Minimal d d -Band Model for the Optical Susceptibility of Non-Centrosymmetric Monolayer Transition Metal Dichalcogenides

Angiolo Huamán · 2025

Read Minimal d d -Band Model for the Optical Susceptibility of Non-Centrosymmetric Monolayer Transition Metal Dichalcogenides by Angiolo Huamán (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-22703.txt.

Introduction

Angiolo Huamán shows that the optical response of atomically thin semiconductors can be captured by a strikingly small model. Monolayer transition metal dichalcogenides lack inversion symmetry and are usually treated with heavy first-principles calculations using full Bloch wavefunctions. Around the band gap, though, their states are mostly d-orbital in character, and Huamán exploits that fact to reduce everything to three energy bands.

Within the minimal model, both the linear and quadratic optical susceptibilities are computed from the d-band structure alone. This simple approach reproduces first-principles results well up to about two electronvolts above the band gap. Huamán offers the model as a cheap starting point for adding many-body effects with only a few bands, where full calculations would be numerically expensive.

Catalog record

CREATORAngiolo Huamán
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE38 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-22703

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Minimal d d -Band Model for the Optical Susceptibility of Non-Centrosymmetric Monolayer Transition Metal Dichalcogenides Angiolo Huamán Abstract The optical response of two-dimensional (2D) materials has been customarily calculated ab initio using plane waves and without separating the most important orbitals contributions. In the family of transition metal dichalcogenides (TMDC) monolayers lacking inversion symmetry, we take advantage of the mostly d d orbital content of the Bloch bands around the semiconductor gap to reduce the calculation of the linear and quadratic optical susceptibilities to a very minimal model. Such a simple approach reproduces well first principles calculations and could be the starting point for the inclusion of many-body effects and spin-orbit coupling (SOC) in TMDCs with only a few energy bands in a numerically inexpensive way. I Introduction The optical response of solids, that is, the radiation fields emitted by them upon illumination, has been historically used to determined their structural and electronic properties, particularly their energy gaps Macfarlane and Roberts 1955 . Even though nowadays we have more precise techniques like ARPES, the study of the optical response, both at the experimental and theoretical level, is still of a fundamental importance. Many different effects comprise the field of light-matter interaction, namely absorption by, transmission through, and reflection from the sample. In experiments dealing with the latter, the reflected patterns carry information about the structural properties of the material. This is more evident when the nonlinear response is detected: this is nonzero only when the undelying crystal lacks an inversion center, while the linear response is present regardless of the crystal geometry. Calculations of the linear and nonlinear optical conductivity in non centrosymmetric materials have been mostly approached to by using first principle methods, such as density functional theory (DFT). However, DFT calculations have shown that the optical bands (those close to the energy gap) in 1 T TMDCs are of a mostly d type, which poses the question of whether a calculation using this minimal basis is feasible and how it compares to more comprehensive approaches. In this work, we build upon the three-band model with d orbitals by Liu et al Liu et al. 2013 and calculate the linear and nonlinear (quadratic) optical susceptibility of non-centrosymmetric monolayer TMDC in the 2H stacking configuration, taking tungsten disulfide (WS 2 ) as an example, although the same calculation can be carried out for the entire family of TMDCs in Liu et al. 2013 . Reflecting an almost complete d character of the energy bands around the energy gap, we see that the optical susceptibility thus obtained matches ab initio calculations for photon energies up to 1.7 eV above the band gap size. Using this simplified model as a basis, many-body effects can be added by using techniques based on tight-binding band structures Trolle et al. 2014 . This paper is organized as follows. Section II gives a brief description of monolayer TMDC in the 1T stacking. In Sec. ( III ) the basics of the formalism of the optical susceptibility in the single-particle approximation are presented. Section IV is an account of the symmetry considerations leading to the reduction to the IBZ. In Sec. VI we present the formalism for the calculation of two-center integrals of the momentum operator. Section VII contains our results. The conclusions we arrived at are presented in