three nanoribbons at 0 eV is much larger than S_c, among which the 2H-zCB_CP-1H reaches a huge S_s of 2 mV K⁻¹ at 200 K. Meanwhile, all three nanoribbons have I_↑ in the opposite direction of I_↓, so the I_s generated are much larger than the I_c, and all exhibit excellent SDSE. Therefore, our 2H-1H asymmetric edge H-passivation scheme can effectively modulate the spin caloritronic transport performance of B₃C₂P₃ nanoribbons, which enhances the application value of the novel material B₃C₂P₃ and provides ideas for designing spin caloritronic devices.

Author contributions

Jing-Jing He: conceptualization, methodology, writing – review & editing. Jia-Bei Dong: investigation, data curation, writing – original draft. Ling-Xiao Liu: methodology, data curation, visualization. Qin-Yue Cao: software, validation. Jun-Yi Gu: visualization, formal analysis. Ying Zhang: methodology, resources. Min Hua: visualization, formal analysis. Jia-Ren Yuan: software, supervision. Xiao-Hong Yan: supervision.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grant No. 62201268, 12264026), the National Key Research and Development Program of China (Grant No. 2022YFA1405200), the Natural Science Foundation of Jiangxi Province, China (Grant No. 20224BAB211013), the Innovation and Entrepreneurship Leading Talent Plan of Jiangxi Province (Grant No. Jxsq2023101068), and the China Scholarship Council (CSC Student ID 202408320264). We are grateful to HZWTECH for the support of computing facilities and technical communication.

