Abstract
Based on first-principles calculations and tight-binding model analysis, we propose monolayer graphdiyne as a candidate material for a two-dimensional higher-order topological insulator protected by inversion symmetry. Despite the absence of chiral symmetry, the higher-order topology of monolayer graphdiyne is manifested in the filling anomaly and charge accumulation at two corners. Although its low energy band structure can be properly described by the tight-binding Hamiltonian constructed by using only the pz orbital of each atom, the corresponding bulk band topology is trivial. The nontrivial bulk topology can be correctly captured only when the contribution from the core levels derived from px,y and s orbitals are included, which is further confirmed by the Wilson loop calculations. We also show that the higher-order band topology of a monolayer graphdyine gives rise to the nontrivial band topology of the corresponding three-dimensional material, ABC-stacked graphdiyne, which hosts monopole nodal lines and hinge states.
Similar content being viewed by others
Introduction
Bulk-boundary correspondence is a fundamental property of topological phases. In conventional topological insulators, the gapped bulk states in d-dimensions support metallic states in (d − 1)-dimensional surfaces. Such a conventional bulk-boundary correspondence, however, is violated in a class of topological crystalline insulators, so-called higher-order topological insulators (HOTIs). The gapless excitations of a HOTI in d-dimensions are localized, instead, in a subspace with a dimension lower than (d − 1), such as corners or hinges, when the global shape of the material preserves the crystalline symmetry relevant to the nontrivial band topology.1,2,3,4,5,6,7,8,9,10,11 Recently, rhombohedral bismuth has been identified as the first example of 3D HOTIs hosting helical hinge states.1 Also there are other candidate materials of 3D HOTIs including SnTe with strain along the (100) direction,2 transition metal dichalcogenides MoTe2 and WTe2, hosting helical hinge states,3 and also Bi2−xSmxSe3 with chiral hinge states.4
In 2D, one can also consider second-order topological insulators (SOTIs) hosting localized corner states.7,8,9,10,11,12 Contrary to 3D HOTIs, however, there are only few theoretical proposals for candidate materials realizing 2D SOTIs.9,10,11,13,14 For instance, phosphorene is one candidate for 2D HOTIs proposed based on simple tight-binding model calculations.15 However, in this system, the corner states originate from the charge polarization,15 which is a 1D topological invariant. Hence, in a strict sense, phosphorene is not a genuine 2D HOTI.13 Another candidate is twisted bilayer graphene, which can also exhibit higher-order band topology at small twist angles. However, to realize the corner states, the twist angle, as well as the chemical potential should be controlled simultaneously.
In general, among the higher-order TIs, only d-dimensional kth-order TIs with 1 < k < d have stable band topology with Wannier obstruction while d-dimensional dth-order TIs with zero-dimensional corner charges are obstructed atomic insulators. Thus, in 3D, a SOTI has Wannier obstruction while a third-order TI (TOTI) is atomic. On the other hand, in 2D, HOTIs are always obstructed atomic insulators. In d-dimensional dth-order TIs, additional chiral or particle-hole symmetry is required to protect corner states which appear as in-gap states separated from bulk states.8–11 Since chiral or particle-hole symmetry usually does not exist in insulators, realizing 2D SOTIs in materials is generally considered difficult.
