betaFeedback

WorksPhysics and Astronomy

General corner charge formula in two-dimensional C_n-symmetric higher-order topological insulators

Abstract

In this paper, we derive a general formula for the quantized fractional corner charge in two-dimensional C_n-symmetric higher-order topological insulators. We assume that the electronic states can be described by the Wannier functions and that the edges are charge neutral, but we do not assume vanishing bulk electric polarization. We expand the scope of the corner charge formula obtained in previous works by considering more general surface conditions, such as surfaces with higher Miller index and surfaces with surface reconstruction. Our theory is applicable even when the electronic states are largely modulated near system boundaries. It also applies to insulators with non-vanishing bulk polarization, and we find that in such cases the value of the corner charge depends on the surface termination even for the same bulk crystal with C_3 or C_4 symmetry, via a difference in the Wyckoff position of the center of the C_n-symmetric crystal.

Introduction

Recent studies have revealed that some topological crystalline insulators exhibit higher-order bulk boundary correspondence, which are known as higher-order topological insulators1–41. For example, three-dimensional second-order topological insulators with inversion symmetry (and time-reversal symmetry) have protected anomalous gapless mode along the hinges9,12–14,16,21–23, and two-dimensional second-order topological insulators with rotation symmetry have protected quantized corner charges24–42. Such fractionally quantized corner charges are generalizations of the quantized surface charge caused by the quantized bulk electric polarization43.

As discussed in the previous works, the quantized fractional corner charge in CnC_n-symmetric crystalline insulators are associated with rotational eigenvalues of the bulk wavefunctions32,36,40,42; nonetheless, they are limited to special cases. For example, in C6C_6-symmetric systems, Ref. [32] considers the case with hexagonal unit cells that cover the whole crystal, while in Ref. [40] systems covered by triangular building blocks are considered, and so the formulas of the corner charge for these two cases are different. In Ref. [41], although a general formula of the quantized corner charge is discussed, the authors assume that the surface is flat and the electronic charge distribution is the same between the finite C6C_6-symmetric system and an infinite system. Most importantly, the formula in Ref. [41] treats the contributions of electrons and ions to corner charges on an equal footing.

Here, a question arises whether the formula for the corner charge in previous works32,36,40,42 holds for any insulating CnC_n symmetric systems; this question is important for application to real materials, which may have deformations of nuclei positions and of electronic states near the boundaries. In the previous works, the corner-charge formula was derived for systems in which nuclei positions and electronic states are perfectly periodic up to the boundaries, and this result was complemented by an argument that surface decoration will leave the corner charge unaffected. Nevertheless, the surface-decoration argument has not been addressed in detail, and it is not clear whether the resulting corner-charge formula holds in general CnC_n-symmetric systems, including those where the nuclei positions and electronic states may be modulated near the boundary.

In the present paper we derive a general corner-charge formula on general grounds with minimal assumptions. Our study shows that even when the positions of the nuclei and the electronic states are largely modulated near the boundaries, the corner charge is quantized to fractional values as long as the system is CnC_n-symmetric and the edge charge density is zero. One of the merits of our method is that it naturally includes surfaces with higher Miller indices and surfaces with surface reconstructions. Our theory also includes insulators with non-vanishing bulk polarization but without edge charges. In particular, in such cases with C3C_3 or C4C_4 symmetries, we find that the value of the corner charge depends on the Wyckoff position of the center of the CnC_n-symmetric crystal. It is in contrast with previous works, where the corner-charge formula assumes zero polarization.

This paper is organized as follows. In Sec. II, we derive the general corner charge formula and clarified the assumptions required for the quantization of the corner charge. We show that as long as the edge is gapped and charge neutral in CnC_n-symmetric finite system, the corner charge is fractionally quantized modulo ∣e∣/n|e|/n. We also discuss the relation of the general formula and the bulk rotational eigenvalues. In Sec. III, we consider systems with nonzero electric polarization, which is beyond the scope of the previous corner charge formula. In Sec. IV, we consider more general surface conditions, such as general edge orientations and surface reconstructions. Conclusion and discussion are given in Sec. V.

Corner charge formula

In this section, we construct a general theory to calculate the corner charges in an insulator with CnC_n-rotation symmetry, based on the notion of filling anomaly. We derive its fractional quantization under the following minimal assumptions. We assume that the system is topologically trivial in the sense that it is adiabatically connected to an atomic limit32,41. It assures that the system has no topological edge states in the gap.

To explain another important assumption, we show the definition of the filling anomaly adopted in this paper:

  • In some bulk insulators, charge neutrality is incompatible with existence of a gap at the Fermi energy for the whole system including the boundaries; in such cases, by adding or subtracting a few electrons from charge neutrality, the Fermi energy is shifted and the system becomes gapped including the boundaries. This deviation of the electron number from charge neutrality is called filling anomaly in this paper.

