Axion-like Particles Overview

• Axion-like particles are pseudo-Nambu–Goldstone bosons that generalize the QCD axion, with decoupled mass and photon-coupling parameters, emerging from string and dark-sector models.
• They facilitate photon–ALP interconversion in magnetic fields, leading to observable spectral modulations and providing targets for precision astrophysical and laboratory experiments.
• Their rich phenomenology spans dark matter, astrophysical signals, and cosmological effects, with upcoming experiments set to explore broad regions of parameter space.

Axion-like particles (ALPs) constitute a theoretically and phenomenologically rich class of pseudo-Nambu–Goldstone bosons arising from spontaneously broken global symmetries, generalizing the QCD axion by relaxing the strict mass-coupling relation enforced by QCD dynamics. ALPs are central in research at the interface of particle physics, cosmology, and astrophysics, as they address a breadth of problems including the strong CP problem, the origin of dark matter, dark radiation, new astrophysical phenomena, and serve as test targets for precision experiments, direct searches, and cosmological probes. Their defining interaction is a dimension-five coupling to two photons, $a F\widetilde F$, permitting photon–ALP interconversion in external electromagnetic fields and enabling a spectrum of observational consequences. The following sections synthesize central aspects of ALP theory and phenomenology, incorporating key developments from effective field theory, string and field-theory axiverse constructions, astrophysical propagation effects, cosmological constraints, and detection strategies.

1. Theoretical Definition and Lagrangian Structure

The defining low-energy effective Lagrangian for a generic axion-like particle is

$$\mathcal{L} = \frac{1}{2}(\partial_\mu a)(\partial^\mu a) - \frac{1}{2} m_a^2 a^2 - \frac{1}{4}g_{a\gamma\gamma}\,a\,F_{\mu\nu}\widetilde{F}^{\mu\nu}$$

where $a(x)$ is a (pseudo-)scalar field, $m_a$ its mass, and $g_{a\gamma\gamma}$ the photon–ALP coupling (in natural units $\hbar = c = 1$). QCD axions present a strict relation between $m_a$ and $g_{a\gamma\gamma}$, namely $$m_{a,\text{QCD}} \sim 6\times10^{-6}\,\mathrm{eV}(10^{12}\,\mathrm{GeV}/f_a)$$ and $g_{a\gamma\gamma} \sim \alpha/(2\pi f_a)$ (modulo anomaly coefficients), but generic ALPs in string theory and other BSM frameworks decouple these parameters (Marsh, 2017, Cicoli, 2013).

Explicit ALP–photon conversions in a magnetic field $$\mathcal{L} = \frac{1}{2}(\partial_\mu a)(\partial^\mu a) - \frac{1}{2} m_a^2 a^2 - \frac{1}{4}g_{a\gamma\gamma}\,a\,F_{\mu\nu}\widetilde{F}^{\mu\nu}$$0 motivate the alternative form $$\mathcal{L} = \frac{1}{2}(\partial_\mu a)(\partial^\mu a) - \frac{1}{2} m_a^2 a^2 - \frac{1}{4}g_{a\gamma\gamma}\,a\,F_{\mu\nu}\widetilde{F}^{\mu\nu}$$1. The full phenomenologically relevant ALP effective field theory can also include derivative couplings to SM fermions and anomalous couplings to $$\mathcal{L} = \frac{1}{2}(\partial_\mu a)(\partial^\mu a) - \frac{1}{2} m_a^2 a^2 - \frac{1}{4}g_{a\gamma\gamma}\,a\,F_{\mu\nu}\widetilde{F}^{\mu\nu}$$2 and $$\mathcal{L} = \frac{1}{2}(\partial_\mu a)(\partial^\mu a) - \frac{1}{2} m_a^2 a^2 - \frac{1}{4}g_{a\gamma\gamma}\,a\,F_{\mu\nu}\widetilde{F}^{\mu\nu}$$3 gauge fields (Arias-Aragón et al., 2022). In the context of extended EFTs, anomaly-matching and discrete gauge symmetries place constraints on viable operator structures, notably in truly axion-like scenarios (DFSZ, KSVZ completions), wherein fermion and gauge couplings are not independent (Arias-Aragón et al., 2022).

