The 1978 theory of Ambegaokar, Halperin, Nelson, and Siggia (AHNS) explains the dissipation of a thin superfluid 4^4He film near its Kosterlitz–Thouless transition through the diffusive motion of quantized vortices, governed by a Langevin equation whose two transport coefficients — the vortex diffusion constant DD and the drag parameter CC — are phenomenological inputs, estimated only as D≲ℏ/m≈10−4 cm2/sD\lesssim\hbar/m\approx10^{-4}\,\mathrm{cm^2/s}. Nearly five decades later these coefficients are still obtained by fitting dissipation data, and the extracted values are nonuniversal: they vary with temperature, film coverage, substrate, and geometry, and in microscale films thermal vortex diffusion is suppressed by orders of magnitude relative to the diffusive picture. The problem is to compute DD and CC from first principles for a specified film — including vortex–vortex collisions, substrate drag, and coupling to excitations such as phonons, ripplons, and rotons — and to establish the regime of validity of the theory's approximations (dilute vortex gas, neglected convective terms, linear response, and the low-frequency diffusion-limited window), in particular the nonlinear regime very close to the transition where the correlation length exceeds the critical vortex-pair radius, and the high-frequency regime beyond the diffusion-limited one.

Background

In a two-dimensional (2D) superfluid, topological order is destroyed not by long-wavelength fluctuations of the phase but by vortices — points of quantized circulation that interact through a logarithmic potential, like charges in a 2D Coulomb gas. In the Berezinskii–Kosterlitz–Thouless (BKT) picture, below the transition temperature TcT_c all vortices are bound into neutral vortex–antivortex pairs, while above TcT_c free vortices proliferate and destroy superfluidity; the superfluid density jumps discontinuously at TcT_c. Thin 4^4He films adsorbed on a substrate are the cleanest realization of this physics. The universal jump was predicted by Nelson and Kosterlitz (Phys. Rev. Lett. 39, 1201 (1977)) and confirmed with torsional-oscillator experiments, in which the film sits on a substrate oscillating at a fixed frequency so that both its inertia (the superfluid density) and its energy dissipation are measured.

Static BKT theory says nothing about dissipation. Ambegaokar, Halperin, Nelson, and Siggia (Phys. Rev. Lett. 40, 783 (1978); full account in Phys. Rev. B 21, 1806 (1980)) supplied the dynamic theory: dissipation arises from the motion of free vortices above TcT_c and from the dynamic polarization of bound pairs below it. The central object is a Langevin equation (a stochastic equation of motion with a thermal random force) for the position of each vortex in the film plane, which in their notation reads

dridt=ni Dℏρ0mkBT z^×(vn−vs(i))+C (vn−vs(i))+vs(i)+ηi(t),\frac{d\mathbf{r}_i}{dt} = n_i\,\frac{D\hbar\rho_0}{mk_BT}\,\hat{\mathbf{z}}\times(\mathbf{v}_n-\mathbf{v}_s^{(i)}) + C\,(\mathbf{v}_n-\mathbf{v}_s^{(i)}) + \mathbf{v}_s^{(i)} + \boldsymbol{\eta}_i(t),

where ni=±1n_i=\pm1 is the vortex circulation sign, ρ0\rho_0 the background superfluid density per film thickness, mm the helium-atom mass, vn\mathbf{v}_n the substrate (normal-component) velocity, vs(i)\mathbf{v}_s^{(i)} the local superfluid velocity at the vortex, z^\hat{\mathbf{z}} the film normal, and ηi\boldsymbol{\eta}_i a Gaussian random force. The theory's content is then an electrostatics-like computation of the frequency-dependent dielectric response of this vortex 'Coulomb gas', with pair relaxation rate ν(r)∼2D/r2\nu(r)\sim2D/r^2 for a pair of separation rr. The dissipated power in the regimes accessible to oscillating-substrate experiments follows, with distinct scaling behaviors governed by the correlation lengths ξ±\xi_{\pm} above and below TcT_c, the diffusion length (D/ω)1/2(D/\omega)^{1/2}, and the critical pair-unbinding radius rc=ℏ/(m∣vn−us∣)r_c=\hbar/(m|\mathbf{v}_n-\mathbf{u}_s|), where usu_s is the spatially averaged superfluid velocity, beyond which a driven pair runs apart (which produces the nonlinear response at large oscillation amplitudes).

