Photonic Topological Insulators Floquet Topological Insulator Chiral Edge States Ssh Model

# Photonic Topological Insulators, SSH Lattices, Floquet Phases, and Chiral Edge States

---

## Executive Summary

As semiconductor device dimensions scale into the sub-2 nm regime, quantum phase coherence and geometric gauge fields in momentum space become central to electron dynamics. The Berry phase γ and its associated field strength—the Berry curvature tensor Ω_n(k)—act as fictitious magnetic fields in crystal momentum space, radically altering carrier transport through anomalous velocity contributions. This article provides a comprehensive theoretical framework of geometric quantum phases in electronic band structures. We derive the Berry connection A_n(k), Berry curvature Ω_n(k), and topological Chern numbers C_n from adiabatic quantum mechanics. We formulate semiclassical wave-packet dynamics incorporating anomalous velocities, analyze the Quantum Anomalous Hall Effect (QAHE) and spin-orbit-coupled transport, and present a complete Python numerical solver mapping 2D Berry curvature and Chern numbers in topological 2D semiconductor channels (e.g., monolayer transition metal dichalcogenides and Dirac semimetals).

---

## Table of Contents

1. Introduction: Geometric Quantum Phases in Condensed Matter
2. Adiabatic Evolution and the Berry Phase Derivation
3. Berry Connection Vector Field and Gauge Invariance
4. Berry Curvature Tensor and Topological Chern Numbers
5. Semiclassical Wave-Packet Dynamics and Anomalous Velocity
6. Topological Insulators and Edge State Transport in Sub-2nm Devices
7. Valleytronics and Anomalous Hall Effects in 2D TMD Channels
8. Python Implementation: 2D Berry Curvature Mapping and Chern Integrator
9. Quantum Device Applications: Dissipationless Topo-FETs and Quantum Sensors
10. Experimental Metrology: Angle-Resolved Photoemission (ARPES) and Transport Spectroscopy
11. References & Further Reading

---

## 1. Introduction: Geometric Quantum Phases in Condensed Matter

Classical transport theory treats electronic wave-packets as quasi-particles governed strictly by energy dispersion E_n(k) via group velocity v_g = (1)/(ℏ) ∇_k E_n(k). However, quantum mechanics dictates that an electron state |u_n(k)⟩ is a complex vector in Hilbert space. As an electron moves through momentum space under applied electric or magnetic fields, the phase of its wavefunction evolves non-trivially.

In 1984, Sir Michael Berry demonstrated that an eigenstate undergoing cyclic adiabatic evolution acquires a geometric phase in addition to the standard dynamical phase. In crystalline solids, this Berry phase manifests as a geometric gauge field in Brillouin zone momentum space.

In sub-2 nm semiconductor nodes—where 2D channel materials (e.g., MoS_2, WSe_2, silicene, monolayer bismuthene) and spin-orbit-coupled heterostructures are integrated into Gate-All-Around (GAA) and 3D architectures—the Berry curvature plays a governing role in driving anomalous Hall currents, valley-polarized transport, and dissipationless edge conduction.

---

## 2. Adiabatic Evolution and the Berry Phase Derivation

Consider a quantum system governed by a Hamiltonian H(R) that depends on a set of continuous parameters R(t) = (R_1(t), R_2(t), …) changing adiabatically over time.

Let |n(R)⟩ be an instantaneous orthonormal eigenbasis of H(R):

According to the adiabatic theorem, a system initially in state |n(R(0))⟩ remains in the n-th instantaneous eigenstate |n(R(t))⟩, accumulating a complex phase factor:

where the first exponential is the dynamical phase, and γ_n(t) is the geometric phase.

Substituting |ψ(t)⟩ into the time-dependent Schrödinger equation iℏ (∂)/(∂ t)|ψ(t)⟩ = H(R(t))|ψ(t)⟩ and taking the inner product with ⟨ n(R(t))| yields:

Integrating over a closed path C in parameter space (such that R(T) = R(0)) gives the Berry Phase γ_n:

where A_n(R) is the Berry connection.

---

## 3. Berry Connection Vector Field and Gauge Invariance

### 3.1 Definition of the Berry Connection

In crystal momentum space (R → k), the Berry connection A_n(k) is defined as:

where |u_n(k)⟩ is the periodic Bloch function of the n-th energy band.