2. Origin and Parameter Space

ALPs generically emerge in UV completions including string compactifications and dark-sector field-theory models:

• String axiverse: Compactification of extra-dimensional string theories yields up to $$\mathcal{L} = \frac{1}{2}(\partial_\mu a)(\partial^\mu a) - \frac{1}{2} m_a^2 a^2 - \frac{1}{4}g_{a\gamma\gamma}\,a\,F_{\mu\nu}\widetilde{F}^{\mu\nu}$$4 ALPs with masses and decay constants logarithmically distributed between $$\mathcal{L} = \frac{1}{2}(\partial_\mu a)(\partial^\mu a) - \frac{1}{2} m_a^2 a^2 - \frac{1}{4}g_{a\gamma\gamma}\,a\,F_{\mu\nu}\widetilde{F}^{\mu\nu}$$5 eV and $$\mathcal{L} = \frac{1}{2}(\partial_\mu a)(\partial^\mu a) - \frac{1}{2} m_a^2 a^2 - \frac{1}{4}g_{a\gamma\gamma}\,a\,F_{\mu\nu}\widetilde{F}^{\mu\nu}$$6 eV, $$\mathcal{L} = \frac{1}{2}(\partial_\mu a)(\partial^\mu a) - \frac{1}{2} m_a^2 a^2 - \frac{1}{4}g_{a\gamma\gamma}\,a\,F_{\mu\nu}\widetilde{F}^{\mu\nu}$$7 GeV (Cicoli, 2013, Marsh, 2017, Alexander et al., 2024). Instanton and moduli-fixing control the precise spectrum.
• Field-theory axiverse: QCD-like dark sectors with $$\mathcal{L} = \frac{1}{2}(\partial_\mu a)(\partial^\mu a) - \frac{1}{2} m_a^2 a^2 - \frac{1}{4}g_{a\gamma\gamma}\,a\,F_{\mu\nu}\widetilde{F}^{\mu\nu}$$8 light flavors and $$\mathcal{L} = \frac{1}{2}(\partial_\mu a)(\partial^\mu a) - \frac{1}{2} m_a^2 a^2 - \frac{1}{4}g_{a\gamma\gamma}\,a\,F_{\mu\nu}\widetilde{F}^{\mu\nu}$$9 colors produce $a(x)$0 pseudo-Goldstone bosons ("$a(x)$1-axiverse"), with all axion-like states sharing a common decay constant $a(x)$2 (Alexander et al., 2024).
• Composite and glueball ALPs: Confining Yang–Mills sectors with heavy portal fermions generate "Glueball ALPs" (GALPs), whose effective photon, gluon, and nucleon couplings are suppressed by dimension-eight operators scaling as $a(x)$3 where $a(x)$4 is the dark confinement scale (Carenza et al., 2024).

ALP mass $a(x)$5 and coupling $a(x)$6 scan an extensive, non-universal band in the $a(x)$7 plane:

• $a(x)$8 is typical for dark-matter–motivated ALPs, with $a(x)$9 (Marsh, 2017).
• In string and composite field-theory realizations, dark-matter–eligible ALP bands depend on relic-density production (misalignment, decay of heavy moduli), requiring $m_a$0 for $m_a$1 axions (Cicoli, 2013, Marsh, 2017, Cadamuro, 2012).

3. Astrophysical and Cosmological Signatures

Photon–ALP Mixing and High-Energy Propagation

A core observational consequence arises from photon–ALP interconversion in astrophysical and cosmological magnetic fields, described by a mixing matrix

$m_a$2

with $m_a$3, $m_a$4, $m_a$5 ($m_a$6 is photon/ALP energy) (Batista et al., 2023, Batković et al., 2021). The energy-dependent conversion probability over a domain of length $m_a$7 is

$m_a$8

where $m_a$9 and $g_{a\gamma\gamma}$0. The transition to the strong-mixing regime occurs above a critical energy

$g_{a\gamma\gamma}$1

where the conversion probability becomes quasi energy-independent, and observable spectral features can emerge.

