betaFeedback

WorksPhysics and Astronomy

Magneto-optical anisotropies of 2D antiferromagnetic MPX3_3 from first principles

Abstract

Here we systematically investigate the impact of the spin direction on the electronic and optical properties of transition metal phosphorus trichalcogenides (MPX3_3, M=Mn, Ni, Fe; X=S, Se) exhibiting various antiferromagnetic arrangement within the 2D limit. Our analysis based on the density functional theory and versatile formalism of Bethe-Salpeter equation reveals larger exciton binding energies for MPS3_3 (up to 1.1 eV in air) than MPSe3_3(up to 0.8 eV in air), exceeding the values of transition metal dichalcogenides (TMDs). For the (Mn,Fe)PX3_3 we determine the optically active band edge transitions, revealing that they are sensitive to in-plane magnetic order, irrespective of the type of chalcogen atom. We predict the anistropic effective masses and the type of linear polarization as an important fingerprints for sensing the type of magnetic AFM arrangements. Furthermore, we identify the spin-orientation-dependent features such as the valley splitting, the effective mass of holes, and the exciton binding energy. In particular, we demonstrate that for MnPX3_3 (X=S, Se) a pair of non equivalent K+ and K- points exists yielding the valley splittings that strongly depend on the direction of AFM aligned spins. Notably, for the out-of-plane direction of spins, two distinct peaks are expected to be visible below the absorption onset, whereas one peak should emerge for the in-plane configuration of spins. These spin-dependent features provide an insight into spin flop transitions of 2D materials. Finally, we propose a strategy how the spin valley polarization can be realized in 2D AFM within honeycomb lattice.

Introduction

The subtle interplay between the spin, charge, orbital, and lattice degrees of freedom driven by the electron correlation is one of the key aspects in condensed matter physics behind novel electronic phases of matter and intriguing physical phenomena. In particular, the electronic properties can be modified whenever the spin direction is altered, as spin-orbit coupling (SOC) depends on the spin direction. For instance, spin-valley coupling serves as the fundamental mechanism in optically-controlled valley polarization1, spin-Hall and valley-Hall effects2. Contrast to most of the findings in non-magnetic 2D materials[3, 4, 5], where pseudo-spins are involved, 2D magnets exhibit active carrier spins, enabling studies of magnetism in reduced dimensions[6].

Unlike ferromagnets (FMs), the antiferromagnets (AFMs) are commonly found in nature and they are permitted in each magnetic symmetry group, however they are less utilized than FMs[7]. Currently, AFM materials are considered as promising candidates for future spintronic applications, due to unique properties including insensitivity to external magnetic fields, lack of stray fields, and ultrafast spin dynamics in the terahertz regime[8]. The spontaneously long-range ordered of microscopic magnetic moments, resulting in zero net magnetization, which makes the AFMs insensitive to external magnetic fields.

Thus, a control of AFM state requires very high magnetic fields and demands unconventional means of detection[7]. Many properties such as optical, electronic, and vibrational, rely on the magnetic ordering[9, 10, 11], as well as magnetic moment orientation[12]. In particular, a giant impact of the spin direction on the band structure, have been recently demonstrated for 2D ferromagnetic CrI3_3 material[13]. Although, the spin-direction properties are reported for conventional and layered FMs[14, 15], the research on 2D AFM materials is very limited and scarce[6, 16, 17].

One of essential feature for manipulating AFM state is the magnetocrystalline anisotropy (MAE) which is evident in spin flip or spin flop transitions. The latter one requires relatively weak MAE and it is reported in literature for AFM compounds exhibiting various electronic states such as topological insulators[18], layered materials[19], and conventional semiconductors[20]. In addition, the experimental techniques including indirect means of magnetic phases such as anisotropic magnetoresistance (AMR), anomalous Hall effect (AHE) or second harmonic generation (SGH)[7, 21, 22], are employed to detect the AFM orientation in 2D materials. Interestingly, the magnetic @page 1/20 starts_mid=0 ends_mid=1

Introduction

The subtle interplay between the spin, charge, orbital, and lattice degrees of freedom driven by the electron correlation is one of the key aspects in condensed matter physics behind novel electronic phases of matter and intriguing physical phenomena. In particular, the electronic properties can be modified whenever the spin direction is altered, as spin-orbit coupling (SOC) depends on the spin direction. For instance, spin-valley coupling serves as the fundamental mechanism in optically-controlled valley polarization1, spin-Hall and valley-Hall effects2. Contrast to most of the findings in non-magnetic 2D materials[3, 4, 5], where pseudo-spins are involved, 2D magnets exhibit active carrier spins, enabling studies of magnetism in reduced dimensions[6].

Unlike ferromagnets (FMs), the antiferromagnets (AFMs) are commonly found in nature and they are permitted in each magnetic symmetry group, however they are less utilized than FMs[7]. Currently, AFM materials are considered as promising candidates for future spintronic applications, due to unique properties including insensitivity to external magnetic fields, lack of stray fields, and ultrafast spin dynamics in the terahertz regime[8]. The spontaneously long-range ordered of microscopic magnetic moments, resulting in zero net magnetization, which makes the AFMs insensitive to external magnetic fields.

Thus, a control of AFM state requires very high magnetic fields and demands unconventional means of detection[7].

Many properties such as optical, electronic, and vibrational, rely on the magnetic ordering[9, 10, 11], as well as magnetic moment orientation[12]. In particular, a giant impact of the spin direction on the band structure, have been recently demonstrated for 2D ferromagnetic CrI3_3 material[13]. Although, the spin-direction properties are reported for conventional and layered FMs[14, 15], the research on 2D AFM materials is very limited and scarce[6, 16, 17].

One of essential feature for manipulating AFM state is the magnetocrystalline anisotropy (MAE) which is evident in spin flip or spin flop transitions. The latter one requires relatively weak MAE and it is reported in literature for AFM compounds exhibiting various electronic states such as topological insulators[18], layered materials[19], and conventional semiconductors[20]. In addition, the experimental techniques including indirect means of magnetic phases such as anisotropic magnetoresistance (AMR), anomalous Hall effect (AHE) or second harmonic generation (SGH)[7, 21, 22], are employed to detect the AFM orientation in 2D materials. Interestingly, the magnetic phase transition of controlled anisotropic phenomenon in layered magnets provides crucial understanding of fundamental magnetism in reduced dimensions.

The present work is motivated by the lack of systematic studies regarding the impact of AFM orientation of magnetic moments on the optoelectronic properties of 2D materials. Hence, in this paper we put particular attention to pinpoint the magnetic fingerprints in indirect properties that can engineer the AFM ordering. We conduct a theoretical analysis based on density functional theory (DFT) and effective Bethe-Salpeter equation (BSE) to identify various spin direction features, which can be utilised in spin-processing functionalities. This study focuses on spin angle evolution of the effective masses of carriers, optically active band edge transitions, exciton binding energies, which remain almost unexplored in the context of van der Waals (vdW) AFM crystals.

Here, we present a systematic study of the impact of spin orientation on the electronic and optical properties of series of the monolayers of MPX3_3, where M= Mn, Ni, Fe, and X= S, Se, assuming collinear arrangement of the magnetic moments. The results are presented as following: first we examine the magnetic ground state, determining easy and hard axes of magnetization at the level of PBE+U+SOC approach. Next, we consider the electronic features such as band extrema, effective masses, valley splitting and excitonic properties including the excitonic binding energy. Although, the electronic properties have been widely reported for particular spin arrangement and direction, the impact of the orientation of the collinearly ordered spins on opto-electronic properties MPX3_3 are largely missing. Finally, band edge excitons have been systematically predicted for this class of materials for the first time.

