betaFeedback

WorksPhysics and Astronomy

Magnetic-slope crossovers in the QCD bulk-to-shear viscosity ratio

Abstract

How a magnetic field alters dissipation in quark–gluon matter is almost unconstrained from first principles, since real-time transport is not directly accessible to lattice simulation. Using a top-down MQGP-inspired M/string-theory construction of a QCD-like thermal plasma, together with a kinetic-theory identity within a specified single-mode quasiparticle truncation and a compatibility construction using available lattice equation-of-state data, we propose a normalization-insensitive sign prediction that the field slope of the longitudinal bulk-to-shear viscosity ratio ζ_\parallel/η_\parallel reverses sign twice with temperature—below and above the deconfinement pseudo-critical temperature T_0, at T_1<T_0<T_2 (the lower reversal T_1 conditional on a stated pion-gas rate inequality). A kinetic-theory identity with phase-appropriate carriers, a quark–gluon plasma above T_0 and a chiral pion gas below, reproduces the same pattern, and lattice speed-of-sound data can be shown to be consistent as well. The bottom-up VQCD in the deconfined phase supports a negative slope for large temperatures; the slope crossover is not resolvable from the available ζ_\parallel/s, η_\parallel/s plots.

Introduction. A magnetic field is one of the few external handles that can be applied to strongly interacting matter both in the laboratory and in the early universe. Non-central heavy-ion collisions generate the strongest fields known, transiently of order eB∼mπ2eB \sim m_{\pi}^{2}; magnetars sustain comparable fields for astrophysical times. What such a field does to the equilibrium properties of QCD matter—the equation of state, the chiral condensate, the transition temperature—is by now well mapped by lattice simulation. What it does to dissipation is not.

The obstruction is structural rather than technical. Viscosities are real-time response coefficients, defined by Kubo formulae involving the analytic continuation of Euclidean correlators; extracting them from lattice data requires an ill-posed inversion, and doing so in a background field has not been achieved. Transport in magnetized QCD is therefore a domain where a normalization-insensitive sign prediction—as opposed to a fit or a model—has real value, because there is currently nothing to fit to. We are explicit about the logical hierarchy: the top-down construction proposes the sign pattern; the kinetic identity explains the sign mechanism within a controlled single-mode truncation; the lattice equation of state demonstrates compatibility (not a determination of ζ/η\zeta/\eta); and VQCD checks the high-TT negative-slope trend only.

We make such a prediction here. The quantity we track is the ratio of bulk-to-shear viscosity, measured along the field direction. Bulk viscosity ζ\zeta measures resistance to uniform compression and vanishes for a scale-invariant fluid; shear viscosity η\eta measures resistance to shape deformation. Their ratio is a clean diagnostic of how far QCD sits from conformality.

Our claim is about the slope of this ratio in the field, at fixed temperature. We find that

∂∂B(ζ∥η∥)∣T(1)\left.\frac{\partial}{\partial B}\left(\frac{\zeta_{\parallel}}{\eta_{\parallel}}\right)\right|_{T} \tag*{(1)}

changes sign twice as TT is raised through the transition region: it is positive at low TT, turns negative at a temperature above T1<T0T_1 < T_0 (T0T_0 being the reduced transition temperature in the presence of a weak magnetic field) still inside the confined phase, and—after the deconfinement transition at T0T_0—turns positive/negative at temperatures below/above T2>T0T_2 > T_0 in the deconfined phase. With standard inputs, T1≃134 MeVT_1 \simeq134\ {\rm MeV}, T0≃143 MeVT_0 \simeq143\ {\rm MeV} and T2≃187 MeVT_2 \simeq187\ {\rm MeV}.

Two features make this worth stating as a Letter. First, it is a qualitative statement—a pattern of signs—and therefore insensitive to the overall normalizations of any holographic computation. Second, it is reached from two directions with no common dynamical input—a top-down M\mathcal{M}-theory dual and a kinetic-theory identity—while lattice equation-of-state data can be made consistent with both (a compatibility construction, not an independent determination of ζ/η\zeta/\eta). Interestingly, a bottom-up VQCD holographic model is shown also to be consistent in the sense of obtaining a negative large-temperature slope in the deconfined phase even though the crossover in the same phase is not graphically resolvable from the available ζ∥/η∥\zeta_{\parallel}/\eta_{\parallel} plots.

