Antimatter in the universe and Milky Way

The paper describes a theoretical model that predicts a significant population of primordial antimatter in the Milky Way and presents observational data that support this prediction. The model, developed in refs. [16, 17], proposes a mechanism for primordial black hole (PBH) creation based on the Affleck-Dine baryogenesis scenario, where a scalar field χ with nonzero baryonic number plays a key role. During inflation, quantum fluctuations can drive χ to large values along flat directions of its potential, creating bubbles with high baryon number density. These bubbles can later form PBHs with a log-normal mass spectrum, a prediction confirmed by observations of gravitational waves from LIGO-Virgo-KAGRA. The same mechanism can also create compact stellar-like objects, disperse hydrogen and helium clouds, and, if the baryonic chemical potential is negative, antistars.

The paper presents several observational pieces of evidence for galactic antimatter:

1. Cosmic positrons: A strong 511 keV gamma-ray line from the Galactic bulge indicates frequent annihilation of nonrelativistic electron-positron pairs, with a rate of 10 43 s-1. The origin of these positrons is debated, but one possibility is that they are primordial, produced in small antimatter domains in the early universe.

2. Cosmic antinuclei: AMS-02 has observed anti-helium (3He and 4He) and anti-deuterium events. The observed ratio of anti-deuterium to anti-helium is of order unity, which is in huge contrast to standard big bang nucleosynthesis predictions. This suggests that these antinuclei are not secondary products of cosmic ray interactions but are primordial, created inside antimatter bubbles with high baryon density during a non-standard BBN process in non-expanding compact stellar-like objects.

3. Antistars: A catalog of 14 antistar candidates has been identified, not associated with any known gamma-ray source classes and with spectra compatible with baryon-antibaryon annihilation. A supplementary method for identifying antistars via X-ray lines from excited pp̄ and He p̄ atoms is proposed. The unexpectedly high flux of antinuclei and the observation of antistar candidates strongly support the hypothesis that antihelium cosmic rays are created by antistars, possibly through anti-nova outbursts or thermonuclear anti-SN Ia explosions. Additionally, a very powerful gamma-ray burst (GRB 221009A) with a bright MeV emission line, interpreted as electron-positron annihilation, may have emerged from a star-antistar annihilation.

The paper concludes that the DS/DKK mechanism (from refs. [16, 17]) is well verified by data on the BH mass spectrum and on the existence of antimatter in the Galaxy, especially antistars. It states that Antimatter in our backyard is predicted and found and that canonical ΛCDM cosmology is saved by PBHs.

Improvements to AI Systems Based on This Paper:

1. Anomaly Detection in Gamma-Ray and Cosmic-Ray Spectra

• Improvement: Train AI models to identify weak, diffuse spectral lines (e.g., 511 keV) and rare cosmic-ray events (anti-helium, anti-deuterium) by incorporating theoretical priors from non-standard baryogenesis models.

• Capability: The AI can automatically flag candidate antimatter signatures in large survey datasets (e.g., Fermi-LAT, AMS-02) with higher sensitivity than current background-only models, reducing false negatives for primordial antimatter.

2. Multi-Messenger Source Classification

• Improvement: Integrate the model’s predictions (e.g., antistar X-ray lines from excited pp̄/He p̄ atoms, gamma-ray bursts from star-antistar annihilation) into a Bayesian classifier that combines gravitational-wave, gamma-ray, and cosmic-ray data.

• Capability: The AI can distinguish between known astrophysical sources (pulsars, AGN) and exotic candidates (antistars, antimatter clouds) by cross-correlating spectral, temporal, and spatial features, enabling automated discovery of new antimatter objects.

3. Simulation-to-Observation Inference for Primordial Black Hole Mass Spectra

• Improvement: Use the log-normal PBH mass spectrum predicted by the Affleck-Dine mechanism as a prior in AI-driven gravitational-wave parameter estimation (e.g., for LIGO-Virgo-KAGRA events).

• Capability: The AI can infer whether observed black hole mergers are consistent with primordial origin, and simultaneously constrain the baryonic chemical potential (positive vs. negative) from the inferred antimatter fraction.

4. Cosmological Parameter Estimation with Non-Standard BBN

• Improvement: Extend AI-based likelihood-free inference (e.g., normalizing flows) to include the model’s non-standard big bang nucleosynthesis in compact stellar-like objects, which predicts anti-deuterium/anti-helium ratios of order unity.

• Capability: The AI can jointly fit cosmic microwave background, light-element abundances, and cosmic-ray antinuclei data, breaking degeneracies between standard ΛCDM parameters and baryogenesis parameters (e.g., χ field dynamics).

5. Predictive Alert System for Antimatter Annihilation Events

• Improvement: Build a reinforcement-learning agent that uses the model’s predictions of antistar distributions and anti-nova/anti-SN Ia rates to prioritize follow-up observations of gamma-ray bursts (e.g., GRB 221009A-like events) with MeV emission lines.

• Capability: The AI can issue real-time alerts to telescopes for targeted spectroscopy, increasing the chance of capturing the annihilation signature before it fades.

6. Uncertainty-Aware Data Fusion for Rare Event Searches

• Improvement: Implement a hierarchical Bayesian model that treats the existence of antimatter domains as a latent variable, combining weak signals from multiple independent datasets (positron line, antinuclei, antistar candidates) with explicit uncertainty propagation.

• Capability: The AI can quantify the posterior probability of primordial antimatter in the Milky Way as a whole, rather than per-event, providing a robust statistical verdict that current single-channel analyses lack.

7. Generative Modeling of Antimatter Backgrounds

• Improvement: Train a generative adversarial network (GAN) or diffusion model on simulated antimatter bubble distributions (from the DS/DKK mechanism) to produce realistic synthetic maps of gamma-ray and cosmic-ray signatures.

• Capability: The AI can generate mock datasets for detector design (e.g., future AMS-02 upgrades) and for training other AI models to recognize faint antimatter signals against complex astrophysical noise.

8. Automated Theory–Observation Consistency Checker

• Improvement: Develop a natural language processing (NLP) system that reads new astrophysical papers and automatically compares their observational claims (e.g., antistar candidates) against the DS/DKK model’s predictions, flagging concordances or tensions.

• Capability: The AI can serve as a living literature review, updating the model’s validity score in real time and suggesting which new observations would most strongly confirm or falsify the theory.

Theoretical model leading to a significant Galaxy population by primordial antimatter is described. Observational data confirming this "crazy" prediction are presented.