Based on this definition of the filling anomaly, we assume that the finite-sized CnC_n-symmetric crystal considered is gapped. These two assumptions, namely topological triviality and existence of the gap, guarantee that the occupied states are described in term of localized orbitals. We note that the gap in this finite crystal should survive in the large system size limit. For example, in a hexagonal-shaped crystal with C6C_6 symmetry, suppose the edge and bulk are gapped, and the corners support one in-gap bound state per each corner, with the Fermi energy at these in-gap corner states. Then a tiny hybridization gap O(e−L/λ)O(e^{-L/\lambda}) appears within the sixfold multiplet of the corner states, where LL is the distance between neighboring corners and λ\lambda is a penetration depth of the corner states along the edges. In this case, the occupied states at the corners cannot be described as localized orbitals, but are extended over the system size, and it violates our assumptions. Thus, these assumptions justify the description of occupied electronic states as localized orbitals.

In the previous works32,36,40–42, the formula of the quantized corner charge is derived for cases with the electronic charge distribution being perfectly periodic even near the system boundaries. It is then argued that the quantization will remain even with surface decorations onto the system while preserving the CnC_n symmetry and the bulk gap. Nonetheless, details of surface-decoration argument have not been addressed, and it is not clear in which kinds of systems the corner charge remains quantized to be the value determined by bulk electronic states. In real materials, electronic states are determined as solutions of the Schrödinger equation, and nuclei positions are determined as minima of the total energy; this necessarily leads to deviations of electronic wavefunctions and nuclei positions near the boundaries. In many cases, details of the nuclei positions and electronic states near the boundaries are complicated and not easily discussed on general grounds. Thus, a general proof for CnC_n-symmetric insulators is desired. In the following, we give a general proof for this problem.

In this paper, we assume that the charges of the respective ions are integer multiples of the electronic charge ∣e∣|e|. It naturally holds in real materials. On the other hand, some theoretical models may have non-integer ionic charges, but such models are outside of our theory.

Derivation of the general corner charge formula

We consider a crystal of finite size which preserves the CnC_n symmetry. We assume that the system is insulating, in the sense that the Fermi level is within the energy gap for both the bulk and the boundaries. In some cases, degenerate corner states exist at the Fermi energy, which violates the above assumption, when charge neutrality is assumed. However, we can add or remove several electrons so that no states are at the Fermi energy. This deviation from charge neutrality is called filling anomaly32.

In order to calculate the filling anomaly, we divide the system conceptually into two CnC_n-symmetric regions (I) and (II), as shown in Figure 1. The regions (I) and (II) can be regarded as surface and bulk regions, respectively, and we explain this division in the following. The width of the surface region (I), dedged_{\mathrm{edge}}, is taken to be large enough so that both the electronic states and the ionic positions in the region (II) are the same as those at the bulk.

Figure 1. (Color online) A conceptual picture of the region division for (a) C3C_3, (b) C4C_4 and (c) C6C_6-symmetric systems. In each system, the width of the region (I), dedged_{\mathrm{edge}}, is taken to be large enough so that both the electronic states and the ionic positions in the region (II) are the same as those at the bulk.

To calculate the filling anomaly, we proceed with the following steps. First, we classify the ions into two groups according to which region they belong to, so that no ions are at the junctures of two regions.

Second, we classify the electronic states. We assume that the occupied states of this system can be described by a CnC_n-symmetric basis set {ϕi}\{\phi_i\} composed of exponentially localized basis orbitals43. Well within the crystal, the electronic states can be well described by bulk Wannier functions WnlW_{n l}, and we take these Wannier functions WnlW_{n l} to be among {ϕi}\{\phi_i\}. The remaining ϕi\phi_i's are additional localized basis orbitals ϕi(add)\phi_i^{(\mathrm{add})} near the boundary. We classify them into regions (I) and (II) so that the set of orbitals in region (I) and that in region (II) are CnC_n symmetric. By construction the numbers of ions and electronic localized basis orbitals in the region (I) are integer multiples of nn. Here, we impose a condition that all the additional localized basis orbitals φi(add)\varphi_i^{(\mathrm{add})} are classified into the region (I), and we choose the region (I) large enough so that this condition is satisfied. Thus, even if the electronic states near the boundary are largely deformed, our theory is still applicable, by incorporating these deformed wavefunctions into region (I). We note that this construction of localized basis functions is similar to that used in the argument in Ref. [43] to show that the bulk polarization in terms of the Zak phase gives surface charge density. We note that some degrees of freedom remain in this division into two regions, but they do not affect the following argument.

