Completing the backreaction analysis of inflationary magnetogenesis: the self-consistent anisotropic phase, its primordial spectra, and strong-coupling viability in the Kanno–Soda–Watanabe scenario

In Ratra-type inflationary magnetogenesis with action $-\frac14 f^2(\phi)F_{\mu\nu}F^{\mu\nu}$, the amplified electric field backreacts on the inflationary background. Kanno, Soda and Watanabe (2009) showed that this backreaction does not end inflation but drives a second, weakly anisotropic inflationary phase ($\Sigma/H\sim10^{-2}$, attractor $f\propto a^{-2}$) that suppresses Mpc-scale magnetic fields by $\sim10^{-24}$. Their analysis assumed a coherent anisotropic stress of fixed direction, leading-order shear, and instantaneous reheating; left the curvature and tensor spectra of the anisotropic phase uncomputed; and set aside the strongly coupled negative-$c$ branch. Later literature has settled parts of this list: the leading-order statistical anisotropy of primordial fluctuations in anisotropic inflation is now computed and constrained by Planck to percent level (generically in tension with observations), and the reheating-history dependence is well explored. The open problem is the completion of this analysis: the fully self-consistent nonlinear solution of the anisotropic backreaction phase — including the time evolution of the direction of the anisotropic electromagnetic stress — with the complete scalar, tensor, and residual magnetic spectra derived from it, and the determination of whether magnetogenesis in this framework, with self-consistent backreaction and in the presence of the strong-coupling problem that excludes the negative-$c$ branch, can be observationally viable.

Background

Magnetic fields of order microgauss are measured in galaxies, and Faraday-rotation and gamma-ray observations indicate that even the intergalactic medium hosts weak fields on megaparsec scales. Since the galactic dynamo can only amplify a pre-existing seed, the origin of primordial magnetism is an open problem of cosmology. Inflation is an attractive generator because it stretches vacuum fluctuations to super-horizon scales, but in a spatially flat Friedmann–Robertson–Walker (FRW) universe the Maxwell action is conformally invariant (unchanged under local rescalings of the spacetime metric, so vacuum electromagnetic fluctuations are not stretched to observable scales) and no such amplification occurs. The standard remedy breaks conformal invariance with a time-dependent gauge kinetic function, $S=\int d^4x\sqrt{-g}\left[-\frac14 f^2(\phi)F_{\mu\nu}F^{\mu\nu}\right]$, where $\phi$ is the inflaton; for a power-law coupling $f\propto a^{-2c}$ ($a$ being the scale factor), a scale-invariant magnetic spectrum at the end of inflation requires $c=3/2$.

This mechanism faces a structural obstacle. The coupling that amplifies magnetic modes also amplifies the electric field, whose energy density grows rapidly; moreover, since the effective gauge coupling is $e/f(\phi)$ (with $e$ the ordinary electromagnetic gauge coupling), any branch in which $f$ passes through very small values during inflation renders the effective theory strongly coupled — perturbation theory becomes invalid even though the classical background looks smooth. Demozzi, Mukhanov and Rubinstein (2009) sharpened this into a dilemma: the negative-$c$ branch suffers from strong coupling, while on the positive-$c$ branch the electric energy density grows until backreaction on the inflationary background must be included.

Kanno, Soda and Watanabe (2009) performed that backreaction analysis for $c>0$. Building on their anisotropic-hair solution (Watanabe, Kanno and Soda 2009) — a homogeneous vector-field stress whose anisotropic expansion rate $\Sigma$ settles into a slow-roll attractor (a slowly evolving solution to which the dynamics converge) of ratio $\Sigma/H\sim10^{-2}$ to the Hubble rate $H$ instead of decaying, so the spacetime is of Bianchi I type (spatially flat, but expanding at different rates along different axes, quantified by the shear $\sigma$) — they showed that backreaction does not end inflation. Instead the gauge kinetic function is dragged to the attractor $f\propto a^{-2}$, the expansion becomes weakly anisotropic, and the magnetic spectrum acquires a break at the scale $a_b$ at which backreaction begins; for $c=3/2$ the present-day Mpc-scale amplitude is suppressed by $\sim10^{-24}$ relative to the no-backreaction estimate (under instantaneous reheating — conversion of the inflaton's energy into radiation immediately at the end of inflation — at a first-phase Hubble rate $H_I\sim10^{-6}M_p$, with $M_p$ the Planck mass). Inflation survives, but magnetogenesis is essentially lost, and the model predicts a period of anisotropic expansion whose imprint on primordial fluctuations — a quadrupolar, direction-dependent modulation of the power spectrum, conventionally parametrized by a parameter $g_\ast$ — is in principle observable in the cosmic microwave background (CMB), whose statistical isotropy is confirmed to roughly the percent level.