Sec. VIII . II TMDC monolayers in the three-band approximation Monolayers of TMDCs can have an allotropic form with a D_{3h} point group, thus lacking an inversion center, as schematically shown in Figs. 1 (a) and (b). The lattice vectors are chosen as \bm{a}_{1}=a\,\hat{\bm{x}} and \bm{a}_{2}=a(\hat{\bm{x}}+\sqrt{3}\,\hat{\bm{y}})/2 , a being the lattice parameter. The transition metal atoms occupy the lattice sites while there are two chalcogens per primitive cell at positions (1/3)(\bm{a}_{1}+\bm{a}_{2})\pm h\hat{\bm{z}}/2 , where h is the chalcogen-chalcogen distance along the z direction. The reciprocal lattice vectors are \bm{b}_{1}=(4\pi/(a\sqrt{3}))(\sqrt{3}\,\hat{\bm{x}}-\hat{\bm{y}})/2 and \bm{b}_{2}=(4\pi/(a\sqrt{3}))\hat{\bm{y}} , and the hexagonal Brillouin zone (BZ) is depicted in Fig. 1 (c). The mirror plane at the xy plane permits a decoupling between the two subsets of d- orbitals \{d_{xy},d_{z^{2}},d_{x^{2}}\} and \{d_{yz},d_{xz}\} , with different parity with respect to the operation z\rightarrow-z . Also, ab initio calculations have shown that the energy bands around the gap are mostly composed of orbitals of the former set, which allows the use of an effective three-band model as introduced by Liu et al Liu et al. 2013 , whose details are given in Appendix A . Before going into the calculation of the optical susceptibility, in the next section we briefly review the main concepts and formulae. III Optical susceptibility in the single particle approximation We will work in the single-particle approximation and with an electromagnetic field with no spatial dependence, described by the vector potential \bm{A}(t)=\bm{A}(\omega)e^{-\mathrm{i}\omega t}+cc . Although not explicitly written, \omega is to be understood as a complex frequency \omega+\mathrm{i}\eta , with \eta a small positive quantity that guarantees that the vector potential \bm{A}(t) vanishes at a remote past ( t\rightarrow-\infty ). The underlying monolayer TMDC under study is described by the static Hamiltonian H_{0}=\bm{p}^{2}/2m+V(\bm{r}) [ V(\bm{r}) is the self-consistent potential with the spatial symmetry of the crystal] with eigenvalues and Bloch eigenvectors satisfying H_{0}\psi_{n\bm{k}}=\epsilon_{n}(\bm{k})\psi_{n\bm{k}} , and that will be constructed using DFT with the SIESTA implementation. In the dipole approximation, the time-dependent Hamiltonian of the material coupled to the optical field, H(t) , can be obtained using the Peierls substitution and keeping only terms linear in the vector potential \bm{A}(t) : Figure 1: (a) Top view of monolayer WS 2 with trigonal prismatic configuration 2H and lattice vectors \bm{a}_{1} and \bm{a}_{2} (lattice parameter a=|\bm{a}_{1}|=|\bm{a}_{2}| ). Sulfur layers lie one on top of the other. (b) Trigonal unit cell of WS 2 exhibiting its D_{3h} symmetry. (c) Hexagonal first Brillouin zone (BZ), with the irreducible Brillouin zone (IBF, gray shaded region) and time-reversed irreducible region (tIBZ, hatched area). Regions (2) through (6) are obtained from the IBZ by applying the symmetry operations indicated. \displaystyle H(t)=\big[\bm{p}+ \displaystyle e\bm{A}(t)\big]^{2}/2m+V(\bm{r}) \displaystyle\simeq H_{0}+H_{1}(t)+\mathcal{O}(A^{2}), (1) with H_{1}(t)=(e/m)\bm{A}(t)\bm{\cdot}\bm{p} ( -e is the electron charge). The theory of the linear and nonlinear optical susceptibilities is abundant Cabellos et al. 2009 ; Ghahramani et al. 1991 ; Sipe and Ghahramani 1993 ; Moss et al. 1986 , so in what follows we will summarize the main concepts and equations. III.1 Linear optical susceptibility For monochromatic illumination, the single-body linear optical susceptibility tensor \chi_{ij}^{(1)}(\omega) in SI units is dimensionless and given in terms of the Fourier components of the induced polarization \bm{P}(t) and the applied electric field \bm{E}(t) Hemstreet and Fong 1972 ; Cabellos et al. 2009 : P_{i}(\omega)=\varepsilon_{0}\chi_{ij}^{\scriptscriptstyle(1)}(\omega)E_{j}(\omega), (2) with \bm{E}(\omega)=\mathrm{i}\omega\bm{A}(\omega) . The optical susceptibility is obtained by solving the density matrix in a perturbative way, and from it the macroscopic induced current \bm{J}(t) . The macroscopic polarization is obtained through \bm{P}(t)=\int_{-\infty}^{t}d\tau\bm{P}(\tau) Jackson 1998 , up to the linear order in \bm{A}(t) , as shown elsewhere Cabellos et al. 2009 . Considering only interband transitions we get the following expression: {\chi}^{\scriptscriptstyle(1)}_{ij}(\omega)=\frac{e^{2}\hbar^{2}}{\varepsilon_{0}m^{2}}\frac{1}{V}\sum_{\bm{k}}\sum_{v,c}\frac{p^{i}_{vc}p^{j}_{cv}}{\epsilon^{2}_{cv}(\epsilon_{cv}-\hbar\omega-\mathrm{i}\eta)}, (3) with \epsilon_{cv}=\epsilon_{c}-\epsilon_{v} , p^{i}_{vc}=\langle\psi_{v}|p^{i}|\psi_{c}\rangle , p^{i} is the i- th component of the momentum operator, m is the electron mass and V is the volume of the sample. The imaginary part in \omega has been explicitly written. The \bm{k}- dependence on the

Read the complete text