Computational Details

The calculations were performed in the framework of density functional theory (DFT) using the generalized gradient approximation within the PBE flavor[23], as implemented in VASP software[24]. The ion–electron interactions were described by the projector augmented wave (PAW) method[25]. Plane-wave basis cutoff and Γ\Gamma centered Monkhorst-Pack[26] k-point grid were set to 500 eV and 10×6×110 \times6 \times1, respectively. A Gaussian smearing of 0.05 eV was employed for the Brillouin zone (BZ) integration. The interlayer vdW forces were treated within Grimme scheme using D3 correction[27]. A vacuum of the thickness equal to 20 Å was added to mimic the isolated monolayer. Most of the results were obtained using PBE+U method based on Dudarev’s approach, with the effective on-site Hubbard U parameter (Ueff_{eff} =U-J, where J=1eV) assumed for 3d3d orbitals. Note that Ueff_{eff} is hereafter denoted as U. To check its impact on various properties two values of U= 3, 5 eV were employed. The SOC within the non-collinear treatment of magnetism was taken into account on the top of the PBE+U scheme. The position of the atoms and unit cell were fully optimized within the PBE+U approach. In order to predict the magnetic easy and hard axis, the spins of the magnetic ions have been rotated from out-of-plane to in-plane directions with polar angle step ΔΘ\Delta\Theta=15∘^\circ. For every Θ\Theta the atomic positions and the lattice parameters have been fully optimized. The effective masses of carriers at the band edges were examined using finite difference method[28]. The direct inter-band momentum matrix elements were computed from the wave function derivatives using Density Functional Perturbation Theory (DFPT)[29] in order to determine the optically active transitions, as discussed in[30]. The macroscopic optical dielectric constants frequency ϵ∞\epsilon_{\infty} were obtained using DFPT in the independent particle approach with (DFT-TD) and without local field effects (IP)[29]. Exciton calculations were performed with the effective BSE formalism[31, 32, 33, 34, 35] using the effective masses and dielectric constants obtained from the DFT calculations. We solve the effective BSE numerically using the parameters given in Ref.[9].

Results

Magnetic ground state & magnetic easy axis

First, we examine the impact of the rotation of spins on the energy profile to determine the magnetic ground state and magnetic easy/hard axes for all employed MPX3_3 systems. We assume a collinear alignment of the spins and angle of rotation Θ\Theta (see Figure 1 (a)) varied from the out-of plane to in-plane configuration. The rotation of the spins within the basal plane yields up to two orders of magnitude smaller energy changes compared to out-of plane ones, hence in-plane rotations are not further considered.

Magneto-optical anisotropies and magnetic ground-state spin arrangements of MPX3 compounds

Figure 1. (a) The rotation angle Θ\Theta of collinearly aligned AFM spins directed towards the (001) plane. (b-d) Polar plots of energy difference ΔE\Delta E between the particular direction of the spins and the lowest energy configuration for Hubbard U=5U=5 eV. (e) The magnetocrystalline energy (MAE) is defined here, as a difference between the energies of spins parallel to hard (highest energy) and easy axes (lowest energy) of magnetization. (f-h) Structural arrangement of the spins exhibited by the magnetic ground state for all employed systems. The direction of the spins are parallel to the easy axis / plane of magnetization. The ΔE\Delta E and the MAE are given in meV per magnetic ion.

The magnetic ground state exhibits antiferromagnetic Neel (AFM-N) and antiferromagnetic zigzag (AFM-z) arrangement of the spins for Mn and M=Ni, Fe, respectively (see Figure 1), in line with other reports[9, 10, 36, 37]. The MPX3_3 materials are reported to be robust antiferromagnets[10] and even high concentration of the substitutional dopants could not alter the magnetic ordering of the host[10, 38]. Additionally, changing the angle of spin alignment Θ\Theta requires at least one order of magnitude lower energy (tens of meV per magnetic ion) than change of the AFM ordering (at least tens of meV per magnetic ion). The computed energy difference (ΔE\Delta E) between the out-of-plane and in-plane directions can be one order of magnitude greater for Se than S compounds (see Figure 1 (b-d)), which is expected for heavier atoms, exhibiting larger SOC coupling[38]. Surprisingly, FePS3_3 exhibits a larger MAE than FePSe3_3 (see Figure 1(e)). The value of UU does affect neither the type of AFM ordering nor the direction of the magnetic easy axis. However, the ΔE\Delta E depends on the Hubbard UU parameters. Namely, the smaller values are generally obtained for larger UU, except for FePS3_3 monolayer (see Fig. S2). For the employed materials, an increase in the effective UU resulted in larger lattice constants by up to 1.5 % and larger magnetic moments (see Table S2).

MPX3\mathrm{MPX}_{3}:MnPS3\mathrm{MnPS}_{3}MnPSe3\mathrm{MnPSe}_{3}FePS3\mathrm{FePS}_{3}NiPS3\mathrm{NiPS}_{3}NiPSe3\mathrm{NiPSe}_{3}NiPS3\mathrm{NiPS}_{3} − bulkCoPS3\mathrm{CoPS}_{3}
Ueff[eV]U_{\mathrm{eff}}[\mathrm{eV}]:4.95.35.15.65.65.65.7

Table S2.

Let us now consider the easy axes of magnetization, which are presented in Figure 1(f, g) for all employed structures, with the hard axes predicted to be orthogonal to the corresponding easy axis/plane. In particular, for FePX3_3 the magnetic easy axis points along the cc crystallographic direction as reported in previous studies[36, 39, 40, 41, 42, 43]. Hence,

this kind of systems are considered to be Ising-type antiferromagnets, with a strong uniaxial magnetic anisotropy.

For the MnPX3_3 compounds, we predict that the easy plane coincides with monolayer plane, unlike the recent experimental report for MnPS3_3, which has demonstrated that the spins are slightly tilted from the cc-axis[44]. The direction of the ordered spins for MPX3_3 compounds result from the interplay between the single-ion anisotropy and magnetic dipolar interactions (MDIs), as already discussed in a recent publication[38]. In addition, the spin-orbit splitting is negligible for MnPS3_3, which manifests itself as the lowest MAE for all employed structures (see Figure 1(b,e)). Thus, the MDI, generally regarded to be weak[45], might be decisive in determining the direction of the MnPS3_3 spins. Note, that the MDIs are not accounted for in our PBE+U+SOC approach, and will be considered elsewhere. On the other hand, in the case of the MnPSe3_3, the MDIs could be neglected due to larger lattice parameters and SOC. Hence, the magnetic easy axis of MnPSe3_3 coincides with the monolayer plane, in agreement with other reports in the literature[41, 43, 46], and confirmed by spin flop transition reported upon the nonmagnetic substitution in MnPS3_3 thin films[38]. Regarding NiPS3_3, the in-plane position of the magnetic moments are preferable and the magnetic ordering can be described by the XY Heisenberg Hamiltonian[39, 40, 47, 48]. Our results predict the easy axis of magnetization within the monolayer frame, without any deviation in the z direction as reported recently for monolayer[49]. However, we do not exclude the possibility that the inclusion of the MDIs interaction can facilitate the rotation of the spins towards the out of the plane direction. For the case of NiPSe3_3, we determined the easy axis of magnetization to be in out-of-plane direction in line with recent report[50]. Similar results regarding the easy axis/plane of magnetization have been reported for the corresponding bulk systems[38].

Electronic properties

electronic band gaps

Our results reveal that the structures containing Mn and Fe exhibit direct band gaps located at K high symmetry point (see Figure 2 (a)), in line with previous reports[9, 37, 51], while the NiPS3_3 system is an indirect semiconductor confirmed by experiments[47]. In particular, for FePS3_3 and NiPSe3_3 there is a strong contribution of the 3d states to the bands in vicinity of Fermi level. Hence, for these systems, significant changes of the position of the band extrema (VBM, CBM) and their curvatures upon the Hubbard U parameter are visible (see Table S2 and Table S3 and discussions therein).

