Beyond-subleading corrections to the color-superconducting gap in dense QCD and the domain of validity of the weak-coupling asymptotics

At asymptotically large quark chemical potential $\mu$, QCD becomes a weakly coupled color superconductor, and the Cooper-pairing gap $\Delta$ is computable from one-gluon exchange with hard-dense-loop resummation. Schäfer and Wilczek (Phys. Rev. D 60, 114033 (1999)) solved the resulting energy-dependent (Eliashberg) gap equation for two massless quark flavors, confirming the non-analytic leading behavior $\Delta \sim \mu\, g^{-5}\exp(-3\pi^2/(\sqrt{2}g))$ with $g$ the running QCD coupling, while stating that the overall prefactor was fixed only up to a factor of order one and posing the open questions: the extension to three flavors (where color-flavor locking completely Higgses the color group and the Meissner effect modifies gluon exchange), the functional form of the higher-order corrections, and the numerical bound on $g$ — equivalently on $\mu$ — for perturbation theory to be quantitatively reliable. Later work (1999–2004) completed the next-to-leading (subleading) order: the full prefactor is now known for the two-flavor (2SC) and color-flavor-locked (CFL) phases, gauge independence at this order is established, and the condensate's feedback on the gluon propagator was shown to enter only beyond subleading order. The surviving open core is the tier below: the corrections of relative order $g$ — sub-subleading terms of the gap equation, comprising subleading vertex corrections, the imaginary part of the gap function, the Meissner/plasmon modification of gluon exchange, and the running-coupling scale ambiguity — have not been computed in a combined treatment (only the plasmon modification has been evaluated, in isolation), and consequently the coupling (density) range within which the weak-coupling asymptotics quantitatively predict $\Delta(\mu)$ is not known. Authoritative reviews state the result is rigorous only for $\mu > 10^8$ MeV, while recent comparisons with lattice QCD at finite isospin density suggest empirical applicability down to $\mu \sim 1$ GeV. The problem is to compute the beyond-subleading corrections in a gauge-independent manner and, from their size, determine the validity domain of the asymptotic gap formula — including whether a controlled prediction of $\Delta$ at phenomenologically relevant densities is possible within perturbation theory at all.

Background

QCD, the theory of quarks and gluons, becomes weakly coupled at high energy density (asymptotic freedom). At large baryon density — equivalently, at large quark chemical potential $\mu$, the energy cost of adding quark matter to the system — the Fermi surface of the quarks is unstable to Cooper pairing, in close analogy with the BCS instability of electrons in a superconducting metal. The attractive channel is provided by gluon exchange in the color-antitriplet channel (the antisymmetric combination of two quark color charges, the channel in which one-gluon exchange between two quarks is attractive). The resulting color-superconducting phases are believed to be the ground states of QCD at asymptotically large $\mu$; the pairing gap $\Delta$ (the energy needed to break a pair, and the scale of the condensate) is small compared to $\mu$ but, at the densities possibly realized inside compact stars ($\mu$ of a few hundred MeV), is expected to be of order MeV to tens of MeV and to control transport and equation-of-state physics relevant to neutron stars.

Determining $\Delta$ from first principles requires solving the linearized gap equation with the gluon propagator corrected by the dense medium. Static electric gluon exchange is Debye screened (the medium screens longitudinal electric fields, with the Debye mass $m_D \sim g\mu$), but static magnetic gluon exchange is not screened, which produces an infrared enhancement; the magnetic interaction is eventually cut off dynamically by Landau damping (the scattering of soft gluons off the degenerate quarks, which screens magnetic fields at soft frequencies). Incorporating these one-loop medium effects — the hard dense loop (HDL) gluon self-energy, the resummed propagator valid at momenta $\Delta \ll k \ll \mu$ — and treating the energy dependence of the gap function (the Eliashberg equation, the finite-temperature/density generalization of the BCS gap equation) changes the characteristic BCS exponential $\exp(-1/g^2)$ into $\exp(-1/g)$. Son (1999) derived the leading weak-coupling result for two massless flavors, and the source of the present problem, the paper of Schäfer and Wilczek (1999), confirmed the exponent by an independent method and solved the full energy-dependent gap equation numerically, presenting the result $\Delta_0 \simeq 256\pi^4\,\mu\, g^{-5}\exp(-3\pi^2/(\sqrt{2}g))$ with the overall coefficient stated to be correct only up to a factor of order one.