The Berry connection is real-valued because ∇_k ⟨ u_n | u_n ⟩ = 0 ⇒ ⟨ u_n | ∇_k u_n ⟩ + ⟨ ∇_k u_n | u_n ⟩ = 0.

### 3.2 Gauge Transformations

Under a local phase transformation |u_n(k)⟩ → e^i θ(k) |u_n(k)⟩, the Berry connection transforms as a vector potential:

While A_n(k) is gauge-dependent, the line integral over a closed loop ∮_C A_n(k) · dk (the Berry phase γ_n) is gauge-invariant modulo 2π.

---

## 4. Berry Curvature Tensor and Topological Chern Numbers

### 4.1 The Berry Curvature Tensor

Applying Stokes' Theorem to the closed loop integral transforms the line integral into a surface integral over a surface S bounded by C:

where Ω_n(k) is the Berry curvature vector field:

In component form, the antisymmetric Berry curvature tensor Ω_n, αβ(k) is:

Using perturbation theory, the Berry curvature can be expressed without explicit derivatives of Bloch functions:

This expression highlights that Berry curvature peaks sharply near band anti-crossings or Dirac points where energy gaps E_n - E_m become minimal.

### 4.2 Chern Numbers and Topological Invariants

For a closed 2D 2-torus (the 2D First Brillouin Zone BZ), the total integral of the Berry curvature over the entire BZ must be an integer multiple of 2π:

The integer C_n is the Chern number (or TKNN invariant, named after Thouless, Kohmoto, Nightingale, and den Nijs).

• Materials with C_n = 0 are topologically trivial (ordinary insulators/semiconductors).
• Materials with C_n ≠ 0 are topologically non-trivial (Chern insulators / Quantum Anomalous Hall states).

---

## 5. Semiclassical Wave-Packet Dynamics and Anomalous Velocity

### 5.1 Modified Equations of Motion

In the presence of electric field E and magnetic field B, the semiclassical equations of motion for an electron wave-packet centered at (r, k) in band n are modified by the Berry curvature:

The second term in the velocity equation, -dotk × Ω_n(k), is the anomalous velocity.

### 5.2 The Intrinsic Anomalous Hall Effect

Under a pure electric field E (with B = 0), ℏ dotk = -e E. The electron velocity becomes:

Integrating over all occupied states in the Brillouin zone yields an intrinsic transverse Hall current perpendicular to E:

The corresponding intrinsic Anomalous Hall Conductivity σ_xy in 2D is:

For a completely filled band (f_0 = 1), σ_xy = C_n (e²)/(h), demonstrating that transverse conductivity is quantized in units of e²/h.

---

## 6. Topological Insulators and Edge State Transport in Sub-2nm Devices

Topological Insulators (TIs) possess a bulk energy gap but exhibit gapless, dissipationless edge states protected by Time-Reversal Symmetry (TRS).

Under TRS:
- E_n(-k) = E_n(k)
- Ω_n(-k) = -Ω_n(k)

Therefore, the total Chern number integrated over the whole BZ for a TRS system is strictly zero (C = 0). However, spin-orbit coupling splits states into spin-up (↑) and spin-down (↓) channels with opposite Berry curvatures Ω_↑(k) = -Ω_↓(k).

This defines the Z_2 Spin Chern Number:

This gives rise to the Quantum Spin Hall Effect (QSHE), where spin-up and spin-down electrons travel in opposite directions along channel edges without backscattering, enabling ultra-low energy dissipation transistors (Topo-FETs).

---

## 7. Valleytronics and Anomalous Hall Effects in 2D TMD Channels

Monolayer Transition Metal Dichalcogenides (MoS_2, WS_2, WSe_2) possess a hexagonal lattice lacking inversion symmetry, producing two distinct, degenerate conduction band valleys at the K and K' points of the Brillouin Zone.

Due to broken inversion symmetry (I), the Berry curvature is non-zero at K and K':

Under an in-plane electric field E_x, carriers in the K valley experience an anomalous velocity towards +y, while carriers in the K' valley deflect towards -y. This spatial separation of valley index creates the Valley Hall Effect, forming the foundation of Valleytronic Logic Devices.