However, even if the chiral symmetry of 2D SOTIs is broken, the material still inherits the nontrivial band topology of 2D SOTIs. The resulting state is called a 2D TOTI.16,17,18 The key idea is that even if the in-gap states are merged into the bulk states, the nontrivial band topology manifests in the filling anomaly.5,11,12 Namely, the half-filling condition can never be satisfied as long as the symmetry associated with the second-order band topology is preserved. The half-filling condition can be satisfied only when extra electrons or holes are added, which naturally leads to charge accumulation or depletion at corners. Therefore, the hallmark of 2D HOTIs lacking chiral symmetry (or simply 2D TOTIs), is the existence of filling anomaly at half-filling and the accumulation or depletion of a half-integral charge at corners after adding extra electrons or holes.16,17,18
In this work, we theoretically propose monolayer graphdiyne (MGD) as the first realistic candidate material for 2D HOTIs. Namely, MGD is a 2D SOTI when chiral symmetry exists, but in real systems, it appears as a 2D TOTI due to the lack of chiral symmetry. Based on first-principles calculations and tight-binding model analysis, we show that a MGD is a HOTI characterized by a 2D topological invariant w2, that is quantized when the system is invariant under inversion P symmetry. Although corner states are not guaranteed to be in the gap due to the lack of chiral symmetry in MGD, the nontrivial higher-order band topology is manifested in filling anomaly and the corner charge accumulation. Also, it is found that although the low energy excitation can be well-described by using the state derived from the pz orbitals of each atom, the relevant band topology of such a simplified model is trivial. To describe the nontrivial band topology, it is essential to include the contribution from the core levels derived from px,y and s orbitals. Finally, it is shown that the nontrivial w2 of MGD gives rise to the topological band structure of the corresponding three-dimensional material, ABC stacked graphdiyne, hosting nodal lines with Z2 monopole charges.19,20 Due to the nonzero w2 in a range of 2D momentum subspaces with fixed vertical momentum kz, the ABC stacked graphdiyne host hinge states for the corresponding range of kz.
Results
Topological invariant of monolayer graphdiyne
MGD is an experimentally realized planar carbon system composed of sp2-sp carbon network of benzene rings connected by ethynyl chains.21,22 Figure 1a describes the unit cell of MGD composed of 18 carbon atoms. Since spin-orbit coupling (SOC) is negligible, MGD can be regarded as a spinless fermion system. The lattice has inversion P, time-reversal T, a two-fold rotation about the x-axis C2x, a six-fold rotation about the z-axis C6z, and a mirror Mz: z → −z symmetries. Also, due to the bipartite lattice structure, MGD has chiral (or sublattice) symmetry when only the nearest neighbor hopping between different sublattices is considered, similar to the case of graphene. Fig. 1b shows the band structure along high symmetry directions obtained by first-principles calculations. Here, the blue color indicates the bands derived from pz orbitals while the red color denotes the bands derived from s, px, py orbitals. Since pz orbitals are odd whereas s, px, py orbitals are even under Mz symmetry, the energy spectrum from pz orbitals is not hybridized with that from s, px, py orbitals. One can see that the low-energy band structure has approximate chiral symmetry and can be effectively described by using only pz orbitals. However, to capture the higher-order band topology, the core electronic states derived from s, px, py orbitals should be included as shown below.
In general, a 2D P-symmetric spinless fermion system carries three Z2 topological invariants, w1x, w1y, and w2.20 In terms of the sewing matrix Gmn(k) = 〈um-k∣P∣unk〉, w1a (a = x, y) is given by its 1D winding number as
which is equivalent to the quantized Berry phase Φa, in a way that Φa = πw1a (mod 2π).23,24 w1a can be determined by using relation \({(-1)}^{{w}_{1a}}={\prod }_{i=1}^{2}\xi ({\Gamma }_{ia})\) where ξ(Γia) (i = 1, 2) is the product of parity eigenvalues of occupied states at the time-reversal invariant momentum (TRIM) Γia, and Γ1a, Γ2a are two TRIMs along the reciprocal lattice vector Ga. Since w1a is equivalent to the quantized charge polarization along the Ga direction, it can be considered as a 1D topological invariant.
On the other hand, w2 is a 2D topological invariant given by the 2D winding number of the sewing matrix G(k) (modulo 2),6 and measures the higher-order band topology of P-symmetric spinless fermion systems.5,6,13 w2 can be determined by
where \({N}_{\,\text{occ}\,}^{-}({\Gamma }_{i})\) is the number of occupied bands with odd parity at the TRIM Γi and the square bracket [α] indicates the greatest integer value of α25,26,27,28,29,30.
Using the band structure from first-principles calculations and the corresponding parity eigenvalues at TRIMs, one can easily show that MGD is a higher-order topological insulator with w1x = w1y = 0 and w2 = 1. One interesting feature of the MGD band structure is that we get w2 = 0 when only the bands derived from pz orbitals are considered. Namely, although the low-energy band structure itself can be well-described by using only pz orbitals, the higher-order topology can be correctly captured only when the core electronic states derived from s, px, py orbitals are included. This result can be confirmed by tight-binding analysis, as well as first-principles calculations. Since the bands derived from pz orbitals do not hybridize with those derived from other orbitals because of Mz symmetry, one can compute the relevant topological invariants separately.