The weak-coupling expansion has a non-standard structure organized by the size of contributions to the gap equation, writing $\phi_0$ for the gap function at the Fermi surface. Leading contributions come from almost-static magnetic gluon exchange (of order $g^2\phi_0$ with two powers of the large logarithm $\ln(\mu/\Delta)$) and fix the exponent $3\pi^2/(\sqrt{2}g)$; subleading contributions — static electric and non-static magnetic exchange, of order $g\phi_0$ — fix the prefactor; sub-subleading contributions, of order $g^2\phi_0$ with no compensating large logarithm, would correct the prefactor at relative order $g$. The subleading order was completed during 1999–2004. The quark self-energy (wave-function renormalization) suppresses the gap by the factor $e^{-b_0'} = \exp[-(\pi^2+4)/8] \simeq 0.177$ (Brown–Liu–Ren; Wang–Rischke), and the complete next-to-leading-order result for $N_f$ massless flavors is, in the notation of the 2008 review of Alford, Schmitt, Rajagopal and Schäfer (its Eqs. (44)–(45)),

$$\Delta = 512\pi^4 \left(\frac{2}{N_f}\right)^{5/2} e^{-b_0'}\, e^{-\zeta}\, e^{-d}\; \mu\, g^{-5}\, \exp\left(-\frac{3\pi^2}{\sqrt{2}\,g}\right),$$

where $e^{-\zeta}$ is a condensate-structure factor ($2^{-1/3}$ for the CFL phase, $1$ for 2SC) and $e^{-d} = 1$ for spin-zero condensates. At this order the result was also proven gauge independent on the quasiparticle mass shell (Pisarski–Rischke; Gerhold–Rebhan; Hou–Wang–Rischke): the gap extracted off the mass shell depends on the gauge-fixing parameter, a well-known subtlety of this problem.

For three massless quark flavors the expected ground state is the color-flavor-locked (CFL) phase, in which the diquark condensate — the condensate of quark–quark Cooper pairs — locks color and flavor rotations and all eight gluons acquire Meissner masses, the Meissner effect being the Anderson–Higgs mechanism by which gauge fields become massive in a superconductor, so that static magnetic fields are screened. The source paper listed this as a necessary extension: the condensate modifies the gluon propagator that mediates the pairing interaction, and CFL has two independent order parameters (a dominant color-antisymmetric one, $\Delta_{\bar 3}$, and a subdominant color-symmetric one, $\Delta_6$; the review finds the admixture ratio $\Delta_6/\Delta_{\bar 3} = \sqrt{2}\ln 2\, g/(36\pi)$, parametrically small at weak coupling). These extensions were carried out at leading and subleading order during 1999–2004: the CFL gap has the same exponent as the two-flavor (2SC) gap and a prefactor smaller by $2^{-1/3}$, a consequence of the two-gap quasiparticle structure of CFL.

What remains unsettled is the tier below subleading order and, with it, the quantitative reliability of the asymptotic formula. Because $\Delta$ depends exponentially on $g$, and the running coupling $g(\mu)$ decreases only logarithmically with $\mu$, the formula is rigorously trustworthy only where $g$ is small: the 2008 review states it is valid for $\mu > 10^8$ MeV — an estimate originating with Rajagopal and Shuster, who used the residual gauge dependence of the subleading-order calculation as a diagnostic and found that the neglected contributions are controlled only at such densities — astronomically far from the densities of compact stars, where $g \sim 2$–$4$; naively extrapolated to $\mu \simeq 400$–$500$ MeV it gives $\Delta \simeq 20$ MeV, comparable to model estimates, but with uncontrolled error. The source paper phrased the question as whether perturbation theory requires $g < 1$ or some weaker condition like $g < \pi$. Answering it requires the sub-subleading (relative order $g$) corrections, whose individual sources are identified in the literature — subleading vertex corrections, the imaginary part of the gap function generated by the finite lifetime of quasiparticles (Feng–Hou–Li–Ren; Reuter), the Meissner/plasmon modification of gluon exchange in the paired phase (plasmons being the collective oscillation modes of the gauge fields in the medium), and the ambiguity in the renormalization scale of the running coupling — but which have not been combined into a consistent order-$g$ result: only isolated pieces exist (the imaginary part is known to be suppressed by one power of $g$ relative to the real part, and the CFL plasmon effect has been evaluated by itself), and their sizes and signs within a single gauge-independent calculation are not established.

