betaFeedback

WorksPhysics and Astronomy

Turning Number and the Geometry of Disorder-Induced Topology

Abstract

The geometry of momentum-space Hamiltonian maps derived from the zero-frequency single-particle Green's function provides a route to a refined topological characterization of quantum systems beyond the noninteracting clean limit. The central idea is to complement standard topological invariants by invariants that characterize the geometry of the map itself. As a paradigmatic example, we study the one-dimensional chiral Su-Schrieffer-Heeger model with disorder, where the topological Hamiltonian defines a planar Bloch loop. In addition to its conventional winding number, this loop possesses an integer turning number that characterizes the winding of its tangent vector. We identify turning-number transitions that occur while the topological Hamiltonian remains finite and spectrally gapped, driven solely by a vanishing tangent vector of the Bloch loop, and obtain an extended phase diagram with additional topological structure beyond the winding-number classification. Asymptotic analysis predicts the emergence and disorder dependence of the corresponding transition regime, in agreement with the numerical results. Our findings establish the geometry of momentum-space Hamiltonian maps as an additional source of topological structure beyond conventional classifications.

Motivation. The traditional classification of phases of matter is based on symmetry and spontaneous symmetry breaking, as embodied in Landau’s paradigm [1, 2, 3]. Topological phases extend this framework by allowing distinct phases that cannot be distinguished by local order parameters or symmetry breaking.

A major achievement in the theory of topological matter is the systematic classification of gapped noninteracting lattice-fermion systems [4, 5, 6]. For each spatial dimension and each combination of time-reversal, particle-hole, and chiral symmetries, gapped Bloch Hamiltonians are assigned topological invariants that are unchanged under continuous deformations preserving both the bulk gap and the relevant symmetries. Different values of these invariants therefore identify distinct topological phases. Conversely, within the stable classification, Hamiltonians with the same invariants can be continuously deformed into one another without closing the gap or breaking the relevant symmetries. Thus, within this framework, the classification is complete.

This completeness, however, refers to classification under continuous deformations and therefore does not resolve geometric properties of the Hamiltonian map that can change without changing its homotopy class. Can such geometric information itself carry robust topological structure? This question suggests a route distinct from existing refinements beyond the stable tenfold-way classification, including finite-band or fragile topology [7, 8], multigap topology [9], and crystalline classifications [10]. These approaches retain additional band, gap, or symmetry information while leaving the underlying notion of continuous deformation essentially unchanged.

For interacting [11, 12] or disordered systems [13], topological phases can likewise be characterized by Green-function topological invariants, including the Ishikawa-Matsuyama invariant and its generalizations [14, 15, 16, 17, 18]. These are homotopy invariants of the single-particle Green’s function G(k,iω)G(k,i\omega) and remain unchanged under continuous deformations as long as neither poles nor zeros of GG are encountered in frequency-momentum space. Unlike in the noninteracting or clean case, however, these invariants do not provide a complete classification of, e.g., interacting topological phases [19, 20, 21]. Nevertheless, they have been widely used [20, 22, 23, 24], not least because, under suitable regularity conditions, they can be evaluated through the topological Hamiltonian [16, 17, 18], Htop(k)=−G−1(k,iω=0)\boldsymbol{H}_{\mathrm{top}}(k)=-G^{-1}(k,i\omega=0). The topological Hamiltonian makes it possible to pose the same geometric question for interacting systems: beyond its conventional homotopy invariants, what additional topological structure is encoded in Htop(k)\boldsymbol{H}_{\mathrm{top}}(k)?

Green-function zeros provide a striking example of topological structure beyond conventional quasiparticle band descriptions. In interacting systems, they arise naturally from singularities of the self-energy and can drive topological transitions alongside Green-function poles [16, 17]. Recently, analogous zero-driven transitions have also been found in noninteracting systems with strong disorder [13]. After disorder averaging, translational invariance is restored and the resulting Green’s function G‾(k,ω)\overline{G}(k,\omega) acquires a nontrivial self-energy, making disordered systems a very transparent setting for studying the geometry of Green-function-defined Hamiltonians without the additional complexity of many-body correlations.

Main result. Here we pursue this idea for the disordered one-dimensional Su-Schrieffer-Heeger (SSH) model, with two sublattices, intracell and intercell hoppings t1t_1 and t2t_2, and disorder strength WW. Chiral symmetry of Htop(k)=d(k)⋅σ\boldsymbol{H}_{\mathrm{top}}(k)=\boldsymbol{d}(k)\cdot\boldsymbol{\sigma}, where σ\boldsymbol{\sigma} denotes the Pauli matrices, constrains the Bloch vector d(k)\boldsymbol{d}(k) to a plane. As kk traverses the BZ, d(k)\boldsymbol{d}(k) therefore traces a closed loop. Besides its winding number ν\nu around the origin, this loop possesses a second integer-valued quantity, its turning number τ\tau, which measures the winding of its tangent vector ∂kd(k)\partial_k\boldsymbol{d}(k). The turning number is invariant under regular homotopy, i.e., under smooth deformations that preserve the immersion condition ∂kd(k)≠0\partial_k d(k) \ne0, unlike ordinary homotopy, where the relevant condition is d(k)≠0d(k) \ne0.

Figure 1 summarizes our central result. While ν\nu distinguishes the conventional topological sectors, τ\tau reveals additional structure within them. In the clean limit, the usual transition at t1=t2t_1 = t_2 is driven by a Green-function pole and changes ν\nu from 0 to 1 without changing τ\tau. At strong disorder, the corresponding transition along the same (solid white) line is instead driven by Green-function zeros; although ν\nu again changes from 0 to 1, the two transitions are distinguished by τ\tau. Moreover, τ\tau exhibits an additional transition regime, located around the yellow line, within a fixed winding-number sector, revealing structure that is invisible to the conventional homotopy invariant. Details of the computations, including an additional phase with τ=0\tau= 0 (light blue) and the region with scattered τ\tau values, are discussed below. For the white dashed(-dotted) Weyl lines, see also Sec. A of the Supplemental Material (SM) [25].

Clean system

The clean SSH model [26] describes spinless fermions on a chain of LL unit cells with A and B sublattice sites: H^0=∑i=1L(t1ai†bi+t2ai+1†bi+H.c.)\hat{H}_0 = \sum_{i=1}^{L}(t_1a_i^\dagger b_i + t_2a_{i+1}^\dagger b_i + \mathrm{H.c.}), where t1t_1 and t2t_2 are the intra- and intercell hoppings. Periodic boundary conditions (aL+1≡a1a_{L+1} \equiv a_1) are assumed.

For t2>t1t_2 > t_1, the system is in a chiral-symmetry-protected topological phase with winding number ν=1\nu= 1, supporting localized zero-energy edge states under open boundary conditions.

The Bloch Hamiltonian reads H0(k)=d0(k)⋅σH_0(k) = d_0(k) \cdot\boldsymbol{\sigma}, with d0(k)=(t1+t2cos⁡k,t2sin⁡k)d_0(k) = (t_1 + t_2 \cos k, t_2 \sin k), σ=(σx,σy)\boldsymbol{\sigma} = (\sigma_x,\sigma_y), and k∈(−π,π]k \in(-\pi,\pi]. The Bloch vector defines a smooth map d0:S1→R2∖{0}d_0 : S^1 \to\mathbb{R}^2 \setminus\{0\}, whose oriented winding around the origin is counted by ν∈Z\nu\in\mathbb{Z}. Mathematically, ν\nu labels the homotopy classes π1(R2∖{0})≅Z\pi_1(\mathbb{R}^2 \setminus\{0\}) \cong\mathbb{Z}. The loops and corresponding band structures are shown in Figure 2.

Figure 2. Band structure ϵ±(k)\epsilon_\pm(k) for −π<k≤π-\pi< k \le\pi (top) and Bloch vector d0(k)d_0(k) (bottom) in the clean limit W=0W = 0. From left to right: (t1,t2)=(1,0),(1,0.5),(1,1),(0.5,1),(0,1)(t_1,t_2) = (1,0),(1,0.5),(1,1),(0.5,1),(0,1).

The winding number becomes ill-defined when the loop crosses the origin, corresponding to a closing of the band separation Δ(k)=2∣d0(k)∣=2t12+t22+2t1t2cos⁡k\Delta(k) = 2\lvert d_0(k)\rvert= 2\sqrt{t_1^2 + t_2^2 + 2t_1t_2\cos k}. This occurs at t2=t1t_2 = t_1 and kc=πk_c = \pi. The Green’s function G0(k,ω)=(ω−H0(k))−1G_0(k,\omega) = (\omega-H_0(k))^{-1} then satisfies G0(kc,ω)=1/ωG_0(k_c,\omega) = 1/\omega, implying a pole at ω=0\omega= 0 and a divergent det⁡G0(kc,ω)\det G_0(k_c,\omega) as ω→0\omega\to0.

Adding disorder

The winding number also admits a formulation directly in terms of the single-particle Green’s function [16, 22]. For a nonsingular Green’s function at zero frequency, the winding number defined in this way equals that obtained from the topological Hamiltonian [17]. It can change only when G(k,0)G(k,0) becomes singular, i.e., when det⁡G(k,0)\det G(k,0) vanishes or diverges for some kk. While poles already occur in clean noninteracting systems, Green-function zeros require interactions or, analogously, disorder averaging.

We consider binary potential disorder, i.e., H^=H^0+V^\hat{H} = \hat{H}_0+\hat{V} with V^=∑i(ϵiAai†ai+ϵiBbi†bi)\hat{V} = \sum_i(\epsilon_{iA}a_i^\dagger a_i + \epsilon_{iB}b_i^\dagger b_i). The ϵiA\epsilon_{iA} and ϵiB\epsilon_{iB} are independent and identically distributed, taking values ±W/2\pm W/2 with equal probability. H^0\hat{H}_0 has chiral symmetry with respect to Γ^=diag⁡(1A,−1B)\hat{\Gamma} = \operatorname{diag}(1_A,-1_B), i.e., the anticommutator [Γ^,H^0]+=0[\hat{\Gamma},\hat{H}_0]_+ = 0. Although a generic disorder realization breaks chiral symmetry, we have Γ^(H^0+V^)Γ^=−(H^0−V^)\hat{\Gamma}(\hat{H}_0+\hat{V})\hat{\Gamma} = -(\hat{H}_0-\hat{V}). Since V^\hat{V} and −V^-\hat{V} occur with equal probability, the disorder-averaged Green’s function recovers the chiral constraint at zero frequency. Translational symmetry is likewise restored by disorder averaging. For ease of notation, we denote its Fourier transform again by G(k,ω)G(k,\omega).

