--- title: Spin Bott Index in Quantum Spin Hall Systems url: https://www.emergentmind.com/topics/spin-bott-index type: topic --- # Spin Bott Index in Quantum Spin Hall Systems The spin Bott index is a real-space topological invariant used primarily to identify time-reversal-invariant quantum spin Hall phases in finite, disordered, amorphous, quasicrystalline, and other non-periodic systems. It is constructed by decomposing the occupied subspace into two effective spin sectors and taking half the difference of their Bott indices, thereby replacing Brillouin-zone-based \(\mathbb Z_2\) diagnostics when translational symmetry is absent; in a distinct literature on real Bott towers, the same phrase is not standard, but the explicit Bott-matrix formula for the spin obstruction \(w_2\) plays the closest analogous role [1810.00081, 1810.00070, 1609.05630]. ## 1. Conceptual scope and terminological usage In condensed-matter theory, the spin Bott index was introduced because the standard \(\mathbb Z_2\) invariant for quantum spin Hall insulators is ordinarily defined from Bloch bands, Berry-curvature constructions, hybrid Wannier charge centers, parity eigenvalues, or Pfaffians, all of which rely on translational symmetry and a Brillouin zone. In quasicrystals and other non-periodic systems, Bloch momentum \(k\) is not a good quantum number, Brillouin-zone formulas fail, and the relevant electronic states may be critical rather than Bloch-like. The spin Bott index addresses this by working directly in real space on finite samples [1810.00081, 1810.00070]. The construction generalizes the ordinary Bott index, which had already been used for Chern topology in nonperiodic systems. The ordinary Bott index measures the noncommutativity of projected position operators, while the spin Bott index introduces a spin-sector decomposition of the occupied subspace so that opposite spin sectors can carry opposite topological winding, as expected in a quantum spin Hall phase. In that sense it is a real-space analogue of spin Chern or \(\mathbb Z_2\) diagnostics rather than a net Chern marker [1810.00081]. A separate mathematical usage arises in the topology of real Bott towers and generalized real Bott manifolds. The relevant papers do not define a named invariant called “spin Bott index”; instead, they derive necessary-and-sufficient spin criteria from explicit formulas for the second Stiefel–Whitney class \(w_2\), expressed in terms of the Bott matrix or associated column vectors. A plausible implication is that the phrase can be used informally for these parity obstructions, but this is an interpretive rather than standard terminological usage [1609.05630, 2111.09585]. ## 2. Real-space construction for quantum spin Hall phases The starting point is the projector onto the occupied subspace, \[ P=\sum_i^{N_{\mathrm{occ}}} |\psi_i\rangle\langle\psi_i|. \] From it one defines the projected spin operator \[ P_z=P\hat s_z P,\qquad \hat s_z=\frac{\hbar}{2}\sigma_z. \] If \([H,\hat s_z]=0\), the spectrum of \(P_z\) consists of the two values \(\pm \hbar/2\). When spin is not strictly conserved, the eigenvalues spread, but the construction remains valid provided the spectrum still splits into two groups separated by zero; this separation is the spin gap of \(P_z\) [1810.00081, 1810.00070]. Let \[ P_z|\pm\phi_i\rangle=S_\pm|\pm\phi_i\rangle. \] The positive- and negative-eigenvalue subspaces define the effective spin-sector projectors \[ P_\pm=\sum_i^{N_{\mathrm{occ}}/2} |\pm\phi_i\rangle\langle\pm\phi_i|. \] For each sector one then forms projected position operators using rescaled coordinates \(X,Y\in[0,1)\), \[ U_\pm=P_\pm e^{i2\pi X}P_\pm+(I-P_\pm),\qquad V_\pm=P_\pm e^{i2\pi Y}P_\pm+(I-P_\pm). \] The complementary terms \(I-P_\pm\) are included to make the matrices closer to unitary and improve numerical stability [1810.00081]. Because the finite-sample projected operators are not exactly unitary, the standard implementation uses a singular value decomposition. For any projected operator \(M\), \[ M=Z\Sigma W^\dagger,\qquad \tilde M=ZW^\dagger. \] Applying this to \(U_\pm\) and \(V_\pm\) gives stabilized unitary matrices \(\tilde U_\pm\) and \(\tilde V_\pm\). The sector Bott indices are then \[ B_\pm=\frac{1}{2\pi}\operatorname{Im}\left\{\operatorname{tr}\left[\log\left(\tilde V_\pm \tilde U_\pm \tilde V_\pm^\dagger \tilde U_\pm^\dagger\right)\right]\right\}, \] and the spin Bott index is \[ B_s=\frac{1}{2}(B_+-B_-). \] In the condensed-matter usage, \(B_s\approx 1\) indicates a quantum spin Hall phase and \(B_s\approx 0\) a trivial insulator. The construction requires both a bulk energy gap and an open spin gap of \(P_z\); if spin mixing closes the spin gap, the decomposition into \(P_+\) and \(P_-\) ceases to be well defined [1810.00081, 1810.00070]. ## 3. Bott-index algebra, logarithmic formulation, and stability The spin Bott index inherits its basic algebraic structure from the ordinary Bott index. For a pair of unitary operators \(U,V\) satisfying the spectral condition \(\|[U,V]\|<2\), equivalently \(-1\notin \sigma(UVU^*V^*)\), the Bott index is defined by \[ \operatorname{Bott}(U,V)=\frac{1}{2\pi i}\operatorname{Tr}\log(UVU^*V^*), \] with \(\log\) taken as the principal logarithm. In the infinite-dimensional formulation, additional trace-class conditions such as \((U-1)(V-1)\in\mathcal I^1\) and \((V-1)(U-1)\in\mathcal I^1\) ensure that the logarithm is trace class and that the index is well defined [2112.01339]. The index is integer-valued. Writing the eigenvalues of the commutator unitary as \(e^{i\theta_j}\), \(\theta_j\in(-\pi,\pi)\), one obtains \[ 1=\prod_j e^{i\theta_j}=e^{\operatorname{Tr}\log(UVU^*V^*)}, \] hence \[ \frac{1}{2\pi i}\operatorname{Tr}\log(UVU^*V^*)\in\mathbb Z. \] For finite periodic lattices, the Bott index can be constructed from projected coordinate unitaries, and rigorous results identify it with the Chern number: \[ \textrm{Bott}\left( e^{2 \pi i P \frac{X}{L} P },e^{2 \pi i P \frac{Y}{L} P } \right)=\mathrm{Ch}(P). \] This relation explains why a spin-resolved difference of Bott indices can reproduce the quantum spin Hall classification in crystalline and disordered settings [1708.05912, 2112.01339]. Homotopy invariance is a central structural property. Under continuous unitary deformations that preserve the spectral admissibility condition, the Bott index remains constant. Time-dependent analyses of projected Bott invariants on large finite tori show that, for local bounded Hamiltonians that are initially gapped, the Bott index of the time-evolved Fermi projection is constant in the thermodynamic limit, and any change at finite size is a finite-size effect whose characteristic time diverges with system size. This suggests that spin-resolved Bott constructions inherit stability under local driving so long as the occupied projector remains well defined and the spin-gap condition persists [1708.05916]. ## 4. Diagnostic content, model systems, and observable signatures The spin Bott index was benchmarked on the Kane–Mele model, where it agrees with the standard \(\mathbb Z_2\) invariant from hybrid Wannier charge center flow. Reported representative values include \(B_s=0.9957\approx 1\) in a quantum spin Hall phase and \(B_s=2.1581\times 10^{-5}\approx 0\) in a trivial insulator. The same framework remains applicable when Rashba coupling breaks exact \(s_z\) conservation, provided the spin gap of \(P_z\) stays open [1810.00081]. In the disorder Kane–Mele model, \[ H_{\mathrm{disorder}}=H_{\mathrm{KM}}+W\sum_i \epsilon_i c_i^\dagger c_i,\qquad \epsilon_i\in[-1,1), \] the spin Bott index captures a disorder-driven transition consistent with the topological Anderson insulator scenario and previous conductance results. In the reported numerics, the transition occurs around \(W\sim 1.0t\). This establishes the invariant as a disorder-compatible real-space diagnostic rather than a purely crystalline one [1810.00081]. The Penrose-type quasicrystal lattice provides the canonical non-periodic example. There the absence of a periodic unit cell and Brillouin zone makes conventional band-topology tools unavailable, but direct real-space computation yields \(B_s=0.9974\approx 1\) in one study and \(B_s=1\) as the hallmark value in another. The corresponding phase is supported by a bulk gap under periodic boundary conditions, mid-gap boundary states under open boundary conditions, edge localization of the in-gap wavefunctions, and a quantized two-terminal conductance plateau \[ G=\frac{2e^2}{h}. \] The edge states are reported to be robust against boundary-shape changes, which is particularly significant for quasicrystalline geometry [1810.00081, 