Two coefficients, however, are pure inputs: the vortex diffusion constant DD (combining the random-force strength and the drag) and the dimensionless drag parameter CC in the equation above. The authors did not derive them; they argued only that D≲ℏ/m≈10−4 cm2/sD\lesssim\hbar/m\approx10^{-4}\,\mathrm{cm^2/s}, and they neglected the convective term vs(i)\mathbf{v}_s^{(i)} (the advection of a vortex by the local superflow of the others) and the vortex-core structure, working with a dilute gas of point-like 'renormalized' vortices in the low-frequency, diffusion-limited regime. A microscopic calculation of DD and CC would have to account for the interactions of a moving vortex line with the solid substrate (through the film's normal component and its roughness), with the elementary excitations of the film — phonons, ripplons (quantized capillary waves of the free surface), and rotons (the short-wavelength excitation near the minimum of the helium dispersion relation) — and with other vortices through collisions.

The subsequent history makes the status of these coefficients concrete. Bishop and Reppy (Phys. Rev. B 22, 5171 (1980)) confirmed the BKT picture in oscillating-substrate experiments but observed finite-velocity effects 'not explained by current dynamic theories', exposing the nonlinear regime. Petschek and Zippelius (Phys. Rev. B 23, 3483 (1981)) reanalyzed the same phenomenological Langevin equation including convective motion, showing by renormalization-group methods that dynamic screening renormalizes DD (with a universal cusp approaching a finite value at TcT_c) while the convective parameter is unchanged — a formal renormalization, not a microscopic determination of the bare coefficients. Experiments extract DD by fitting dissipation peaks: Agnolet, McQueeney, and Reppy (Phys. Rev. B 39, 8934 (1989)) found the dynamic response nonuniversal as a function of film coverage; related studies of vortex diffusivity in thin films include McQueeney et al. (Jpn. J. Appl. Phys. 26, Suppl. 26-3, 79 (1987)), Ben-Ezra and Glaberson (Physica B 284–288, 122 (2000)), and Huber (Phys. Lett. A 79, 331 (1980)). In a microscale 2D helium film on a silicon chip, Sachkou et al. (Science 366, 1480 (2019)) found coherent vortex dynamics dominating, with thermal vortex diffusion suppressed by five orders of magnitude in the published paper (six in the preprint) — showing that the drag on a vortex is strongly geometry- and substrate-dependent. The framework has been extended to other 2D superfluids (e.g., the dynamic Kosterlitz–Thouless theory of two-dimensional ultracold gases by Wu, Zhang, and Zhai, Phys. Rev. A 102, 043311 (2020)) and reviewed by Minnhagen (Rev. Mod. Phys. 59, 1001 (1987)); in all of these the vortex transport coefficients remain fitting parameters, as in the most recent applications to nanochannel-confined films (Novotný et al., arXiv:2603.02388 (2026)), where DD is obtained from fits and varies with system and temperature.

Problem Statement

For a two-dimensional superfluid (a thin 4^4He film of thickness and substrate specified by the solver) near its Kosterlitz–Thouless transition, determine from first principles the two transport coefficients of the Ambegaokar–Halperin–Nelson–Siggia vortex equation of motion: the vortex diffusion constant DD and the drag parameter CC, currently only estimated as D≲ℏ/m≈10−4 cm2/sD\lesssim\hbar/m\approx10^{-4}\,\mathrm{cm^2/s}. The calculation must incorporate the microscopic mechanisms of vortex drag in a film — including interaction with the substrate through the normal component and surface roughness, coupling to the film's excitations (phonons, ripplons, rotons), vortex-core structure, and vortex–vortex collisions — and connect the resulting microscopic coefficients to the coarse-grained 'renormalized' vortices of the AHNS theory. Together with this, establish the regime of validity of the theory's approximations: the neglected convective (advection of a vortex by the local superflow) terms, the dilute-vortex-gas and linear-response assumptions, the nonlinear regime very close to TcT_c where the correlation length ξ−\xi_{-} exceeds the critical unbinding radius rcr_c (so the free-vortex generation rate requires modification), and frequencies beyond the diffusion-limited window ω≪Dξ+−2\omega\ll D\xi_{+}^{-2}. Predicted D(T)D(T) and C(T)C(T) for specified films must reproduce the measured dissipation of oscillating-substrate, third-sound, and resonator experiments (third sound being a surface wave of a superfluid helium film in which the normal component is clamped to the substrate), and must explain the nonuniversal, coverage-, geometry-, and temperature-dependent values extracted from them (including the suppression of thermal vortex diffusion in microscale films). The first-principles determination of D(T)D(T) and C(T)C(T), the regime-of-validity analysis, and its validation against the experiments above are to be established together.