Excitation spectrum

Numerical results have been obtained by averaging over N=1000N = 1000 independent configurations for a system with L=400L = 400 unit cells. Figure 3 (left) shows the evolution of the spectral density A(k,ω)≡AAA(k,ω)=ABB(k,ω)=(−1/π)Im⁡GAA(k,ω+i0+)A(k,\omega) \equiv A_{AA}(k,\omega) = A_{BB}(k,\omega) = (-1/\pi)\operatorname{Im}G_{AA}(k,\omega+i0^+) with increasing WW, keeping t1=1.25t_1 = 1.25 and t2=0.75t_2 = 0.75 fixed. The imaginary part of the retarded diagonal disorder self-energy ΣAA(k,ω+i0+)\Sigma_{AA}(k,\omega+i0^+) is shown in the right panels (note the different scales). The high-frequency expansion yields the sum rule L−1∑k∫−∞∞(−1/π)Im⁡ΣAA(k,ω+i0+)dω=W2/4L^{-1}\sum_k\int_{-\infty}^{\infty}(-1/\pi)\operatorname{Im}\Sigma_{AA}(k,\omega+i0^+)\mathrm{d}\omega= W^2/4.

Figure 3. Imaginary part of the disorder-averaged Green’s function (left) and of the disorder self-energy (right) for different disorder strengths WW. Parameters: t1=0.75t_1 = 0.75, t2=1.25t_2 = 1.25, L=400L = 400, N=1000N = 1000.

As seen in the figure, A(k,ω)A(k,\omega) undergoes a pronounced reconstruction. The original clean bands at W=0W = 0 evolve into a number of weakly dispersive structures. At W=3W = 3, these structures, together with their spectral weights, still loosely resemble the clean-limit dispersions, although they have been shifted towards higher absolute frequencies and strongly broadened. At low frequencies around ω=0\omega= 0, a broad low-energy spectral structure develops. Upon further increasing WW, this feature is progressively depleted around ω=0\omega= 0, with the gap opening first near the BZ boundary and subsequently extending over the entire BZ. At W=5W = 5, a full spectral gap has formed again. The insulating state reached at large WW, however, is qualitatively different from the clean W=0W = 0 insulator. In particular, the reopening of the spectral gap is accompanied by two weakly dispersive branches of Green-function zeros, i.e., poles of the self-energy (lower-right panel of Figure 3), at low frequencies.

Topology carried by zeros

These zero bands in fact carry a nontrivial topology. We have numerically computed the winding number via ν=12π∫BZdk ∂karg⁡[dx(k)+idy(k)]\nu= \frac{1}{2\pi}\int_{\mathrm{BZ}} dk\,\partial_k \arg[d_x(k) + i d_y(k)], where d(k)\mathbf{d}(k) is obtained from Htop(k)=d(k)⋅σ=−G−1(k,ω=0)H_{\mathrm{top}}(k) = \mathbf{d}(k)\cdot\boldsymbol{\sigma} = -G^{-1}(k,\omega= 0). In the clean limit at W=0W = 0, ν\nu jumps from 00 to 11 at t2,c=t1=0.75t_{2,c} = t_1 = 0.75, where the gap closes at kc=πk_c = \pi, see Figure 2. For strong disorder, e.g., W=5W = 5, the situation appears the same at the level of ν\nu: The winding number again jumps from ν=0\nu= 0 to ν=1\nu= 1 at t2,c=t1t_{2,c} = t_1. However, this transition is qualitatively different from the clean-limit case, since there is no gap closure at t2,ct_{2,c}. Rather, the two zero branches meet at ω=0\omega= 0 and kc=πk_c = \pi, i.e., det⁡G(kc,ω=0)=0\det G(k_c,\omega= 0) = 0 for t2=t2,ct_2 = t_{2,c}. Comparing A(k,ω)A(k,\omega) at W=0W = 0 with (−1/π)Im⁡ΣAA(k,ω)(-1/\pi)\operatorname{Im}\Sigma_{AA}(k,\omega) at W=5W = 5 for the noncritical value t2=1.25t_2 = 1.25, highlights the close analogy between the pole and zero structures, see the upper-left and lower-right panels of Figure 3. In Sec. B of the SM [25], we demonstrate by a real-space calculation that the zero bands indeed carry the topology of the system. This complements the findings of Wagner et al. [27], who demonstrated a Mott insulator whose nontrivial topology is carried by Green-function zeros, stimulating several related studies [28, 29, 30, 31, 32]; see also Refs. [33, 34] and, for a two-dimensional disordered system, Ref. [13]. The numerical results for ν\nu in the entire WW-t2t_2 plane at fixed t1=0.75t_1 = 0.75 are shown in Sec. C of the SM. The constant-ν\nu phases are also indicated by the labels in Figure 1.

Turning number phase diagram

Figure 1. Turning number τ\tau (color code) as a function of t2/t1t_2/t_1 and WW for t1=0.75t_1 = 0.75. Left of the dashed and right of the dashed-dotted line, Weyl’s inequality guarantees an open spectral gap.

Turning number

The winding number does not distinguish between the transitions at weak and strong WW, i.e., it does not indicate whether poles or zeros carry the topology. We therefore propose to additionally consider the turning number τ\tau [35], defined as the winding number of d˙(k)≡∂kd(k)\dot{\mathbf{d}}(k) \equiv\partial_k\mathbf{d}(k). Thus, τ\tau counts the windings of the tangent vector d˙(k)\dot{\mathbf{d}}(k) rather than of d(k)\mathbf{d}(k) itself. Mathematically, τ=deg⁡(d˙/∣d˙∣)∈Z\tau= \deg(\dot{\mathbf{d}}/|\dot{\mathbf{d}}|) \in\mathbb{Z}, provided that d˙(k)≠0\dot{\mathbf{d}}(k) \ne0 throughout the BZ. It is therefore a regular-homotopy invariant of the planar loop d(k)\mathbf{d}(k) itself and, in particular, is independent of the position of the loop relative to the origin or any other fixed point. Numerically, we evaluate τ=12π∫BZdk ∂karg⁡[∂kdx(k)+i∂kdy(k)]\tau= \frac{1}{2\pi}\int_{\mathrm{BZ}} dk\,\partial_k \arg[\partial_k d_x(k) + i\partial_k d_y(k)]. Compared with ν\nu, this amounts to replacing d(k)\mathbf{d}(k) by ∂kd(k)\partial_k\mathbf{d}(k).

Applications of the turning number have so far mainly concerned the geometry of immersed momentum-space contours, such as figure-eight Fermi surfaces in thin-film Weyl semimetals [36], spiral Fermi surfaces in quasicrystalline systems and twisted bilayer graphene [37], and higher-turning-number spiral Fermi surfaces in twisted multilayer systems [38]. Turning-number-based immersion classes have also been used to construct superselection sectors for anyonic systems [39]. By contrast, here the turning number characterizes the geometry of the map k↦Htop(k)k \mapsto H_{\mathrm{top}}(k).

In the clean limit W=0W = 0, see Figure 2, τ\tau is insensitive to the topological transition. Upon varying t2/t1t_2/t_1, the planar loop d0(k)\mathbf{d}_0(k) passes through the origin d0=0\mathbf{d}_0 = 0 at t2,c=t1t_{2,c} = t_1, where ν\nu changes from ν=0\nu= 0 to ν=1\nu= 1. The tangent vector, however, remains nonzero throughout, including at t2,ct_{2,c}. As the BZ is traversed once in the positive direction, the tangent vector winds once in the positive direction. Consequently, τ=+1\tau= +1 independently of t2/t1t_2/t_1.

For strong disorder W=3.5W = 3.5, by contrast, the loop passes through “infinity” rather than through d=0\mathbf{d} = 0. This is visualized in Figure 4 (top panels) using an inverse stereographic projection onto a sphere representing the compactified d\mathbf{d} plane. The north pole (N, red dot) and south pole (S, green dot) represent ∣d∣=∞|\mathbf{d}| = \infty and d=0\mathbf{d} = 0, respectively. Exactly at t2=t1t_2 = t_1, the loop reaches N at k=πk = \pi. The middle panels in Figure 4 display the evolution of the loop in the d\mathbf{d} plane. With increasing t2t_2, but still t2<t1t_2 < t_1, the loop grows. For t2→t2,c=t1t_2 \to t_{2,c} = t_1, the loop becomes singular and extends to infinity, with dx(π)→−∞d_x(\pi) \to-\infty, since a zero of det⁡G(k,ω)\det G(k,\omega) crosses ω=0\omega= 0 at k=πk = \pi. For t2>t1t_2 > t_1, the loop reappears from infinity on the opposite (dx>0d_x > 0) side and closes again; upon further increasing t2t_2, it shrinks. Importantly, the loops for t2<t1t_2 < t_1 and t2>t1t_2 > t_1 differ in both invariants: the winding number jumps from ν=0\nu= 0 to ν=1\nu= 1, while the turning number changes from τ=−1\tau= -1 to τ=+1\tau= +1. The latter qualitatively distinguishes the strong-disorder transition from its clean-limit counterpart and, for t2<t1t_2<t_1, also distinguishes the clean and strong-disorder regimes despite their common winding number ν=0\nu=0.

Loops in the $d_x$-$d_y$ plane and inverse stereographic projections onto the sphere $S^2$

Figure 4. Loop d(k)d(k) in the dxd_x-dyd_y plane (middle) and its inverse stereographic projection onto the sphere S2S^2 (top panels) for various t2t_2 at W=3.5W=3.5. Bottom: The same but for W=2.5W=2.5. Parameters: t1=0.75t_1=0.75, L=600L=600, N=1000N=1000.

Refined topological phase diagram. Figure 1 shows the entire WW-t2t_2 topological phase diagram at fixed t1=0.75t_1=0.75 as obtained for L=200L=200 unit cells and N=2000N=2000 disorder configurations (see also SM [25], Sec. C). The winding and turning numbers have been obtained numerically as described above. There is a parameter region where ν\nu and τ\tau are ill-defined as topological invariants, since the disorder-averaged spectral function has finite weight at zero frequency, A(k,0)>0A(k,0)>0 for some kk, and the zero-frequency retarded Green’s function, and hence Htop(k)H_{\mathrm{top}}(k), is no longer Hermitian. The numerical winding formulas nevertheless still return integer values. These values, however, are no longer topologically protected and fluctuate strongly between neighboring points in parameter space, rather than forming extended regions of constant ν\nu and τ\tau. Weyl’s inequality [40] provides a sufficient criterion for an open spectral gap and hence for ν\nu to be well defined; see SM [25], Sec. A, and the white dashed and dashed-dotted lines in Figure 1.