Spin-orientation electronic properties

Figure 2. Spin-orientation electronic properties. (a) Band extrema for all employed monolayers as a function of spin direction (color scale). (b) Enhancement of the electronic band gaps ΔEgap\Delta E_{\mathrm{gap}} for particular spin directions in respect to out-of-plane spins. (c) Valley splitting ΔVB\Delta^{VB}, ΔCB\Delta^{CB} are defined ΔVB=EK−VB−EK+VB\Delta^{VB}=E_{K^-}^{VB}-E_{K^+}^{VB}, ΔCB=EK−CB−EK+CB\Delta^{CB}=E_{K^-}^{CB}-E_{K^+}^{CB}, respectively for Mn and Fe contained monolayers, given in meV. Note, that for the clarity of picture regarding MnPS3_3, the −ΔCB-\Delta^{CB} is plotted. (d) In-plane components of effective mass tensor for electrons (positive masses) and holes (negative masses) of MnPX3_3 in respect to spin direction and for two non-equivalent ±K\pm K valleys.

ML of MPX3\mathrm{MPX}_{3}UeffU_{\mathrm{eff}} [eV]magn. statelattice constant [Å]μm\mu_m [μB\mu_B]Band gap [eV]VBM CBM
MnPS3\mathrm{MnPS}_{3}5AFM-N6.1084.6102.502 (D)K
MnPS3\mathrm{MnPS}_{3}3AFM-N6.0984.4912.242 (D)K
MnPSe3\mathrm{MnPSe}_{3}5AFM-N6.4244.6061.846 (D)K
MnPSe3\mathrm{MnPSe}_{3}3AFM-N6.3854.4751.662K
NiPS3\mathrm{NiPS}_{3}6.45AFM-zig5.8561.6212.288KY
NiPS3\mathrm{NiPS}_{3}5AFM-zig5.8391.5221.954KY
NiPS3\mathrm{NiPS}_{3}3AFM-zig5.8191.3781.572KY
NiPSe3\mathrm{NiPSe}_{3}6.45AFM-zig6.1841.5661.883(D)K
NiPSe3\mathrm{NiPSe}_{3}5AFM-zig6.1681.4571.6564KY
NiPSe3\mathrm{NiPSe}_{3}3AFM-zig6.1461.2911.258(0.282;0.5;0)Y
FePS3\mathrm{FePS}_{3}5.3AFM-zig5.9883.7042.423(D)K
FePS3\mathrm{FePS}_{3}2.6AFM-zig5.9633.4221.736Γ\Gamma(0.273;0;0)
FePSe3\mathrm{FePSe}_{3}5.3AFM-zig6.3043.6871.825(D)K
FePSe3\mathrm{FePSe}_{3}3AFM-zig6.2743.5271.482(D)K

Table S3.

U dependence

For all employed compounds, except MnPX3_3, the U parameter significantly affects the curvature of the conduction bands as reported previously[37], and hence the effective masses of holes. Hence, all features are examined assuming two Hubbard U parameters U=3, 5 eV). In particular, the large value of the effective masses of electrons exhibited for Ni contained systems, indicating a flat character of the conduction bands, which can be further enhanced by adopting larger values of U (see Table S3). By applying the Hubbard U to 3d states, one can shift the 3d states away from the Fermi level. In particular, for MnPX3_3 structures the conduction bands are mainly build from p-states, and further shifting 3d states do no affect the band curvature nor the band gap value, whereas for FePX3_3 compounds the conduction bands consist mainly of 3d states, hence further enlargement of U strongly impacts electronic features around band edges. Additionally, our results corroborate the previous reports on strong dependence of Hubbard U on fundamental energy band gap (see Table S2)[37].

Effective masses

For all of employed monolayers, the heaviest effective mass is obtained for NiPX3_3, whereas the lightest one is demonstrated for MnPSe3_3 (see Table S3 and Fig. S3). Generally, the lower effective masses are exhibited by the electrons than holes, and for Se contained structures than corresponding S systems (by about 30%), except for NiPX3_3 structures where these trends are opposite. Additionally, the anisotropic in-plane components of effective masses (m1_1 ≠\ne m2_2) are exhibited by monolayer with AFM-z ground state, whereas for MnPX3_3 monolayers with AFM-N ordering, the in-plane components are isotropic (m1_1=m2_2). Note, that the AFM-z magnetic arrangement breaks the hexagonal symmetry, as it consist of spins ferromagnetically aligned along zigzag chain (see Figure 1(g)), whereas the AFM-N type of magnetic order is commensurate with the structural symmetry. Hence, the anisotropic effective masses, and thus, anisotropic transport properties can be regarded as a magnetic marker distinguishing different type of antiferromagnetic ordering within hexagonal lattice. In addition, the inclusion of the SOC does not affect the curvature of examined systems, except for MnPSe3_3 where slight changes are shown for K+ and K- valleys (see Table S4).

Table S4. In-plane components of the effective mass tensor (m1m_1, m2m_2). The m1m_1 and m2m_2 are given in mem_e unit and they are determined along principal axes (square brackets). The direction of the principal axes are given in Cartesian coordinates in reciprocal space.

ML of MPX3_3UeffU_{\rm eff} [eV]eff. mass of holes m1m_1 [princ. axis], m2m_2 [princ. axis]eff. mass of electrons m1m_1 [princ. axis], m2m_2 [princ. axis]
MnPS3_35-0.957 [0.97 0.23 0] -0.955 [-0.23 0.97 0]0.577 [-0.39 0.93 -0.1] 0.588 [0.93 0.39 0.1]
3-1.244 [0.97 -0.23 0] -1.246 [0.23 0.97 0]0.691 [0.99 -0.13 0] 0.689 [0.13 0.99 0]
MnPSe3_35-0.767 [-0.39 0.93 0] -0.768 [0.93 0.39 0]0.546 [0.93 -0.39 0] 0.546 [0.93 0.39 0]
3-1.023 [0.97 -0.23 0] -1.026 [0.23 0.97 0]0.625 [0.95 -0.32 0] 0.623 [0.32 0.95 0]
NiPS3_35-1.183 [0 1 0] -1.722 [1 0 0]4.724 [0 1 0] 4.262 [1 0 -0.01]
3-1.550 [0 1 0] -16.592 [1 0 -0.1]2.577 [1 0 -0.1] 2.447 [0 1 0]
NiPSe3_35-1.293 [0 1 0] -2.209 [1 0 0]3.865 [0.1 1 0] 1.718 [1.0 -0.1 -0.1]
3-1.380 [1 0 0] -4.628 [0 1 -0.1]1.869 [0.1 1 0] 1.031 [1 -0.1 -0.1]
FePS3_35.3-0.927 [0 1 0.1] -0.969 [1 0 0]1.053 [0.1 1 0] 0.938 [1 -0.1 0]
FePSe3_35.3-0.772 [0 1 -0.1] -0.788 [1 0 0]0.578 [-0.39 0.93 -0.1] 0.577 [0.93 0.39 0.1]
3-1.056 [-0.01 1 0] -1.334 [1 0.01 -0.01]1.087 [0.04 1 0] 1.045 [1 -0.04 0]

Table S4.

Spin-orientation-dependent electronic features in MPX3_3 (M=Mn, Fe; X=S,Se)

Now, let us make closer inspection on the band edges of MLs exhibiting direct transitions. The electronic features are collected in Table S3. Note, that in the absence of the SOC, the band extrema are doubly degenerated for all employed systems. The presence of SOC in the structures containing Mn preserves the spin degeneracy of the band extrema (VBM, CBM), however causing the valley splitting at +K and -K, which is expected for the honeycomb lattice with AFM-N arrangements of the spins with SOC included[52]. Namely, the SOC preserves the spin degeneracy but leads to a renormalization of the valleys (+K and -K are not equivalent) for AFM-N magnetic ground state. The size of the valley splitting (Δ\Delta) depend on the chalcogen atoms (larger values for Se atoms), Hubbard U, and spin directions (see Figure 2 and Tables I, Table S6). In particular, the largest valley splitting is attained for the out-of-plane direction of the spins (ΔCB=39\Delta^{CB}=39 meV for MnPSe3_3 and U=3 eV) and lower value of U as presented in Figure 2(c), whereas no polarity of the valleys are observed for the in-plane direction of the spins. Namely, the rotation of the spins towards out-of-plane directions results in enhancement of the valley splitting. Hence, the band gap of MnPSe3_3 can be changed by up to 35 meV as presented in Figure 2(b, Figure 2c), whereas for the rest of the compounds the minor changes are observed (see Figure 2(b)). Additionally, the effective masses of holes for MnPSe3_3 are sensitive to the direction of the spins as presented in Figure 2(d). Interestingly, in the case of MnPS3_3, the band gap changes its character from direct to indirect when the spins are rotated from an out-of-plane to an in-plane directions.