The orbital dependence of w2 can also be confirmed by Wilson loop calculations.13,20,31,32 The Wilson loop is defined as a path ordered product of the exponential of Berry connections,33,34
where \({[{F}_{i}]}_{mn}=\langle {u}_{m}(\frac{2\pi (i\,+\,1)}{N},{k}_{2})| {u}_{n}(\frac{2\pi i}{N},{k}_{2})\rangle\), and its spectrum is gauge-invariant.33 Computed along the k1 direction parallel to the reciprocal lattice vector G1 from (k1, k2), the set of Wilson loop eigenvalues \(\{{e}^{i{\nu }_{i}({k}_{2})}\}\) indicates the position of Wannier centers at given k2, and the corresponding total charge polarization is given by \({p}_{1}({k}_{2})=\frac{1}{2\pi }{\sum }_{i=1}^{{N}_{{\mathrm{occ}}}}{\nu }_{i}({k}_{2})\). The Wilson loop spectrum of MGD shows that the charge polarization is zero in both k1 and k2 directions, which is consistent with the parity eigenvalue analysis (Fig. 1c). Also, the single linear-crossing at (ky, ν) = (0, π) indicates w2 = 1.20 Projecting the Wilson loop operator into the pz (s, px, and py) orbital basis, we observe w2 = 0 (w2 = 1), which confirms that the higher-order band topology originates from the core electronic levels. The orbital dependence of w2 can be further confirmed by computing the Wannier center of each orbital explicitly and also by computing the nested Wilson loop that measures the higher-order band topology.5,7 In the presence of PT symmetry, the Wilson loop spectrum can be divided into two sub-bands that are separated by a gap, each centered at ν = 0 or ν = π, respectively. When the Wilson loop operator at given k2 is diagonalized as
then, using the set of eigenvectors corresponding to the sub-band centered at ν = π, the nested Wilson loop operator \(\tilde{W}\) along the G2 vector is defined as
where \({[{\tilde{F}}_{i}]}_{mn}=\langle {\nu }_{m}(\frac{2\pi (i\,+\,1)}{N})| {\nu }_{n}(\frac{2\pi i}{N})\rangle\). Since the determinant of the nested Wilson loop operator \({\mathrm{det}}(\tilde{W})\) is identical to \({(-1)}^{{w}_{2}}\),20 we find that \({\mathrm{det}}(\tilde{W})\) is nontrivial. Projecting the Wilson loop operator into the pz orbital basis, we observe that the corresponding nested Wilson loop spectrum is trivial. On the other hand, within the s, px, and py orbital basis, the nested Wilson loop spectrum is nontrivial, which confirms that the higher-order band topology originates from the core electronic levels.
Filling anomaly and corner charges
The higher-order band topology of MGD with w2 = 1 induces a pair of anomalous corner states.3 To see this, we consider a finite-size structure invariant under P symmetry. When the system is chiral symmetric, there are two robust zero-energy in-gap states related by P whose locations are fixed at My-invariant corners. One way to see this is to use the topological classification proposed by Geier et al.9 When a 2D system has commuting mirror and chiral symmetries, it belongs to BDI\({}^{{M}_{++}}\) class that supports nontrivial topology, and the corresponding insulator has zero modes at mirror invariant corners. In MGD, the My symmetry commutes with chiral symmetry since My symmetry does not mix different sublattices. On the other hand, when chiral symmetry and Mx symmetry anti-commute, the system belongs to BDI\({}^{{M}_{+-}}\) class, which is topologically trivial, so there are no protected corner charges at Mx invariant corners.