Outside the region bounded by the Weyl lines, we find ν=0\nu=0 for t2/t1<1t_2/t_1<1 and ν=1\nu=1 for t2/t1>1t_2/t_1>1. The transition is pinned to t2=t1t_2=t_1 by the symmetry of the disorder-averaged problem under interchange of the intra- and intercell hoppings, t1↔t2t_1\leftrightarrow t_2 (SM [25], Sec. D). The turning number, by contrast, is not constrained by this hopping-exchange symmetry, and its transitions can occur away from t2=t1t_2=t_1.

While the winding number ν\nu is the topological invariant characterizing one-dimensional chiral systems in class AIII, the turning number τ\tau classifies the corresponding planar loops of the topological Hamiltonian under regular homotopy, i.e., under continuous deformations for which the tangent vector never vanishes (the “immersion” condition). It therefore resolves additional geometric structure within a fixed winding-number sector.

Changes of τ\tau between gapped regions can occur across a gapless region, as for t2<t1t_2<t_1, or across a divergence of HtopH_{\mathrm{top}} associated with a Green-function zero at t2=t1t_2=t_1. More interestingly, however, τ\tau can also change while HtopH_{\mathrm{top}} remains finite and gapped, namely when the immersion condition is lost, ∂kd(k)=0\partial_kd(k)=0 for some kk. Such a transition has no counterpart in the winding-number classification.

Immersion-driven turning-number transition. In fact, such a transition is predicted by a perturbative calculation of the complexified dd vector, d+=dx+idyd_+=d_x+\mathrm{i}d_y, for weak t2→0t_2\to0 and strong W→∞W\to\infty. The corresponding asymptotic expansions depend on the order in which these limits are taken. Expanding first in t2/t1t_2/t_1 and subsequently taking W≫t1W\gg t_1, one finds d+(k)≃−W416t13+t2eikd_+(k)\simeq-\frac{W^4}{16t_1^3}+t_2e^{\mathrm{i}k}. On the other hand, taking W→∞W\to\infty first and subsequently expanding in t2/t1t_2/t_1 yields, for t2<t1t_2<t_1, d+(k)≃−W416t13+W4t2216t15e−ikd_+(k)\simeq-\frac{W^4}{16t_1^3}+\frac{W^4t_2^2}{16t_1^5}e^{-\mathrm{i}k} (see SM [25], Secs. E and F for details). We note that the respective kk-independent leading terms are identical. Importantly, however, the leading kk-dependent corrections have opposite windings. The crossover between the two asymptotic regimes is therefore expected when the amplitudes of these corrections become comparable. Equating the corresponding coefficients gives the crossover scale t2,c≃4t13W2t_{2,c}\simeq\frac{4t_1^3}{W^2}. This provides an estimate for the onset of the turning-number transition regime in the WW-t2t_2 phase diagram, indicated by the yellow line in Figure 1.

The agreement with the numerically obtained τ\tau values is nearly quantitative for W≳3W\gtrsim3, see Figure 1. For W≳4W\gtrsim4, however, the numerical accuracy deteriorates, leading to scattered data. Since perturbation theory, at least to the orders considered, cannot distinguish between a direct τ=−1→+1\tau=-1\to+1 transition and a transition via an intermediate τ=0\tau=0 phase, we performed t2t_2-dependent calculations at fixed W=2.5W=2.5 (see black dots in the inset in Figure 1) for larger systems with L=600L=600 unit cells. The bottom panels of Figure 4 indeed demonstrate the existence of a τ=0\tau=0 (figure-eight) phase for t2,c1≈0.203<t2<0.224≈t2,c2t_{2,c1}\approx0.203<t_2<0.224\approx t_{2,c2}. At the two critical values, the dd loops exhibit cusp singularities at the time-reversal-invariant momenta k=0k=0 and π\pi, respectively (see also SM [25], Sec. G). Altogether, these results show that the regular-homotopy classification by τ\tau reveals geometric distinctions of the topological Hamiltonian that are invisible to its conventional homotopy classification by ν\nu.

Perspectives. Our results suggest several immediate directions. A particularly important one is to search for analogous τ\tau-resolved phases and immersion-driven transitions in interacting one-dimensional systems, thereby testing whether the geometric refinement found here extends beyond disorder. A complementary direction is to develop the gauge-covariant formulation of this geometric refinement, in which the immersion condition can be expressed in terms of a covariant derivative of HtopH_{\mathrm{top}}. Another direction concerns a bulk-boundary correspondence for τ\tau, for example through the spectrum of ∂kHtop\partial_k H_{\mathrm{top}} or a related current operator. The immersion transition is expected to leave observable signatures, as the associated change in the geometry of the Hamiltonian map can modify van Hove singularities and thereby produce characteristic features in optical response. Most importantly, the present construction motivates geometric classifications in higher dimensions, with immersion singularities providing a natural starting point for two-dimensional Hamiltonian maps.

Acknowledgments. This work was supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) through the research unit QUAST, FOR 5249 (project P8), project ID 449872909.

Turning Number and the Geometry of Disorder-Induced Topology — Supplemental Material —

Lars Emmrich; Michael Potthoff

Implications from Weyl’s inequality

Weyl’s inequality [40] provides a simple rigorous bound how disorder can reduce the gap of the SSH model. Let A^=A^0+A^1\hat{A} = \hat{A}_0 + \hat{A}_1, where A^0\hat{A}_0 and A^1\hat{A}_1 are Hermitian operators acting on a linear space of finite dimension DD. Weyl’s inequality implies

∣λn(A^)−λn(A^0)∣≤∥A^1∥,(1)\left|\lambda_n(\hat{A}) - \lambda_n(\hat{A}_0)\right| \le\lVert\hat{A}_1\rVert, \tag*{(1)}

where λn(A^)\lambda_n(\hat{A}) denote the eigenvalues of a Hermitian operator AA, ordered as λ1(A)≥⋯≥λn(A)≥⋯≥λD(A)\lambda_1(A) \ge\cdots\ge\lambda_n(A) \ge\cdots\ge\lambda_D(A). Furthermore, ∥A^∥=sup⁡∥ψ∥=1∥A^ψ∥\lVert\hat{A}\rVert= \sup_{\lVert\psi\rVert=1} \lVert\hat{A}\psi\rVert is the operator norm of A^\hat{A}. Equivalently, ∥A^∥=max⁡n∣λn(A)∣\lVert\hat{A}\rVert= \max_n \lvert\lambda_n(A)\rvert.

Let H^=H^0+V^\hat{H} = \hat{H}_0 + \hat{V}, where H^0\hat{H}_0 and V^\hat{V} are the clean Hamiltonian and the disorder potential, respectively. Here, for binary disorder, we have ∥V^∥=W/2\lVert\hat{V}\rVert= W/2. With Eq. (1) we can conclude that the unoccupied states have energies EnE_n with

En≥Δ2−∥V^∥=Δ2−W2,(2)E_n \ge\frac{\Delta}{2} - \lVert\hat{V}\rVert= \frac{\Delta}{2} - \frac{W}{2}, \tag*{(2)}

where Δ\Delta denotes the gap in the clean limit. Analogously, the occupied states have energies such that

En≤−Δ2+∥V^∥=−Δ2+W2.(3)E_n \le-\frac{\Delta}{2} + \lVert\hat{V}\rVert= -\frac{\Delta}{2} + \frac{W}{2}. \tag*{(3)}

Hence the disordered system has a spectral gap satisfying

Δ(W)≥Δ−W.(4)\Delta(W) \ge\Delta- W . \tag*{(4)}

In the clean limit, H^0\hat{H}_0, the gap is given by

Δ=2∣∣t1∣−∣t2∣∣.(5)\Delta= 2\left||t_1| - |t_2|\right| . \tag*{(5)}

Thus

Δ(W)≥2∣∣t1∣−∣t2∣∣−W.(6)\Delta(W) \ge2\left||t_1| - |t_2|\right| - W . \tag*{(6)}

In particular, the gap is guaranteed to remain open as long as

W<2∣∣t1∣−∣t2∣∣.(7)W < 2\left||t_1| - |t_2|\right| . \tag*{(7)}

One may apply Eq. (1) once more, but with exchanged roles of H^0\hat{H}_0 and V^\hat{V}. This is meaningful in the strong-disorder regime W≫∣t1∣,∣t2∣W \gg|t_1|, |t_2|.

The “unperturbed” operator V^\hat{V} has only the two eigenvalues ±W/2\pm W/2. Treating H^0\hat{H}_0 as the perturbation, Weyl’s inequality, see Eq. (1), implies that every eigenvalue of V^\hat{V} can be shifted by at most ∥H^0∥\lVert\hat{H}_0\rVert. Consequently, the lower group of eigenvalues of H^\hat{H} is contained in the interval

[−W2−∥H^0∥,−W2+∥H^0∥],(8)\left[-\frac{W}{2} - \lVert\hat{H}_0\rVert, -\frac{W}{2} + \lVert\hat{H}_0\rVert\right], \tag*{(8)}

whereas the upper group is contained in

[+W2−∥H^0∥,+W2+∥H^0∥].(9)\left[+\frac{W}{2} - \lVert\hat{H}_0\rVert, +\frac{W}{2} + \lVert\hat{H}_0\rVert\right]. \tag*{(9)}

Hence the gap between the two spectral regions satisfies

Δ(W)≥W−2∥H^0∥.(10)\Delta(W) \ge W - 2\lVert\hat{H}_0\rVert. \tag*{(10)}

A finite gap is therefore guaranteed whenever

W>2∥H^0∥.(11)W > 2\lVert\hat{H}_0\rVert. \tag*{(11)}

For the clean SSH Hamiltonian with intracell and intercell hoppings t1t_1 and t2t_2, the band dispersions are given by

ϵ±(k)=±∣t1+t2e−ik∣,(12)\epsilon_{\pm}(k) = \pm\left|t_1 + t_2 e^{-ik}\right| , \tag*{(12)}

and thus

∥H^0∥=max⁡k∣t1+t2e−ik∣=∣t1∣+∣t2∣.(13)\lVert\hat{H}_0\rVert= \max_k\left|t_1 + t_2 e^{-ik}\right| = |t_1| + |t_2| . \tag*{(13)}

Therewith, we find

Δ(W)≥W−2(∣t1∣+∣t2∣),(14)\Delta(W) \ge W - 2\left(|t_1| + |t_2|\right) , \tag*{(14)}

and thus

W>2(∣t1∣+∣t2∣)(15)W > 2\left(|t_1| + |t_2|\right) \tag*{(15)}

is a sufficient condition for the spectrum to be gapped. It should be stressed that the Weyl bounds give sufficient conditions for the persistence of the spectral gap, but not the actual critical disorder strengths. We can combine Eq. (7) and Eq. (15): If

W<2∣∣t1∣−∣t2∣∣(16)W < 2\left||t_1| - |t_2|\right| \tag*{(16)}

the gap is certainly open. If 2∣t1∣−∣t2∣≤W≤2(∣t1∣+∣t2∣)2\lvert t_1\rvert- \lvert t_2\rvert\le W \le2(\lvert t_1\rvert+\lvert t_2\rvert) Weyl’s inequality gives no conclusion. If

W>2(∣t1∣+∣t2∣)(17)W > 2(\lvert t_1\rvert+\lvert t_2\rvert) \tag*{(17)}

the gap is certainly open again.