References

  1. V. Ambegaokar, B. I. Halperin, D. R. Nelson, and E. D. Siggia, Dissipation in Two-Dimensional Superfluids, Phys. Rev. Lett. 40, 783 (1978). DOI: 10.1103/PhysRevLett.40.783, https://doi.org/10.1103/PhysRevLett.40.783
  2. V. Ambegaokar, B. I. Halperin, D. R. Nelson, and E. D. Siggia, Dynamics of superfluid films, Phys. Rev. B 21, 1806 (1980). DOI: 10.1103/PhysRevB.21.1806, https://doi.org/10.1103/PhysRevB.21.1806
  3. David R. Nelson and J. M. Kosterlitz, Universal Jump in the Superfluid Density of Two-Dimensional Superfluids, Phys. Rev. Lett. 39, 1201 (1977). DOI: 10.1103/PhysRevLett.39.1201, https://doi.org/10.1103/PhysRevLett.39.1201
  4. D. J. Bishop and J. D. Reppy, Study of the superfluid transition in two-dimensional 4^4He films, Phys. Rev. B 22, 5171 (1980). DOI: 10.1103/PhysRevB.22.5171, https://doi.org/10.1103/PhysRevB.22.5171
  5. R. Petschek and A. Zippelius, Renormalization of the vortex diffusion constant in superfluid films, Phys. Rev. B 23, 3483 (1981). DOI: 10.1103/PhysRevB.23.3483, https://doi.org/10.1103/PhysRevB.23.3483
  6. P. Minnhagen, The two-dimensional Coulomb gas, vortex unbinding, and superfluid-superconducting films, Rev. Mod. Phys. 59, 1001 (1987). DOI: 10.1103/RevModPhys.59.1001, https://doi.org/10.1103/RevModPhys.59.1001
  7. G. Agnolet, D. McQueeney, and J. Reppy, Kosterlitz-Thouless transition in helium films, Phys. Rev. B 39, 8934 (1989). DOI: 10.1103/PhysRevB.39.8934, https://doi.org/10.1103/PhysRevB.39.8934
  8. D. F. McQueeney, G. Agnolet, G. K. S. Wong, and J. D. Reppy, Vortex Dynamics in Thin 4^4He Films, Jpn. J. Appl. Phys. 26, Suppl. 26-3, 79 (1987). DOI: 10.7567/jjaps.26s3.79, https://doi.org/10.7567/jjaps.26s3.79
  9. D. L. Huber, Vortex diffusion in superfluid helium films, Phys. Lett. A 79, 331 (1980). DOI: 10.1016/0375-9601(80)90362-X, https://doi.org/10.1016/0375-9601(80)90362-X
  10. S. Ben-Ezra and W. Glaberson, Vortex diffusivity in thin superfluid helium films, Physica B 284–288, 122 (2000). DOI: 10.1016/S0921-4526(99)02139-0, https://doi.org/10.1016/S0921-4526(99)02139-0
  11. Mitsunori Hieda, Kenji Matsuda, Tsuyoshi Kato, Taku Matsushita, and Nobuo Wada, Extremely High Frequency Dependence of Two-Dimensional Superfluid Onset, J. Phys. Soc. Jpn. 78, 033604 (2009). DOI: 10.1143/JPSJ.78.033604, https://doi.org/10.1143/JPSJ.78.033604
  12. Yauhen P. Sachkou, Christopher G. Baker, Glen I. Harris, Oliver R. Stockdale, Stefan Forstner, Matthew T. Reeves, Xin He, David L. McAuslan, Ashton S. Bradley, Matthew J. Davis, and Warwick P. Bowen, Coherent vortex dynamics in a strongly interacting superfluid on a silicon chip, Science 366, 1480 (2019); preprint arXiv:1902.04409. DOI: 10.1126/science.aaw9229, https://arxiv.org/abs/1902.04409
  13. Oliver R. Stockdale, M. Reeves, Xiaoquan Yu, G. Gauthier, Kwan Goddard-Lee, W. Bowen, T. Neely, and M. Davis, Universal dynamics in the expansion of vortex clusters in a dissipative two-dimensional superfluid, Phys. Rev. Research 2, 033138 (2020). DOI: 10.1103/PhysRevResearch.2.033138, https://doi.org/10.1103/PhysRevResearch.2.033138
  14. Zhi-Gang Wu, Shi-Zhong Zhang, and H. Zhai, Dynamic Kosterlitz-Thouless theory for two-dimensional ultracold atomic gases, Phys. Rev. A 102, 043311 (2020). DOI: 10.1103/PhysRevA.102.043311, https://doi.org/10.1103/PhysRevA.102.043311
  15. Filip Novotný, Marek Talíř, Balázs Szalai, and Emil Varga, Kosterlitz-Thouless transition in uniformly confined 4^4He, arXiv:2603.02388 (2026). DOI: 10.48550/arXiv.2603.02388, https://arxiv.org/abs/2603.02388
  16. P. W. Adams and W. I. Glaberson, Vortex dynamics in superfluid helium films, Phys. Rev. B 35, 4633 (1987). DOI: 10.1103/PhysRevB.35.4633, https://doi.org/10.1103/PhysRevB.35.4633