The source paper's own limitations define the problem addressed here: the anisotropic stress was approximated as a coherent background vector of fixed direction and the shear treated only at leading perturbative order; the curvature and tensor spectra of the anisotropic second phase were not computed; the numerical estimates assumed instantaneous reheating, a single inflationary energy scale, and chaotic inflation (large-field slow-roll inflation driven by a power-law inflaton potential); and the negative-$c$ branch was excluded outright because of the strong-coupling problem, with no ultraviolet completion (a weakly coupled higher-energy theory whose low-energy limit reproduces the effective theory) offered that would permit amplification while remaining weakly coupled.

Problem Statement

Within the framework $S=\int d^4x\sqrt{-g}\left[-\frac14 f^2(\phi)F_{\mu\nu}F^{\mu\nu}\right]$ — slow-roll inflation driven by the inflaton $\phi$, coupled to the electromagnetic field $F_{\mu\nu}$ through a time-dependent gauge kinetic function $f(\phi)$, with $a$ the scale factor — the Kanno–Soda–Watanabe analysis of chaotic (large-field, power-law-potential) inflation showed that once the electric field amplified by this coupling backreacts on the background, the universe enters a second inflationary phase of weakly anisotropic expansion: the anisotropic expansion rate $\Sigma$ (characterizing the difference of expansion rates along the spatial axes, of the same order as the shear) settles at $\Sigma/H\sim10^{-2}$, with $H$ the Hubble rate, and the gauge kinetic function is dragged to the attractor $f\propto a^{-2}$. On Mpc scales the present-day magnetic spectrum is suppressed by $\sim10^{-24}$ relative to the estimate ignoring backreaction. Complete the program this analysis left open, with the following jointly supporting components:

(1) Construct the fully self-consistent solution of this backreaction phase, going beyond the assumptions of a coherent anisotropic stress of fixed direction and leading-order perturbative shear — that is, including the nonlinear evolution of the shear and of the direction of the anisotropic electromagnetic stress, and their feedback on the amplification of magnetic modes across the break scale $a_b$ (the scale at which backreaction begins and the magnetic spectrum acquires a break). Here the direction of the anisotropic electromagnetic stress means the principal axes of the gauge-invariant stress–energy tensor of the amplified electromagnetic field: in the source, the anisotropic stress is modeled by a single homogeneous background vector whose spatial orientation is fixed by a gauge choice, so the physically meaningful relaxation of the fixed-direction assumption concerns this gauge-invariant stress rather than the coordinate direction of that vector.

(2) Derive from this self-consistent background the complete primordial spectra — the curvature and gravitational-wave power spectra, including their quadrupolar statistical anisotropy (parametrized by the conventional parameter $g_\ast$, the fractional amplitude of the direction-dependent modulation of the power spectrum, and its bispectrum counterparts) and the residual magnetic spectrum — in a form that can be confronted with cosmic microwave background (CMB) statistical-isotropy limits and magnetic-field amplitude requirements.

(3) Settle the viability question for this mechanism: determine whether magnetogenesis with self-consistent backreaction in this framework can yield observationally viable magnetic fields on Mpc scales, given the statistical-anisotropy and bispectrum constraints that exclude the generic parameter space, the reheating-history dependence of the estimates, and the strong-coupling problem — the obstruction that excludes the negative-$c$ branch of the power-law coupling $f\propto a^{-2c}$ (on which $f$ passes through very small values during inflation and the effective theory breaks down).

These three components jointly answer one question: is inflationary magnetogenesis with electromagnetic backreaction, in the Kanno–Soda–Watanabe setting, self-consistent and observationally viable? An answer obtained within this setting resolves the question; constructing a weakly coupled extension of the mechanism is optional rather than required.