Edge poles and edge zeros.

In the strong-disorder regime and for t2>t1t_2 > t_1, the imaginary part of the disorder self-energy, (−1/π)Im⁡ΣAA(k,ω+iϵ)(-1/\pi)\operatorname{Im}\Sigma_{AA}(k,\omega+\mathrm{i}\epsilon), exhibits two weakly dispersing sharp features in the low-frequency regime, as shown in the bottom-right panel of Figure 3. The apparent finite spectral width of these features results exclusively from the finite broadening parameter. They therefore correspond to poles of the self-energy and, consequently, to zero bands of the Green’s function G(k,ω)G(k,\omega). In the parameter regime considered here, the winding number obtained from the topological Hamiltonian is nontrivial and given by ν=1\nu=1.

The fact that this nontrivial topology is carried by the zero bands can be further corroborated by considering the disordered SSH model in real space with open boundaries, as shown in Figure 5. Specifically, we consider a chain of L=20L=20 sites consisting of two halves: an almost clean part on the left, with W=0.1W=0.1 on sites i=1,…,10i=1,\ldots,10, and a strongly disordered part on the right, with W=5W=5 on sites i=11,…,20i=11,\ldots,20. The sites i=1i=1 and i=20i=20 form the outer boundaries and are coupled to their respective neighboring sites by the weaker hopping t1<t2t_1<t_2.

Figure 5. Top: Imaginary part of the local, disorder-averaged Green’s function, (−1/π)Im⁡Gii(ω+iϵ)(-1/\pi)\operatorname{Im}G_{ii}(\omega+\mathrm{i}\epsilon), with i=1,…,20i=1,\ldots,20 labeling the sites of an SSH chain with disorder strength W=0.1W=0.1 in the left half and W=5W=5 in the right half (left and right of the dashed red line). Bottom: Imaginary part of the corresponding disorder self-energy, (−1/π)Im⁡Σii(ω+iϵ)(-1/\pi)\operatorname{Im}\Sigma_{ii}(\omega+\mathrm{i}\epsilon). Parameters: t1=0.75t_1=0.75, t2=1.25t_2=1.25, L=20L=20, N=1000N=1000, Nω=200N_\omega=200, and ϵ=Δω\epsilon=\Delta\omega.

On the left-hand side, the nontrivial topology is represented by poles of the Green’s function. For ν=1\nu=1, the bulk-boundary correspondence predicts a zero-energy edge pole. This is clearly visible in the top panel of Figure 5 at the left boundary, i=1i=1, with a spectral weight that rapidly decreases with increasing distance from the edge. By contrast, no corresponding low-energy spectral weight is found at the interface site i=10i=10.

Conversely, on the strongly disordered right-hand side, the nontrivial topology is represented by an edge zero. It appears as a pole of the self-energy with large spectral weight at the right boundary, i=20i=20, in the lower panel of Figure 5. Again, the weight rapidly decreases with increasing distance from the boundary, while no corresponding low-energy self-energy peak is found at the interface site i=11i=11.

To make the interface physics explicit, we denote the hopping between sites i=10i=10 and i=11i=11 by tintt_{\mathrm{int}}. Upon switching off this hopping, tint=0t_{\mathrm{int}}=0, a second edge pole appears on the left part at i=10i=10 (not shown), while a second edge zero appears on the right part at i=11i=11 (not shown), as expected from the bulk-boundary correspondence for the two topologically nontrivial subsystems. Upon restoring tintt_{\mathrm{int}} to its bulk value, the pole and zero at the interface hybridize and the zero-energy interface singularities disappear. This realizes the “topological matter-antimatter annihilation” [27], i.e., a Green-function pole annihilates with a Green-function zero.

The effect is analogous to the pole-zero annihilation found in Ref. [27] at the interface between a noninteracting topological SSH+U+U chain and a strongly interacting topological SSH+U+U chain.

Topological phase diagram obtained from the winding number.

Figure 6 shows the result of a numerical computation of the winding number ν\nu, as defined via ν=12π∫BZdk ∂karg⁡[dx(k)+idy(k)]\nu=\frac{1}{2\pi}\int_{\mathrm{BZ}}\mathrm{d}k\,\partial_k\arg[d_x(k)+\mathrm{i}d_y(k)], in the WW-t2t_2 plane for fixed t1=0.75t_1=0.75. The phases with ν=0\nu=0 for t2<t1t_2<t_1 and ν=1\nu=1 for t2>t1t_2>t_1 in the clean limit (see Eq. (16)) or at weak disorder are clearly separated from ν=0\nu=0 and ν=1\nu=1 phases in the strong-disorder limit by a gapless phase.

Figure 6. Winding number ν\nu (color code) as function of t2/t1t_{2}/t_{1} and WW for t1=0.75t_{1}=0.75. Orange dot: critical point at W=0W=0. Orange line: topological transition line. Left of the dashed and right of the dashed-dotted white lines, Weyl’s inequality guarantees that the spectral gap is open, see Eqs. (16) and (17). Parameters: L=200L=200, N=2000N=2000.

The exact boundaries of the gapless phase are difficult to determine numerically. However, Weyl’s inequality [40] is very useful to identify the region, where the gap is certainly open and thus where ν\nu is well-defined as a topological invariant. This is discussed in Sec. A of the SM, see Eqs. (16) and (17) in particular. Left of the dashed and right of the dashed-dotted line, Weyl’s inequality guarantees that the spectral gap is open.

Formally, the winding of an individual closed d(k)\mathbf{d}(k) loop, as computed from ν=12π∫BZdk ∂karg⁡[dx(k)+idy(k)]\nu=\frac{1}{2\pi}\int_{\mathrm{BZ}}\mathrm{d}k\,\partial_k\arg[d_x(k)+\mathrm{i}d_y(k)], remains integer-valued as long as the loop avoids the origin. However, for a gapless phase, these integers are not topologically protected and fluctuate strongly between neighboring points in parameter space, rather than forming extended regions of constant ν\nu.

Chiral symmetry and the t1=t2t_{1}=t_{2} line.

The clean SSH Hamiltonian possesses (sublattice) chiral symmetry,

ΓH0(k)Γ=−H0(k),(18)\Gamma H_{0}(k)\Gamma=-H_{0}(k), \tag*{(18)}

where the chiral matrix Γ\Gamma is unitary and Γ2=1\Gamma^{2}=1. While individual disorder configurations generically break chiral symmetry, the symmetry is restored by the disorder ensemble, and hence

ΓG(k,ω)Γ=−G(k,−ω)(19)\Gamma G(k,\omega)\Gamma=-G(k,-\omega) \tag*{(19)}

for the disorder-averaged Green’s function at complex frequencies. We note in passing, that this immediately implies

ΓHtop(k)Γ=−Htop(k)(20)\Gamma H_{\mathrm{top}}(k)\Gamma=-H_{\mathrm{top}}(k) \tag*{(20)}

for the topological Hamiltonian Htop(k)=−G−1(k,iω=0)H_{\mathrm{top}}(k)=-G^{-1}(k,i\omega=0).

Taking the determinant of the 2×22\times2 matrices on both sides of Eq. (19), and using Γ2=1\Gamma^{2}=1, we get

det⁡G(k,ω)=det⁡G(k,−ω).(21)\det G(k,\omega)=\det G(k,-\omega). \tag*{(21)}

Now let ωz,n(k)\omega_{z,n}(k) denote a Green-function zero, det⁡G(k,ωz,n(k))=0\det G(k,\omega_{z,n}(k))=0, then

0=det⁡G(k,ωz,n(k))=det⁡G(k,−ωz,n(k)).(22)0=\det G(k,\omega_{z,n}(k))=\det G(k,-\omega_{z,n}(k)). \tag*{(22)}

Hence, for every zero at positive frequency there must be a zero at the corresponding negative frequency at the same momentum:

ωz,nˉ(k)=−ωz,n(k).(23)\omega_{z,\bar{n}}(k)=-\omega_{z,n}(k). \tag*{(23)}

For two zero branches, labeled by ++ and −-, this reads

ωz,+(k)=−ωz,−(k).(24)\omega_{z,+}(k)=-\omega_{z,-}(k). \tag*{(24)}

Analogously, we have

ωp,+(k)=−ωp,−(k)(25)\omega_{p,+}(k)=-\omega_{p,-}(k) \tag*{(25)}

for the Green-function poles. Consequently, if one of the two chiral-symmetry-related zero bands reaches the Fermi level at any momentum kck_{\mathrm{c}}, i.e., ωz,+(kc)=0\omega_{z,+}(k_{\mathrm{c}})=0, then its partner necessarily reaches the Fermi level at the same momentum:

ωz,+(kc)=ωz,−(kc)=0.(26)\omega_{z,+}(k_{\mathrm{c}})=\omega_{z,-}(k_{\mathrm{c}})=0. \tag*{(26)}

We now consider the unitary transformation U^\hat{U} acting on the annihilators as

ai↦U^aiU^†=bi,bi↦U^biU^†=ai+1.(27)a_i\mapsto\hat{U}a_i\hat{U}^{\dagger}=b_i,\qquad b_i\mapsto\hat{U}b_i\hat{U}^{\dagger}=a_{i+1}. \tag*{(27)}

Under U^\hat{U}, the clean Hamiltonian, with periodic boundary conditions aL+1≡a1a_{L+1}\equiv a_1, transforms as

H^0=∑i=1L(t1ai†bi+t2ai+1†bi+H.c.)↦U^H^0U^†=∑i=1L(t1bi+1†ai+1+t2bi+1†ai+1+H.c.)=∑i=1L(t2bi†ai+t1bi†ai+1+H.c.)(28)\begin{aligned} \hat{H}_0&=\sum_{i=1}^{L}(t_1a_i^\dagger b_i+t_2a_{i+1}^\dagger b_i+\mathrm{H.c.})\\ &\mapsto\hat{U}\hat{H}_0\hat{U}^{\dagger}=\sum_{i=1}^{L}(t_1b_{i+1}^\dagger a_{i+1}+t_2b_{i+1}^\dagger a_{i+1}+\mathrm{H.c.})\\ &=\sum_{i=1}^{L}(t_2b_i^\dagger a_i+t_1b_i^\dagger a_{i+1}+\mathrm{H.c.}) \tag*{(28)} \end{aligned}

with bL+1≡b1b_{L+1}\equiv b_1. Hence U^H^0(t1,t2)U^†=H^0(t2,t1)\hat{U}\hat{H}_0(t_1,t_2)\hat{U}^{\dagger}=\hat{H}_0(t_2,t_1), i.e., intra- and intercell hopping amplitudes are interchanged.

