A delayed 400 GeV photon from GRB 221009A and implication on the intergalactic magnetic field

The strength of the intergalactic magnetic field (IGMF) is a fundamental but challenging-to-measure parameter that may trace primordial magnetism and cosmic evolution. A proposed method to probe the IGMF uses time-delayed gamma-ray signals from extragalactic TeV transients (GRBs and blazars), where primary TeV gamma rays are absorbed by the diffuse infrared background, produce ultra-relativistic e⁺e⁻ pairs, and generate delayed secondary GeV–TeV photons via inverse Compton scattering on the CMB. The magnetic deflection of pairs dominates the arrival-time delay unless the IGMF is extremely weak, enabling constraints or measurements of the IGMF from delayed emission. Previous GRB studies set limits B_IGMF ≥ 10⁻²¹–10⁻¹⁷ G. GRB 221009A, at low redshift (z = 0.151) with extraordinarily powerful TeV emission, provides an ideal case to search for such cascade emission and probe the IGMF. This work analyzes long-term Fermi-LAT data toward GRB 221009A to identify delayed high-energy photons and assess their origin and implications for the IGMF.

The paper reviews the theoretical basis for using delayed gamma-ray emission to probe the IGMF, originating from TeV photon absorption on the infrared background, pair production, and inverse Compton upscattering of CMB photons. The method was proposed by Plaga and applied to GRB 940217 and other energetic bursts (e.g., GRB 130427A, GRB 190114C), yielding lower limits on B_IGMF in the range 10⁻²¹–10⁻¹⁷ G. It also references constraints from TeV blazar observations. For GRB 221009A, multiple observatories reported exceptional high-energy emission: Fermi-GBM trigger, Fermi-LAT GeV emission starting ≈200 s post-trigger, and LHAASO detections above 10 TeV within ≈2000 s, establishing the context for searching for delayed cascade signatures at later times.

Data analysis: The study focuses on Fermi-LAT observations from T0 + 0.05 to T0 + 250 days (MET 687018224–708609605), in the 500 MeV–1 TeV energy range within 15° of the Swift/UVOT localization (RA = 288.265°, Dec = 19.774°) of GRB 221009A. FRONT+BACK conversion-type data with SOURCE class are used, excluding zenith angles >90° and applying standard quality cuts (DATA_QUAL = 1 && LAT_CONFIG == 1). Spectral analysis: Using Fermitools with P8R3_SOURCE_V3 IRFs, the source model includes Galactic diffuse (gll_iem_v07.fits), isotropic diffuse (iso_P8R3_SOURCE_V3_v1.txt), and all 4FGL incremental catalog sources within 25°. GRB 221009A is modeled as a point source at the UVOT position with a power-law spectrum. Unbinned likelihood fits are performed in three time intervals: 0.05–0.3 d, 0.3–1 d, and 0.3–250 d. SEDs are derived in 10 logarithmic energy bins from 500 MeV to 1 TeV, fixing the spectral index to 2 per bin while fitting normalization. Event characterization: A 397.7 GeV photon at T0 + 33554 s (≈0.4 d) is identified without accompanying GeV photons. Spatially, it aligns with UVOT, LHAASO-WCDA, and VLBA localizations. Background above 100 GeV within 0.5° over 14 years yielded only two events, indicating low background. The gtsrcprob tool gives association probability 0.9999937 (≈4.4σ). The event is ULTRACLEAN class, minimizing cosmic-ray contamination. Analytical estimate: The arrival time of cascade photons is t_arr = max(Δt_TeV, Δt_IR, Δt_IC, Δt_B), with Δt_IR ≈ 10(1+z)(ε_γ/10 TeV)²(n_IR/0.1 cm⁻³)⁻¹ s, Δt_IC ≈ 0.038(1+z)(ε_γ/10 TeV)³ s, and Δt_B ≈ 7×10⁵ (ε_γ/10 TeV)^(−5/4) (B_IGMF/10⁻¹⁷ G)⁻¹ (1+z)^(−16) s, assuming the magnetic coherence length exceeds the IC cooling radius. Interpreting the 400 GeV secondary photon implies a primary energy ε_γ ≈ 12 TeV, leading to an estimated B_IGMF ≈ 4 × 10⁻¹⁷ G for a ≈0.4 d delay. Monte Carlo simulation (cascade): ELMAG 3.03 is used to simulate intergalactic electromagnetic cascades with a turbulent IGMF (RMS 4 × 10⁻¹⁷ G, correlation length 1 Mpc) and standard EBL (Dominguez et al. 2011). The intrinsic spectrum is taken from LHAASO (average within 2000 s), approximated as a power law dN/dE ∝ E⁻²·⁴ between 200 GeV and 7 TeV. Simulations inject 6,000,000 primary gamma rays at z = 0.151 with energies 100 GeV–100 TeV, jet half-opening angle 0.8°, and default ELMAG parameters. Arrival-time-binned cascade spectra are produced for 0.05–0.3 d, 0.3–1 d, and 0.3–250 d. Model comparison and probabilities: Expected cascade and SSC fluxes are compared to LAT SEDs. Expected counts in 100 GeV–1 TeV are obtained by multiplying model fluxes by LAT exposure toward GRB 221009A and integrating, then computing Poisson probabilities for ≥1 detected photon. SSC afterglow predictions are derived from multi-band afterglow modeling (narrow/structured jet scenario) and compared with observed SEDs. Intrinsic spectrum uncertainty: A power-law with exponential cutoff (PLEcut) at 20 TeV is also tested to account for uncertainty beyond 13 TeV, with corresponding cascade simulations. Future observatories: Sensitivity comparisons with H.E.S.S., MAGIC, and CTA (50 h) are made against simulated long-time cascade spectra for B_IGMF = 4 × 10⁻¹⁷ G to assess detectability and parameter constraints.