What is being computed. In a magnetic field the rotational symmetry of the medium is reduced from SO(3)SO(3) to SO(2)SO(2) about the field axis, and a single shear and a single bulk viscosity no longer suffice. The full classification of the resulting transport coefficients for a relativistic magnetized fluid is discussed in [1]: the shear sector splits into longitudinal and transverse coefficients η∥,η⊥\eta_{\parallel},\eta_{\perp} together with two non-dissipative Hall viscosities, and the bulk sector is governed by four coefficients of which only three are independent, by an Onsager relation.

We track one specific combination in kinetic theory — the longitudinal, parity-even projection

R(B,T)≡ζ∥(B,T)η∥(B,T)=R0(T)+R2(T)(BTc2)2+⋯(2)\mathcal{R}(B,T) \equiv\frac{\zeta_{\parallel}(B,T)}{\eta_{\parallel}(B,T)} = R_0(T) + R_2(T)\left(\frac{B}{T_c^2}\right)^2 + \cdots \tag*{(2)}

obtained by projecting the viscosity tensors along the field direction. The longitudinal projection returns a genuine Kubo coefficient rather than a bare tensor label: the Eling–Oz horizon datum responds to BB only through the single scalar rh/0(B)r_{h/0}(B), so the induced compression is isotropic at leading order and the cross viscosity ζ∥⊥\zeta_{\parallel\perp}—which would enter through the anisotropic (ℓ=2\ell=2) part of the horizon expansion—is NN-suppressed beyond O(1/N)O(1/N) by the same fractional-anisotropy estimate that controls the projection anisotropy [Supplemental Material [2], Sec. I]. Writing the expansion in the dimensionless variable B/Tc2B/T_c^2 makes the comparison between frameworks unambiguous. In the equilibrium background (Ω\Omega(vorticity) = EE(electric field) = 0) the field BB is PP-even and TT-odd—it is the only axial vector present—so a (pseudo)scalar (e.g. (B⋅a)nB(B\cdot a)^{n_B}, aμa^\mu being the acceleration - a vector)’s parity is (−1)nB(-1)^{n_B} and R\mathcal{R}, built from the even-in-BB (longitudinal magnetization and viscosity) structures of the Hernández–Kovtun classification, is parity-even; hence ∂BR∣B=0=0\partial_B\mathcal{R}|_{B=0}=0 identically. By “field slope” we therefore mean ∂BR\partial_B\mathcal{R} evaluated at fixed small B>0B>0, where ∂BR=2BR2(T)/Tc4+O(B3)\partial_B\mathcal{R}=2B R_2(T)/T_c^4+O(B^3), so its sign is that of the weak-field curvature coefficient R2(T)R_2(T)—the quantity that actually changes sign, and the one the figures resolve at finite BB. The M\mathcal{M}/string-theory top-down dual instead yields a series in B/Tc2B/T_c^2 in which a parity-odd linear-in-BB term does occur—admitted by the genuinely parity-odd type-IIB axion and warped five-form background, not by BB itself, and carried on a separate projection—in analogy with [3], referring to (S5) and the discussion following (S27) of the supplemental file, interestingly, the ratio of the O(B)O(B) and O(B2)O(B^2) coefficients can be interpreted as the product of the cube roots respectively of the real and imaginary parts of the periods of the non-compact and compact three-cycles of a deformed conifold and is additionally large NN-suppressed (with this ratio, numerically, working out to a little under ≈1%\approx1\%); the lattice cs2c_s^2-vs-TT data at several BBs are shown to be compatible with this structure (a parity-even quadratic term with a percent-level parity-odd linear admixture) while ensuring a smooth crossover at T0T_0 from lattice EOS data; given that (for simplicity) we work in the Nf=3N_f=3 chiral limit inn the string/M\mathcal{M}-theory dual, we obtain a first-order Hawking-Page phase transition therein (the double magnetic slope crossover is common to both though). All definitions and the weak-field power counting, including the (−1)nB(-1)^{n_B} parity rule and the retained even-in-BB structures, are given in Sec. I of the Supplemental Material [2].

A kinetic-theory identity. The organizing tool is an identity that is exact within the single-mode variational quasi-particle truncation, where the bulk and shear channels share the same scalar weight, so common phase-space and Landau-level factors cancel in the ratio (this cancellation is a property of the truncation, not a theorem of magnetized QCD transport) for how the ratio responds to the field. Writing X=ζ,ηX=\zeta,\eta in the standard variational form X=NX2/ΓXX=\mathcal{N}_X^2/\Gamma_X—a source projection NX\mathcal{N}_X squared over a collision rate ΓX\Gamma_X—and differentiating the logarithm of the ratio gives

∂ln⁡(ζ/η)∂(qfB)2=2(νζ−νη)−(γζ−γη),(3)\frac{\partial\ln(\zeta/\eta)}{\partial(q_fB)^2}=2(\nu_\zeta-\nu_\eta)-(\gamma_\zeta-\gamma_\eta), \tag*{(3)}

where νX\nu_X and γX\gamma_X are the field log-derivatives of the source and of the rate, respectively. Equation (3) is an identity that holds in both phases, and all the Physics is in which quasi-particles supply NX\mathcal{N}_X and ΓX\Gamma_X.

