Band Flip Transition: Mechanisms & Applications

• Band Flip Transition is a family of phenomena where control parameters reorder band assignments through gap closure and reopening.
• Key mechanisms include band inversion, exchange of band-edge characters, and occupation transfer, observed across superconducting, photonic, and mechanical systems.
• Practical insights reveal how engineered transitions enable control over topological states and nonequilibrium dynamics in diverse experimental platforms.

Searching arXiv for the cited papers and closely related "band flip" uses across domains to ground the article. A band flip transition is a parameter-driven reordering of band, band-edge, or band-character assignments in a spectrum. In the cited literature, the term is used for several closely related phenomena: a transition from single-band to two-band superconductivity in oxide interfaces, a band inversion accompanied by a topological phase transition, an exchange of a leaky resonance edge and a bound-state-in-the-continuum (BIC) edge in photonic lattices, a dimerization-driven SSH-type topological transition in mechanics, and a dissipation-driven transfer of occupation from dispersive bands into a flat band (Singh et al., 2018, Mondal et al., 2023, Lee et al., 2018, Chaunsali et al., 2017, Hu et al., 1 Apr 2025). Taken together, these works suggest that “band flip transition” is not a single universal mechanism, but a family of transitions in which a control parameter changes which band, orbital sector, symmetry representation, or edge state governs the low-energy or observable physics.

1. Conceptual meaning and scope

The most common use of the term is a band inversion or band reordering across a gap closing and reopening. In the dielectric photonic-crystal slab studied in “Probing Phase Transition of Band Topology via Radiation Topology” (Ji et al., 2022), the band-topology phase transition occurs because the odd-parity $p$-like modes and even-parity $d$-like modes exchange order at the $\Gamma$ point when the hole-center displacement is tuned through the critical condition $R=P/3$. In the AB-stacked bilayer Haldane model, anisotropic hopping drives the original Dirac points toward each other until they merge at the intermediate $\mathbf M$ point, and in the presence of Haldane flux the system undergoes multiple topological phase transitions with discontinuous Chern-number changes (Mondal et al., 2023).

A second established use is an exchange of band-edge character rather than a conventional bulk-band inversion. In leaky-mode photonic lattices, the second stop band has two edges: one is a radiative guided-mode-resonance edge and the other is a nonradiative BIC edge. A band flip occurs when these two edge states transit across the gap and swap which side of the gap they occupy (Lee et al., 2018). In this setting, the flip is tied to the competition between Bragg processes generated by the first two Fourier harmonics of the spatial modulation rather than to an electronic band inversion in the condensed-matter sense.

A third usage concerns changes in band participation in a collective ordered state. In the $(110)$-oriented LaAlO$_3$/SrTiO$_3$ interface, electrostatic doping drives a transition from superconductivity dominated by a single $d_{xz}/d_{yz}$-derived band to a two-band superconducting state involving a newly populated $d_{xy}$ band (Singh et al., 2018). The paper explicitly interprets this as an effective band flip in the quantum well because the occupancy hierarchy of the interfacial subbands changes with gate voltage.

A fourth usage extends the idea from spectral ordering to occupation transfer in open systems. In “Dissipation-Driven Transition of Particles from Dispersive to Flat Bands” (Hu et al., 1 Apr 2025), engineered bond dissipation selectively eliminates dispersive-band occupation and funnels particles into flat-band dark states. This is a band flip in the sense of band occupation rather than eigenvalue ordering.

2. Core mechanisms

A recurring mechanism is gap closure followed by reopening with exchanged character. In the photonic-crystal slab, the gap closes and reopens at the critical geometry $d$0, and the associated far-field polarization vortex charge changes from $d$1 to $d$2 or from $d$3 to $d$4 as the band topology changes from trivial to non-trivial (Ji et al., 2022). In the granular-chain realization of an SSH-like lattice, the gap closes at $d$5 when the alternating stiffnesses become equal, then reopens after the strong and weak bonds have interchanged, with the acoustic-band Zak phase changing from $d$6 to $d$7 across the transition (Chaunsali et al., 2017).

Another mechanism is competition between coupled channels that determine band-edge order. In leaky-mode photonic lattices, the second stop band is controlled mainly by the superposition of a second-order Bragg reflection generated by the first Fourier harmonic and a first-order Bragg reflection generated by the second Fourier harmonic. In the Kazarinov–Henry description, the gap closes when

$d$8

and the subsequent reopening reverses the ordering of the BIC edge and the leaky guided-mode-resonance edge (Lee et al., 2018).

A distinct mechanism is orbital-selective filling under electrostatic doping. In the $d$9 oxide interface, self-consistent Poisson–Schrödinger quantum-well simulations indicate that only the lower $\Gamma$0 band is occupied at low doping, whereas the $\Gamma$1 band becomes populated at higher doping. The experimental signatures of this threshold are nonlinear Hall response, deviation between $\Gamma$2 and $\Gamma$3, and a qualitative change in the temperature dependence of the normalized superfluid stiffness $\Gamma$4 (Singh et al., 2018).

Open-system realizations replace gap engineering with dark-state selection by dissipation. The Lindblad dynamics

$\Gamma$5

are supplemented by bond-dissipation operators of the form

$\Gamma$6

which annihilate flat-band states when the dissipation matches the compact-localized-state geometry. The resulting steady state is supported entirely on the flat band (Hu et al., 1 Apr 2025).

3. Electronic and superconducting realizations

In “Single-band to two-band superconductivity transition in two-dimensional oxide interfaces” (Singh et al., 2018), the physical system is a 10 unit-cell LaAlO$\Gamma$7 film grown on a $\Gamma$8-oriented SrTiO$\Gamma$9 substrate, forming a two-dimensional electron gas at the interface. The orientation is central because it changes the ordering and confinement of the Ti $R=P/3$0 $R=P/3$1 orbitals. The authors argue that, unlike the more familiar $R=P/3$2 case, the $R=P/3$3-derived band is occupied first and the $R=P/3$4-derived band appears only at stronger electrostatic doping. Resonant microwave transport is used to extract the superfluid stiffness

$R=P/3$5

with the interface embedded in a parallel RLC resonant circuit whose response shifts when the kinetic inductance of the superconducting two-dimensional electron gas changes.

The experimental evidence for the transition combines normal-state Hall transport and superconducting-state superfluid response. In the underdoped regime, the Hall resistance is linear in magnetic field and the normalized stiffness curves collapse onto a single BCS-like form, consistent with single-gap superconductivity. In the overdoped regime, the Hall response becomes nonlinear and the stiffness develops a change in curvature, a pronounced low-temperature tail, and a much steeper drop near the transition. These features are reproduced by a two-band self-consistent BCS model with

$R=P/3$6

indicating two distinct gaps and weak interband coupling (Singh et al., 2018).

The notable paradox is that the superconducting transition temperature decreases when the second band is populated. The paper emphasizes that this is counter to simple weak-coupling BCS intuition, because adding a band with finite density of states would usually be expected to enhance or at least sustain $R=P/3$7. To reconcile this behavior, the authors propose opposite-sign order parameters on the two bands, namely an $R=P/3$8 state in which interband impurity scattering is pair-breaking. In this usage, the band flip transition is not merely a normal-state Lifshitz-like threshold; it reorganizes the structure of the superconducting condensate itself (Singh et al., 2018).

A different electronic realization appears in “Electronic topological transition in sliding bilayer graphene” (Son et al., 2010). There, extremely small lateral interlayer sliding produces an electronic topological transition because the sliding generates an effective non-Abelian vector potential in the low-energy Dirac Hamiltonian. The four Dirac cones near the $R=P/3$9 point shift differently under sliding, and the result depends strongly on the sliding direction. For one class of directions, two cones carrying opposite Berry phases approach, merge at a critical shift, and annihilate, opening a gap in that channel. For another class, three cones merge and evolve into a new anisotropic Dirac cone. The paper therefore treats the transition as a reorganization of band connectivity and Fermi-surface topology enforced by Berry-phase conservation rather than by ordinary rigid-band motion.

4. Topological and photonic realizations

In the band-engineered bilayer Haldane model, anisotropic nearest-neighbor hopping $\mathbf M$0 and Haldane flux $\mathbf M$1 together generate a sequence of topological phase transitions (Mondal et al., 2023). Without flux, increasing $\mathbf M$2 moves the two Dirac points toward each other until they merge at $\mathbf M$3 at the intermediate $\mathbf M$4 point, producing semi-Dirac dispersion that is linear along one direction and quadratic along the other. With Haldane flux, the bands acquire nonzero Berry curvature and Chern numbers, and the merger becomes the locus of discontinuous topological changes such as $\mathbf M$5, $\mathbf M$6, $\mathbf M$7, and $\mathbf M$8. Edge spectra in nanoribbon geometry and the appearance or disappearance of anomalous Hall plateaus support these assignments. Here the band flip transition is explicitly topological: bands exchange Chern character when the bulk gap closes and reopens.

The photonic-crystal slab of “Probing Phase Transition of Band Topology via Radiation Topology” (Ji et al., 2022) implements a photonic analog of a quantum spin Hall-type transition. The slab is a SiN$\mathbf M$9 photonic-crystal slab with a graphene-like hexagonal lattice of triangular air holes, and the control parameter is the hole-center displacement $(110)$0, equivalently the ratio $(110)$1. For $(110)$2 the topology is trivial; for $(110)$3 it is non-trivial because the odd $(110)$4-like and even $(110)$5-like modes invert at $(110)$6. A defining feature of this work is the connection between band topology and radiation topology. The far-field polarization vortex charge

$(110)$7

changes from $(110)$8 to $(110)$9 or from $_3$0 to $_3$1 when the band topology flips. The reported charge exchange provides an experimentally accessible criterion for the band-topology transition through angle-resolved photoluminescence and Stokes-phase reconstruction.

“Band flips and bound-state transitions in leaky-mode photonic lattices” (Lee et al., 2018) isolates a different photonic mechanism. In a symmetric one-dimensional periodic photonic layer, the second stop band has an antisymmetric upper edge with no radiation loss, identified as the BIC, and a symmetric lower edge with maximal radiation loss, identified as the guided-mode-resonance edge. As fill factor $_3$2 and index modulation $_3$3 are varied, these two edges can meet and exchange sides. The transition is governed analytically by the condition $_3$4, and numerically the band gap is observed to open, shrink to zero at a critical modulation, and reopen with reversed edge assignment. This realization makes clear that a band flip need not mean inversion of bulk Bloch-band irreducible representations; it can instead be an exchange of radiative and nonradiative band-edge roles.

5. Mechanical and open-system realizations

“Demonstrating an in-situ topological band transition in cylindrical granular chains” (Chaunsali et al., 2017) provides a mechanical SSH-like realization in which the tunable parameter is the contact angle between precompressed cylindrical particles. The Hertzian contact law

$_3$5

linearizes under static precompression to an effective spring constant

$_3$6

Alternating contact angles create alternating couplings $_3$7 and $_3$8. As $_3$9 is tuned through $_3$0, the gap closes and reopens because the relative ordering of the two stiffnesses swaps. The acoustic-band Zak phase changes from $_3$1 for $_3$2 to $_3$3 for $_3$4, and in a finite chain the nontrivial regime supports a boundary-localized in-gap mode. When two topologically distinct dimerizations are joined, a finite-frequency interface mode appears and remains inside the common band gap. In this setting the band flip transition is a reversal of dimerization pattern and bulk topology.

In open quantum systems with flat bands, the transition is formulated in terms of steady-state occupation rather than dispersion reordering (Hu et al., 1 Apr 2025). The canonical examples are the cross-stitch lattice and the sawtooth lattice, both of which support compact localized states on the flat band. The central criterion is that the jump operators annihilate the flat-band eigenstates,

$_3$5

so that the flat band becomes a dark sector of the Liouvillian. For $_3$6 in the cross-stitch model, the steady state is a pure flat-band state for both $_3$7 and $_3$8, differing only in the phase pattern of the compact-localized-state superposition. For $_3$9, the zero-mode multiplicity increases and the steady state becomes a higher-dimensional flat-band manifold. This work extends the band-flip idea from spectral inversion to engineered nonequilibrium state preparation.

A plausible implication is that these mechanical and dissipative realizations broaden the meaning of band flip transition beyond fermionic band theory. In both cases, the decisive event is not merely the motion of eigenvalues but the reassignment of robust physical content: boundary localization in the mechanical chain and dark-state support in the open quantum lattice.

6. Related usages, distinctions, and limitations

The literature also contains usages in which “flip” refers to a spectral reconfiguration rather than a literal band inversion. In the switchable induced-transmission filter enabled by VO$d_{xz}/d_{yz}$0, the metallic phase of an ultrathin VO$d_{xz}/d_{yz}$1 layer suppresses most transmission bands except the one located at the electric-field node, so the device switches from a broad long-wave-infrared transmission window across $d_{xz}/d_{yz}$2–$d_{xz}/d_{yz}$3 to a narrow passband around $d_{xz}/d_{yz}$4–$d_{xz}/d_{yz}$5 (Wan et al., 2021). The paper explicitly states that the transition is not a literal shifting of one band across the spectrum; instead it is a broad multi-band transmission state collapsing into a narrow single-band transmission state.

By contrast, some superficially similar “flip” phenomena are explicitly not band flips. “Direct visualization of surface spin-flip transition in MnBi$d_{xz}/d_{yz}$6Te$d_{xz}/d_{yz}$7” (Ge et al., 2022) studies a first-order magnetic surface spin-flip transition of the topmost MnBi$d_{xz}/d_{yz}$8Te$d_{xz}/d_{yz}$9 layer at a field lower than the bulk spin-flip field. The paper states that this is not a literal electronic band inversion in the measured data, even though the altered magnetic boundary conditions are highly relevant for topological surface states and quantized transport in ultra-thin films. This distinction matters because the word “flip” can refer either to band topology or to magnetic-layer reversal.

Taken together, these distinctions show that the term is context-dependent. In some works it denotes a topological band inversion with gap closure and reopening; in others it denotes an exchange of band-edge roles, an orbital-threshold transition in a multiband condensate, or a dissipative inversion of band occupation. The common structure is a control-parameter-induced reassignment of which spectral sector carries the defining physical function, but the microscopic meaning of “band” and “flip” must be read from the model under discussion.