MAE is the energy difference between the AFM aligned spins along out-of-plane (along easy axis) and in-plane directions (along hard axis), given in meV per magnetic ion. Note, that for in-plane directions the electronic structure is identical as without inclusion of the SOC. The Δin−outVB,CB(K±)\Delta^{VB,CB}_{in-out}(K^{\pm}) denotes the difference between the VBM (CBM) at given K point. The EginE_g^{in} (EgoutE_g^{out}) electronic band gap for the in-plane (out-of-plane) direction of the AFM aligned spins.

U [eV]MAEΔin−outVB(K+)\Delta^{VB}_{in-out}(K^{+}) [meV]Δin−outCB(K+)\Delta^{CB}_{in-out}(K^{+}) [meV]Δin−outVB(K−)\Delta^{VB}_{in-out}(K^{-}) [meV]Δin−outCB(K−)\Delta^{CB}_{in-out}(K^{-}) [meV]EginE_g^{in} [meV]EgoutE_g^{out} [meV]
MnPS3_330.041005.59.32.2032.194
50.028006.53.92.4832.479
7-006.91.92.6672.665
MnPSe3_330.315034.417.801.6891.654
50.223027.915.401.8531.825
7-016.412.601.9601.944

Table S6.

Regarding the monolayer of FePX3_3, the presence of SOC and local inversion symmetry breaking in FeX6_6 octahedra lift the degeneracy of the band extrema (few meV), irrespective to the spin direction (see column 4 in Table I). In particular, larger spin polarization δ\delta is obtained for CBM than VBM (δCB=9.1\delta^{CB}=9.1 meV, δVB=1.8\delta^{VB}=1.8 meV), and for higher values of U for FePS3_3. In contrast to the FePS3_3, the opposite trends are obtained for FePSe3_3 (for the details see Table S6). Hence, the spin splitting of the bands is obtained for hexagonal lattices exhibiting in-plane structural anisotropy, as reported for MnPS3_3 (see Fig. 5(d) in [51]). In this case the spin polarization occurred, along-side with renormalization of the +K/-K valleys as expected for AFM-N. In addition, one of the consequences of the local inversion symmetry breaking exploited in Fe monolayers is a lattice distortion reflected in the ratio of the lattice parameters deviated from hexagonal symmetry (b/a≠3b/a \ne\sqrt{3})[51, 55].

U. In particular, the larger values of χ∥\chi_{\parallel} are attained for the smaller Hubbard U, which correlates with the strong impact of the U on the DOS (see Fig. S2 (b)).

MPX3_3 (magn. state)spin direction (Θ\Theta)K+/K- (ΔVB\Delta^{VB}, ΔCB\Delta^{CB})Spin Deg. (δVB\delta^{VB}, δCB\delta^{CB})Hex. symm.(pol.) Inten. [(eVA˚)2][(\mathrm{eV}\mathring{\mathrm{A}})^2]
MnPS3_3 (AFM-N)0∘0^{\circ}✓\checkmark (-6.4, -4.4)X✓\checkmark(z) 0.59
90∘90^{\circ}X (0, 0)X✓\checkmark(z) 0.62
MnPSe3_3 (AFM-N)0∘0^{\circ}✓\checkmark (-17.7, 33.1)X✓\checkmark(z) 0.13
90∘90^{\circ}X (0, 0)X✓\checkmark(z) 0.18; (x,y) 0.05
FePS3_3 (AFM-z)0∘0^{\circ}X✓\checkmark (1.8, 9.1)broken0.71 (y)
90∘90^{\circ}X✓\checkmark (1.5, 7.4)broken0.73 (y)
FePSe3_3 (AFM-z)0∘0^{\circ}X✓\checkmark (7.4, 3.1)broken(y) 0.27; (x) 0.04
90∘90^{\circ}X✓\checkmark (9.3, 5)broken(y) 0.29; (x) 0.02

Table 1. Spin-dependent electronic features for direct band edge transition occurring at ±K\pm K valleys (VBM→\rightarrowCBM) in the presence of SOC (PBE+U(5eV)+SOC approach). The valley splitting of the VBM (ΔVB\Delta^{VB}) and CBM (ΔCB\Delta^{CB}) are defined as ΔVB=EK−VB−EK+VB\Delta^{VB}=E^{VB}_{K-}-E^{VB}_{K+} and ΔCB=EK−CB−EK+CB\Delta^{CB}=E^{CB}_{K-}-E^{CB}_{K+}, respectively. The spin degeneracy of the band edges δVB\delta^{VB} and δCB\delta^{CB} are calculated as δVB=∣E↑VB−E↓VB∣\delta^{VB}=|E^{VB}_{\uparrow}-E^{VB}_{\downarrow}| and δCB=∣E↑CB−E↓CB∣\delta^{CB}=|E^{CB}_{\uparrow}-E^{CB}_{\downarrow}|, respectively. We refer intensity (Inten.) to the oscillator strength of the band edge transitions. The optically active transitions coupled to linear polarization of light (x,y,z) are collected in the last column. The intensity is given by ∣ℏmee^⋅p⃗cv∣2\left|\sqrt{\frac{\hbar}{m_e}}\hat{e}\cdot\vec{p}_{cv}\right|^2 in which e^={x^,y^,z^}\hat{e}=\{\hat{x},\hat{y},\hat{z}\} is the light polarization (pol.) and p⃗cv\vec{p}_{cv} is the matrix element between CBM and VBM.

Excitonic properties

Dielectric properties

We start our discussion by analyzing the dielectric properties of the employed systems. Generally, the in-plane components of static dielectric constants (εxx,εyy\varepsilon_{xx},\varepsilon_{yy}) are isotropic for all studied systems, except for monolayers of (Ni,Fe)PX3_3, where they differ by up to 4% (see SI and Table S7, Table S8). Due to the fact that dielectric tensor is well defined for the bulk materials, we calculate the 2D polarizability χ∥\chi_{\parallel} (dielectric screening length, for the details see SM). Our results reveal that the dielectric screening length increases as the atomic number of the metal also increases χ∥Mn>χ∥Fe>χ∥Ni\chi_{\parallel}^{Mn}>\chi_{\parallel}^{Fe}>\chi_{\parallel}^{Ni}, exhibiting the smallest and highest values equal to 3.25 Å (MnPSe3_3, U=5 eV) and 6.13 Å (NiPSe3_3, U=5eV), respectively. In addition, the systems containing Se exhibit ∼40%\sim40\% larger screening properties than their corresponding S systems. The screening properties of MPX3_3 compounds are generally smaller than for other vdW structures such as MoSe2_2 (8.23 Å). Note that for the (Ni,Fe)PX3_3 systems, the dielectric screening lengths strongly depend on the Hubbard

Electronic band edge features such as valley splitting (ΔVB=EK−VB−EK+VB\Delta^{VB}=E^{VB}_{K^{-}}-E^{VB}_{K^{+}}, ΔCB=EK−CB−EK+CB\Delta^{CB}=E^{CB}_{K^{-}}-E^{CB}_{K^{+}}), and spin degeneracy (δVB=EK−VB−EK+VB\delta^{VB}=E^{VB}_{K^{-}}-E^{VB}_{K^{+}}, δCB=EK−CB−EK+CB\delta^{CB}=E^{CB}_{K^{-}}-E^{CB}_{K^{+}}, for monolayers of MnPX3_3 and FePX3_3, respectively for in-plane and out-of-plane directions of AFM aligned spins.

valley splittingMnPSe3_3 for U=3eVMnPSe3_3 for U=5eVMnPS3_3 for U=3eVMnPS3_3 for U=5eV
ΔVB\Delta^{VB}ΔCB\Delta^{CB}ΔVB\Delta^{VB}ΔCB\Delta^{CB}ΔVB\Delta^{VB}ΔCB\Delta^{CB}ΔVB\Delta^{VB}ΔCB\Delta^{CB}
in-plane spins-5.70000000
out-of-plane spins-21.538.6-17.733.1-6.6-10.3-6.4-4.4
Spin degeneracy of the band edgesFePS3_3 for U=2.6eVFePS3_3 for U=5.3eVFePSe3_3 for U=3eVFePSe3_3 for U=5.3eV
δ\deltaδVB\delta^{VB}δCB\delta^{CB}δVB\delta^{VB}δCB\delta^{CB}δVB\delta^{VB}δCB\delta^{CB}δVB\delta^{VB}δCB\delta^{CB}
in-plane spins07.51.89.17.45.57.43.1
out-of-plane spins06.41.57.411.58.29.35

Table S7.