In momentum space,

H^0=∑k∈BZ(ak†,bk†)H0(k)(akbk),(29)\hat{H}_0=\sum_{k\in\mathrm{BZ}}(a_k^\dagger,b_k^\dagger)H_0(k)\begin{pmatrix}a_k\\b_k\end{pmatrix}, \tag*{(29)}

with

H0(k)=(0t1+t2e−ikt1+t2eik0),(30)H_0(k)=\begin{pmatrix}0&t_1+t_2e^{-ik}\\t_1+t_2e^{ik}&0\end{pmatrix}, \tag*{(30)}

the transformation acts as

H0(k)↦U(k)H0(k)U(k)†,U(k)=(01eik0),(31)H_0(k)\mapsto U(k)H_0(k)U(k)^\dagger,\qquad U(k)=\begin{pmatrix}0&1\\e^{ik}&0\end{pmatrix}, \tag*{(31)}

with the analogous result

U(k)H0(t1,t2;k)U(k)†=H0(t2,t1;k).(32)U(k)H_0(t_1,t_2;k)U(k)^\dagger=H_0(t_2,t_1;k). \tag*{(32)}

The effect of the unitary transformation on the disorder potential is

V^=∑i(ϵiAai†ai+ϵiBbi†bi)↦∑i(ϵiAbi†bi+ϵiBai+1†ai+1).(33)\hat{V}=\sum_i(\epsilon_{iA}a_i^\dagger a_i+\epsilon_{iB}b_i^\dagger b_i) \mapsto\sum_i(\epsilon_{iA}b_i^\dagger b_i+\epsilon_{iB}a_{i+1}^\dagger a_{i+1}). \tag*{(33)}

Since the parameters ϵiA\epsilon_{iA}, ϵiB\epsilon_{iB} for all ii are independent and identically distributed, the disorder ensemble is invariant. For the disorder-averaged Green’s function we thus have

G‾(t2,t1;k,ω)=U(k)G‾(t1,t2;k,ω)U†(k),(34)\overline{G}(t_2,t_1;k,\omega)=U(k)\overline{G}(t_1,t_2;k,\omega)U^\dagger(k), \tag*{(34)}

and therefore

Htop(t2,t1;k)=U(k)Htop(t1,t2;k)U†(k).(35)H_{\mathrm{top}}(t_2,t_1;k)=U(k)H_{\mathrm{top}}(t_1,t_2;k)U^\dagger(k). \tag*{(35)}

Chiral symmetry, Eq. (20), implies Htop(k)=dx(k)σx+dy(k)σyH_{\mathrm{top}}(k)=d_x(k)\sigma_x+d_y(k)\sigma_y:

Htop(k)=(0q∗(k)q(k)0)(36)H_{\mathrm{top}}(k)=\begin{pmatrix}0&q^*(k)\\q(k)&0\end{pmatrix} \tag*{(36)}

with q(k)=dx(k)+idy(k)q(k)=d_x(k)+id_y(k). With Eq. (35), this yields

q(t2,t1;k)=eikq∗(t1,t2;k).(37)q(t_2,t_1;k)=e^{ik}q^*(t_1,t_2;k). \tag*{(37)}

The winding number can be expressed in terms of q(k)q(k) as

ν(t1,t2)=12πi∫BZdk ∂kln⁡q(t1,t2;k).(38)\nu(t_1,t_2)=\frac{1}{2\pi i}\int_{\mathrm{BZ}}dk\,\partial_k\ln q(t_1,t_2;k). \tag*{(38)}

Since k↦eikk\mapsto e^{ik} has winding number +1+1 and since complex conjugation reverses the winding, Eq. (37) implies

ν(t2,t1)=1−ν(t1,t2).(39)\nu(t_2,t_1)=1-\nu(t_1,t_2). \tag*{(39)}

There is no analogous constraint for the turning number, since differentiating Eq. (35) with respect to kk produces an extra term due to the kk dependence of the unitary transformation. Hence, there need not be a symmetry relation for the τ\tau phase diagram with respect to exchanging t1t_1 with t2t_2.

Note that both the winding and the turning number are evaluated in a fixed real-space orbital representation of the SSH model. The unitary transformation considered here relates different points of the parameter space in this same representation, rather than serving as a passive change of orbital basis.

We also note that ν\nu and τ\tau are evaluated in the natural real-space orbital representation of the SSH model, with a fixed unit-cell convention and real hopping amplitudes. Under a momentum-dependent change of orbital basis, H~(k)=U(k)H(k)U†(k)\widetilde{H}(k)=U(k)H(k)U^\dagger(k), ordinary kk derivatives are not covariant. Introducing Ak=i(∂kU)U†A_k=i(\partial_kU)U^\dagger and the covariant derivative DkH~=∂kH~+i[Ak,H~]D_k\widetilde{H}=\partial_k\widetilde{H}+i[A_k,\widetilde{H}], one instead has DkH~=U(∂kH)U†D_k\widetilde{H}=U(\partial_kH)U^\dagger. Hence the singular condition reads as DkH=0D_kH=0, and therefore the location of an immersion-driven transition, is representation independent. In the natural orbital representation used throughout this work, the connection vanishes and DkH=∂kHD_kH=\partial_kH.

Concluding, Eq. (39) explains that if ν=0\nu=0 for t2<t1t_2<t_1, we must have ν=1\nu=1 (if defined) for t2>t1t_2>t_1 and vice versa, see Figure 1. Furthermore, if ν\nu were well-defined for t1=t2t_1=t_2, the relation would imply ν=1/2\nu=1/2, which is impossible since ν∈Z\nu\in\mathbb{Z}. Therefore, for any disorder strength WW, where the winding number is well-defined on both sides of the line, ν\nu must be ill-defined on the t1=t2t_1=t_2 line, and thus this line is a topological-transition line.

Perturbative approach to the small-t2t_2 regime

We consider the SSH model with independent binary disorder, i.e., with on-site energies ϵiA=ziAW/2\epsilon_{iA}=z_{iA}W/2 and ϵiB=ziBW/2\epsilon_{iB}=z_{iB}W/2, where ziA=±1z_{iA}=\pm1 and ziB=±1z_{iB}=\pm1 with equal probability. Here, we treat the intercell hopping t2t_2 perturbatively.

For t2=0t_2=0, the system decomposes into independent dimers. At zero frequency, the Green’s function of a dimer is

g=4zAzBW2−4t12(zBW/2t1t1zAW/2).(40)\boldsymbol{g}=\frac{4}{z_Az_BW^2-4t_1^2}\begin{pmatrix}z_BW/2&t_1\\t_1&z_AW/2\end{pmatrix}. \tag*{(40)}

Averaging over the four configurations zA,zB=±1z_A,z_B=\pm1 gives

g‾=a(W)σx,(41)\overline{\boldsymbol{g}}=a(W)\sigma_x, \tag*{(41)}

with

a≡a(W)=16t13W4−16t14.(42)a\equiv a(W)=\frac{16t_1^3}{W^4-16t_1^4}. \tag*{(42)}

To determine the correction of first order in t2t_2, we consider the intercell hopping term in reciprocal space

V(k)=t2(cos⁡k σx+sin⁡k σy).(43)V(k)=t_2(\cos k\,\sigma_x+\sin k\,\sigma_y). \tag*{(43)}

To first order in t2t_2, the resolvent expansion yields

G‾(k,0)=g‾+g‾V(k)g‾+O(t22),(44)\overline{G}(k,0)=\overline{\boldsymbol{g}}+\overline{\boldsymbol{g}}V(k)\overline{\boldsymbol{g}}+O(t_2^2), \tag*{(44)}

where the factorization of the disorder average in the second term follows from the statistical independence of the on-site energies in neighboring unit cells. Hence, with a=a(W)a=a(W) from Eq. (42),

G‾(k,0)=[a+a2t2cos⁡k]σx−a2t2sin⁡k σy+O(t22).(45)\overline{G}(k,0)=\left[a+a^2t_2\cos k\right]\sigma_x-a^2t_2\sin k\,\sigma_y+O(t_2^2). \tag*{(45)}

Using the identity (gxσx+gyσy)−1=(gxσx+gyσy)/(gx2+gy2)(g_x\sigma_x+g_y\sigma_y)^{-1}=(g_x\sigma_x+g_y\sigma_y)/(g_x^2+g_y^2), we find the topological Hamiltonian

Htop(k)=−G‾(k,0)−1=dx(k)σx+dy(k)σy(46)H_{\mathrm{top}}(k)=-\overline{G}(k,0)^{-1}=d_x(k)\sigma_x+d_y(k)\sigma_y \tag*{(46)}

with

dx(k)=−a+a2t2cos⁡ka2+2a3t2cos⁡k+a4t22+O(t22),(47)d_x(k)=-\frac{a+a^2t_2\cos k}{a^2+2a^3t_2\cos k+a^4t_2^2}+O(t_2^2), \tag*{(47)}

and

dy(k)=a2t2sin⁡ka2+2a3t2cos⁡k+a4t22+O(t22).(48)d_y(k)=\frac{a^2t_2\sin k}{a^2+2a^3t_2\cos k+a^4t_2^2}+O(t_2^2). \tag*{(48)}

For a regular closed planar curve, the turning number is the winding number of its tangent vector and can be written as