1810.00070]. A parameter-space phase diagram was constructed in terms of the on-site \(s\)-\(p\) offset \(\Delta=\epsilon_s-\epsilon_p\), the spin-orbit coupling strength \(\lambda\), and the bonding-length scale \(d_0\). The three identified regimes are normal insulator, quantum spin Hall insulator with \(B_s=1\), and weak metal. In the \(d_0\)-dependent analysis, the spin Bott index switches to \(1\) at a critical value around \[ d_0^c\approx 0.89, \] after which a later gap closing leads to the weak-metal regime. The reported interpretation is that increasing hybridization drives an \(s\)-\(p\) band inversion, spin-orbit coupling reopens the gap into a topological phase, and sufficiently strong hopping eventually overwhelms that gap [1810.00070]. ## 5. Higher-order superconductors and related real-space Bott developments The Bott-index framework has been extended from quantum spin Hall insulators to time-reversal-invariant higher-order topological superconductors hosting Majorana Kramers pairs. In that setting an ordinary chiral Bott index vanishes by time-reversal symmetry, because the relevant eigenphases occur in \((e^{i\lambda},e^{-i\lambda})\) pairs. The remedy is a spin-resolved Bott construction that separates the two Kramers-related sectors and thereby detects corner topology in real space under open boundary conditions and for arbitrary sample shapes [2407.20334]. For conserved \(s_z\), the paper introduces \[ \mathcal N=\frac{1}{4\pi i}\mathrm{Tr}\!\left[\mathcal C\log\!\left(MQM^\dagger Q\right)\right],\qquad \mathcal C=s_z C, \] and also an equivalent spin-twisted form \[ M_z=e^{2\pi i f(\mathbf r)s_z\tau_0},\qquad \bar{\mathcal N}=\frac{1}{4\pi i}\mathrm{Tr}\!\left[C\log\!\left(M_zQM_z^\dagger Q\right)\right]. \] When spin is conserved, the invariant decomposes as \[ \mathcal N=N_+-N_-. \] When \(s_z\) is broken, the formulation based on the spin-projected operator \[ P_z=Ps_zP,\qquad P=\frac{\mathbbm 1-Q}{2}, \] remains valid as long as the spin-projection gap of \(P_z\) stays open. The resulting series of spin Bott indices does not merely distinguish topological from trivial phases; it encodes which corners host Majorana Kramers pairs and tracks their redistribution under boundary cleavage [2407.20334]. The same real-space Bott philosophy has also been broadened in directions adjacent to, but distinct from, the spin Bott index. The Bott metric uses the same plaquette operator as the Bott index and extracts the real part rather than the imaginary part of the trace-log, \[ B=\frac{1}{2\pi}\Im\,\mathrm{Tr}\log(W),\qquad M=-\frac{1}{2\pi}\Re\,\mathrm{Tr}\log(W), \] so that topology is encoded in the phase of the loop while quantum geometry is encoded in its contraction. That framework is compatible with spinful Hilbert spaces because internal labels may include spin, although no separate spin Bott metric is derived there [2604.04447]. A further neighboring development is the extension of Bott-index methods to topological pumping in quasiperiodic systems. That work does not define a spin Bott index explicitly, but it states that the method can be used to study topological spin pumping in spinful electron systems or in topological magnets. This suggests a natural route from static spin Bott diagnostics to time-dependent spin-resolved transport settings [2105.05654]. ## 6. Spin obstruction in real and generalized Bott manifolds In the topology of real Bott towers, the phrase “spin Bott index” is best understood cautiously. The relevant invariant is not named that way in the literature; rather, the spin obstruction is exactly the second Stiefel–Whitney class \(w_2\), computed explicitly from the Bott matrix. A real Bott tower \(Y(C)\) is encoded by an upper triangular Bott matrix \[ C=(c_{i,j})\in M_n(\mathbb Z_2),\qquad c_{i,i}=1,\quad c_{i,j}=0\text{ for }i>j, \] and the cohomology ring is \[ H^*(Y_n;\mathbb Z_2)\cong \mathbb Z_2[y_1,\dots,y_n]/J,\qquad y_i^2=y_i\sum_{j=i+1}^n c_{i,j}y_j. \] The total Stiefel–Whitney class is \[ w(Y_n)=\prod_{i=1}^{2n}(1+x_i), \] and the orientability criterion is \[ Y_n \text{ is orientable } \iff \sum_{j=i+1}^n c_{i,j}=0 \pmod 2\quad \text{for every }1\le i\le n-1. \] Under this orientability hypothesis, Theorem 4.10 gives the spin condition: \[ \sum_{r=j+1}^{n}\sum_{s=k+1}^{n} c_{j,r}c_{k,s} + c_{j,k}\sum_{r,s=k+1\atop r