---

## 8. Python Implementation: 2D Berry Curvature Mapping and Chern Integrator

The following complete Python code computes the 2D Berry curvature tensor Ω_z(k_x, k_y) and the topological Chern number for a 2D Dirac-Haldane semiconductor model:

"""
Berry Curvature & Topological Chern Number Numerical Solver
Computes 2D momentum-space Berry curvature maps and integrates 
topological invariants for sub-2nm 2D semiconductor channels.
"""

import numpy as np
import matplotlib.pyplot as plt

# --- Model Parameters (2-Band Haldane/Qi-Wu-Zhang Model) ---
u_param = 1.0  # Mass parameter (u > 0: Chern = 1, u < 0: Chern = 0)
t_hop = 1.0    # Hopping amplitude

def hamiltonian_2d(kx, ky, u, t):
    """
    Constructs 2x2 Hamiltonian H(k) = d_x*sigma_x + d_y*sigma_y + d_z*sigma_z
    """
    dx = t * np.sin(kx)
    dy = t * np.sin(ky)
    dz = u - t * np.cos(kx) - t * np.cos(ky)
    
    # Pauli matrices
    H = dx * np.array([[0, 1], [1, 0]]) + \
        dy * np.array([[0, -1j], [1j, 0]]) + \
        dz * np.array([[1, 0], [0, -1]])
    return H, np.array([dx, dy, dz])

def compute_berry_curvature_grid(N_k=100, u=1.0):
    """
    Computes Berry curvature Omega_z(kx, ky) across the 2D Brillouin Zone
    using Fukui-Hatsugai-Suzuki discretized link-variable algorithm.
    """
    kx_vals = np.linspace(-np.pi, np.pi, N_k, endpoint=False)
    ky_vals = np.linspace(-np.pi, np.pi, N_k, endpoint=False)
    dk = 2 * np.pi / N_k
    
    # Compute ground-state eigenvectors |u(kx, ky)>
    psi = np.zeros((N_k, N_k, 2), dtype=complex)
    
    for i, kx in enumerate(kx_vals):
        for j, ky in enumerate(ky_vals):
            H, d_vec = hamiltonian_2d(kx, ky, u, t_hop)
            energies, vecs = np.linalg.eigh(H)
            # Ground state eigenvector (lowest energy)
            psi[i, j, :] = vecs[:, 0]
            
    # Compute Link Variables U_x and U_y
    # U_x(k) = <psi(k)|psi(k + dk_x)> / |<psi(k)|psi(k + dk_x)>|
    Ux = np.zeros((N_k, N_k), dtype=complex)
    Uy = np.zeros((N_k, N_k), dtype=complex)
    
    for i in range(N_k):
        for j in range(N_k):
            ip1 = (i + 1) % N_k
            jp1 = (j + 1) % N_k
            
            # Inner products
            dot_x = np.vdot(psi[i, j], psi[ip1, j])
            dot_y = np.vdot(psi[i, j], psi[i, jp1])
            
            Ux[i, j] = dot_x / np.abs(dot_x)
            Uy[i, j] = dot_y / np.abs(dot_y)
            
    # Field strength F_xy(k) = ln( Ux(k) Uy(k+dx) Ux(k+dy)^-1 Uy(k)^-1 )
    F_xy = np.zeros((N_k, N_k))
    for i in range(N_k):
        for j in range(N_k):
            ip1 = (i + 1) % N_k
            jp1 = (j + 1) % N_k
            
            plaquette = Ux[i, j] * Uy[ip1, j] * np.conj(Ux[i, jp1]) * np.conj(Uy[i, j])
            F_xy[i, j] = np.imag(np.log(plaquette))
            
    # Berry curvature Omega_z = F_xy / (dk_x * dk_y)
    Omega_z = F_xy / (dk**2)
    
    # Chern number is sum of F_xy divided by 2*pi
    chern_number = np.sum(F_xy) / (2 * np.pi)
    
    return kx_vals, ky_vals, Omega_z, chern_number

# --- Main Simulation Execution ---
print("=========================================================")
print("Berry Curvature Mapping & Chern Number Integrator")
print("=========================================================
")

N_grid = 120
u_topological = 1.0

kx_grid, ky_grid, Omega_z_map, C_num = compute_berry_curvature_grid(N_k=N_grid, u=u_topological)

print(f"Brillouin Zone Grid Resolution : {N_grid} x {N_grid}")
print(f"Topological Mass Parameter u   : {u_topological}")
print(f"Calculated Chern Number C      : {C_num:.6f}")
print(f"Quantized Integer Topological C: {int(np.round(C_num))}")

# Plotting 2D Berry Curvature Map
Kx, Ky = np.meshgrid(kx_grid, ky_grid)

fig, ax = plt.subplots(figsize=(9, 7))
contour = ax.contourf(Kx, Ky, Omega_z_map.T, 100, cmap='magma')
cbar = fig.colorbar(contour, ax=ax)
cbar.set_label(r'Berry Curvature $\Omega_z(\mathbf{k})$ ($\AA^2$)', fontsize=12)

ax.set_xlabel(r'$k_x (\pi/a)$', fontsize=13)
ax.set_ylabel(r'$k_y (\pi/a)$', fontsize=13)
ax.set_title(f'2D Momentum-Space Berry Curvature Distribution (Chern = {int(np.round(C_num))})', fontsize=14, pad=12)
ax.grid(True, linestyle=':', alpha=0.5)

plt.tight_layout()
plt.savefig("berry_curvature_2d_map.png", dpi=150)
print("\nVisualization saved: berry_curvature_2d_map.png")

---

## 9. Quantum Device Applications: Dissipationless Topo-FETs and Quantum Sensors

1. Topological Field-Effect Transistors (Topo-FETs): Electric fields shift the mass parameter u, driving a phase transition between a topological insulating state (C=1, edge conduction ON) and a trivial band insulator (C=0, OFF state). This voltage-controlled topological phase transition operates with sub-60 mV/dec subthreshold swing.
2. Dissipationless Interconnects: Chiral edge states in Quantum Anomalous Hall channels carry current without backscattering from non-magnetic impurities, eliminating Joule heating in ultra-dense IC interconnect networks.
3. Quantum Berry Phase Sensors: High-sensitivity magnetic and strain sensors measure Berry phase shifts induced by local lattice deformations or stray magnetic fields.

---

## 10. Experimental Metrology: Angle-Resolved Photoemission (ARPES) and Transport Spectroscopy

Characterization of Berry curvature and topological states relies on advanced metrology:
- Circular Dichroism ARPES (CD-ARPES): Uses circularly polarized light to photoemit electrons; the intensity difference between left- and right-circular polarization maps the Berry curvature distribution across momentum space.
- Quantum Hall & Anomalous Hall Transport Spectroscopy: Measures transverse Hall conductance σ_xy under varying temperature and gate bias to confirm quantized plateaus at e²/h.

---

## 11. References & Further Reading

1. Berry, M. V. (1984). "Quantal Phase Factors Accompanying Adiabatic Changes." *Proceedings of the Royal Society of London. Series A*, 392(1802), 45–57.
2. Xiao, D., Chang, M. C., & Niu, Q. (2010). "Berry phase effects on electronic properties." *Reviews of Modern Physics*, 82(3), 1959–2007.
3. Hasan, M. Z., & Kane, C. L. (2010). "Colloquium: Topological insulators." *Reviews of Modern Physics*, 82(4), 3045–3067.
4. Fukui, T., Hatsugai, Y., & Suzuki, H. (2005). "Chern numbers in discretized Brillouin zone: Efficient method of computing (spin) Hall conductances." *Journal of the Physical Society of Japan*, 74(6), 1674–1677.
5. Thouless, D. J., Kohmoto, M., Nightingale, M. P., & den Nijs, M. (1982). "Quantized Hall Conductance in a Two-Dimensional Periodic Potential." *Physical Review Letters*, 49(6), 405–408.

## Photonic implementation of the SSH lattice

The Su-Schrieffer-Heeger model is the smallest lattice that makes bulk-boundary correspondence visible. In a dimerized waveguide array, intracell coupling v and intercell coupling w alternate along propagation. With sublattice basis (A,B), the Bloch Hamiltonian is

The eigenvalues are

The gap closes at k = π when v = w. For w > v, the vector d(k) = (v+w cos k, w sin k) winds once around the origin as k traverses the Brillouin zone. For v > w it does not enclose the origin. This winding cannot change without closing the bulk gap as long as chiral symmetry is preserved.

The Zak phase of the lower band is

With a consistent unit cell and inversion-symmetric gauge, γ_Zak is π in the topological phase w > v and zero in the trivial phase v > w. The phase depends on unit-cell origin, while the physically measurable difference between two dimerizations and their interface state does not.

An open chain in the topological phase supports exponentially localized end modes. At zero energy, amplitudes on one sublattice obey A_(n+1) = -(v/w)A_n, so the localization length in lattice periods is

Finite chains hybridize their two end states and split them away from zero by an amount exponentially small in chain length. On-site disorder breaks chiral symmetry and can move an edge state away from midgap; coupling disorder that preserves the bipartite structure retains the spectral symmetry.

## Coupled-mode photonics

Paraxial propagation along z maps to a Schrödinger-like equation, i∂_z a = H a, so propagation distance acts as time. Evanescent coupling falls approximately exponentially with waveguide separation, making alternating v and w lithographically accessible. In ring-resonator arrays the evolution coordinate is physical time, and coupling phases can be engineered with link resonators or modulation.

Loss makes the Hamiltonian non-Hermitian. Uniform loss shifts all propagation constants without changing eigenvectors, but sublattice-dependent loss changes the topology and can produce exceptional points. Fabrication disorder modifies both on-site propagation constants and couplings. A topological claim should state which symmetry remains and which disorder ensemble was tested.

## Floquet photonic topological insulators

Periodic modulation creates an effective band structure even when the instantaneous lattice is topologically trivial. Over one drive period T, evolution is

Floquet eigenstates satisfy U(T)|u_n⟩ = exp(-iε_nT)|u_n⟩, where quasienergy ε is defined modulo 2π/T. Therefore both the zero and π/T quasienergy gaps can host protected boundary modes. An effective Hamiltonian H_F = (i/T)log U(T) depends on the logarithm branch; gap-specific winding invariants avoid losing anomalous Floquet topology.

Helically curved waveguides realize a synthetic gauge field. In the comoving frame, the curvature produces a periodic vector potential that gives hopping a time-dependent Peierls phase. In two dimensions the drive can open a gap at Dirac points and yield nonzero Chern bands with unidirectional edge transport.

A numerically stable Floquet calculation divides the period into short intervals and multiplies unitary step propagators in time order. Directly exponentiating the time-averaged Hamiltonian misses micromotion and fails when the drive is not in the high-frequency limit.

## Chiral edge dispersion and bulk-boundary correspondence

For a two-dimensional strip that remains periodic along x and finite along y, diagonalize H(k_x). Bulk bands fill broad regions while states localized at opposite boundaries cross the gap. The group velocity v_edge = dω/dk_x sets propagation direction. A nonzero net number of chiral crossings equals the sum of Chern numbers below the gap.

A single chiral channel cannot elastically backscatter from static disorder because no counterpropagating state exists at the same frequency on the same boundary. In scattering language, the reflection block is empty for an ideal one-channel edge. Disorder can shift phase, delay the pulse, or scatter into bulk states if it closes the mobility gap. It can also couple opposite edges in a narrow sample. Protection is therefore against a defined class of perturbations, not against absorption, nonlinearity, radiation loss, or arbitrary symmetry breaking.

For reciprocal time-reversal-invariant photonic systems, counterpropagating pseudospin channels may exist. Their protection depends on a crystalline or synthetic symmetry and is weaker than a truly chiral Chern edge. Calling every robust-looking boundary mode chiral obscures this distinction.

## Synthetic frequency dimensions

Equally spaced resonator modes can act as lattice sites indexed by frequency. Electro-optic modulation couples adjacent modes, and modulation phase supplies a synthetic hopping phase. Multiple modulation tones create long-range links or higher-dimensional synthetic lattices without additional physical area.

Input-output coupling makes frequency-space boundaries and observables different from real-space waveguides. Dispersion, finite comb bandwidth, and modulation harmonics limit translational invariance. Measure both converted power and phase to reconstruct synthetic-band transport.

## Python SSH and Floquet tight-binding solver

import numpy as np
import matplotlib.pyplot as plt
from scipy.linalg import expm

def ssh_open(ncell, v, w):
    h = np.zeros((2*ncell,2*ncell),dtype=complex)
    for n in range(ncell):
        h[2*n,2*n+1] = h[2*n+1,2*n] = v
        if n<ncell-1:
            h[2*n+1,2*n+2] = h[2*n+2,2*n+1] = w
    return h

def ssh_bloch(k,v,w):
    off = v+w*np.exp(-1j*k)
    return np.array([[0,off],[np.conj(off),0]],complex)

ncell=30
ratios=np.linspace(.25,1.75,180)
spectrum=[]
for r in ratios:
    spectrum.append(np.linalg.eigvalsh(ssh_open(ncell,r,1.0)))
spectrum=np.array(spectrum)

h=ssh_open(ncell,.55,1.0)
energy,vec=np.linalg.eigh(h)
edge_index=np.argsort(np.abs(energy))[:2]
edge_density=np.sum(np.abs(vec[:,edge_index])**2,axis=1)

# Two-step Floquet drive: swap strong and weak bonds.
ks=np.linspace(-np.pi,np.pi,501)
T=1.0; quasienergy=[]
for k in ks:
    h1=ssh_bloch(k,.25,1.15)
    h2=ssh_bloch(k,1.15,.25)
    u=expm(-1j*h2*T/2)@expm(-1j*h1*T/2)
    eps=-np.angle(np.linalg.eigvals(u))/T
    quasienergy.append(np.sort(eps))
quasienergy=np.array(quasienergy)

# Discrete winding of q(k)=v+w exp(-ik).
q=.55+1.0*np.exp(-1j*ks)
winding=np.sum(np.diff(np.unwrap(np.angle(q))))/(2*np.pi)

fig,ax=plt.subplots(1,3,figsize=(14,4))
ax[0].plot(ratios,spectrum,color='C0',lw=.35)
ax[0].axvline(1,color='k',ls=':')
ax[0].set(xlabel='v/w',ylabel='open-chain eigenvalue')
ax[1].bar(np.arange(2*ncell),edge_density)
ax[1].set(xlabel='site',ylabel='two midgap-state density')
ax[2].plot(ks/np.pi,quasienergy)
ax[2].set(xlabel='k/pi',ylabel='Floquet quasienergy')
fig.tight_layout();plt.show()
print('winding number:',winding)

The finite-chain spectrum shows midgap modes only on the topological side. The density plot checks localization rather than identifying an edge state by energy alone. For a two-dimensional Floquet model, extend the Bloch Hamiltonian to (k_x,k_y), compute Wilson loops or lattice-gauge Berry curvature, and inspect strip eigenvectors at both boundaries.

## Experimental observables

Near-field imaging measures boundary localization and propagation around defects. Fourier-space imaging reconstructs dispersion. Phase-resolved measurements are needed for winding and Berry-phase reconstruction; intensity alone cannot determine a complex eigenvector. In resonators, transmission poles reveal edge modes, but localized defect resonances can imitate them unless their spectral flow and bulk invariant are also measured.

Benchmark edge propagation against a trivial waveguide with matched loss and group velocity. Introduce controlled disorder over many realizations. Report insertion loss, delay, bandwidth, bend geometry, sample width, and the fraction of power remaining in the edge subspace.

## Validation checklist

• Declare the unit cell used for the Zak phase.
• Verify the bulk gap and invariant before labeling a boundary resonance topological.
• Plot eigenvector localization as well as eigenvalues.
• Test disorder that preserves the protecting symmetry separately from disorder that breaks it.
• In Floquet systems, inspect both zero and π/T quasienergy gaps.
• Use time-ordered propagation rather than only an averaged Hamiltonian.
• Distinguish chiral Chern edges from reciprocal pseudospin or valley-Hall edges.
• Check opposite-edge hybridization versus strip width.
• Report loss and usable bandwidth; topology does not cancel absorption.
• For synthetic dimensions, calibrate modulation amplitude and phase on every link.

Photonic topology is valuable when it gives a measured system-level advantage: routing through tight bends, fabrication-tolerant delay, controlled frequency conversion, or stable mode selection. The proof requires a bulk invariant, a localized boundary state, and perturbation tests consistent with the claimed protection.