To be specific, we can find the location of corner states by studying the low-energy Hamiltonian of MGD near the Γ point:
where Γ1 = σxτx, Γ2 = σyτx, Γ3 = τz, and Γ4 = σzτx. Corresponding symmetry representations are T = σxτzK, P = τz, Mx = σxτz, My = σx, and chiral symmetry S is described by S = Γ5 = τy. Considering a disk-shaped geometry, the low-energy Hamiltonian in real space becomes \(H=-i{\tilde{\Gamma }}_{1}(\theta ){\partial }_{r}-i{\tilde{\Gamma }}_{2}(\theta ){r}^{-1}{\partial }_{\theta }+\Delta (r){\Gamma }_{3}\), where \({\tilde{\Gamma }}_{1}(\theta )\,=\,{\mathrm{cos}}\theta {\Gamma }_{1}\,+\,{\mathrm{sin}}\theta {\Gamma }_{2}\) and \({\tilde{\Gamma }}_{2}(\theta )=-{\mathrm{sin}}\theta {\Gamma }_{1}\,+\,{\mathrm{cos}}\theta {\Gamma }_{2}\). We assume that Δ(r) > 0 inside the bulk and Δ(r) < 0 outside the bulk. The edge Hamiltonian is obtained by applying the projection operator \(P(\theta )=\frac{1}{2}(1+i{\tilde{\Gamma }}_{1}(\theta ){\Gamma }_{3})\), leading to \({H}_{\text{edge}}=-i{R}^{-1}{\tilde{\sigma }}_{z}{\partial }_{\theta }\) where \({\tilde{\sigma }}_{z}=P{\tilde{\Gamma }}_{2}(\theta ){P}^{-1}\). On the edge, position-dependent mass terms that open the edge gap are allowed. Mass terms that are commuting with P(θ) and anti-commuting with \({\tilde{\Gamma }}_{2}(\theta )\) and S are \({H}_{m}={m}_{4}(R,\theta ){\Gamma }_{4}+i{m}_{25}(R,\theta ){\tilde{\Gamma }}_{2}{\Gamma }_{5}\). The corresponding edge Hamiltonian is \({H}_{\text{edge}}=-i{R}^{-1}{\tilde{\sigma }}_{z}{\partial }_{\theta }+{\tilde{m}}_{1}(R,\theta ){\tilde{\sigma }}_{x}\), where \({\tilde{m}}_{1}={m}_{4}+{m}_{25}\), \({\tilde{\sigma }}_{x}=P{\Gamma }_{4}{P}^{-1}=Pi{\tilde{\Gamma }}_{2}{\Gamma }_{5}{P}^{-1}\). Here, PT symmetry requires \({\tilde{m}}_{1}(\theta +\pi )=-{\tilde{m}}_{1}(\theta )\) while My requires \({\tilde{m}}_{1}(x,y)=-{\tilde{m}}_{1}(x,-y)\). Thus, the mass gap is closed at My-invariant corners leading to localized corner charges.
Due to the degeneracy of the two in-gap states, the half-filling condition cannot be satisfied as long as P is preserved, which is known as the filling anomaly.5,11,12 When chiral symmetry is broken, the two degenerate in-gap states can be shifted from zero energy and even merged to the valence (conduction) bands generating a single hole (electron) at half-filling. Namely, the presence of an odd number of holes (electrons) in the valence (conduction) band is the manifestation of the higher-order band topology of generic 2D HOTIs lacking chiral symmetry.5,16,17,18 To resolve the filling anomaly while keeping P, one needs to add an extra electron (or hole), which leads to the accumulation or depletion of a half-integral charge at each My-invariant corner.
Explicitly, let us illustrate how the number of states can be counted in a finite-size HOTI with w2 = 1 following ref. 5 When the total number of states is N and chiral symmetry exists, there are N∕2 − 1 occupied and unoccupied states, respectively, since two in-gap states appear at zero-energy. When chiral symmetry is broken while inversion is preserved, the two in-gap states related by inversion are absorbed into either the valence or the conduction bands simultaneously. Then the number of states below the gap is ngap = N∕2 ± 1. Therefore ngap = (\(\frac{N}{2}+\) an odd integer) when w2 = 1, whereas ngap = (\(\frac{N}{2}+\) an even integer) when w2 = 0. When the valence (conduction) band of a system with w2 = 1 is completely occupied (unoccupied) by adding electrons or holes additionally, a half-integral electric charge should be accumulated (or depleted) at two corners related by inversion.
