Non-Markovian Electroweak Baryogenesis: Memory Effects on CP-Violating Transport and Gravitational Waves

This paper extends electroweak baryogenesis to a non-Markovian framework using the Schwinger--Keldysh formalism, demonstrating that memory effects significantly alter CP-violating transport dynamics and the resulting baryon asymmetry while also influencing the stochastic gravitational-wave signal.

The Gist

The Big Picture: Why Are We Here?

Imagine the universe right after the Big Bang. It was a hot, chaotic soup of particles. The paper asks a fundamental question: Why is there more matter (us, stars, planets) than antimatter?

According to the laws of physics, the Big Bang should have created equal amounts of both, which would have annihilated each other instantly, leaving nothing but light. But we exist, so something tipped the scales. This paper investigates a specific mechanism called Electroweak Baryogenesis (EWBG) to explain how that tipping happened.

The Standard Story: The "Fast-Forward" Bubble

In the standard version of this story, the universe cooled down and underwent a phase transition, like water turning into ice.

• The Bubble: Imagine bubbles of "new physics" (like ice) forming in the hot soup (water). These bubbles expand, sweeping through the universe.
• The Wall: The edge of the bubble is a "wall." As particles hit this wall, they interact in a way that slightly favors matter over antimatter.
• The Standard Assumption (Markovian): The old theory assumes these particles are like hyper-active ping-pong balls. They bounce around so fast that they instantly forget where they were a split second ago. They react to the wall immediately and forget it immediately. This is called a "Markovian" process—no memory.

The New Idea: The "Heavy Memory" Particles

This paper argues that the "hyper-active" assumption might be wrong in certain scenarios.

The Analogy: The Sticky Floor
Imagine walking across a room.

• Standard View: You are walking on a smooth, slippery floor. You take a step, and you are instantly ready for the next one. Your past steps don't affect your current balance.
• This Paper's View: Imagine the floor is covered in thick, sticky mud. When you take a step, your foot sinks in. It takes time to pull your foot out. Your current movement is heavily influenced by where your foot was a moment ago. You have memory.

In the universe, some particles (mediators of CP violation) might be like that sticky foot. If the "bubble wall" moves at a certain speed, these particles don't have time to "shake off" their previous interactions before hitting the next part of the wall. They carry a memory of the past.

What Happens When Particles Have Memory?

The authors used complex math (Schwinger–Keldysh framework and Kadanoff–Baym equations) to simulate this "sticky" universe. Here is what they found:

1. The "Sweet Spot" Moves
In the standard story, there is a "Goldilocks" speed for the bubble wall: not too slow, not too fast, just right to create the most matter.

• With Memory: Because the particles are "sticky" and slow to react, the bubble wall needs to move slower to be effective. If it moves too fast, the sticky particles can't keep up, and the matter-creation process fails. The "sweet spot" shifts toward slower speeds.

2. The "Non-Monotonic" Surprise
This is the most unique finding.

• Standard Logic: If you make the process "slower" or "less efficient," you get less matter. It's a straight line down.
• Memory Logic: The paper found that for certain speeds, adding a little bit of "memory" (making the particles slightly stickier) actually increases the amount of matter created before it starts decreasing again.
• Analogy: Imagine trying to fill a bucket with a hose. If the hose is too fast, water splashes out. If it's too slow, it takes forever. But with memory, there's a weird middle ground where slowing the water down slightly actually helps the bucket fill up more efficiently for a moment, before slowing down too much and failing again. This "up-then-down" curve cannot be explained by the old "no memory" theory.

3. The "Fingerprint" of Memory
The authors show that you cannot simply tweak the old math to fake this result. The "memory" changes the relationship between different forces in the universe in a specific, correlated way. It's like changing the engine of a car; you can't just paint the hood a different color and call it a new engine. The internal mechanics are genuinely different.

The Ripple Effect: Gravitational Waves

When these bubbles expand and collide, they create ripples in space-time called Gravitational Waves (like sound waves in a pond, but for gravity).