Magn.U [eV]methodεxx\varepsilon_{xx}εyy\varepsilon_{yy}χxx∥\chi^{\parallel}_{xx}χyy∥\chi^{\parallel}_{yy} [Å]
MnPS35IP2.352.353.253.25
5TD-DFT2.162.162.802.80
MnPSe35IP2.602.604.504.50
5TD-DFT2.422.423.993.99
NiPS35IP2.842.824.284.23
5TD-DFT2.702.683.943.90
NiPSe35IP3.313.266.136.00
5TD-DFT3.183.135.765.65
FePS35.3IP3.203.213.513.52
5.3TD-DFT2.962.953.133.10
FePSe35.3IP2.812.784.904.84
5.3TD-DFT2.632.614.434.38

Table S8. Static dielectric constants εij\varepsilon_{ij} (electronic contribution) and 2D polarizability χ∥\chi^{\parallel} calculated for the monolayer systems and U=3 eV. The static dielectric properties are calculated by means of density functional perturbation theory in the independent particle approach (IP) neglecting local field effects and including local fields effect (TD-DFT)

Band edge excitons

Let us now examine the band edge excitons using the relevant information from DFT (meff,χ∥m_{eff},\chi_{\parallel}) combined with the effective BSE[31, 32]. We show the exciton binding energies (EBE_B) in Figure 3(a) (for the details of calculations see Ref. 9). All employed structures exhibit large EBE_B exceeding 1 eV and 0.8 eV for bare monolayers of MP3_3 and MPSe3_3, respectively, outperforming the values in TMDCs[9] (the exact values are collected in Table S9, Table S10). The excitons have been recently experimentally reported for few layers of NiPS3_3[47, 56, 57], and FePS3_3[51], however, their origin is still under a hot debate. Generally the EBE_B are larger by about 30% for S contained structures than corresponding Se compounds, mainly due to larger effective masses and smaller dielectric screening for MP3_3. The EBE_B decreases as the effective dielectric constant of the environment increases, preserving the same trend for all monolayers. This can be explained within the oversimplified exciton picture (hydrogen model) for which EBE_B is proportional to effective mass and inversely proportional to the square of dielectric screening.

Exciton binding energies and spin-orientation dependence in MPX3 monolayers

Figure 3. (a) Large exciton binding energies obtained for all considered monolayers with magnetic ground state with respect to the dielectric screening of the surroundings. The values for air/vacuum, SiO2_2{sup}53 substrate, and hBN encapsulation{sup}54 are denoted by vertical lines. (b) Evolution of the EBE_{\mathrm{B}} with respect to the spin direction for two values of Hubbard UU (3 and 5 eV). (c) The energy difference between the EBE_{\mathrm{B}} calculated for −K-K and +K+K valleys as a function of the spin direction angle Θ\Theta (within DFT+UU=5eV and DFT-TD approaches), for hBN encapsulated and freestanding layers.

Magn.U [eV]methodεxx\varepsilon_{xx}εyy\varepsilon_{yy}χxx∥\chi^{\parallel}_{xx}χyy∥\chi^{\parallel}_{yy} [Å]
MnPS33IP2.442.443.453.45
3TD-DFT2.262.263.023.02
MnPSe33IP3.023.024.824.82
3TD-DFT2.822.824.334.33
NiPS33IP3.153.105.125.01
3TD-DFT3.022.984.824.72
NiPSe33IP4.144.027.517.21
3TD-DFT3.022.984.824.72
FePSe33IP3.063.035.565.49
3TD-DFT2.912.895.155.09

Table S9. Static dielectric constants εij\varepsilon_{ij} (electronic contribution) and 2D polarizability χ∥\chi^{\parallel} calculated for the monolayer systems and Hubbard U=3eV. The static dielectric properties are calculated by means of density functional perturbation theory in the independent particle approach (IP) neglecting local field effects and including local fields effect (TD-DFT)

systemUU [eV]spin direction Θ\Theta [°]EbK+E_b^{K+} in vacuumEbK+E_b^{K+} in hBNEbK−E_b^{K-} in vacuumEbK−E_b^{K-} in hBN
MnPS3MnPS_{3}U=3U=301032.19344.461034.46346.02
501032.47344.671034.39345.83
901033.37345.341033.48345.45
U=5U=501023.98322.491025.60323.21
501023.98322.491025.60323.21
901024.69322.791024.54322.85
MnPSe3MnPSe_{3}U=3U=30770.25266.92787.82278.42
50773.76269.31785.75276.89
90780.62273.32779.96273.13
U=5U=50773.71254.23785.49261.16
50775.00254.61781.97259.01
90778.86256.97779.77257.45

Table S10. Exciton binding energies at different K valleys and environments (vacuum, hBN) as a function of spin direction Θ\Theta for MnPX3MnPX_{3} (X=SX=S, Se).

Optical transitions and selection rules

Besides the binding energy, it is also relevant to determine the selection rules of the direct band edge transitions, which are summarized in Table S3. All of the transition are optically allowed (non-zero oscillator strength) exhibiting linear polarization of light. In particular, for monolayers with AFM-N phase (MnPX3_3) the polarization of light is along zz direction, whereas for the AFM-z (FePX3_3) pointing along yy direction. Interestingly, similar conclusions are observed assuming various AFM metastable phases within the same magnetic material{sup}9. Therefore, the polarization of light is sensitive to magnetic order irrespective of the type of transition metal and chalcogen atoms. This is in line with recently reported linear polarization of the sharp emission, that aligned perpendicular to the spin orientation{sup}57. Hence, the polarization of light might be a tool to distinguish the type of AFM ordering. All of these direct band edge excitons are optically active transitions and couple to zz-polarized light. Comparing with the widely studied TMDCs, these transitions in MPX3_3 systems have an intensity two orders of magnitude smaller than the bright (A and B) transitions but have comparable intensity to the dark (D) transitions{sup}30,58`.

Spin-orientation-dependent excitonic features in MnPX3

Interestingly, the excitonic properties are sensitive to the direction of the spins. Owing to the spin direction dependent effective masses observed in MnPX3_3 presented in Figure 2 (d), we discuss below only these type of monolayers. The effect of the spin direction on EBE_{\mathrm{B}} is on the order of 1-2 meV in MnPS3_3 but on the order of 10 meV in MnPSe3_3 monolayer, as presented in Figure 3 (b), irrespective of the value used for the Hubbard UU parameter. Furthermore, opposite K-valleys (at +K+K and −K-K) show different effective masses and, consequently, different values of EBE_{\mathrm{B}}, as shown in Figure 3(c). Hence, the EBE_{\mathrm{B}} depends on the orientation of the AFM aligned spins. Particularly, due to the sizeable valley-dependent curvature of the VBM (see Figure 2 (d)) observed for the MnPSe3_3, the EBE_{\mathrm{B}} of excitons at ±K\pm K valleys differ by up to EB−−EB+=E_{\mathrm{B}}^{-}-E_{\mathrm{B}}^{+}= 12 meV (for MnPSe3_3 in vacuum).