τ=12πi∫02πdk ∂k2d+(k)∂kd+(k).(49)\tau=\frac{1}{2\pi i}\int_0^{2\pi}dk\,\frac{\partial_k^2d_+(k)}{\partial_kd_+(k)}. \tag*{(49)}

where d+(k)=dx(k)+idy(k)d_{+}(k)=d_x(k)+id_y(k). With Eq. (47) and Eq. (48) we find the compact expression

d+(k)=−1a+a2t2eik=−1f(k)(50)d_{+}(k)=-\frac{1}{a+a^2t_2e^{ik}}=-\frac{1}{f(k)} \tag*{(50)}

with f(k)≡a+a2t2eikf(k)\equiv a+a^2t_2e^{ik} and suppressing corrections O(t22)\mathcal{O}(t_2^2). Now

∂kd+(k)=∂kf(k)f(k)2(51)\partial_k d_{+}(k)=\frac{\partial_k f(k)}{f(k)^2} \tag*{(51)}

and

∂k2d+(k)∂kd+(k)=∂kln⁡[∂kd+(k)]=∂kln⁡[∂kf(k)f(k)2]=∂k2f(k)∂kf(k)−2∂kf(k)f(k).(52)\frac{\partial_k^2d_{+}(k)}{\partial_kd_{+}(k)}=\partial_k\ln[\partial_kd_{+}(k)]=\partial_k\ln\left[\frac{\partial_kf(k)}{f(k)^2}\right] =\frac{\partial_k^2f(k)}{\partial_kf(k)}-2\frac{\partial_kf(k)}{f(k)}. \tag*{(52)}

Therefore,

τ=12πi∫02πdk∂k2f(k)∂kf(k)−212πi∫02πdk∂kf(k)f(k).(53)\tau=\frac{1}{2\pi i}\int_0^{2\pi}dk\frac{\partial_k^2f(k)}{\partial_kf(k)}-2\frac{1}{2\pi i}\int_0^{2\pi}dk\frac{\partial_kf(k)}{f(k)}. \tag*{(53)}

To evaluate the first integral, we compute

∂k2f(k)∂kf(k)=−a2t2eikia2t2eik=i,(54)\frac{\partial_k^2f(k)}{\partial_kf(k)}=\frac{-a^2t_2e^{ik}}{ia^2t_2e^{ik}}=i, \tag*{(54)}

and thus

12πi∫02πdk∂k2f(k)∂kf(k)=1.(55)\frac{1}{2\pi i}\int_0^{2\pi}dk\frac{\partial_k^2f(k)}{\partial_kf(k)}=1. \tag*{(55)}

The second integral is a winding number

ν′=12πi∫02πdk∂kf(k)f(k).(56)\nu'=\frac{1}{2\pi i}\int_0^{2\pi}dk\frac{\partial_kf(k)}{f(k)}. \tag*{(56)}

around the origin of the circle

k↦f(k)=a+a2t2eik.(57)k\mapsto f(k)=a+a^2t_2e^{ik}. \tag*{(57)}

With Eq. (50) we see that the winding number of the topological Hamiltonian ν=−ν′\nu=-\nu'.

Combining the two terms yields

τ=1+2ν.(58)\tau=1+2\nu. \tag*{(58)}

This circle has center aa and radius ∣a2t2∣\lvert a^2t_2\rvert. Therefore,

ν=0if ∣a(W)t2∣<1,ν=−1if ∣a(W)t2∣>1,(59)\begin{aligned} \nu&=0 &&\text{if } \lvert a(W)t_2\rvert<1,\\ \nu&=-1 &&\text{if } \lvert a(W)t_2\rvert>1, \tag*{(59)} \end{aligned}

and consequently

τ=+1if ∣a(W)t2∣<1,τ=−1if ∣a(W)t2∣>1.(60)\begin{aligned} \tau&=+1 &&\text{if } \lvert a(W)t_2\rvert<1,\\ \tau&=-1 &&\text{if } \lvert a(W)t_2\rvert>1. \tag*{(60)} \end{aligned}

For ∣a(W)t2∣=1\lvert a(W)t_2\rvert=1, the circle passes through the origin at kc=0k_c=0 or kc=πk_c=\pi. This means that G‾(kc,0)\overline{G}(k_c,0) vanishes and that Htop(kc)H_{\mathrm{top}}(k_c) diverges, so that the turning number is not defined.

For t1,t2>0t_1,t_2>0 and t2<t1t_2<t_1, the two solutions of ∣a(W)t2∣=1\lvert a(W)t_2\rvert=1 [see Eq. (42)] are

W±=2t1(1±t2t1)1/4.(61)W_{\pm}=2t_1\left(1\pm\frac{t_2}{t_1}\right)^{1/4}. \tag*{(61)}

At W=W−W=W_- one has at2=−1at_2=-1 and the Green-function zero occurs at k=0k=0, whereas at W=W+W=W_+ one has at2=+1at_2=+1 and it occurs at k=πk=\pi. Thus,

ν=0,τ=+1if W<W−,ν=−1,τ=−1if W−<W<W+,ν=0,τ=+1if W+<W.(62)\begin{aligned} \nu&=0,\quad\tau=+1 &&\text{if }W<W_-,\\ \nu&=-1,\quad\tau=-1 &&\text{if }W_-<W<W_+,\\ \nu&=0,\quad\tau=+1 &&\text{if }W_+<W. \tag*{(62)} \end{aligned}

A change of turning number (and of the winding number) is predicted only within an interval of width O(t2)\mathcal{O}(t_2) around W=2t1W=2t_1. This narrow interval requires caution. At W=2t1W=2t_1, the isolated-dimer Green function is singular and a(W)a(W) diverges. The perturbative expansion in t2t_2 is therefore nonuniform precisely in the region where ∣a(W)t2∣\lvert a(W)t_2\rvert becomes of order unity. Accordingly, the values W±W_{\pm} obtained from the truncated Green function cannot be regarded as quantitatively controlled phase boundaries. Nevertheless, the calculation explicitly illustrates how a change of turning number can be associated with a zero of the disorder-averaged Green function and hence with a divergence of the topological Hamiltonian.

Importantly, however, for fixed W≠2t1W\ne2t_1 and sufficiently small t2>0t_2>0, one obtains ν=0\nu=0 and τ=+1\tau=+1. This fixes the value of τ\tau close to the t2=0t_2=0 axis in the phase diagram Figure 1. For W→0W\to0, the result is consistent with the clean-limit result for t2<t1t_2<t_1. This is expected, since there is no gap closure allowing for a change of ν\nu. Regarding the turning number, it suggests that there is also no loss of immersion allowing for a change of τ\tau.

Perturbative approach for strong disorder

We consider the strong-disorder limit W≫t1,t2W\gg t_1,t_2. Let TT with elements Trr′T_{rr'} denote the hopping matrix of the clean SSH model, where r=(i,α)r=(i,\alpha) refers to the site α=A,B\alpha=A,B in the unit cell i=1,…,Li=1,\ldots,L. Furthermore, VV with elements Vrr′=δrr′ϵrV_{rr'}=\delta_{rr'}\epsilon_r is the disorder term, where ϵr=zrW/2\epsilon_r=z_rW/2 and zr=±1z_r=\pm1 with equal probabilities. The real-space representation of the Green’s function then reads

G(ω)=1ω−T−V.(63)G(\omega)=\frac{1}{\omega-T-V}. \tag*{(63)}

Defining the (negative) on-site Green’s function at zero frequency as grr′=−δrr′1ϵr=−δrr′2zrWg_{rr'}=-\delta_{rr'}\frac{1}{\epsilon_r}=-\delta_{rr'}\frac{2z_r}{W}, we can expand the Green’s function as

G(0)=(g−1−T)−1=g+gTg+gTgTg+gTgTgTg+⋯ .(64)\begin{aligned} G(0)&=(g^{-1}-T)^{-1}\\ &=g+gTg+gTgTg+gTgTgTg+\cdots. \tag*{(64)} \end{aligned}

Upon disorder averaging,

gr‾=0,gr2‾=4W2.(65)\overline{g_r}=0,\qquad\overline{g_r^2}=\frac{4}{W^2}. \tag*{(65)}

Furthermore,

grgr′‾=0(r≠r′),(66)\overline{g_r g_{r'}}=0 \qquad(r\ne r'), \tag*{(66)}

for local and uncorrelated disorder. The second term in the expansion Eq. (64) vanishes, since TT has no diagonal entries. All terms with an odd power in g\mathbf{g} must vanish as well. The leading nonvanishing contribution connecting two neighboring sites is thus of third order in the hopping and corresponds to traversing the same bond three times,

grTrr′gr′Tr′rgrTrr′gr′r=16W4Trr′3(67)g_r T_{rr'} g_{r'} T_{r'r} g_r T_{rr'} g_{r'r}=\frac{16}{W^4}T_{rr'}^3 \tag*{(67)}

with Trr′=Tr′rT_{rr'}=T_{r'r}. After Fourier transformation, we thus have to leading order in 1/W1/W

G‾(k,0)=16W4[(t13+t23cos⁡k)σx+t23sin⁡k σy]+O(1W6).(68)\overline{G}(k,0)=\frac{16}{W^4}\left[(t_1^3+t_2^3\cos k)\sigma_x+t_2^3\sin k\,\sigma_y\right]+\mathcal{O}\left(\frac{1}{W^6}\right). \tag*{(68)}

Exploiting the identity (gxσx+gyσy)−1=(gxσx+gyσy)/(gx2+gy2)(g_x\sigma_x+g_y\sigma_y)^{-1}=(g_x\sigma_x+g_y\sigma_y)/(g_x^2+g_y^2), we get the topological Hamiltonian

Htop(k)=−G‾(k,0)−1=dx(k)σx+dy(k)σy,(69)H_{\mathrm{top}}(k)=-\overline{G}(k,0)^{-1}=d_x(k)\sigma_x+d_y(k)\sigma_y, \tag*{(69)}

where

dx(k)=−W416t13+t23cos⁡kt16+t26+2t13t23cos⁡k,d_x(k)=-\frac{W^4}{16}\frac{t_1^3+t_2^3\cos k}{t_1^6+t_2^6+2t_1^3t_2^3\cos k},
dy(k)=−W416t23sin⁡kt16+t26+2t13t23cos⁡k.(70)d_y(k)=-\frac{W^4}{16}\frac{t_2^3\sin k}{t_1^6+t_2^6+2t_1^3t_2^3\cos k}. \tag*{(70)}

We note that apart from the overall factor W4/16W^4/16 the Bloch vector d(k)\mathbf{d}(k) becomes WW independent as W→∞W\to\infty. Introducing

d+(k)=dx(k)+idy(k),(71)d_+(k)=d_x(k)+i d_y(k), \tag*{(71)}

we find the simple expression

d+(k)=−W4161t13+t23e−ik=−W4161f(k)(72)d_+(k)=-\frac{W^4}{16}\frac{1}{t_1^3+t_2^3e^{-ik}}=-\frac{W^4}{16}\frac{1}{f(k)} \tag*{(72)}

with f(k)=t13+t23e−ikf(k)=t_1^3+t_2^3e^{-ik}. The winding number is obtained via

ν=−12πi∫02πdk ∂kf(k)f(k)=12πi∮∣z∣=1dzz+t13/t23,(73)\nu=-\frac{1}{2\pi i}\int_0^{2\pi}dk\,\frac{\partial_k f(k)}{f(k)} =\frac{1}{2\pi i}\oint_{|z|=1}\frac{dz}{z+t_1^3/t_2^3}, \tag*{(73)}

with z=e−ikz=e^{-ik} and where the contour is mathematically positively oriented. With the residue theorem, we get ν=+1\nu=+1 if the pole z0=−t13/t23z_0=-t_1^3/t_2^3 lies inside the circle and ν=0\nu=0 else:

ν=+1if t2>t1,\nu=+1\qquad\text{if }t_2>t_1,
ν=0if t2<t1,(74)\nu=0\qquad\text{if }t_2<t_1, \tag*{(74)}

where t1,t2>0t_1,t_2>0. This is the same result as in the clean limit.