• The Paper's Claim: Because the "sticky" particles change how the bubble wall moves and how long the collision lasts, the resulting gravitational waves might be louder and last longer than the standard theory predicts.
• The Catch: While the signal might be stronger, the paper notes that for many of the "viable" scenarios (where we get the right amount of matter), the signal is still likely too faint for our current or near-future detectors (like LISA) to hear. However, it opens a new window: if we do detect a signal, its specific shape could tell us if the universe had this "memory" effect.

Summary of Constraints

The paper doesn't just say "anything goes." It puts strict limits on this idea:

• The Wall Speed: It must be slow enough for the "sticky" particles to react, but not so slow that the bubble stops moving entirely.
• The Memory Time: The "stickiness" (memory timescale) has a limit. If it's too long, the physics breaks down or the bubble wall becomes unstable.
• The Phase: The specific "phase" (a quantum property) of the particles must be just right to compensate for the memory effects.

The Bottom Line

This paper proposes that the early universe might have been "stickier" than we thought. Particles didn't just bounce off the bubble walls; they lingered and remembered their past interactions. This "memory" changes the rules of how matter was created, shifting the optimal conditions for the universe to exist and potentially leaving a louder, distinct echo in the gravitational waves of today. It suggests that to understand why we are here, we might need to listen to the universe's "echoes" with a new set of ears.

Technical Summary

Technical Summary: Non-Markovian Electroweak Baryogenesis

Problem Statement
Electroweak baryogenesis (EWBG) relies on transport theory to explain the observed matter-antimatter asymmetry during a first-order electroweak phase transition (FOPT). Standard calculations employ a Markovian approximation, assuming that particles mediating CP violation equilibrate rapidly compared to the bubble wall crossing time ($\tau_{rel} \ll \tau_{wall}$). This assumption leads to local source terms and diffusion equations. However, in extended Higgs sectors or dark-sector models where CP-violating species are near threshold ($M \lesssim \mathcal{O}(\text{few}) \times T$) or have suppressed couplings, the in-medium relaxation time $\tau_{rel}$ can become comparable to the wall-crossing time. In this regime, the plasma retains memory of past CP-violating interactions, rendering standard local transport equations insufficient. Previous discussions of this breakdown lacked a systematic treatment within a consistent non-equilibrium framework.

Methodology
The authors develop a non-Markovian extension of EWBG using the Schwinger–Keldysh (SK) real-time effective field theory and the Kadanoff–Baym (KB) hierarchy.

• Framework: Starting from the KB equations for real-time two-point functions, the authors perform a Wigner transformation and a gradient expansion to derive kinetic equations.
• Memory Kernels: By integrating out slowly relaxing degrees of freedom, they derive collision integrals containing temporal nonlocality. The collision term is expressed as a memory integral weighted by a retarded self-energy kernel $K(\tau)$.
• Approximation: In the weakly coupled limit, the spectral function is dominated by a single quasiparticle pole, yielding an exponential memory kernel $K(\tau) = \Gamma_0 e^{-\Gamma_0 \tau}$. The memory timescale is defined as the first moment of this kernel, $\tau_{mem} = 1/\Gamma_0$, though it is treated as an independent phenomenological parameter to account for multi-particle cuts and higher-loop corrections.
• Derivation: The authors analytically evaluate the CP-violating source and diffusion equations using Laplace transforms. They demonstrate that the non-Markovian dynamics effectively replace the microscopic relaxation rates $\Gamma_i$ with an effective rate $\Gamma_i^{eff} = \Gamma_i / (1 + \Gamma_i \tau_{mem})$.