References

1. Y. Ni, K. L. Yao, H. H. Fu, G. Y. Gao, S. C. Zhu and S. L. Wang, Sci. Rep., 2013, 3, 1380.
2. H. Hao, H. Li, T. Jia, Y. Zhou and X. Zheng, Phys. Chem. Chem. Phys., 2024, 26, 12652–12660.
3. S. Pei, R. Cao, Y.-H. Zhou, X. Zheng and C. Wang, New J. Phys., 2024, 26, 043014.
4. L. Kang, Y. Deng, X. Zheng, P. Jiang, Z. Feng, W. Wang and G. Zhao, Appl. Phys. Lett., 2024, 124, 132902.
5. L. Song, R. Ye, C. Su, C. Wei, D. Chen, X. Liu, X. Zheng and H. Hao, Phys. Rev. B, 2024, 109, 094105.
6. J. Xu, M. Li, Q. Wang, X. Zhang, J. Fei, Y. Shi, B. Zhang and M.-Q. Long, J. Phys. D: Appl. Phys., 2023, 56, 445301.
7. K. Uchida, S. Takahashi, K. Harii, J. Ieda, W. Koshibae, K. Ando, S. Maekawa and E. Saitoh, Nature, 2008, 455, 778–781.
8. J. W. Li, B. Wang and J. M. Wang, Phys. E, 2019, 112, 109–114.
9. X. Deng, Z. Zhang, G. Tang, Z. Fan and C. Yang, Carbon, 2014, 66, 646–653.
10. M. G. Zeng, Y. P. Feng and G. C. Liang, Nano Lett., 2011, 11, 1369–1373.
11. Y. J. Dong, Y. Z. Wu, X. F. Wang, X. F. Yang and Y. S. Liu, J. Phys. Chem. C, 2019, 123, 29126–29132.
12. Q.-B. Liu, D.-D. Wu and H.-H. Fu, Phys. Chem. Chem. Phys., 2017, 19, 27132–27139.
13. K. Zberecki, R. Swirkowicz, M. Wierzbicki and J. Barnaś, Phys. Chem. Chem. Phys., 2014, 16, 12900–12908.
14. S. Kharwar, S. Singh and N. K. Jaiswal, Phys. B, 2021, 615, 413072.
15. P. Jiang, X. Tao, H. Hao, L. Song, X. Zheng and Z. Zeng, RSC Adv., 2017, 7, 28124–28129.
16. H. H. Fu, D. D. Wu, Z. Q. Zhang and L. Gu, Sci. Rep., 2015, 5, 10547.
17. Y. H. Zhou and X. H. Zheng, J. Phys.: Condens. Matter, 2019, 31, 315301.
18. K. Zberecki, R. Swirkowicz and J. Barnas, Phys. Rev. B: Condens. Matter Mater. Phys., 2014, 89, 165419.
19. Y. Chang, A. Nie, S. Yuan, B. Wang, C. Mu, J. Xiang, B. Yang, L. Li, F. Wen and Z. Liu, Nanotechnology, 2018, 30, 035701.
20. Y. Liu, Y. Yang, Y. Qu, Y.-Q. Li, M. Zhao and W. Li, Phys. Chem. Chem. Phys., 2020, 22, 23268–23275.
21. X. Y. Tan, L. N. Zhang and L. L. Liu, Chem. Phys. Lett., 2020, 748, 137386.
22. K. Zberecki, R. Swirkowicz, M. Wierzbicki and J. Barnas, Phys. Chem. Chem. Phys., 2015, 17, 1925–1933.
23. J. Guan, D. Liu, Z. Zhu and D. Tománek, Nano Lett., 2016, 16, 3247–3252.
24. A. A. Kistanov, S. A. Shcherbinin, S. V. Ustiuzhanina, M. Huttula, W. Cao, V. R. Nikitenko and O. V. Prezhdo, J. Phys. Chem. Lett., 2021, 12, 3436–3442.
25. W. C. Tan, Y. Cai, R. J. Ng, L. Huang, X. Feng, G. Zhang, Y. W. Zhang, C. A. Nijhuis, X. Liu and K. W. Ang, Adv. Mater., 2017, 29, 1700503.
26. X. Huang, Y. Cai, X. Feng, W. C. Tan, D. M. N. Hasan, L. Chen, N. Chen, L. Wang, L. Huang, T. J. Duffin, C. A. Nijhuis, Y.-W. Zhang, C. Lee and K.-W. Ang, ACS Photonics, 2018, 5, 3116–3123.
27. A. López-Castillo, Int. J. Quantum Chem., 2012, 112, 3152–3157.
28. J. Dong, H. Li and L. Li, NPG Asia Mater., 2013, 5, e56.
29. S. Huber, V. Medvedev, E. Gullikson, B. Padavala, J. H. Edgar, R. W. van de Kruijs, F. Bijkerk and D. Prendergast, Phys. Chem. Chem. Phys., 2017, 19, 8174–8187.
30. W. L. Wang, X. D. Bai, K. H. Liu, Z. Xu, D. Golberg, Y. Bando and E. G. Wang, J. Am. Chem. Soc., 2006, 128, 6530–6531.
31. J. Wu and Y.-X. Yu, J. Phys. Chem. C, 2022, 126, 12460–12471.
32. H. Al-Jayyousi, M. K. Eswaran, A. Ray, M. Sajjad, J. A. Larsson and N. Singh, ACS Omega, 2022, 7, 38543–38549.
33. X. Fu, J. Lin, X. Cheng, W. Liao, J. Guo, X. Li and L. Li, Mater. Today Commun., 2023, 35, 106175.
34. P. Jiang, X. X. Tao, H. Hao, L. L. Song, X. H. Zheng, L. Zhang and Z. Zeng, 2D Mater., 2017, 4, 035001.
35. L. Lei, Y.-H. Zhou, X. Zheng, W. Wan and W. Wang, Phys. Chem. Chem. Phys., 2024, 26, 3253–3262.
36. P. E. Blöchl, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 50, 17953.
37. J. Taylor, H. Guo and J. Wang, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 63, 245407.
38. C. He, Y.-L. Li, Z.-G. Sun, J.-F. Wang and M.-Y. Chen, Phys. E, 2022, 143, 115333.
39. X. Zheng, Z. Zheng, S. Yang, C.-S. Liu, L. Zhang and H. Hao, Phys. Rev. B, 2024, 110, 115439.
40. X. Zheng, S. Yang, Z. Zheng, C.-S. Liu, W. Wang and L. Zhang, Appl. Phys. Lett., 2024, 125, 223503.
41. L. Song, C. Wei, D. Chen, R. Ye, C. Su, Y. Yang, X. Zheng and H. Hao, Comput. Mater. Sci., 2024, 245, 113290.
42. Z. Han, X. Zheng, C.-S. Liu, L. Zhang, W. Wang and Z. Zeng, IEEE Trans. Electron Dev., 2024, 71, 7169–7176.
43. H. Yu, Y. Li, D. Wei, G. Guo, Z. Feng, J. Ye, Q. Luo, Y. Ma, Y. Tang and X. Dai, Micro Nanostruct., 2022, 165, 207188.
44. J. Yang, X. Yu, Z. Yu and L. Zhang, Europhys. Lett., 2021, 135, 46001.
45. U. Sivan and Y. Imry, Phys. Rev. B: Condens. Matter Mater. Phys., 1986, 33, 551–558.
46. J. Aarons, M. Sarwar, D. Thompsett and C.-K. Skylaris, J. Chem. Phys., 2016, 145, 220901.
47. I. Zutic, J. Fabian and S. Das Sarma, Rev. Mod. Phys., 2004, 76, 323–410.
48. G. Q. Ding, M. Wei, G. Surucu, Z. Y. Liang and X. T. Wang, Appl. Surf. Sci., 2019, 491, 750–756.
49. Y. Jiang, Y. D. Guo, L. Y. Lin and X. H. Yan, Nanoscale, 2022, 14, 10033–10040.
50. P. Jiang, X. Tao, L. Kang, H. Hao, L. Song, J. Lan, X. Zheng, L. Zhang and Z. Zeng, J. Phys. D: Appl. Phys., 2018, 52, 015303.
51. H. Wang, M. Wang, C. Qian, X. Hong, D. Zhang, Y. Liu and X. Yang, Phys. Lett. A, 2017, 381, 1738–1744.
52. X.-J. Gao, P. Zhao and G. Chen, Chem. Phys. Lett., 2018, 699, 250–254.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4cp03033h

---