References

1. S. Kanno, J. Soda, M. Watanabe, 'Cosmological Magnetic Fields from Inflation and Backreaction', J. Cosmol. Astropart. Phys. 2009 (12) (2009) 009. DOI: 10.1088/1475-7516/2009/12/009, https://doi.org/10.1088/1475-7516/2009/12/009 (arXiv:0908.3509)
2. V. Demozzi, V. Mukhanov, H. Rubinstein, 'Magnetic fields from inflation?', J. Cosmol. Astropart. Phys. 2009 (08) (2009) 025. DOI: 10.1088/1475-7516/2009/08/025, https://doi.org/10.1088/1475-7516/2009/08/025 (arXiv:0907.1030)
3. M. Watanabe, S. Kanno, J. Soda, 'Inflationary Universe with Anisotropic Hair', Phys. Rev. Lett. 102 (2009) 191302. DOI: 10.1103/PhysRevLett.102.191302, https://doi.org/10.1103/PhysRevLett.102.191302 (arXiv:0902.2833)
4. M. Watanabe, S. Kanno, J. Soda, 'The Nature of Primordial Fluctuations from Anisotropic Inflation', Prog. Theor. Phys. 123 (2010) 1041. DOI: 10.1143/ptp.123.1041, https://doi.org/10.1143/ptp.123.1041 (arXiv:1003.0056)
5. A. E. Gümrükçüoğlu, B. Himmetoglu, M. Peloso, 'Scalar-Scalar, Scalar-Tensor, and Tensor-Tensor Correlators from Anisotropic Inflation', Phys. Rev. D 81 (2010) 063528. DOI: 10.1103/PhysRevD.81.063528, https://doi.org/10.1103/PhysRevD.81.063528 (arXiv:1001.4088)
6. J. Soda, 'Statistical Anisotropy from Anisotropic Inflation', Class. Quantum Grav. 29 (2012) 083001. DOI: 10.1088/0264-9381/29/8/083001, https://doi.org/10.1088/0264-9381/29/8/083001 (arXiv:1201.6434)
7. J. Ohashi, J. Soda, S. Tsujikawa, 'Observational signatures of anisotropic inflationary models', J. Cosmol. Astropart. Phys. 2013 (12) (2013) 009. DOI: 10.1088/1475-7516/2013/12/009, https://doi.org/10.1088/1475-7516/2013/12/009 (arXiv:1308.4488)
8. N. Bartolo, S. Matarrese, M. Peloso, A. Ricciardone, 'Anisotropic power spectrum and bispectrum in the $f(\phi)F^2$ mechanism', Phys. Rev. D 87 (2013) 023504. DOI: 10.1103/PhysRevD.87.023504, https://doi.org/10.1103/PhysRevD.87.023504 (arXiv:1210.3257)
9. Planck Collaboration (Y. Akrami et al.), 'Planck 2018 results. VII. Isotropy and Statistics of the CMB', Astron. Astrophys. 641 (2020) A7. DOI: 10.1051/0004-6361/201935201, https://doi.org/10.1051/0004-6361/201935201 (arXiv:1906.02552)
10. V. Demozzi, C. Ringeval, 'Reheating constraints in inflationary magnetogenesis', J. Cosmol. Astropart. Phys. 2012 (05) (2012) 009. DOI: 10.1088/1475-7516/2012/05/009, https://doi.org/10.1088/1475-7516/2012/05/009 (arXiv:1202.3022)
11. K. Dimopoulos, A. Ghoshal, T. Papanikolaou, 'Is inflationary magnetogenesis sensitive to the postinflationary history?', Phys. Rev. D 111 (2025) 083550. DOI: 10.1103/PhysRevD.111.083550, https://doi.org/10.1103/PhysRevD.111.083550 (arXiv:2412.10367)
12. R. J. Z. Ferreira, R. K. Jain, M. S. Sloth, 'Inflationary magnetogenesis without the strong coupling problem', J. Cosmol. Astropart. Phys. 2013 (10) (2013) 004. DOI: 10.1088/1475-7516/2013/10/004, https://doi.org/10.1088/1475-7516/2013/10/004 (arXiv:1305.7151)
13. R. J. Z. Ferreira, R. K. Jain, M. S. Sloth, 'Inflationary magnetogenesis without the strong coupling problem II: constraints from CMB anisotropies and B-modes', J. Cosmol. Astropart. Phys. 2014 (06) (2014) 053. DOI: 10.1088/1475-7516/2014/06/053, https://doi.org/10.1088/1475-7516/2014/06/053 (arXiv:1403.5516)
14. D. Nandi, 'Inflationary magnetogenesis: solving the strong coupling and its non-Gaussian signatures', J. Cosmol. Astropart. Phys. 2021 (08) (2021) 039. DOI: 10.1088/1475-7516/2021/08/039, https://doi.org/10.1088/1475-7516/2021/08/039 (arXiv:2103.03159)
15. S. Tripathy, D. Chowdhury, H. V. Ragavendra, R. K. Jain, L. Sriramkumar, 'Circumventing the challenges in the choice of the nonconformal coupling function in inflationary magnetogenesis', Phys. Rev. D 107 (2023) 043501. DOI: 10.1103/PhysRevD.107.043501, https://doi.org/10.1103/PhysRevD.107.043501 (arXiv:2211.05834)
16. S. Pal, D. Maity, T. Q. Do, 'Magnetogenesis from an anisotropic universe', Phys. Rev. D 109 (2024) 083507. DOI: 10.1103/PhysRevD.109.083507, https://doi.org/10.1103/PhysRevD.109.083507 (arXiv:2311.07685)
17. R. Durrer, O. Sobol, S. Vilchinskii, 'Backreaction from gauge fields produced during inflation', Phys. Rev. D 108 (2023) 043540. DOI: 10.1103/PhysRevD.108.043540, https://doi.org/10.1103/PhysRevD.108.043540 (arXiv:2303.04583)
18. T. Fujita, S. Yokoyama, 'Critical constraint on inflationary magnetogenesis', J. Cosmol. Astropart. Phys. 2014 (03) (2014) 013. DOI: 10.1088/1475-7516/2014/03/013, https://doi.org/10.1088/1475-7516/2014/03/013 (arXiv:1402.0596)
19. A. Kandus, K. E. Kunze, C. G. Tsagas, 'Primordial magnetogenesis', Phys. Rep. 505 (2011) 1. DOI: 10.1016/j.physrep.2011.03.001, https://doi.org/10.1016/j.physrep.2011.03.001 (arXiv:1007.3891)