Third, we calculate the total charge of the finite crystal in terms of mod n∣e∣n|e|. First, the total charge contribution from the region (I) should be 00 (mod n∣e∣n|e|) because the number of the nuclei and the electronic basis orbitals within region (I) is an integer multiple of nn. Next, we calculate total charges of electrons and ions in the region (II). First, in the region (II), all the orbitals are Wannier orbitals WnlW_{n l}, and therefore positions of the ions and electrons are classified in terms of Wyckoff positions. Because the whole crystal is CnC_n symmetric, the center of the crystal should be located at the Wyckoff position with CnC_n local symmetry. Let 1a1a be the Wyckoff position with CnC_n symmetry located at the center of the crystal. Then, the ions not located at 1a1a contribute nM∣e∣nM|e| (M∈ZM \in\mathbb{Z}) to the net charge in the region (II), while those at 1a1a contribute nM′∣e∣+n1a(ion)∣e∣nM'|e|+n_{1a}^{(\mathrm{ion})}|e| (M′∈ZM' \in\mathbb{Z}), where n1a(ion)∣e∣n_{1a}^{(\mathrm{ion})}|e| is the charge of the ion at the 1a1a position. Therefore, the total ionic charge in the region (II) is equal to n1a(ion)∣e∣n_{1a}^{(\mathrm{ion})}|e| mod n∣e∣n|e| and so is that in the whole crystal. Similarly, the total electron charge is equal to −n1a(e)∣e∣-n_{1a}^{(\mathrm{e})}|e| (mod n∣e∣n|e|) where n1a(e)n_{1a}^{(\mathrm{e})} is the number of electron Wannier orbitals at the 1a1a position. From the above discussion, the total charge of the system in terms of modulo n∣e∣n|e| is calculated as follows,

Qtot≡(n1a(ion)−n1a(e))∣e∣(modn∣e∣)(1)Q_{\mathrm{tot}} \equiv\left(n_{1a}^{(\mathrm{ion})}-n_{1a}^{(\mathrm{e})}\right)|e| \pmod{n|e|} \tag*{(1)}

where n1a(ion)∣e∣n_{1a}^{(\mathrm{ion})}|e| is an ionic charge at 1a1a, and n1a(e)n_{1a}^{(\mathrm{e})} is the number of electronic Wannier functions at 1a1a. This formula represents a filling anomaly in the crystal.

Fourth, we discuss how QtotQ_{\mathrm{tot}} is distributed in the crystal. As a first step, we define the moving average of the total charge density ρˉ(x)\bar{\rho}(\boldsymbol{x}) as follows:

ρˉ(x)=1∣Ω∣∫Ωρ(x+x′) dx′.(2)\bar{\rho}(\boldsymbol{x})=\frac{1}{|\Omega|}\int_{\Omega}\rho(\boldsymbol{x}+\boldsymbol{x}')\,d\boldsymbol{x}'. \tag*{(2)}

Here, ρ(x)\rho(\boldsymbol{x}) is the total charge density including ions and electrons, Ω\Omega is a CnC_n-symmetric unit cell whose center is at the origin and ∣Ω∣|\Omega| is the area of Ω\Omega. From the definition, ρˉ(x)\bar{\rho}(\boldsymbol{x}) have the following properties:

ρˉ(x)=0(x∈bulk),(3)\bar{\rho}(\boldsymbol{x})=0 \quad(\boldsymbol{x}\in\text{bulk}), \tag*{(3)}
∫crystaldx′ ρˉ(x′)=Qtot.(4)\int_{\mathrm{crystal}}d\boldsymbol{x}'\,\bar{\rho}(\boldsymbol{x}')=Q_{\mathrm{tot}}. \tag*{(4)}
ρˉ(Cn⋅x)=ρˉ(x)(5)\bar{\rho}(C_n\cdot\boldsymbol{x})=\bar{\rho}(\boldsymbol{x}) \tag*{(5)}

We note that as long as ρˉ(x)\bar{\rho}(\boldsymbol{x}) satisfies Eqs. (3), (4) and (5), we can also use other definitions for moving averages. In a sufficiently large system, we expect that electronic states in the middle part of the edges far away from the corners are almost the same as those on an edge of a semi-infinite system with an open boundary condition in one direction. To calculate the corner charge based on this expectation, we divide the system into three types of regions: the (A) corner, (B) edge, and (C) bulk regions in a CnC_n-symmetric way as schematically shown in Figure 2. The edge region (B) should be chosen to satisfy the following conditions. First, it should be sufficiently far away from the corners, such that the electronic states in the edge region (B) are the same as those on an edge of the corresponding semi-infinite system. Second, it should have a width equal to an integer multiple of the period along the edge, NedgeaedgeN_{\mathrm{edge}}a_{\mathrm{edge}}, and a sufficiently long depth into the bulk, dedged_{\mathrm{edge}}, such that in the remaining bulk region (C), the electronic states are the same as those for a bulk infinite system. Here, we assume that one edge region includes NedgeN_{\mathrm{edge}} edge unit cells. Since ρˉ(x)\bar{\rho}(\boldsymbol{x}) takes nonzero values only in the (A) corner and (B) edge regions, its integral can be calculated as the sum of the total charges in these regions:

Figure 2. (Color online) A conceptual picture of the region division for calculating the corner charge for a C4C_4-symmetric system as an example. The total charge in each region is calculated by integrating the moving average of the total charge density ρˉ(x)\bar{\rho}(\boldsymbol{x}) in each region. In the bulk region (C), ρˉ(x)=0\bar{\rho}(\boldsymbol{x})=0, and the total charge in the bulk region QbulkQ_{\mathrm{bulk}} is zero. The edge region (B) includes an integer number of edge unit cells. The charge of each edge unit cell is calculated by the product of the edge charge density σ=P⋅n\sigma=\boldsymbol{P}\cdot\boldsymbol{n} and the edge period aedgea_{\mathrm{edge}}. The total charge is represented as summation of the edge charges and the corner charges: Qtot=n(Qedge+Qcorner)Q_{\mathrm{tot}}=n(Q_{\mathrm{edge}}+Q_{\mathrm{corner}}). Therefore, when QedgeQ_{\mathrm{edge}} vanishes, Qcorner≡Qtot/nQ_{\mathrm{corner}}\equiv Q_{\mathrm{tot}}/n (mod ee).

∫crystaldx′ ρˉ(x′)=n(Qedge+Qcorner),(6)\int_{\mathrm{crystal}}d\boldsymbol{x}'\,\bar{\rho}(\boldsymbol{x}')=n\left(Q_{\mathrm{edge}}+Q_{\mathrm{corner}}\right), \tag*{(6)}

where QedgeQ_{\mathrm{edge}} and QcornerQ_{\mathrm{corner}} are the total charges at one of the (B) edge regions and one of the (C) corner regions, respectively. Then QedgeQ_{\mathrm{edge}} can be calculated as:

Qedge=Nedgeσedgeaedge.(7)Q_{\mathrm{edge}}=N_{\mathrm{edge}}\sigma_{\mathrm{edge}}a_{\mathrm{edge}}. \tag*{(7)}

Here, NedgeN_{\mathrm{edge}} is the number of edge unit cells within one edge region, σedge\sigma_{\mathrm{edge}} is the edge charge density, and aedgea_{\mathrm{edge}} is the period of the edge unit cell.

Finally, we calculate the corner charge. We assume that the edge is charge neutral, σedge=0\sigma_{\mathrm{edge}}=0. Then the extra charge QtotQ_{\mathrm{tot}} is distributed into the nn corners. Since the crystal is assumed to be CnC_n-symmetric, the distributed charges are equal among the corners, and this gives a corner charge formula:

Qcorner=Qc-1a(n)≡(n1a(ion)−n1a(e))∣e∣n(mod∣e∣).(8)Q_{\mathrm{corner}}=Q_{c\text{-}1a}^{(n)}\equiv\frac{\left(n_{1a}^{(\mathrm{ion})}-n_{1a}^{(\mathrm{e})}\right)|e|}{n}\pmod{|e|}. \tag*{(8)}

Here, Qc-1a(n)Q_{c\text{-}1a}^{(n)} means the corner charge when the center of the crystal is located at the 1a Wyckoff position.

If the edge charge density σedge≠0\sigma_{\mathrm{edge}}\ne0, the extra charge QtotQ_{\mathrm{tot}} is distributed not only to the corners, but also to the edges. Since the separation of charges between the corners and the edges depends on the choice of the corner regions, the corner charge is not well defined.

Comparison with previous works

References

  1. [32]32
  2. [49]49 A. Kogar, M. S. Rak, S. Vig, A. A. Husain, F. Flicker, Y. I. Joe, L. Venema, G. J. MacDougall, T. C. Chiang, Eduardo Fradkin, Jasper van Wezel, Peter Abbamonte, Science 358, 1314 (2017).
  3. [50]50 P. Lee, T. Rice, and P. Anderson, Solid State Communications 88, 1001 (1993).
  4. [51]51 L. Trifunovic, Phys. Rev. Research 2, 043012 (2020).
  5. [52]52 S. Ren, I. Souza, and D. Vanderbilt, Phys. Rev. B 103, 035147 (2021).

Paper details

vlnc:W_vpeeq6w8
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

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