Schematic diagrams of direct transitions and absorption onsets for MnPSe3 and MnPS3

Figure 4. Schematic diagrams (U=3U=3 eV) of direct transitions for (a, b) MnPSe3_3 and (c, d) MnPS3_3 exhibiting in-plane and out-of-plane directions of the AFM aligned spins, as schematically denoted at the bottom of the picture (g, h). Schematic absorption onsets (c, f) with depicted in coloured exciton peaks A, B and C. Namely, green (A) red (B) and blue (C), peaks indicate band edge excitons for in-plane and out-of-plane spin configurations of AFM aligned spins at +KK and −KK valleys, respectively. Black arrows close to bands indicate spin degenerated bands. The position of the peaks are in order but not in scale. All of the presented direct transition are coupled to "z" polarized light with the intensities listed in Table S3. (i) The first BZ with labeled high symmetry k-points.

Combining the information of the optical selection rules and exciton binding energies, we present in Figure S8 the schematic diagram of optically active band edge transition considering the in-plane and out-of the plane direction of spins. Our calculations reveal that the emissions from the opposite +KK and −KK valleys are energetically below the onset of absorption spectrum. For the in-plane directions of the spins, the energy difference between +KK and −KK is rather small (separated about 1 meV and 0.1 meV for MnPSe3_3 and MnPS3_3, respectively, as shown in Figure S8,Figure S8)). On the other hand, for out-of-plane spins, the exciton emission from ±KK valleys are separated by 42 meV and 6 meV for MnPSe3_3 and MnPS3_3, respectively. The separation of the exciton peaks at ±K\pm K valleys (see difference between the position of the B and C peaks in Figure S8) is calculated as:

Schematic diagrams of direct transitions and absorption onsets for MnPSe₃ and MnPS₃

Figure S8. Schematic diagrams (U=5U=5 eV) of direct transitions for (a, b) MnPSe₃ and (c, d) MnPS₃ exhibiting in-plane and out-of-plane directions of the AFM aligned spins, as schematically denoted at the bottom of the picture (e, f). On the right side of the pictures schematic absorption onsets with depicted in coloured exciton peaks. Namely, blue, red and green peaks indicate band edge excitons for out-of-plane spin configurations at +K and −K valleys, and in-plane configuration of AFM aligned spins, respectively. Black arrows close to bands indicate spin degenerated bands. The position of the peaks are in order but not in scale. All of the presented direct transition are coupled to "z" polarized light with the intensities listed in Table S3. (g) The first BZ with labeled high symmetry k-points.

ΔB=[Eg(K−)−EB(K−)]−[Eg(K+)−EB(K+)]=ΔEB+ΔVVB−ΔCCB.(1)\Delta_B = [E_g(K-) - E_B(K-)] - [E_g(K+) - E_B(K+)] = \Delta E_B + \Delta V^{VB} - \Delta C^{CB}. \tag*{(1)}

where ΔEB=EB(K−)−EB(K+)\Delta E_B = E_B(K-) - E_B(K+) is the difference between the exciton binding energies at ∓K\mp K valleys and Eg(K±)E_g(K\pm) the electronic band gap at a given KK valley. We predict that one peak should be observed for in-plane oriented spins (green peak in Figure S8) whereas two peaks emerge for the spins deflected from the monolayer plane (red and blue peaks in Figure S8). Therefore, the energy ΔB\Delta_B is a robust magnetic fingerprint of the AFM spin directions. Note that the energy separation of ±K\pm K valleys depends on the dielectric environment, as presented in Figure 3(c), but should be visible in hBN encapsulated samples, specially in MnPSe3_3 since the valley splitting is dominated by the electronic counterparts ΔCB\Delta^{CB} and ΔVB\Delta^{VB}.

Conclusions

Here, we systematically examine the MPX3_3 materials emphasizing the role of spin reorientation, magnetic arrangement and electron correlation effects in various properties. In this regard, the chemical trends in respect to the type of the chalcogen atom as well as transition metal are examined. We have demonstrated that the electronic features such as band gaps, effective masses, dielectric screening and exciton binding energies strongly depends on the type of chalcogen atom. Notably, larger effective masses are attained for S than Se contained monolayers (by about 30 %), whereas smaller dielectric screening (up to 40%) are reached for MPS3_3. In general, the larger effective masses and smaller dielectric screening length are obtained for employed materials than for widely examined TMDs, resulting in larger exciton binding energies of direct transitions of MPX3_3 than corresponding monolayers of TMDcs. Generally, all examined band edge direct transitions of (Mn,Fe)PX3_3 monolayers turned out to be optically active coupled to linear polarized light, with type of polarization sensitive to magnetic arrangement. Regarding the direct band edge transitions, the MnPX3_3 are of particular importance as could be exploited as promising valley electronics materials. In particular, we have shown that the valley splitting at the ±K\pm K can be effectively controlled by the direction of the magnetic moments. In particular, the sizable valley splitting occurs for the out-of-plane direction of the magnetic moments in monolayer of MnPSe3_3 (smaller effect observed for MnPS3_3), resulting in valley dependent gaps. Namely, two distinct peaks are expected to be visible below the absorption onset for the out-of-plane AFM aligned spins, whereas one peak for the in-plane case. Hence, the number of the peaks and the separation of the peaks can be regarded as magnetic fingerprint of the orientation of AFM aligned spins. The change in the separation of the peaks points towards the plausible spin rotation.

On the other hand, the ±K\pm K valleys obtained for MnPX3_3 are spin degenerated, which might hinder their spintronic applications. Hence, we have also propose a novel way how the spin valley polarization can be attained. We have suggested that particular deformation of the hexagonal lattice of MnPX3_3 could lead to the spin resolved valley splitting. Our results give insight into the valley splitting realization in 2D antiferromagnets. In addition, the anisotropic effective masses and the type of linear polarization can be regarded as magnetic markers probing the type of AFM arrangements. Finally, the spin dependent features have been identified such as valley splitting of VBM and CBM, the effective mass of holes and exciton binding energies. These features can be referred as sensitive parameters that provide insight into spin flop transitions.

Acknowledgment

M.B. acknowledges financial support from the University of Warsaw under the "Excellence Initiative - Research University" project. P.E.F.J. and J.F. acknowledge the financial support of the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) SFB 1277 (Project-ID 314695032, projects B07 and B11), SPP 2244 (Project No. 443416183), and of the European Union Horizon 2020 Research and Innovation Program under Contract No. 881603 (Graphene Flagship). T.W. acknowledges financial support of National Science Centre, Poland under grant no. 2021/41/N/ST3/04516. Access to computing facilities of the Interdisciplinary Center of Modeling (ICM), University of Warsaw are gratefully acknowledged. We gratefully acknowledge the Polish high-performance computing infrastructure PLGrid (HPC Centers: ACK Cyfronet AGH) for providing computer facilities and support within computational grants no. PLG/2022/015685 and PLG/2023/016571. We acknowledge ACK Cyfronet AGH (Poland) for awarding this project access to the LUMI supercomputer, owned by the EuroHPC Joint Undertaking, hosted by CSC (Finland) and the LUMI consortium through P1-Grid organization (Poland), under the grant entitled: "Electronic, optical and thermoelectric properties of selected layered

Magneto-optical anisotropies of 2D antiferromagnetic MPX3_3 from first principles

Miłosz Rybak; Paulo E. Faria Junior; Tomasz Woźniak; Paweł Scharoch; Jaroslav Fabian; Magdalena Birowska

materials and selected heterostructures".

Determination of Hubbard U from Ab initio

Response functions found from a linear fit of the number of d-electrons as a function of the additional potential V

Figure S5. Response functions found from a linear fit of the number of d-electrons as a function of the additional potential V: (a) for MPX3(M=Mn,NiMPX_3(M = Mn, Ni, (b) for M′PX3(M′=Co,Fe)M'PX_3(M' = Co, Fe)

To estimate the Hubbard UeffU_{\mathrm{eff}} parameter, we adopt the linear response method proposed by method of Anisimov and co-workers in the framework of a plane-wave pseudopotential approach implemented in VASP by Cococcioni et al. [59]. The U parameter is found from the linear fits for both non-selfconsistent (NSCF) as and selfconsistent (SCF) response functions as defined

U≈(∂NSCF∂VI)−1−(∂NNSCF∂VI)−1.U \approx\left(\frac{\partial N^{\mathrm{SCF}}}{\partial V_I}\right)^{-1} - \left(\frac{\partial N^{\mathrm{NSCF}}}{\partial V_I}\right)^{-1}.