Cosmological Observables

ALPs generated nonthermally via the misalignment mechanism, thermal freeze-in, or heavy particle decay contribute to the energy budget and radiation content of the early universe. Constraints derive from:

• Relic density: For random initial misalignment $g_{a\gamma\gamma}$2, the present-day abundance for a quadratic potential is (Marsh, 2017)

$g_{a\gamma\gamma}$3

• ΔN_eff and dark radiation: Decay or production of relativistic ALPs can contribute to effective neutrino number, $g_{a\gamma\gamma}$4; modulus decay in string models can yield $g_{a\gamma\gamma}$5 (Cicoli, 2013).
• CMB and BBN: ALP decay at $g_{a\gamma\gamma}$6 causes CMB spectral distortions (FIRAS limits $g_{a\gamma\gamma}$7), modifies recombination and primordial element yields, and modifies reionization optical depth (Cadamuro, 2012).
• 21-cm cosmology: Ultralight ALPs ($g_{a\gamma\gamma}$8 eV; "fuzzy DM") can cool baryons via thermal contact after BEC formation, deepening the 21-cm absorption trough, or heat the CMB via resonant $g_{a\gamma\gamma}$9 conversion, affecting the effective $\hbar = c = 1$0 and $\hbar = c = 1$1 signal at $\hbar = c = 1$2 (Das, 2024).

4. Experimental Searches and Astrophysical Constraints

The search for ALPs is multi-pronged:

• Helioscopes (CAST, IAXO): Target solar ALPs ($\hbar = c = 1$3 for $\hbar = c = 1$4 eV). IAXO aims to probe $\hbar = c = 1$5 (Marsh, 2017, Cicoli, 2013).
• Haloscopes (ADMX, HAYSTAC): Probe relic dark-matter ALPs in the $\hbar = c = 1$6eV mass range with sensitivities $\hbar = c = 1$7 GeV$\hbar = c = 1$8 (Marsh, 2017, Cicoli, 2013).
• Light-Shining-Through-Wall (LSW, ALPS-II, OSQAR): Probe sub-meV mass ALPs, with current/future sensitivities $\hbar = c = 1$9 GeV$m_a$0 (Marsh, 2017).
• Gamma-ray observations: Imaging Atmospheric Cherenkov Telescopes (IACTs: H.E.S.S., MAGIC, CTA, LHAASO) and space-based missions (Fermi-LAT, e-ASTROGAM) search for spectral irregularities, increased transparency, oscillatory modulation, and photon deficits in GeV–TeV band AGN and SN spectra; existing data exclude $m_a$1 GeV$m_a$2 for $m_a$3 neV (Batista et al., 2023, Batković et al., 2021, Galanti, 2019, Dominguez et al., 2011, Roncadelli et al., 2017).
• Diffuse backgrounds: Neutrino–gamma-ray connections and anisotropic photon fluxes via Milky Way $m_a$4 conversion probe $m_a$5 GeV$m_a$6 up to $m_a$7 eV (LHAASO, HAWC, CTA) (Vogel et al., 2017).
• Beam dump and collider experiments: MeV–GeV mass ALPs are accessible through rare decays (e.g., mesons, proton bremsstrahlung), displaced decays in beam dump searches (DarkQuest, SHiP), and through associated production (LHC; $m_a$8 or $m_a$9jets) for ALPs with couplings to gauge bosons (Blinov et al., 2021, Chenarani et al., 27 May 2025).

Astrophysical and cosmological constraints supplement laboratory searches. Stellar-cooling, SN 1987A neutrino burst duration, and constraints from extra-galactic and galaxies' X-ray and gamma-ray backgrounds carve out viable parameter space, excluding $g_{a\gamma\gamma}$0 GeV$g_{a\gamma\gamma}$1 for $g_{a\gamma\gamma}$2 keV (globular-cluster stars, CAST), and $g_{a\gamma\gamma}$3 GeV$g_{a\gamma\gamma}$4 at $g_{a\gamma\gamma}$5 MeV (SN 1987A) (Galanti, 2019, Carenza et al., 2024, Cadamuro, 2012).