Key Contributions

1. Derivation of Non-Markovian Transport Equations: The paper derives a closed-form expression for the CP-violating source and the associated diffusion equations, showing that memory effects introduce a temporally nonlocal structure that cannot be captured by simple local reparameterizations.
2. Effective Rate Hierarchy: A central finding is that memory effects deform the entire hierarchy of interaction rates (Yukawa, sphaleron, Higgs number violation) simultaneously but non-uniformly. The ratio of effective rates changes as $\Gamma_{ss}^{eff}/\Gamma_{Y}^{eff} = (\Gamma_{ss}/\Gamma_{Y}) \cdot (1+\Gamma_{Y}\tau_{mem})/(1+\Gamma_{ss}\tau_{mem})$. This deformation is shown to be irreducible to a simple rescaling of a single parameter in a Markovian framework.
3. Analytic Source Modification: The CP-violating source $S_{CP}^{NM}$ is derived in closed form, retaining the functional shape of the Markovian source but with a shifted peak velocity and suppressed amplitude controlled by $\Gamma_{eff}$.

Results

• Shift in Optimal Wall Velocity: The presence of memory effects shifts the optimal wall velocity ($v_w^*$) for maximal baryon production to smaller values: $v_w^{*,NM} = L_w \Gamma_0 / (1 + \Gamma_0 \tau_{mem})$. As $\tau_{mem}$ increases, the plasma requires a slower wall to integrate the CP-violating source coherently.
• Non-Monotonic Dependence: For sub-optimal wall velocities ($v_w < v_w^*$), the baryon asymmetry $Y_B$ exhibits a non-monotonic dependence on $\tau_{mem}$. Initially, increasing $\tau_{mem}$ shifts the peak toward the fixed $v_w$, enhancing $Y_B$. Beyond a calculable turnover point, amplitude suppression dominates, causing $Y_B$ to decrease. This behavior is a unique signature of non-Markovian dynamics and cannot be reproduced by Markovian reparameterization.
• Parameter Space Constraints:
  • The viable parameter space in the $(\tau_{mem}, v_w)$ plane contracts and shifts toward smaller velocities as $\tau_{mem}$ increases.
  • Hydrodynamic stability of the deflagration front imposes an upper bound $\tau_{mem} \lesssim 100/T$ (for $v_w \gtrsim 0.05$).
  • In the $(\delta_{CP}, \tau_{mem})$ plane, the observed baryon asymmetry selects a diagonal band where the CP phase $\delta_{CP}$ must grow linearly with $\tau_{mem}$ to compensate for transport suppression, up to the physical limit $\delta_{CP} \leq \pi$.
• Gravitational Wave (GW) Correlation: The authors assess the impact on stochastic GW signals. Memory effects reduce plasma friction, extending the effective source duration and increasing the efficiency of energy conversion into sound waves. This enhances the GW amplitude $\Omega_{GW}$. However, a joint analysis reveals an anti-correlation: larger $\tau_{mem}$ suppresses $Y_B$ (at fixed $v_w$) while enhancing $\Omega_{GW}$. A significant portion of the viable parameter space yields GW signals below the sensitivity of near-future detectors like LISA, though a window exists for $\tau_{mem} \gtrsim 3/T$ and $\alpha \gtrsim 0.08$.

Significance and Claims
The paper establishes non-Markovian transport as a well-motivated extension of the standard EWBG framework. The authors claim that:

• Memory effects introduce a new physical parameter, the memory timescale $\tau_{mem}$, which is testable through baryon asymmetry measurements, collider CP probes, and GW observations.
• The non-Markovian deformation of the transport rate hierarchy is a genuine physical effect, distinct from simple Markovian reparameterization, leading to qualitatively different predictions (e.g., non-monotonic $Y_B$ dependence).
• While theoretical uncertainties regarding hydrodynamic coefficients ($\gamma, \eta$) remain at the order-of-magnitude level for GW predictions, the qualitative trends and the existence of a jointly testable window are robust.
• The results constrain the allowed memory timescale from hydrodynamic stability and the physical range of the CP-violating phase, demonstrating that the non-Markovian regime occupies a finite, phenomenologically viable region of parameter space compatible with current experimental constraints.