Problem Statement

Two linked objectives remain open on the question posed by Schäfer and Wilczek (1999), after the subleading order was completed by later work:

(a) Compute, for the two-flavor color-superconducting (2SC) and color-flavor-locked (CFL) ground states of QCD at asymptotically large quark chemical potential $\mu$ with massless quarks, the corrections of relative order $g$ (with $g$ the running QCD coupling) to the color-superconducting gap — the sub-subleading terms of the weak-coupling gap equation (with $\phi_0$ the gap function at the Fermi surface, the terms of order $g^2\phi_0$ that carry no compensating large logarithm $\ln(\mu/\Delta)$ and hence correct the known subleading result at relative order $g$). The sources identified in the literature that must all be accounted for are: (i) vertex corrections to the quark–gluon interaction beyond the hard-dense-loop vertex; (ii) the imaginary part of the gap function induced by the finite lifetime of quasiparticles; (iii) the modification of the exchanged gluon propagator by the superconducting condensate itself (Meissner masses and, in the CFL phase, light plasmon modes); and (iv) the effect of the running of the coupling and of the choice of its renormalization scale. The result must be gauge independent on the quasiparticle mass shell and must reduce to the known subleading-order formula when the order-$g$ terms are dropped.

(b) From the size of these corrections, determine the domain of validity of the asymptotic gap formula — the question the source paper posed as whether perturbation theory requires $g < 1$ or a weaker condition such as $g < \pi$: the largest coupling $g$ (equivalently, the smallest quark chemical potential $\mu$) for which the leading-plus-subleading weak-coupling prediction for $\Delta(\mu)$ is quantitatively reliable. No tolerance criterion defining 'quantitatively reliable' is presupposed: the sharp '$g < 1$ versus $g < \pi$' formulation becomes meaningful only once such a criterion is specified, and specifying and justifying it is itself part of the objective. Objective (b) therefore comprises an explicitly stated and justified reliability criterion; the threshold coupling $g$ (equivalently, the smallest quark chemical potential $\mu$) that this criterion implies; and an assessment of whether a controlled prediction of $\Delta$ at phenomenologically relevant chemical potentials ($\mu \lesssim 1$ GeV, where $g \sim 2$–$4$) is attainable by pushing the perturbative expansion to this order, or whether it requires nonperturbative input at accessible densities.

Objective (a) is the contribution of each source (i)–(iv) to the prefactor at relative order $g$ together with gauge independence of the corrected gap, and objective (b) is the resulting quantitative error estimate and the validity domain of the asymptotic formula. The two objectives are linked: the error of the asymptotic formula is dominated by the uncomputed corrections of relative order $g$ — and, as the power counting anticipates, by terms of order $g\log g$ and higher in $\ln(\Delta/\mu)$ — so a controlled determination of the validity domain in (b) requires the computation in (a); a matching determination against nonperturbative input can locate the domain empirically but does not by itself fix the size of the missing corrections.