Its usefulness is that the two terms have transparent and competing physical content. The source difference is controlled by the conformal-breaking factor 13−cs2\frac{1}{3}-c_s^2, since only the bulk source carries it:

2(νζ−νη)=−4χB13−cs2,χB≡∂cs2∂(qfB)2.(4)2(\nu_\zeta-\nu_\eta)=-\frac{4\chi_B}{\frac{1}{3}-c_s^2}, \qquad\chi_B\equiv\frac{\partial c_s^2}{\partial(q_fB)^2}. \tag*{(4)}

Because 13−cs2\frac{1}{3}-c_s^2 is small, it is the driver of the effect. The rate difference γζ−γη\gamma_\zeta-\gamma_\eta acts as the competitor, and the sign of ∂BR\partial_B\mathcal{R} is decided by which of the two wins. A crossover occurs whenever the balance tips. Derivations of (3) and (4), including the cancellations that make the common phase-space factors drop out of the difference, are in Sec. II of the Supplemental Material [2].

The top-down input. The top-down input comes from a UV-complete type IIB string/type IIA mirror/M\mathcal{M}-theory uplift of a QCD-like thermal plasma in the MQGP limit [6, 7] [5]—a holographic model with large (M)N(\mathcal{M})N color (fractional) D3D3-branes, finite intermediate coupling and wrapped flavor D7D7-branes designed to reproduce several realistic QCD-like observables in appropriate regimes in which the magnetic field appears as a supergravity field localized near the black-brane horizon and the flavor sector is realized by wrapped branes. The dual has been checked against lattice and PDG data across glueball [8] and meson spectra [9], the trace anomaly [10] and several transport coefficients [11, 12] [5]. So it has not been (fine) tuned to ζ\zeta.

For the viscosity ratio the relevant statement is due to Eling and Oz [13]: at the two-derivative level ζ/η\zeta/\eta is fixed by the rate at which the internal moduli of the geometry flow along the family of black-brane solutions labeled by temperature, evaluated at the horizon. Because both viscosities are horizon responses with the same normalization, the ratio is free of the overall factors that plague individual transport coefficients—the reason we predict a ratio rather than ζ\zeta or η\eta separately.

A concern specific to a UV-complete construction is that the dual contains higher-derivative O(R4)O(R^4) corrections, which could in principle modify the Eling-Oz (EO) prescription itself rather than merely deform the background. We show that they do not, at the order worked: the higher-derivative corrections to the horizon entropy (Wald) are suppressed by an additional power of 1/N1/N relative to the terms retained, so evaluating the isotropic two-derivative formula on the O(R4)\mathcal{O}(R^4)-corrected background is the consistent truncation rather than an assumption. A projection (∥\parallel vs. ⊥\perp) anisotropy of the ratio would arise only from an anisotropic bulk metric; here the geometry is isotropic, the field being a constant on the flavor D6 world-volume whose only leading backreaction is a shift of the single scalar horizon datum rh/0(B)r_{h/0}(B) (the horizon radius for T>TcT>T_c, the IR cutoff for T<TcT<T_c). Since ζ\zeta and η\eta depend on BB only through rh/0(B)r_{h/0}(B), the ratio is a function of that one scalar. The residual metric anisotropy is bounded by its fractional size: solving the linearized color Einstein equation with the anisotropic (ℓ=2\ell=2) D6-brane energy-momentum tensor,

δgx3x3gx3x3=(ΔL,D6,ℓ=2−1)x3x3(GTSUGRA−1)x3x3,\frac{\delta g_{x^3x^3}}{g_{x^3x^3}}=\frac{\left(\Delta^{-1}_{L,D6,\ell=2}\right)_{x^3x^3}}{\left(\mathcal{G}^{-1}_{\mathrm{TSUGRA}}\right)_{x^3x^3}},