For the finite-size MGD composed of n × n unit cells shown in Fig. 2a, the total number of states are N = 72n2 when 2s, 2px, 2py, and 2pz orbitals are considered. In MGD, there is a single sp electron localized at every edge of a hexagonal unit cell. If an edge of a unit cell is not shared with an adjacent unit cell, the relevant sp orbital becomes a half-filled zero-energy non-bonding state when chiral symmetry exists. When we consider the geometry of a finite-size structure shown in Fig. 2a, MGD has 8n − 2 non-bonding states along the boundary. When chiral symmetry is broken, 8n − 2 states can be merged into the valence band, so that \({n}_{\text{gap}}=36{n}^{2}+4n-1=\frac{N}{2}+4n-1\). This is what is observed in first-principles calculations and the corresponding density of states is shown in Fig. 3. Thus, when all the states below the gap are occupied, and thus the filling anomaly is lifted, a half-integral charge (modulo an integer charge) is accumulated at two My-invariant corners. This can be easily understood because an odd number of non-bonding states exists only at the My-invariant corners (See Fig. 2a).
To observe the corner states more clearly, we add a hydrogen atom to the carbon atom at every corner, except at two My-invariant corners, and keep P and My symmetries. Then the number of added hydrogen atoms is an integer multiple of four, which does not alter the band topology of MGD. On the other hand, if two hydrogen atoms are added at the My-invariant corners, the total number of bands becomes \(N^{\prime} =N+2\), while ngap remains the same as described in Fig. 3a. Then \({n}_{\text{gap}}=\frac{N^{\prime} }{2}+4n-2\) and w2 becomes trivial. Therefore, the number of added hydrogen atoms should be an integer multiple of four to keep w2 unchanged. Maximally, (8n − 4) hydrogen atoms can be added to the finite-size MGD with two non-bonding states remaining at My-invariant corners. The corresponding density of states is shown in Fig. 3c. Here, a half electric charge is accumulated at each My-invariant corner when the states below the gap are fully occupied as shown in Fig. 2b.
Wannier function description
Here, we show that the higher-order band topology is associated with the chemical bonding of the s, px, and py orbitals across the unit cell boundary. In other words, Wannier functions are located away from the atomic sites and cannot be moved to the atomic sites by a continuous deformation that preserves inversion symmetry. Here, we follow the idea proposed by van Miert and Ortix35 to analyze the Wannier centers in inversion-symmetric systems.
Let us take the inversion-invariant unit cell shown in Fig. 1a whose center is located at the position m. There are four Wyckoff positions in a unit cell labeled as A, B, C, and D—located at 0, \(\frac{1}{2}{{\bf{a}}}_{{\bf{1}}}\), \(\frac{1}{2}{{\bf{a}}}_{{\bf{2}}}\), and \(\frac{1}{2}{{\bf{a}}}_{{\bf{1}}}+\frac{1}{2}{{\bf{a}}}_{{\bf{2}}}\) from the center of the unit cell, respectively—which are invariant under inversion up to lattice vectors a1 and a2. The symmetric Wannier functions centered at Wyckoff positions transform under inversion as follows:
where wα±,m(x) denotes a Wannier function with the coordinate x, centered at the Wyckoff position α of the unit cell with the center at m. The sign ± indicates eigenvalues of the inversion operation with respect to the Wyckoff position. We define the number of symmetric Wannier functions centered at each Wyckoff position with inversion eigenvalues ±1 as NA±, NB±, NC±, and ND±. Let us note that two Wannier functions with inversion eigenvalues +1 and −1 at A can be expressed as a linear combination of two Wannier functions with inversion eigenvalues +1 and −1 at any other Wyckoff position. Thus, only the difference να = Nα+ − Nα− at each Wyckoff position α is a well-defined \({\mathbb{Z}}\) invariant35.
which are related to the inversion eigenvalues at TRIMs in the Brillouin zone as35
In monolayer graphdiyne (MGD), we find that νA = νB = νC = νD = − 1 from the inversion eigenvalues at TRIM. In the viewpoint of chemistry, ∣νA∣ = 1 comes from the electrons on the benzene ring around the unit cell center, ∣νB∣ = ∣νC∣ = ∣νD∣ = 1 originates from the sp bonding across the unit cell boundary.
To relate these indices with the two-dimensional topological invariant w2 defined before, we define electric dipole moments pi = 1,2 and an electric quadrupole moment q12 for the unit cell as follows,
where Xi and Yi are the a1 and a2 components of the Wannier center for the ith occupied Wannier function. px and py are nothing but the polarizations along a1 and a2, respectively. w2 is related to px,y, qxy via w2 = 4(pxpy − qxy) mod 2,13,20 so
because ν = N+ − N− = N+ + N− − 2N− = N − 2N− = N mod 2. This shows that \({w}_{2}=1\ {\rm{mod}}\ 2\) when ∣νB∣ = ∣νC∣ = ∣νD∣ = 1 due to the sp bonding across the unit cell boundary.