References

1. T. Schäfer and F. Wilczek, 'Superconductivity from perturbative one-gluon exchange in high density quark matter', Phys. Rev. D 60, 114033 (1999). DOI 10.1103/PhysRevD.60.114033, arXiv:hep-ph/9906512, https://arxiv.org/abs/hep-ph/9906512. (Source of the open question posed here.)
2. D. T. Son, 'Superconductivity by long-range color magnetic interaction in high-density quark matter', Phys. Rev. D 59, 094019 (1999). DOI 10.1103/PhysRevD.59.094019, arXiv:hep-ph/9812287, https://arxiv.org/abs/hep-ph/9812287
3. R. D. Pisarski and D. H. Rischke, 'Gaps and critical temperature for color superconductivity', Phys. Rev. D 61, 051501 (2000). DOI 10.1103/PhysRevD.61.051501, arXiv:nucl-th/9907041, https://arxiv.org/abs/nucl-th/9907041
4. R. D. Pisarski and D. H. Rischke, 'Color superconductivity in weak coupling', Phys. Rev. D 61, 074017 (2000). DOI 10.1103/PhysRevD.61.074017, arXiv:nucl-th/9910056, https://arxiv.org/abs/nucl-th/9910056
5. I. A. Shovkovy and L. C. R. Wijewardhana, 'On gap equations and color-flavor locking in cold dense QCD with three massless flavors', Phys. Lett. B 470, 189 (1999). DOI 10.1016/S0370-2693(99)01297-6, arXiv:hep-ph/9910225, https://arxiv.org/abs/hep-ph/9910225
6. D. K. Hong, V. A. Miransky, I. A. Shovkovy, and L. C. R. Wijewardhana, 'Schwinger-Dyson approach to color superconductivity in dense QCD', Phys. Rev. D 61, 056001 (2000) [Erratum: Phys. Rev. D 62, 059903 (2000)]. DOI 10.1103/PhysRevD.61.056001, arXiv:hep-ph/9906478, https://arxiv.org/abs/hep-ph/9906478
7. W. E. Brown, J. T. Liu, and H.-c. Ren, 'Perturbative nature of color superconductivity', Phys. Rev. D 61, 114012 (2000). DOI 10.1103/PhysRevD.61.114012, arXiv:hep-ph/9908248, https://arxiv.org/abs/hep-ph/9908248
8. D. H. Rischke, 'Debye screening and the Meissner effect in a three-flavor color superconductor', Phys. Rev. D 62, 054017 (2000). DOI 10.1103/PhysRevD.62.054017, arXiv:nucl-th/0003063, https://arxiv.org/abs/nucl-th/0003063
9. K. Rajagopal and E. Shuster, 'Applicability of weak-coupling results in high density QCD', Phys. Rev. D 62, 085007 (2000). DOI 10.1103/PhysRevD.62.085007, arXiv:hep-ph/0004074, https://arxiv.org/abs/hep-ph/0004074
10. D. H. Rischke, 'Gluon self-energy in a two-flavor color superconductor', Phys. Rev. D 64, 094003 (2001). DOI 10.1103/PhysRevD.64.094003, arXiv:nucl-th/0103050, https://arxiv.org/abs/nucl-th/0103050
11. D. H. Rischke, D. T. Son, and M. A. Stephanov, 'Asymptotic deconfinement in high-density QCD', Phys. Rev. Lett. 87, 062001 (2001). DOI 10.1103/PhysRevLett.87.062001, arXiv:hep-ph/0011379, https://arxiv.org/abs/hep-ph/0011379
12. S. R. Beane, P. F. Bedaque, and M. J. Savage, 'Renormalization group improved gap equation for color superconductors', Nucl. Phys. A 688, 931 (2001). DOI 10.1016/S0375-9474(00)00595-9, arXiv:nucl-th/0004013, https://arxiv.org/abs/nucl-th/0004013
13. Q. Wang and D. H. Rischke, 'How the quark self-energy affects the color-superconducting gap', Phys. Rev. D 65, 054005 (2002). DOI 10.1103/PhysRevD.65.054005, arXiv:nucl-th/0110016, https://arxiv.org/abs/nucl-th/0110016