The turning number can be calculated via

τ=12πi∫02πdk ∂k2f(k)∂kf(k)−22πi∫02πdk ∂kf(k)f(k),(75)\tau=\frac{1}{2\pi i}\int_0^{2\pi}dk\,\frac{\partial_k^2 f(k)}{\partial_k f(k)}-\frac{2}{2\pi i}\int_0^{2\pi}dk\,\frac{\partial_k f(k)}{f(k)}, \tag*{(75)}

see Eq. (53). We have

12πi∫02πdk ∂k2f(k)∂kf(k)=12πi∫02πdk (−i)=−1.(76)\frac{1}{2\pi i}\int_0^{2\pi}dk\,\frac{\partial_k^2 f(k)}{\partial_k f(k)} =\frac{1}{2\pi i}\int_0^{2\pi}dk\,(-i)=-1. \tag*{(76)}

Hence, τ=−1+2ν\tau=-1+2\nu, and thus

τ=+1if t2>t1,\tau=+1\qquad\text{if }t_2>t_1,
τ=−1if t2<t1(77)\tau=-1\qquad\text{if }t_2<t_1 \tag*{(77)}

(for t1,t2>0t_1,t_2>0) in the strong-disorder limit.

For t2≪t1t_2\ll t_1, this result appears at first sight to contradict the expansion in t2t_2 discussed above. In that expansion, the topological-Hamiltonian curve is described by

d+(k)=−1a+a2t2eik,(78)d_+(k)=-\frac{1}{a+a^2t_2e^{ik}}, \tag*{(78)}

where

a=a(W)=16t13W4−16t14.(79)a=a(W)=\frac{16t_1^3}{W^4-16t_1^4}. \tag*{(79)}

Expanding this expression to first order in t2t_2 gives

d+(k)=−1a+t2eik+O(t22).(80)d_+(k)=-\frac{1}{a}+t_2e^{ik}+\mathcal{O}(t_2^2). \tag*{(80)}

Thus, the leading kk-dependent contribution has positive orientation and yields τ=+1\tau=+1. With W≫t1W\gg t_1 and Eq. (79), we have

d+(k)≃−W416t13+t2eik.(81)d_+(k)\simeq-\frac{W^4}{16t_1^3}+t_2e^{ik}. \tag*{(81)}

On the other hand, the strong-disorder expansion gives

d+(k)=−W4161t13+t23e−ik+⋯ .(82)d_+(k)=-\frac{W^4}{16}\frac{1}{t_1^3+t_2^3e^{-ik}}+\cdots. \tag*{(82)}

For t2<t1t_2<t_1, expansion in t2/t1t_2/t_1 yields

d+(k)=−W416t13+W4t2316t16e−ik+O(W4t26t19).(83)d_+(k)=-\frac{W^4}{16t_1^3}+\frac{W^4t_2^3}{16t_1^6}e^{-ik}+\mathcal{O}\left(\frac{W^4t_2^6}{t_1^9}\right). \tag*{(83)}

The leading kk-dependent contribution therefore has negative orientation and yields τ=−1\tau=-1. Although this term is of third order in t2t_2, its coefficient grows as W4W^4, whereas the term linear in t2t_2 obtained from the small-t2t_2 expansion is independent of WW. Retaining the leading kk-dependent contribution from both expansions, we may therefore write asymptotically

d+(k)≃d0+Beik+Ce−ik,(84)d_+(k)\simeq d_0+Be^{ik}+Ce^{-ik}, \tag*{(84)}

with

d0≃−W416t13,B=t2,C=Wt2416t16.(85)d_0 \simeq-\frac{W^4}{16t_1^3}, \qquad B=t_2, \qquad C=\frac{Wt_2^4}{16t_1^6}. \tag*{(85)}

The tangent vector is then

∂kd+(k)≃iBeik−iCe−ik.(86)\partial_k d_+(k) \simeq iB e^{ik}-iC e^{-ik}. \tag*{(86)}

For B>CB>C, the positively oriented contribution dominates and the tangent winds once counterclockwise, i.e., τ=+1\tau=+1. For C>BC>B, the negatively oriented contribution dominates, i.e., τ=−1\tau=-1. The change of turning number occurs when B=CB=C, namely at the critical intercell hopping

t2,c≃4t13W2,W≫t1.(87)t_{2,c}\simeq\frac{4t_1^3}{W^2}, \qquad W\gg t_1. \tag*{(87)}

Writing the condition in the form

t2,ct1≃4(t1W)2≪1,(88)\frac{t_{2,c}}{t_1}\simeq4\left(\frac{t_1}{W}\right)^2\ll1, \tag*{(88)}

shows that the critical line indeed lies within the small-t2t_2 regime for sufficiently strong disorder.

Right at the transition, B=CB=C, we have

∂kd+(k)=iB(eik−e−ik)=−2Bsin⁡k.(89)\partial_k d_+(k)=iB\left(e^{ik}-e^{-ik}\right)=-2B\sin k. \tag*{(89)}

which vanishes at k=0k=0 and k=πk=\pi. The topological-Hamiltonian curve thus ceases to be an immersion at the transition. In contrast, d+(k)d_+(k) itself remains finite in the regime t2,c≪t1t_{2,c}\ll t_1, since the large constant term d0d_0 dominates over both oscillating contributions. Within this asymptotic analysis, the change from τ=+1\tau=+1 to τ=−1\tau=-1 is therefore associated with a vanishing tangent rather than with a divergence of the topological Hamiltonian.

Consequences of time-reversal symmetry

For a given disorder realization V^\hat{V}, the single-particle Hamiltonian H^=H^0+V^\hat{H}=\hat{H}_0+\hat{V} is time-reversal invariant, [T^,H^]=0[\hat{T},\hat{H}]=0, where, for spinless particles, T^\hat{T} is the antiunitary representation given by complex conjugation in the real-space basis. Hence, for the clean-limit Bloch Hamiltonian in reciprocal space, time-reversal invariance enforces

H0(k)∗=H0(−k).(90)H_0(k)^*=H_0(-k). \tag*{(90)}

Since the disorder probability distribution is invariant under time reversal as well, we find for the disorder-averaged Green’s function

G(k,ω)∗=G(−k,ω∗).(91)G(k,\omega)^*=G(-k,\omega^*). \tag*{(91)}

For the topological Hamiltonian Htop(k)=−G−1(k,0)H_{\mathrm{top}}(k)=-G^{-1}(k,0), this implies

Htop(k)∗=Htop(−k),(92)H_{\mathrm{top}}(k)^*=H_{\mathrm{top}}(-k), \tag*{(92)}

as long as the Green’s function is nonsingular and has no spectral weight at ω=0\omega=0.

The topological Hamiltonian can be decomposed as

Htop(k)=dx(k)σx+dy(k)σy,(93)H_{\mathrm{top}}(k)=d_x(k)\sigma_x+d_y(k)\sigma_y, \tag*{(93)}

where the kk-dependent coefficients dx(k)d_x(k) and dy(k)d_y(k) are real. Note that chiral symmetry (see Sec. D) enforces d0(k)≡0d_0(k)\equiv0 and dz(k)≡0d_z(k)\equiv0. Using σx∗=σx\sigma_x^*=\sigma_x and σy∗=−σy\sigma_y^*=-\sigma_y, time-reversal symmetry, i.e., Eq. (92), gives us

dx(k)=dx(−k),d_x(k)=d_x(-k),
dy(k)=−dy(−k).(94)d_y(k)=-d_y(-k). \tag*{(94)}

The closed curve d(k)=(dx(k),dy(k))\boldsymbol{d}(k)=(d_x(k),d_y(k)) is therefore symmetric under reflection at the dxd_x axis. In particular,

dy(0)=dy(π)=0.(95)d_y(0)=d_y(\pi)=0. \tag*{(95)}

Let v(k)=∂kd(k)\boldsymbol{v}(k)=\partial_k\boldsymbol{d}(k) denote the tangent at kk. Differentiation of Eq. (94) with respect to momentum yields

vx(k)=−vx(−k),v_x(k)=-v_x(-k),
vy(k)=vy(−k).(96)v_y(k)=v_y(-k). \tag*{(96)}

Hence,

vx(0)=vx(π)=0.(97)v_x(0)=v_x(\pi)=0. \tag*{(97)}

The turning number of the planar curve is the winding number of its tangent vector,

τ=12π∮BZdk ∂karg⁡[dx(k)+idy(k)].(98)\tau=\frac{1}{2\pi}\oint_{\mathrm{BZ}} \mathrm{d}k\,\partial_k\arg\left[d_x(k)+i d_y(k)\right]. \tag*{(98)}

It is invariant under regular homotopies and can change only if the curve ceases to be an immersion, i.e., if for some critical momentum and disorder strength

∂kd(kc,Wc)=0.(99)\partial_k\boldsymbol{d}(k_c,W_c)=0. \tag*{(99)}

The consequences of time-reversal symmetry depend qualitatively on whether kck_c is a time-reversal-invariant momentum.