References

  1. [1]J. Hernandez and P. Kovtun, Relativistic magnetohydrodynamics, JHEP 05 (2017) 001 [arXiv:1703.08757 [hep-th]].DOI
  2. [2]See Supplemental Material for the observable definitions and weak-field power counting (Sec. I), the derivation of the kinetic-theory identity (Sec. II), the \(\mathcal{M}\)-theory construction and the \(O(R^4)\) analysis (Sec. III), the confined-phase pion-gas endpoints (Sec. IV), and the lattice construction (Sec. V), which includes Refs. [5, 7, 14–23, 32].
  3. [3]J. M. Maldacena and D. Martelli, “The unwarped, resolved, deformed conifold: fibranes and the baryonic branch of the Klebanov-Strassler theory,” JHEP 01 (2010) 104, arXiv:0906.0591 [hep-th].DOI
  4. [4]The angular regularization factor C_g, used in [5], is defined as: 1/(ε₁⁹ε₂⁴) { −3640ε₁⁴ε₂²(C_θ₁z − C_yz) −1960ε₁²ε₂⁴(C_θ₁z − C_yz) log(ε₂) +96ε₁⁶(272C_θ₁z − 210C_yz) −2374525C_θ₁zε₂⁴ } ≡ C_g ≡ finite, C_MN being constants of integration appearing in the solutions to the EOMs for the O(R⁴) metric corrections and ε₁,₂ are angular regulators: θ₁,₂ → 0 respectively as ε₁,₂ → 0; (x, y, z) being the toroidal analogs of (φ₁, φ₂, ψ) for the construction of delocalized SYZ type IIA mirror. S. S. Kushwah and A. Misra, Bulk viscosity, speed of sound and contact structure at intermediate coupling, Phys. Rev. D 110, no.12, 126010 (2024) [arXiv:2403.10541 [hep-th]].
  5. [6]M. Mia, K. Dasgupta, C. Gale and S. Jeon, Five Easy Pieces: The Dynamics of Quarks in Strongly Coupled Plasmas, Nucl. Phys. B 839 (2010) 187 [arXiv:0902.1540 [hep-th]] ; Toward Large N Thermal QCD from Dual Gravity: The Heavy Quarkonium Potential, Phys. Rev. D 82 (2010) 026004 [arXiv:1004.0387 [hep-th]].DOI
  6. [7]M. Dhuria and A. Misra, Towards MQGP, JHEP 11 (2013) 001 [arXiv:1306.4339 [hep-th]].DOI
  7. [8]K. Sil, V. Yadav and A. Misra, Top-down holographic G-structure glueball spectroscopy at (N)LO in N and finite coupling, Eur. Phys. J. C 77, no.6, 381 (2017) [arXiv:1703.01306 [hep-th]].DOI
  8. [9]V. Yadav, A. Misra and K. Sil, Delocalized SYZ Mirrors and Confronting Top-Down SU(3)-Structure Holographic Meson Masses at Finite g and N_c with P(article) D(ata) G(rоup) Values, Eur. Phys. J. C 77, no.10, 656 (2017) [arXiv:1707.02818 [hep-th]].DOI
  9. [10]A. Misra and C. Gale, The QCD trace anomaly at strong coupling from M-theory, Eur. Phys. J. C 80, no.7, 620 (2020) [arXiv:1909.04062 [hep-th]].DOI
  10. [11]K. Dasgupta, C. Gale, S. Jeon, A. Misra, M. Richard and K. Sil, Bulk Viscosity at Extreme Limits: From Kinetic Theory to Strings, JHEP 07, 145 (2019) [arXiv:1807.04713 [hep-th]]; On bulk viscosity at weak and strong ’t Hooft couplings, Mod. Phys. Lett. A 35, no.27, 2030012 (2020) [arXiv:1807.07950 [hep-th]].DOI
  11. [12]S. S. Kushwah and A. Misra, T_c, Photoproduction, Paramagnetic Anisotropic Plasma, IR Log-Gravitational-DBI Renormalization, and G_2-Structure Induced (Almost) Contact 3-Structures in Hot Strongly Magnetic MQCD at Intermediate Coupling, Fortsch. Phys. 73, no.12, e70044 (2025) [arXiv:2503.07732 [hep-th]]; K. Sil and A. Misra, New Insights into Properties of Large-N Holographic Thermal QCD at Finite Gauge Coupling at (the Non-Conformal/Next-to) Leading Order in N, Eur. Phys. J. C 76, no.11, 618 (2016) [arXiv:1606.04949 [hep-th]]; M. Dhuria and A. Misra, Transport Coefficients of Black MQGP M3-Branes, Eur. Phys. J. C 75, no.1, 16 (2015) [arXiv:1406.6076 [hep-th]].arxiv.org/abs/2503.07732