14. A. Schmitt, Q. Wang, and D. H. Rischke, 'When the transition temperature in color superconductors is not like in BCS theory', Phys. Rev. D 66, 114010 (2002). DOI 10.1103/PhysRevD.66.114010, arXiv:nucl-th/0209050, https://arxiv.org/abs/nucl-th/0209050
15. R. D. Pisarski and D. H. Rischke, 'Gauge invariance of the color-superconducting gap on the mass shell', Nucl. Phys. A 702, 177 (2002). DOI 10.1016/S0375-9474(02)00702-9, arXiv:nucl-th/0111070, https://arxiv.org/abs/nucl-th/0111070
16. A. Gerhold and A. Rebhan, 'Gauge dependence identities for color superconducting QCD', Phys. Rev. D 68, 011502 (2003). DOI 10.1103/PhysRevD.68.011502, arXiv:hep-ph/0305108, https://arxiv.org/abs/hep-ph/0305108
17. D. Hou, Q. Wang, and D. H. Rischke, 'Generalized Ward identity and gauge invariance of the color-superconducting gap', Phys. Rev. D 69, 071501 (2004). DOI 10.1103/PhysRevD.69.071501, arXiv:hep-ph/0401152, https://arxiv.org/abs/hep-ph/0401152
18. A. Schmitt, Q. Wang, and D. H. Rischke, 'Mixing and screening of photons and gluons in a color superconductor', Phys. Rev. D 69, 094017 (2004). DOI 10.1103/PhysRevD.69.094017, arXiv:nucl-th/0311006, https://arxiv.org/abs/nucl-th/0311006
19. T. Schäfer, 'Effective theory of superfluid quark matter', in KIAS-APCTP International Symposium in Astro-Hadron Physics: Compact Stars — Quest for New States of Dense Matter (2004), pp. 330–344. DOI 10.1142/9789812702524_0026, arXiv:hep-ph/0402032, https://arxiv.org/abs/hep-ph/0402032
20. Q. Wang, 'Recent developments in weak-coupling colour superconductivity', J. Phys. G 30, S1251 (2004). DOI 10.1088/0954-3899/30/8/101, arXiv:nucl-th/0404017, https://arxiv.org/abs/nucl-th/0404017
21. H. Malekzadeh and D. H. Rischke, 'Gluon self-energy in the color-flavor-locked phase', Phys. Rev. D 73, 114006 (2006). DOI 10.1103/PhysRevD.73.114006, arXiv:hep-ph/0602082, https://arxiv.org/abs/hep-ph/0602082
22. B. Feng, D.-f. Hou, J.-r. Li, and H.-c. Ren, 'The imaginary part of the gap function in color superconductivity', Nucl. Phys. B 754, 351 (2006). DOI 10.1016/j.nuclphysb.2006.08.021, arXiv:nucl-th/0606015, https://arxiv.org/abs/nucl-th/0606015
23. P. T. Reuter, 'Complex gap in color superconductivity', Phys. Rev. D 74, 105008 (2006). DOI 10.1103/PhysRevD.74.105008, arXiv:nucl-th/0608020, https://arxiv.org/abs/nucl-th/0608020
24. H. Malekzadeh, 'Gluon self-energy effect on color-flavor-locked gap equation', Phys. Rev. D 79, 034010 (2009). DOI 10.1103/PhysRevD.79.034010, arXiv:0810.5235, https://arxiv.org/abs/0810.5235
25. M. G. Alford, A. Schmitt, K. Rajagopal, and T. Schäfer, 'Color superconductivity in dense quark matter', Rev. Mod. Phys. 80, 1455 (2008). DOI 10.1103/RevModPhys.80.1455, arXiv:0709.4635, https://arxiv.org/abs/0709.4635
26. Y. Fujimoto, 'Interplay between the weak-coupling results and the lattice data in dense QCD', arXiv:2408.12514 (2024), https://arxiv.org/abs/2408.12514
27. A. Kurkela, K. Rajagopal, and R. Steinhorst, 'Astrophysical equation-of-state constraints on the color-superconducting gap', Phys. Rev. Lett. 132, 262701 (2024). DOI 10.1103/PhysRevLett.132.262701, arXiv:2401.16253, https://arxiv.org/abs/2401.16253
28. Y. Fujimoto, 'Renormalization group analysis of color superconductivity revisited', arXiv:2508.19728 (2025), https://arxiv.org/abs/2508.19728