Discussion
Since MGD has w2 = 1, when layers of MGD are stacked vertically, the resulting 3D insulator with PT symmetry becomes a 3D weak topological insulator, dubbed a 3D weak Stiefel-Whitney insulator,20 when inter-layer coupling is weak. When inter-layer coupling becomes large enough, an accidental band crossing can happen at a TRIM at which a pair of nodal lines with Z2 monopole charge are created. The existence of two nodal lines at kz = ± kc(kc > 0) was predicted in ABC-stacked graphdiyne,19 and their Z2 monopole charge was also confirmed20 based on first-principles calculations. In ABC-stacked graphdiyne, the 2D subspace with a fixed kz carries w2 = 1 (w2 = 0) when ∣kz∣ < kc (kc < ∣kz∣ < π) since the band inversion is happened at k = (0, 0, π). Since a 2D HOTI with w2 = 1 possesses corner charges, similar corner charges are also expected in ABC-stacked graphdiyne in the subspace with a fixed kz where w2 = 1,3 which leads to hinge modes of the 3D structure shown in Fig. 4a.
To demonstrate the hinge modes of ABC-stacked graphdiyne, we study a tight-binding model by using only pz orbitals to reduce numerical costs. The energy spectrum of a finite-size system with PT and C2x symmetries is shown in Fig. 4b, which clearly shows the presence of hinge modes in the momentum region where w2 = 1 as expected. When only pz orbitals are included, although the low-energy band structure from the tight-binding model is consistent with that from first-principles calculations, the topological property is different. Namely, the tight-binding model predicts w2 = 0 (w2 = 1) when ∣kz∣ < kc (kc < ∣kz∣ < π), which is opposite to the result from the first-principles calculations. Such a discrepancy can be remedied when the core electronic levels are included in the tight-binding calculation, which is confirmed by separate calculations.
To sum up, we have shown that the nontrivial band topology of a 2D HOTI lacking chiral symmetry can be revealed by using the filling anomaly, which can be explicitly demonstrated by counting the number of states of a finite-size system. This idea can be straightforwardly generalized to any d-th order topological insulator in d-dimensions hosting zero-dimensional corner states. We believe that our theoretical study provides a general way to extend the scope of HOTI materials to a wider class.
Methods
DFT calculations
We have carried out the first-principles density-functional theory calculations using the Vienna ab initio simulation package.36 The projector augmented-wave method37 and the exchange-correlation functional of the generalized gradient approximation in the Perdew, Burke, and Ernzerhof38 scheme were used. The self-consistent total energy was evaluated with an 12 × 12 × 1k-point mesh, and the cutoff energy for the plane-wave basis set was 500 eV. For the finite-size structure, we used only Γ point. A single unit cell contains 18 carbon atoms, where a benzene ring is connected to six neighboring ones by linear chains. In the optimized structure, we have L = 9.46 Å, b1 = 1.43 Å, b2 = 1.40 Å, b3 = 1.23 Å, and b4 = 1.34 Å.
Tight-binding calculation
Low-energy band structure of ABC-stacked graphdiyne can be properly described by pz orbital basis. The intralayer coupling is described by the nearest neighbor hopping parameters expressed as the transfer integral \({V}_{{\mathrm{pp\pi}} }={V}_{{\mathrm{pp\pi}}}^{0}{e}^{(-R-{a}_{0})/{r}_{0}}\), where \({V}_{{\mathrm{pp\pi}}}^{0}\approx -2.7\,{\mathrm{eV}}\) is the hopping parameter for graphene’s nearest neighbor interaction (a0 = 0.142 nm and r0 = 0.09 nm).19 The interlayer coupling is described by two hopping parameters of two shortest vertical bonds which are expressed as the transfer integral \({V}_{{\mathrm{pp\sigma}}}={V}_{{\mathrm{pp\sigma}}}^{0}{e}^{(-R-{d}_{0})/{r}_{0}}\), where Vppσ ≈ 0.48 eV is the hopping parameter at the interlayer distance of graphite (d0 = 0.335 nm).19
Data availability
The datasets generated during and/or analyzed during this study are available from the corresponding author on reasonable request.