  12. [13]C. M. Eling and Y. Oz, JHEP 06 (2011) 007 [arXiv:1103.1657 [hep-th]].DOI
  13. [14]G. S. Bali, F. Bruckmann, G. Endrödi, S. D. Katz and A. Schäfer, The QCD equation of state in background magnetic fields, JHEP 08 (2014) 177 [arXiv:1406.0269 [hep-lat]].DOI
  14. [15]T. Demircik, D. Gallegos, U. Gürsoy, M. Järvinen and R. Lier, Phys. Rev. D 110 (2024) 066007 [arXiv:2402.12224 [hep-th]].arxiv.org/abs/2402.12224
  15. [17]K. Goswami, D. Sahu, J. Dey, R. Sahoo and R. Stock, Anisotropy of magnetized quark matter, Phys. Rev. D 109 (2024) 074012 [arXiv:2310.02711 [hep-ph]].DOI
  16. [18]S. Weinberg, Pion scattering lengths, Phys. Rev. Lett. 17 (1966) 616.DOI
  17. [19]G. S. Bali, B. B. Brandt, G. Endrödi and B. Gläßle, Phys. Rev. D 97 (2018) 034505 [arXiv:1707.05600 [hep-lat]].DOI
  18. [20]H.-T. Ding et al., Phys. Rev. D 104 (2021) 014505 [arXiv:2008.00493 [hep-lat]].DOI
  19. [21]V. Yadav and A. Misra, On M-theory dual of large-N thermal QCD-like theories up to O(R^4) and G-structure classification of underlying non-supersymmetric geometries, Adv. Theor. Math. Phys. 26 (2022) no.10, 3801-3894 [arXiv:2004.07259 [hep-th]].DOI
  20. [22]The O(R^4) on-shell eleven-dimensional action reduces to two-derivative fluxed gravity evaluated on the O(R^4)-corrected metric,
  21. [23]K. Dasgupta, C. Gale, S. Jeon, S. S. Kushwah and A. Misra, in preparation.
  22. [24]S. Jain, R. Samanta and S. P. Trivedi, The Shear Viscosity in Anisotropic Phases, JHEP 10, 028 (2015) [arXiv:1506.01899 [hep-th]].DOI
  23. [25]T. Demircik, C. Ecker and M. Järvinen, Astrophys. J. Lett. 907 (2021) no.2, L37 [arXiv:2009.10731 [astro-ph.HE]].DOI
  24. [26]Z. Cao, K. Hattori, M. Hongo, X.-G. Huang and H. Taya, Gyrohydrodynamics: Relativistic spinful fluid with strong vorticity, JHEP 2022 (2022) 071D01 [arXiv:2205.08051 [hep-th]].DOI
  25. [27]A. Misra, Contact Structure-induced Symplectic Nearly Half-Flat SU(3) Structure in Finite-N Thermal M-Theory, arXiv:2608.26267.
  26. [28]S. Jeon and L. G. Yaffe, Phys. Rev. D 53 (1996) 5799 [hep-ph/9512263].
  27. [29]P. B. Arnold, G. D. Moore and L. G. Yaffe, JHEP 01 (2003) 030 [hep-ph/0209353]; JHEP 05 (2003) 051 [hep-ph/0302165].
  28. [30]H. B. Meyer, Calculation of the Bulk Viscosity in SU(3) Gluodynamics, Phys. Rev. Lett. 100 (2008) 162001, arXiv:0710.3717 [hep-lat]; H. B. Meyer, Phys. Rev. D 76, 101701 (2007), arXiv:0704.1801 [hep-lat].DOI
  29. [31]P. Gerber and H. Leutwyler, Nucl. Phys. B 321 (1989) 387.
  30. [32]T. Demircik, D. Gallegos, U. Gürsoy, M. Järvinen and R. Lier, A novel method for holographic transport, arXiv:2311.00042 [hep-th].arxiv.org/abs/2311.00042
  31. [33]G. Yadav, V. Yadav and A. Misra, “McTEQ (M chiral perturbation theory-compatible deconfinement Temperature and Entanglement Entropy up to terms Quartic in curvature) and FM (Flavor Memory)”, JHEP 10, 220 (2021) [arXiv:2108.05372 [hep-th]].

Paper details

vlnc:W_gfkmw73t
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Open AI (ChatGPT), Claude Opus 4.8 (bare)

View statement
Looking for related papers…
Contents