• A 397.7 GeV photon was detected by Fermi-LAT at T0 + 33554 s (≈0.4 days) from the direction of GRB 221009A, with no accompanying low-energy (GeV) photons.
• The event’s spatial position matches independent localizations (UVOT, LHAASO-WCDA, VLBA). Historical background above 100 GeV within 0.5° over 14 years yielded only two events, implying low background.
• Association probability with GRB 221009A is 0.9999937 (≈4.4σ). The photon is an ULTRACLEAN class event, making cosmic-ray misidentification unlikely.
• SEDs show that 0.05–0.3 d emission is dominated by lower energies, while at 0.3–1 d and 0.3–250 d only the single ≈400 GeV photon is detected without lower-energy counterparts.
• SSC afterglow modeling fits 0.05–0.3 d data but underpredicts ≈400 GeV fluxes at later times by ~3 orders of magnitude.
• In a cascade interpretation, the ≈400 GeV photon corresponds to ~12 TeV primaries. An IGMF strength ≈4 × 10⁻¹⁷ G (coherence length ≈1 Mpc) accounts for the ~0.4 d delay, consistent with limits from TeV blazars.
• Monte Carlo simulations with B_IGMF = 4 × 10⁻¹⁷ G yield Poisson probabilities for Fermi-LAT to detect ≥1 photon above 100 GeV of ≈2.0% in 0.3–1 d and ≈20.5% in 0.3–250 d. The SSC model probabilities are ≈0.6% for both intervals, favoring the cascade by factors of ≈3–30.
• Fermi-LAT upper limits in 0.3–250 d disfavor weak IGMF B_IGMF ≈ 1 × 10⁻¹⁸ G due to excessive predicted ~10 GeV flux, while stronger fields ≳10⁻¹⁶ G reduce >100 GeV flux too much to explain the observation.
• Considering a PLEcut intrinsic spectrum with E_cut = 20 TeV, the cascade still predicts higher >100 GeV flux than SSC; the probability for ≥1 >100 GeV photon in 0.3–250 d is ≈9.6%, and B_IGMF ≤ 1 × 10⁻¹⁸ G remains excluded by LAT limits.

The detection of a delayed ≈400 GeV photon from GRB 221009A, without contemporaneous lower-energy LAT emission, is difficult to reconcile with SSC afterglow expectations, which underpredict the >100 GeV flux at 0.3–1 d and 0.3–250 d by large factors. A cascade scenario, in which ~10–12 TeV primaries are absorbed by the EBL producing e⁺e⁻ pairs that upscatter CMB photons, naturally explains a delay of ~0.4 d if B_IGMF ≈ 4 × 10⁻¹⁷ G. Simulations show that while the detection is a low-probability event within a narrow time window (≈2% for 0.3–1 d), the long-term probability (≈20.5% for 0.3–250 d) is reasonable, and significantly higher than SSC expectations. LAT upper limits at ~10 GeV over long timescales rule out very weak fields (≈10⁻¹⁸ G) that would overproduce GeV flux, and fields ≳10⁻¹⁶ G struggle to yield the observed >100 GeV photon, pointing to a preferred B_IGMF near 4 × 10⁻¹⁷ G. This value aligns with independent constraints from TeV blazars. Although the conclusion is based on a single high-energy photon and limited exposure, the combined temporal, spectral, and probabilistic evidence favors the cascade origin and provides a promising measurement of the IGMF strength.

This study reports the most energetic GRB photon detected by Fermi-LAT to date, a ≈400 GeV photon arriving ~0.4 days after GRB 221009A, with no accompanying low-energy emission. Modeling indicates that SSC afterglow emission cannot account for this event, whereas an intergalactic cascade from ~12 TeV primaries can, implying an intergalactic magnetic field strength B_IGMF ≈ 4 × 10⁻¹⁷ G (for a 1 Mpc coherence length). Monte Carlo simulations quantify higher detection probabilities for the cascade scenario than SSC and, together with LAT upper limits, disfavor both very weak (≈10⁻¹⁸ G) and stronger (≳10⁻¹⁶ G) fields. Future deep observations with ground-based Cherenkov telescopes (H.E.S.S., MAGIC, CTA) and next-generation space missions (e.g., VLAST) will enable more precise spectral measurements, distinguish cascade versus SSC emission, and refine or confirm the IGMF measurement.

• The main conclusion is supported by a single late-time high-energy photon, making statistical significance limited despite a high association probability.
• The event is a low-probability occurrence within short time windows, and broader confirmation requires additional detections.
• The intrinsic GRB spectrum beyond 13 TeV is uncertain; assumptions (PL or PLEcut with E_cut = 20 TeV) affect cascade predictions.
• The IGMF estimate assumes a coherence scale of 1 Mpc and a specific EBL model; different assumptions could shift the inferred B_IGMF.
• LAT sensitivity and exposure at >100 GeV are limited; upper limits constrain models but reduce contemporaneous low-energy context around the 400 GeV event.