References
Schindler, F. et al. Higher-order topology in bismuth. Nat. Phys. 14, 918 (2018).
Schindler, F. et al. Higher-order topological insulators. Sci. Adv. 4, eaat0346 (2018).
Wang, Z., Wieder, B. J., Li, J., Yan, B. & Bernevig, B. A. Higher-order topology, monopole nodal lines, and the origin of large fermi arcs in transition metal dichalcogenides XTe2 (X= Mo, W). Phys. Rev. Lett. 123, 186401 (2019).
Yue, C. et al. Symmetry-enforced chiral hinge states and surface quantum anomalous hall effect in the magnetic axion insulator Bi2−x Smx Se3. Nat. Phys. 15, 577 (2019).
Wieder, B. J. & Bernevig, B. A. The axion insulator as a pump of fragile topology. Preprint at http://arXiv.org/abs/1810.02373 (2018).
Ahn, J. & Yang, B.-J. Symmetry representation approach to topological invariants in c2zt-symmetric systems. Phys. Rev. B 99, 235125 (2019).
Benalcazar, W. A., Bernevig, B. A. & Hughes, T. L. Quantized electric multipole insulators. Science 357, 61–66 (2017).
Langbehn, J., Peng, Y., Trifunovic, L., vonOppen, F. & Brouwer, P. W. Reflection-symmetric second-order topological insulators and superconductors. Phys. Rev. Lett. 119, 246401 (2017).
Geier, M., Trifunovic, L., Hoskam, M. & Brouwer, P. W. Second-order topological insulators and superconductors with an order-two crystalline symmetry. Phys. Rev. B 97, 205135 (2018).
Khalaf, E. Higher-order topological insulators and superconductors protected by inversion symmetry. Phys. Rev. B 97, 205136 (2018).
Benalcazar, W. A., Li, T. & Hughes, T. L. Quantization of fractional corner charge in cn-symmetric higher-order topological crystalline insulators. Phys. Rev. B 99, 245151 (2019).
Song, Z., Fang, Z., & Fang, C. (d−2)-dimensional edge states of rotation symmetry protected topological states. Phys. Rev. Lett. 119, 246402 (2017).
Ahn, J., Park, S. & Yang, B.-J. Failure of Nielsen-Ninomiya theorem and fragile topology in two-dimensional systems with space-time inversion symmetry: application to twisted bilayer graphene at magic angle. Phys. Rev. X 9, 021013 (2019).
Lee, E., Furusaki, A. & Yang, B.-J. Fractional charge bound to a vortex in two dimensional topological crystalline insulators. Preprint at http://arXiv.org/abs/1903.02737 (2019).
Ezawa, M. Minimal models for wannier-type higher-order topological insulators and phosphorene. Phys. Rev. B 98, 045125 (2018).
Călugăru, D., Juričić, V. & Roy, B. Higher-order topological phases: a general principle of construction. Phys. Rev. B 99, 041301 (2019).
Okuma, N., Sato, M. & Shiozaki, K. Topological classification under nonmagnetic and magnetic point group symmetry: Application of real-space atiyah-hirzebruch spectral sequence to higher-order topology. Phys. Rev. B 99, 085127 (2019).
Hwang, Y., Ahn, J. & Yang, B.-J. Fragile topology protected by inversion symmetry: diagnosis, bulk-boundary correspondence, and wilson loop. Phys. Rev. B 100, 205126 (2019).
Nomura, T., Habe, T., Sakamoto, R. & Koshino, M. Three-dimensional graphdiyne as a topological nodal-line semimetal. Phys. Rev. Mater. 2, 054204 (2018).
Ahn, J., Kim, D., Kim, Y. & Yang, B.-J. Band topology and linking structure of nodal line semimetals with Z2 monopole charges. Phys. Rev. Lett. 121, 106403 (2018).
Haley, M. M., Brand, S. C. & Pak, J. J. Carbon networks based on dehydrobenzoannulenes: synthesis of graphdiyne substructures. Angew. Chem. Int. Ed. Engl. 36, 836–838 (1997).