5. Particle Phenomenology and Astrophysical Impact

ALPs lead to distinctive and rich phenomenology:

• Spectral signatures in high-energy gamma rays: Photon–ALP oscillations in AGN jets, galaxy clusters, and the intergalactic medium can produce quasi-periodic modulations (oscillatory features, "wiggles"), spectral hardening above the expected EBL cut-off, and degree-scale pair-echo halos (Batista et al., 2023, Batković et al., 2021, Galanti, 2019, Roncadelli et al., 2017).
• Cosmological structure formation: Ultralight ALPs ($g_{a\gamma\gamma}$6 eV) act as fuzzy dark matter, suppressing structure below kpc scales, softening galactic cores, and altering minihalo statistics (Marsh, 2017, Das, 2024).
• Supernova ALP bursts: Both Primakoff and coherent magnetic production in hypernovae/strong-magnetic SNe can give delayed $g_{a\gamma\gamma}$7-ray signals (MeV energies, hours–days time delay), providing access to $g_{a\gamma\gamma}$8 in previously untested regions (Caputo et al., 2021).
• Dark-matter phenomenology: ALP self-interactions, assisted mechanisms (kinetic or large misalignment), and parametric/tachyonic resonance lead to the formation of miniclusters and dense substructure, making gravitational searches (microlensing, GW, 21-cm) sensitive to regions not accessible via electromagnetic couplings (Eröncel, 20 Jan 2025).

6. Future Prospects and Open Issues

The landscape of ALP research is characterized by rapid progress in parameter-space coverage and model discrimination:

• Next-generation laboratory searches (IAXO, ALPS-II, DM-Radio, HAYSTAC) will close coverage gaps for photon couplings in the $g_{a\gamma\gamma}$9eV–meV mass window (Cicoli, 2013, Marsh, 2017).
• Gamma-ray observatories (CTA, SWGO, e-ASTROGAM) will reach $$m_{a,\text{QCD}} \sim 6\times10^{-6}\,\mathrm{eV}(10^{12}\,\mathrm{GeV}/f_a)$$0 GeV$$m_{a,\text{QCD}} \sim 6\times10^{-6}\,\mathrm{eV}(10^{12}\,\mathrm{GeV}/f_a)$$1 at very low masses ($$m_{a,\text{QCD}} \sim 6\times10^{-6}\,\mathrm{eV}(10^{12}\,\mathrm{GeV}/f_a)$$2 eV), test strong-mixing–induced spectral plateaus, and measure pair-echo halo profiles (Batista et al., 2023, Roncadelli et al., 2017).
• 21-cm cosmology and CMB spectral distortions are sensitive to ultralight ALP-induced baryon cooling or CMB heating, with forthcoming experiments (PIXIE, PRISM, DAPPER, FARSIDE) poised to definitively test this regime (Das, 2024).
• Differentiating model origin: The field-theory axiverse ("$$m_{a,\text{QCD}} \sim 6\times10^{-6}\,\mathrm{eV}(10^{12}\,\mathrm{GeV}/f_a)$$3-axiverse") constitutes a clear phenomenological target with a tightly packed mass spectrum and enhanced photon coupling scaling with $$m_{a,\text{QCD}} \sim 6\times10^{-6}\,\mathrm{eV}(10^{12}\,\mathrm{GeV}/f_a)$$4 (Alexander et al., 2024); distinguishing such spectra and their bary-verse companions from the string axiverse remains an open challenge.
• Multi-ALP mixing and hidden sectors: Oscillation among multiple ALPs can inconsistently suppress observable signals, implying that constraints based on single-ALP models may require revision if UV completions feature large hidden sectors (Chadha-Day, 2021). Laboratory and astrophysical searches must systematically incorporate this effect, especially for phenomena relying on long-baseline ALP propagation.

7. Synthesis and Significance

Axion-like particles exemplify the intersection of high-energy theory, astrophysical observables, and laboratory experiment, providing technically-computable and experimentally-accessible signals across broad energy, coupling, and cosmic scales. Their origins in well-motivated theories beyond the Standard Model, diverse cosmological and astrophysical consequences, and robust sensitivity to planned experiments ensure that ALP searches will remain a central component of the search for new physics. The evolving synergy between laboratory, astrophysical, and cosmological probes, and the need for careful consideration of realistic UV completions (string axiverse, field-theory axiverse, glueball ALPs) and multi-ALP mixing effects, underscores the importance of continued theoretical and experimental refinement in axion-like particle phenomenology (Batista et al., 2023, Marsh, 2017, Cicoli, 2013, Cadamuro, 2012, Arias-Aragón et al., 2022).