We first consider a generic critical momentum kc≠0,πk_c\ne0,\pi. Equation (96) implies that, if v(kc,Wc)=0\boldsymbol{v}(k_c,W_c)=0 then necessarily v(−kc,Wc)=0\boldsymbol{v}(-k_c,W_c)=0. A generic immersion singularity at kc≠0,πk_c\ne0,\pi is therefore accompanied by a second singularity at the mirror-related momentum −kc-k_c, and both occur at the same disorder strength WcW_c. One can show that both singularities contribute equally to the change of the turning number. To this end we define the complex tangent as

u(k,W)=vx(k,W)+ivy(k,W).(100)u(k,W)=v_x(k,W)+iv_y(k,W). \tag*{(100)}

Using Eqs. (96), we get

u(k,W)∗=vx(k,W)−ivy(k,W)=−vx(−k,W)−ivy(−k,W)=−u(−k,W).(101)\begin{aligned} u(k,W)^*&=v_x(k,W)-iv_y(k,W)\\ &=-v_x(-k,W)-iv_y(-k,W)=-u(-k,W). \tag*{(101)} \end{aligned}

Note that u(k,W)∗u(k,W)^* is the mirror of the tangent u(k,W)u(k,W) and that, following Eq. (101), complex conjugation reverses the orientation in the complex tangent plane, while the transformation k→−kk \to-k reverses the orientation in momentum space. The two orientation reversals cancel, explaining why the two mirror-related singularities carry the same local index. Consequently, we must have Δτ=±2\Delta\tau= \pm2. Thus, time-reversal symmetry naturally permits a direct transition between τ=+1\tau= +1 and τ=−1\tau= -1 at a single critical disorder strength, without an intervening phase with τ=0\tau= 0.

The situation is different when the immersion is lost at a time-reversal-invariant momentum, kc=0k_c = 0 or kc=πk_c = \pi. Such a momentum is its own mirror partner, so that no second singularity is generated by symmetry. Moreover, ∂kdx(kc)=0\partial_k d_x(k_c) = 0 is already enforced by time-reversal symmetry, see Eq. (97), and the loss of immersion requires

∂kdy(kc,Wc)=0.(102)\partial_k d_y(k_c,W_c) = 0 . \tag*{(102)}

A generic singularity at kc=0k_c = 0 or π\pi therefore changes the turning number by one unit,

Δτ=±1.(103)\Delta\tau= \pm1 . \tag*{(103)}

Consequently, a transition from τ=+1\tau= +1 to τ=−1\tau= -1 mediated exclusively by singularities at time-reversal-invariant momenta generically requires two distinct critical disorder strengths,

τ=+1→Wc1τ=0→Wc2τ=−1.(104)\tau= +1 \xrightarrow{W_{c1}} \tau= 0 \xrightarrow{W_{c2}} \tau= -1 . \tag*{(104)}

Thus, while both types of immersion singularities are locally generic, time-reversal symmetry distinguishes two different scenarios for a total change Δτ=±2\Delta\tau= \pm2: either two separate transitions with an intermediate τ=0\tau= 0 phase when the singularities occur at k=0k = 0 or π\pi, or a single direct transition generated by a simultaneous mirror-related pair at ±kc\pm k_c with kc≠0,πk_c \ne0,\pi.

References

  1. [1]L. D. Landau, On the theory of phase transitions. I., Physik. Zeits. Sowjetunion 11, 26 (1937).
  2. [2]L. D. Landau, On the theory of phase transitions. II., Physik. Zeits. Sowjetunion 11, 545 (1937).
  3. [3]X.-G. Wen, Quantum Field Theory of Many-Body Systems: From the Origin of Sound to an Origin of Light and Electrons (Oxford University Press, 2007).DOI
  4. [4]A. P. Schnyder, S. Ryu, A. Furusaki, and A. W. Ludwig, Classification of topological insulators and superconductors in three spatial dimensions, Phys. Rev. B 78, 195125 (2008).
  5. [5]A. Kitaev, AIP Conf. Proc. 1134, 22 (2009).
  6. [6]S. Ryu, A. P. Schnyder, A. Furusaki, and A. W. Ludwig, Topological insulators and superconductors: tenfold way and dimensional hierarchy, New J. Phys. 12, 065010 (2010).DOI
  7. [7]R. Kennedy and C. Guggenheim, Homotopy theory of strong and weak topological insulators, Phys. Rev. B 91, 245148 (2015).DOI
  8. [8]H. C. Po, H. Watanabe, and A. Vishwanath, Fragile topology and wannier obstructions, Phys. Rev. Lett. 121, 126402 (2018).DOI
  9. [9]B. Jiang, A. Bouhon, Z.-K. Lin, X. Zhou, B. Hou, F. Li, R.-J. Slager, and J.-H. Jiang, Experimental observation of non-abelian topological acoustic semimetals and their phase transitions, Nat. Physics 17, 1239 (2021).DOI
  10. [10]B. Bradlyn, L. Elcoro, J. Cano, M. G. Vergniory, Z. Wang, C. Felser, M. I. Aroyo, and B. A. Bernevig, Topological quantum chemistry, Nature (London) 547, 298 (2017).
  11. [11]S. Rachel, Interacting topological insulators: a review, Rep. Prog. Phys. 81, 116501 (2018).DOI
  12. [12]T. Senthil, Symmetry-protected topological phases of quantum matter, Ann. Rev. Condens. Matter Phys. 6, 299 (2015).DOI
  13. [13]D. Vaish and M. Potthoff, Disorder-induced topological phases in a two-dimensional Chern insulator with strong magnetic disorder, Phys. Rev. Res. 8, 023312 (2026).
  14. [14]K. Ishikawa and T. Matsuyama, Magnetic field induced multi-component QED3 and quantum Hall effect, Z. Phys. C 33, 41 (1986).DOI
  15. [15]G. E. Volovik, The Universe in a Helium Droplet (Oxford University Press, New York, 2003).
  16. [16]V. Gurarie, Single-particle Green’s functions and interacting topological insulators, Phys. Rev. B 83, 085426 (2011).DOI
  17. [17]Z. Wang and S.-C. Zhang, Simplified topological invariants for interacting insulators, Phys. Rev. X 2, 031008 (2012).DOI
  18. [18]Z. Wang and B. Yan, Topological Hamiltonian as an exact tool for topological invariants, J. Phys.: Condens. Matter 25, 155601 (2013).DOI
  19. [19]L. Fidkowski and A. Kitaev, Effects of interactions on the topological classification of free fermion systems, Phys. Rev. B 81, 134509 (2010).
  20. [20]Y.-Y. He, H.-Q. Wu, Z. Y. Meng, and Z.-Y. Lu, Topological invariants for interacting topological insulators. ii. breakdown of single-particle green’s function formalism, Phys. Rev. B 93, 195164 (2016).DOI
  21. [21]J. Zhao, P. Mai, B. Bradlyn, and P. Phillips, Failure of topological invariants in strongly correlated matter, Phys. Rev. Lett. 131, 106601 (2023).
  22. [22]S. R. Manmana, A. M. Essin, R. M. Noack, and V. Gurarie, Topological and interacting one-dimensional fermionic systems, Phys. Rev. B 86, 205119 (2012).DOI
  23. [23]H.-H. Hung, L. Wang, Z.-C. Gu, and G. A. Fiete, Topological phase transition in a generalized Kane-Mele-Hubbard model: A combined quantum Monte Carlo and Green’s function study, Phys. Rev. B 87, 121113(R) (2013).DOI
  24. [24]P. Kumar, T. Mertz, and W. Hofstetter, Interaction-induced topological and magnetic phases in the hofstadter-hubbard model, Phys. Rev. B 94, 115161 (2016).DOI
  25. [25]See Supplemental Material at URL.
  26. [26]W. P. Su, J. R. Schrieffer, and A. J. Heeger, Solitons in polyacetylene, Phys. Rev. Lett. 42, 1698 (1979).
  27. [27]N. Wagner, L. Crippa, A. Amaricci, P. Hansmann, M. Klett, E. J. König, T. Schäfer, D. D. Sante, J. Cano, A. J. Millis, A. Georges, and G. Sangiovanni, Mott insulators with boundary zeros, Nat. Commun. 14, 7531 (2023).
  28. [28]A. Blason and M. Fabrizio, Unified role of green’s function poles and zeros in correlated topological insulators, Phys. Rev. B 108, 125115 (2023).
  29. [29]N. Wagner, D. Guerci, A. J. Millis, and G. Sangiovanni, Edge zeros and boundary spinons in topological Mott insulators, Phys. Rev. Lett. 133, 126504 (2024).
  30. [30]S. Bollmann, C. Setty, U. F. P. Seifert, and E. J. König, Topological Green’s function zeros in an exactly solved model and beyond, Phys. Rev. Lett. 133, 136504 (2024).
  31. [31]E. Pangburn, C. Pépin, and A. Banerjee, Topological charge excitations and Green’s function zeros in paramagnetic mott insulators, Phys. Rev. B 112, 085105 (2025).DOI
  32. [32]R. Flores-Calderón and C. Hooley, Weyl-Mott point: Topological and non-Fermi liquid behavior from an isolated Green’s function zero, Phys. Rev. B 111, 235139 (2025).
  33. [33]T. Morimoto and N. Nagaosa, Weyl Mott insulator, Sci. Rep. 6, 19853 (2016).DOI
  34. [34]E. W. Huang, G. L. Nave, and P. W. Phillips, Discrete symmetry breaking defines the Mott quartic fixed point, Nat. Physics 18, 511 (2022).
  35. [35]H. Whitney, Higher order closed curves in the plane, Compos. Math. 4, 276 (1937).
  36. [36]N. Bovenzi, M. Breitkreiz, T. E. O’Brien, J. Tworzydlo, and C. W. J. Beenakker, Twisted Fermi surface of a thin-film weyl semimetal, New J. Phys. 20, 023023 (2018).DOI
  37. [37]S. Spurrier and N. R. Cooper, Theory of quantum oscillations in quasicrystals: Quantizing spiral Fermi surfaces, Phys. Rev. B 100, 081405(R) (2019).
  38. [38]C. Wang and H. Huang, Decomposing electronic structures in twisted multilayers: Bridging spectra and incommensurate wave functions, Phys. Rev. B 111, 195161 (2025).
  39. [39]B. Shi, Immersed figure-8 annuli and anyons, Annales Henri Poincaré 27, 51 (2026).
  40. [40]R. A. Horn and C. R. Johnson, Matrix Analysis, 2nd ed. (Cambridge University Press, Cambridge, 2013).

Paper details

vlnc:W_b8a2yff8
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