Li, G. et al. Architecture of graphdiyne nanoscale films. Chem. Commun. 46, 3256–3258 (2010).
Zak, J. Berry’s phase for energy bands in solids. Phys. Rev. Lett. 62, 2747 (1989).
Xiao, D., Chang, M.-C. & Niu, Q. Berry phase effects on electronic properties. Rev. Mod. Phys. 82, 1959 (2010).
Song, Z., Zhang, T. & Fang, C. Diagnosis for nonmagnetic topological semimetals in the absence of spin-orbital coupling. Phys. Rev. X 8, 031069 (2018).
Fu, L. & Kane, C. L. Topological insulators with inversion symmetry. Phys. Rev. B 76, 045302 (2007).
Kim, Y., Wieder, B. J., Kane, C. & Rappe, A. M. Dirac line nodes in inversion-symmetric crystals. Phys. Rev. Lett. 115, 036806 (2015).
Hughes, T. L., Prodan, E. & Bernevig, B. A. Inversion-symmetric topological insulators. Phys. Rev. B 83, 245132 (2011).
Turner, A. M., Zhang, Y., Mong, R. S. & Vishwanath, A. Quantized response and topology of magnetic insulators with inversion symmetry. Phys. Rev. B 85, 165120 (2012).
Po, H. C., Vishwanath, A. & Watanabe, H. Symmetry-based indicators of band topology in the 230 space groups. Nat. Commun. 8, 50 (2017).
Zhao, Y. & Lu, Y. pt -symmetric real dirac fermions and semimetals. Phys. Rev. Lett. 118, 056401 (2017).
Fang, C., Chen, Y., Kee, H.-Y. & Fu, L. Topological nodal line semimetals with and without spin-orbital coupling. Phys. Rev. B 92, 081201 (2015).
Yu, R., Qi, X. L., Bernevig, A., Fang, Z. & Dai, X. Equivalent expression of Z2 topological invariant for band insulators using the non-abelian berry connection. Phys. Rev. B 84, 075119 (2011).
Alexandradinata, A., Dai, X. & Bernevig, B. A. Wilson-loop characterization of inversion-symmetric topological insulators. Phys. Rev. B 89, 155114 (2014).
Van Miert, G. & Ortix, C. Higher-order topological insulators protected by inversion and rotoinversion symmetries. Phys. Rev. B 98, 081110 (2018).
Kresse, G. & Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54, 11169 (1996).
Blöchl, P. E. Projector augmented-wave method. Phys. Rev. B 50, 17953 (1994).
Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 3865 (1996).
Acknowledgements
We acknowledge the helpful discussions with Dongwook Kim and Youngkuk Kim. E. L., R. K. and J.A. were supported by IBS-R009-D1. B.-J.Y. was supported by the Institute for Basic Science in Korea (Grant No. IBS-R009-D1) and Basic Science Research Program through the National Research Foundation of Korea (NRF) (Grant No. 0426-20180011), and the POSCO Science Fellowship of POSCO TJ Park Foundation (No. 0426-20180002). This work was supported in part by the U.S. Army Research Office under Grant No. W911NF-18-1-0137.
Author information
Authors and Affiliations
Contributions
E.L. performed tight-binding calculations, R.K. performed first-principles calculations, and E.L. and J.A. did theoretical analysis. B.-J.Y. supervised the project. All authors analyzed the data and wrote the paper.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Lee, E., Kim, R., Ahn, J. et al. Two-dimensional higher-order topology in monolayer graphdiyne. npj Quantum Mater. 5, 1 (2020). https://doi.org/10.1038/s41535-019-0206-8
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41535-019-0206-8
This article is cited by
-
Topological electronic structure and spin texture of quasi-one-dimensional higher-order topological insulator Bi4Br4
Nature Communications (2023)
-
Identifying topological corner states in two-dimensional metal-organic frameworks
Nature Communications (2023)
-
Superconducting dome by tuning through a van Hove singularity in a two-dimensional metal
npj 2D Materials and Applications (2023)
-
Two-dimensional Stiefel-Whitney insulators in liganded Xenes
npj Computational Materials (2022)
-
Topological zero-dimensional defect and flux states in three-dimensional insulators
Nature Communications (2022)