The results of U parameters for dd electrons of transition metal ions are collected in Table S2.

Table S2. Calculated Hubbard UeffU_{\mathrm{eff}} of dd electrons of transition metals for various MPX3\mathrm{MPX}_{3} materials. The values are given in eV.

Tables of collected electronic features.

Table S3. Valence and conduction band edges (VBM, CBM) and electronic band gaps for various monolayer systems obtained for magnetic groundstate denoted as magn. state. The μm\mu_m denotes the spin magnetic moment on metal atom (M = Mn, Ni, Fe).

Projected density of states for MnPS₃ and FePS₃

Figure S6. Projected density of states (PDOS) assuming various exchange-correlation functional for MnPS3_3 (right side) and FePS3_3(left side).

In-plane components of the effective mass tensor

Figure S7. In-plane components of the effective mass tensor for (left picture) all employed monolayers, (right picture) various Hubbard U.

Table S5. MnPSe3_3 with and without SOI included. Note that the collinear spins are assumed. The SO offset is the energy difference between the results obtained with SOC and without it.

methodband edgekk-pointholes: m1m_1 [principle axis]m2m_2 [principle axis]SOSO offset [meV]
without SOVBK/K′-0.767 [-0.38 0.92 0.00]-0.768 [0.92 0.38 0.00]-
with SOC: in-plane spinsVBK-0.752 [0.52 0.86 0.00]-0.825 [0.86 -0.52 0.00]0
VBK′-0.767 [0.74 -0.67 0.00]-0.807 [0.67 0.74 0.00]0
with SOC: out-of-planeVBK-0.752 [0.52 0.86 0.00]-0.825 [0.86 -0.52 0.00]-17.9
VBK′-0.767 [0.74 -0.67 0.00]-0.807 [0.67 0.74 0.00]0.21
methodband edgekk-pointelectrons: m1m_1 [principle axis]m2m_2 [principle axis]SOSO offset [meV]
without SOCBK/K′0.547 [0.92 -0.38 0.00]0.546 [0.38 0.92 0.00]-
with SOC: in-plane spinsCBK0.551 [0.83 -0.56 0.00]0.542 [0.56 0.83 0.00]-6.39
CBK′0.551 [0.59 0.81 0.00]0.541 [0.81 -0.59 0.00]-6.39
with SOC: out-of-plane spinsCBK0.551 [0.83 -0.56 0.00]0.542 [0.56 0.83 0.00]-3.50
CBK′0.551 [0.59 0.81 0.00]0.541 [0.81 -0.59 0.00]-36.60

Table S5.

MLε\varepsilonU=3 eV, IPU=3 eV, TD-DFTU=5 eV, IPU=5 eV, TD-DFT
MnPS₃1.01028.581132.131019.061132.42
MnPS₃1.5814.46890.75797.34879.03
MnPS₃2.0673.53732.35652.41714.90
MnPS₃2.5571.60618.59549.11598.30
MnPS₃3.0494.27532.73470.56510.31
MnPS₃3.5433.05464.87409.35442.17
MnPS₃4.0383.39410.52360.20387.07
MnPS₃4.5342.54365.24319.61342.01
MnPS₃5.0308.07327.64285.84304.97
MnPSe₃1.0778.82843.61778.93850.94
MnPSe₃1.5623.37671.57616.20668.73
MnPSe₃2.0519.65557.50508.78549.51
MnPSe₃2.5444.61475.08431.41463.95
MnPSe₃3.0386.72412.18372.29398.80
MnPSe₃3.5340.69362.07325.94347.92
MnPSe₃4.0303.63321.54288.12306.50
MnPSe₃4.5272.23287.72256.96272.58
MnPSe₃5.0246.07259.44230.95244.25
NiPS₃1.0985.671031.651073.151140.95
NiPS₃1.5825.85862.46886.32939.24
NiPS₃2.0716.31746.68759.50802.73
NiPS₃2.5634.46660.29665.62701.86
NiPS₃3.0569.94592.46592.25623.30
NiPS₃3.5517.45537.29532.90559.77
NiPS₃4.0473.66491.15483.65507.35
NiPS₃4.5436.31451.94442.10462.95
NiPS₃5.0404.01418.21406.42424.93
NiPSe₃1.0670.36940.83805.38844.93
NiPSe₃1.5562.14775.12673.12704.36
NiPSe₃2.0487.23662.72582.05608.00
NiPSe₃2.5430.83579.42513.95536.08
NiPSe₃3.0386.65514.41460.50479.60
NiPSe₃3.5350.42462.08416.84433.64
NiPSe₃4.0320.06418.51380.60395.41
NiPSe₃4.5294.47381.96349.65362.93
NiPSe₃5.0272.19350.44323.07334.86
FePS₃1.01038.181132.84--
FePS₃1.5825.91896.20--
FePS₃2.0685.68740.31--
FePS₃2.5584.24628.15--
FePS₃3.0506.71543.07--
FePS₃3.5445.41475.55--
FePS₃4.0395.42420.84--
FePS₃4.5354.01375.72--
FePS₃5.0319.04338.00--
FePSe₃1.0766.45812.40741.87798.61
FePSe₃1.5626.08661.47590.19632.78
FePSe₃2.0530.94559.68489.66522.38
FePSe₃2.5461.05484.65416.52443.24
FePSe₃3.0406.60426.59360.88382.75
FePSe₃3.5362.81379.87316.85334.78
FePSe₃4.0326.70341.66281.01296.30
FePSe₃4.5296.51309.55251.24264.49
FePSe₃5.0270.58282.16226.68237.58

Table S11. Exciton binding energies for MLs considering different values of the effective dielectric constant of the surroundings ε\varepsilon, calculated using two approaches (IP and TD-DFT) for dielectric screening length of ML.

