How can a superconductor form a pair-density wave?
A pair-density wave (PDW) is a superconducting state where the Cooper pair density varies periodically in space. [2] In the conventional Peierls instability, ions form a repeating pattern. [3] Here we show its superconducting analogue: the Superconducting Peierls Instability.
Click through the three steps below to compare the mechanisms.
Both systems have occupied and empty states that meet at zero energy.
Conventional Peierls Instability
Deviation from equilibrium positions (dashed circles)
Superconducting Peierls Instability
WHY THE DISTORTIONS DEVELOP
The Kohn anomalies
A uniform state resists small distortions: moving the ions or rippling the order parameter costs energy. Scattering of electrons or quasiparticles reduces this restoring stiffness, most strongly at the wavevector joining the zero-energy crossings. In a metal, this softening is the Kohn anomaly. [4] We find its superconducting counterpart. [1]
The slider increases both couplings: \(g\) between electrons and lattice vibrations, and \(M\) between quasiparticles and order-parameter fluctuations. Move it past the marked onset to see each stiffness dip reach zero and become negative.
Conventional Peierls Instability
\(g\) Electron–phonon coupling
bare cost − electronic reduction
Inside \(\chi_0(q)\)
\(E_1=\epsilon_k,\ E_2=\epsilon_{k+q}\).
Superconducting Peierls Instability
\(M\) Order-parameter–quasiparticle coupling
bare cost − quasiparticle reduction
Inside \(\Pi_{\rm mode}(q)\)
\(E_1,E_2\): quasiparticle levels; \(f\): occupation.
Schematic curves: dashed lines show the bare stiffness; solid lines include the electronic or quasiparticle response.
Positive stiffness: the uniform state resists this deformation.
The pattern is favoured once the electronic or quasiparticle energy gain outweighs the cost of distorting the lattice or order parameter.
A CONCRETE REALISATION
The superconducting strip
Does the schematic mechanism occur in a simple model? We test a two-dimensional extended Hubbard model with onsite and nearest-neighbour attraction. Its mixed-symmetry \(s+d+ip_x\) state [5, 6, 7] hosts edge quasiparticles called Andreev bound states (ABS). [8] These states disperse through \(E=0\), providing the zero-energy crossings needed for the instability. Click Edge states to see their dispersion.
The strip is open across x and periodic along y. We initially assume that the order parameters are uniform along y, while allowing them to vary across x.
Choose a component for the 3D view; the profiles show all four nonzero components.
Near edgeFar edge
Across-strip profiles
Andreev bound states
Blue: edge-localised Grey: bulk-like
THE SELF-CONSISTENT RESULT
The edge develops a pair-density wave
Now allow the superconducting order parameters to vary along y. We solve the Bogoliubov–de Gennes equations self-consistently, then compare the result with the uniform reference. [1]
Switch between Uniform and PDW to see what changes. Choose Measure period to follow the order parameter along one edge.
Calculated spectrum: uniform edge
Near edgeFar edgeBulk
The uniform edge has zero-energy crossings.
The calculated state reproduces the schematic mechanism. An edge-localised PDW opens a gap at the quasiparticle crossings, while the interior remains nearly uniform. Its wavevector \(Q\simeq2k_c\) and period \(\pi/k_c\) follow the original crossing separation.
Could these same ingredients drive an instability beyond an edge?
BEYOND THE EDGE
Generalizations of this instability
The Superconducting Peierls mechanism and Superconducting Kohn anomaly require zero-energy quasiparticles connected by finite momentum and an order-parameter response strong enough to soften that mode. These ingredients can occur in higher dimensions, opening a route to bulk pair-density waves.
Our self-consistent strip calculation demonstrates the mechanism at an edge. In the bulk, zero-energy quasiparticles can lie on Bogoliubov Fermi surfaces. These occur in a variety of models, including multiband superconductors [9] and altermagnetic superconductors. [10, 11]
This raises a broader possibility: could they drive a similar reconstruction of the superconducting order parameter in the bulk? This intriguing possibility will be explored in the future!
References
- P. Senarath Yapa et al., arXiv:2609.04164 (2026).
- D. F. Agterberg et al., Annu. Rev. Condens. Matter Phys. 11, 231 (2020).
- R. E. Peierls, Quantum Theory of Solids, Oxford University Press (1955).
- W. Kohn, Phys. Rev. Lett. 2, 393 (1959).
- S. Nayak and S. Kumar, J. Phys.: Condens. Matter 30, 135601 (2018).
- J. Hutchinson and F. Marsiglio, J. Phys.: Condens. Matter 33, 065603 (2021).
- P. Senarath Yapa et al., Physica C 633, 1354719 (2025).
- J. A. Sauls, Phil. Trans. R. Soc. A 376, 20180140 (2018).
- D. F. Agterberg, P. M. R. Brydon, and C. Timm, Phys. Rev. Lett. 118, 127001 (2017).
- M. Wei et al., Phys. Rev. B 109, L201404 (2024).
- S. Hong, M. J. Park, and K.-M. Kim, Phys. Rev. B 111, 054501 (2025).