References

References

  1. [3]N. Tombros, C. Josza, M. Popinciuc, H. T. Jonkman, and B. J. van Wees, Nature 448, 571 (2007).
  2. [4]O. M. J. van ’t Erve, A. L. Friedman, E. Cobas, C. H. Li, J. T. Robinson, and B. T. Jonker, Nature Nanotechnology 7, 737 (2012).
  3. [5]M. V. Kamalakar, C. Groenveld, A. Dankert, and S. P. Dash, Nature Communications 6, 6766 (2015).
  4. [6]M. Gibertini, M. Koperski, A. F. Morpurgo, and K. S. Novoselov, Nature Nanotechnology 14, 408 (2019).
  5. [7]P. Nemec, M. Fiebig, T. Kampfrath, and V. K. Kimel, Nature Physics 14, 229 (2018).
  6. [8]W. Zhang, P. K. J. Wong, R. Zhu, and A. T. S. Wee, InfoMat 1, 479 (2019).
  7. [9]M. Birowska, P. E. Faria Junior, J. Fabian, and J. Kunstmann, Phys. Rev. B 103, L121108 (2021).
  8. [10]C. Autieri, G. Cuono, C. Noce, M. Rybak, K. M. Kotur, C. E. Agrapidis, K. Wohlfeld, and M. Birowska, The Journal of Physical Chemistry C 126, 6791 (2022).
  9. [11]W. M. Linhart, M. Rybak, M. Birowska, P. Scharoch, K. Mosina, V. Mazanek, D. Kaczorowski, Z. Sofer, and R. Kudrawiec, J. Mater. Chem. C 11, 8423 (2023).
  10. [12]J. Liu, S. Meng, and J.-T. Sun, Nano Letters 19, 3321 (2019).
  11. [13]P. Jiang, L. Li, Z. Liao, Y. X. Zhao, and Z. Zhong, Nano Letters 18, 3844 (2018), pMID: 29783842.
  12. [14]P. Sander, Journal of Physics: Condensed Matter 16, R603 (2004).
  13. [15]P. Grünberg, Physics Today 54, 31 (2001), https://pubs.aip.org/physicstoday/article-pdf/54/5/31/11151061/31_1_online.pdf.
  14. [16]X. Martí, I. Fina, and T. Jungwirth, IEEE Transactions on Magnetics 51, 1 (2015).
  15. [17]S. Rahman, J. F. Torres, A. R. Khan, and Y. Lu, ACS Nano 15, 17175 (2021).
  16. [18]P. M. Sass, J. Kim, D. Vanderbilt, J. Yan, and W. Wu, Phys. Rev. Lett. 125, 037201 (2020).
  17. [19]K. Balamurugan, S.-H. Lee, J.-S. Kim, J.-M. Ok, Y.-J. Jo, Y.-M. Song, S.-A. Kim, E. S. Choi, M. D. Le, and J.-G. Park, Phys. Rev. B 90, 104412 (2014).
  18. [20]F. L. A. Machado, P. R. T. Ribeiro, J. Holanda, R. L. Rodríguez-Suárez, A. Azevedo, and S. M. Rezende, Phys. Rev. B 95, 104418 (2017).
  19. [21]T. Jungwirth, X. Marti, P. Wadley, and J. Wunderlich, Nature Nanotechnology 11, 231 (2016).
  20. [22]I. Fina, X. Marti, D. Yi, J. Liu, J. H. Chu, C. Rayan-Serrao, S. Suresha, A. B. Schick, J. Železný, T. Jungwirth, J. Fontcuberta, and R. Ramesh, Nature Communications 5, 4671 (2014).
  21. [23]J. Paier, R. Hirschl, M. Marsman, and G. Kresse, The Journal of Chemical Physics 122, 234102 (2005).
  22. [24]G. Kresse and J. Furthmüller, Computational Materials Science 6, 15 (1996).
  23. [25]N. A. W. Holzwarth, A. R. Tackett, and G. E. Matthews, Comput. Phys. Comm. 135 (2001).
  24. [26]H. J. Monkhorst and J. D. Pack, Phys. Rev. B 13 (1976).
  25. [27]S. Grimme, J. Antony, S. Ehrlich, and H. Krieg, The Journal of Chemical Physics 132, 154104 (2010).
  26. [28]A. Fonari and C. Sutton, Effective Mass Calculator (EMC) (2015).
  27. [29]M. Gajdoš, K. Hummer, G. Kresse, J. Furthmüller, and F. Bechstedt, Phys. Rev. B 73, 045112 (2006).
  28. [30]T. Woźniak, P. E. Faria Junior, G. Seifert, A. Chaves, and J. Kunstmann, Phys. Rev. B 101, 235408 (2020).
  29. [31]S. Albrecht, L. Reining, R. Del Sole, and G. Onida, Phys. Rev. Lett. 80, 4510 (1998).
  30. [32]M. Rohlfing and S. G. Louie, Phys. Rev. B 62, 4927 (2000).
  31. [33]K. Zollner, P. E. Faria Junior, and J. Fabian, Phys. Rev. B 100, 085128 (2019).
  32. [34]K. Zollner, P. E. Faria Junior, and J. Fabian, Phys. Rev. B 101, 085112 (2020).
  33. [35]K. Zollner, P. E. Faria Junior, and J. Fabian, Phys. Rev. B 107, 035112 (2023).
  34. [36]N. Sivadas, M. W. Daniels, R. H. Swendsen, S. Okamoto, and D. Xiao, Phys. Rev. B 91, 235425 (2015).
  35. [37]A. K. Budniak, S. J. Zelweski, M. Birowska, T. Woźniak, T. Bendikov, Y. Kauffmann, Y. Amouyal, R. Kudrawiec, and E. Lifshitz, Advanced Optical Materials 10, 2102489 (2022).
  36. [38]R. Basnet, K. M. Kotur, M. Rybak, C. Stephenson, S. Bishop, C. Autieri, M. Birowska, and J. Hu, Phys. Rev. Res. 4, 023256 (2022).
  37. [39]T. Olsen, Journal of Physics D: Applied Physics 54, 314001 (2021).
  38. [40]A. R. Wildes, V. Simonet, E. Ressouche, G. J. McIntyre, M. Avdeev, E. Suard, S. A. J. Kimber, D. Lancon, G. Pepe, B. Moubaraki, and T. J. Hicks, Phys. Rev. B 92, 224408 (2015).
  39. [41]K. Kurosawa, S. Saito, and Y. Yamaguchi, Journal of the Physical Society of Japan 52, 3919 (1983).
  40. [42]H.-J. Hoo, R. Kremer, and M.-H. Whangbo, Molecules 26 (2021), 10.3390/molecules26051410.
  41. [43]A. Wiedenmann, J. Rossat-Mignod, A. Louisy, R. Brec, and J. Rouxel, Solid State Communications 40, 1067 (1981).
  42. [44]E. Ressouche, M. Loire, V. Simonet, A. Ballou, A. Stunault, and A. Wildes, Phys. Rev. B 82, 100408 (2010).
  43. [45]D. C. Johnston, Phys. Rev. B 93, 014421 (2016).
  44. [46]P. Sen and R. K. Chouhan, Electronic Structure 2, 025003 (2020).
  45. [47]K. Hwangbo, Q. Zhang, Q. Jiang, Y. Wang, J. Fonseca, C. Wang, G. M. Diederich, D. R. Gamelin, D. Xiao, J.-H. Chu, W. Yao, and X. Xu, Nature Nanotechnology 16, 655 (2021).
  46. [48]C. A. Belvin, E. Baldini, I. O. Ozel, D. Mao, H. C. Po, C. J. Allington, S. Son, B. H. Kim, J. Kim, I. Hwang, J. H. Kim, J.-G. Park, T. Senthil, and N. Gedik, Nature Communications 12, 4837 (2021).
  47. [49]K. Kim, S. Y. Lim, J.-U. Lee, S. Lee, T. Y. Kim, K. Park, G. S. Jeon, C.-H. Park, J.-G. Park, and H. Cheong, Nature Communications 10, 345 (2019).
  48. [50]H. Sun, L. Qiu, Y. Han, E. Yi, J. Li, M. Huo, C. Huang, H. Liu, M. Li, W. Wang, D.-X. Yao, B. A. Frandsen, B. Shen, Y. Hou, and M. Wang, Materials Today Physics 36, 101188 (2023).
  49. [51]E. Geraffy, S. Zuri, M. M. Rybak, F. Horani, A. K. Budniak, Y. Amouyal, M. Birowska, and E. Lifshitz, (2022), arXiv:2208.10890.arxiv.org/abs/2208.10890
  50. [52]X. Li, T. Cao, Q. Niu, J. Shi, and J. Feng, Proceedings of the National Academy of Sciences 110, 3738 (2013).
  51. [55]A. Ghosh, M. Birowska, P. K. Ghose, M. Rybak, S. Maity, S. Ghosh, B. Das, K. Dey, S. Bera, S. Bhardwaj, S. Nandi, and S. Datta, Phys. Rev. B 108, L060403 (2023).
  52. [56]S. Kang, K. Kim, B. H. Kim, J. Kim, K. I. Sim, J.-U. Lee, S. Lee, K. Park, S. Yun, T. Kim, A. Nag, A. Walters, M. Garcia-Fernandez, J. Li, L. Chapon, K.-J. Zhou, Y.-W. Son, J. H. Kim, H. Cheong, and J.-G. Park, Nature 583, 785 (2020).
  53. [57]X. Wang, J. Cao, Z. Lu, A. Cohen, H. Kitadai, T. Li, Q. Tan, M. Wilson, C. H. Lui, D. Smirnov, S. Sharifzadeh, and X. Ling, Nature Materials 20, 964 (2021).
  54. [59]M. Cococcioni and S. de Gironcoli, Phys. Rev. B 71, 035105 (2005).

Paper details

vlnc:W_dxckw253
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 3

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents