Quantum dynamics of cosmological particle production: interacting quantum field theories with matrix product states

This paper employs tensor network methods to demonstrate that self-interactions in 1+1 dimensional scalar and gauge theories suppress gravitational particle production and modify entanglement dynamics during cosmological expansion, while also providing a nontrivial numerical validation of bosonization in curved spacetime.

The Gist

Imagine the universe as a giant, stretching rubber sheet. In the very early moments of the Big Bang, this sheet expanded incredibly fast. According to the laws of physics, this rapid stretching shouldn't just move things around; it should actually create new particles out of the empty space itself. This is known as "cosmological particle production."

For decades, physicists have been able to calculate how this works for "free" particles—particles that don't talk to each other. But the real universe is full of particles that interact, bump into, and influence one another. Figuring out how these interactions change the creation of particles in a stretching universe has been a massive, unsolved puzzle.

This paper is like a high-tech simulation lab where the authors built a digital universe to solve this puzzle. Here is what they did and what they found, explained simply:

The Digital Playground

The authors used a powerful mathematical tool called Tensor Networks (think of it as a super-efficient way to organize a massive spreadsheet of quantum possibilities) to simulate two specific types of "toy universes" in a simplified 1+1 dimensional world (one dimension of space, one of time).

1. The $\lambda\phi^4$ Theory: Imagine a field of springs. If you pull one, it affects its neighbors. This represents a scalar field (like the "inflaton" field thought to drive the Big Bang) that has a self-interaction (the springs are connected).
2. The Schwinger Model: This is a bit more complex. It involves electrons (fermions) and electric fields. However, there's a magical trick in physics called bosonization that says this messy system of electrons and fields is mathematically identical to a single scalar field with a "cosine" wiggly potential. It's like saying a complex orchestra playing a symphony sounds exactly the same as a single flute playing a specific, wavy note.

The authors set up these digital universes to start in a calm state, then suddenly "stretch" the space (simulating the expansion of the universe), and watched what happened.

The Big Discovery: Interactions Act Like a Brake

The most important finding is about what happens when particles interact with each other during this expansion.

• The Free Case (No Interaction): When the authors simulated particles that didn't talk to each other, the stretching space created a lot of new particles. This matched the known mathematical predictions perfectly.
• The Interacting Case: When they turned on the interactions (making the particles "talk" to each other), something surprising happened: The production of new particles dropped significantly.

The Analogy: Imagine a crowd of people in a room.

• Free Case: If everyone is ignoring each other and the room suddenly expands, everyone gets scattered, and new "energy" is created everywhere.
• Interacting Case: If everyone is holding hands (interacting), when the room expands, they resist the stretching. They stick together, and fewer new "scattered" particles are created. The interaction acts like a brake on the creation of matter.

The "Bosonization" Check

One of the most exciting technical achievements was verifying the "bosonization" trick in a curved, expanding universe.

• The authors took the complex electron-and-field model (Schwinger) and the simple scalar field model ($\lambda\phi^4$).
• They expanded both.
• They found that the complex electron model behaved exactly like the simple scalar model with a cosine interaction.
• Why this matters: It proves that this mathematical "translation" trick works even when the universe is stretching and warping, not just in flat, calm space. This gives physicists confidence that they can use the simpler models to study complex real-world scenarios.

The Entanglement Mystery

The paper also looked at entanglement, which is a quantum connection where two particles remain linked no matter how far apart they are.

• In the simple scalar model ($\lambda\phi^4$), the interactions suppressed particle creation, which also meant less entanglement was generated.
• In the Schwinger model, it was more complicated. Even though fewer particles were created, the ones that were created became more strongly connected to each other. It's as if the "brake" on creation was applied, but the few particles that did get made were holding hands even tighter.

Summary

In short, this paper used advanced computer simulations to show that when particles interact with each other, they make it harder for the expanding universe to create new matter. They also proved that a specific mathematical trick (bosonization) works perfectly in these dynamic, expanding environments. This provides a new, non-perturbative (meaning it doesn't rely on approximations) way to understand how the early universe might have generated the matter we see today.

Technical Summary

1. Problem Statement

The paper addresses the challenge of understanding real-time dynamics of interacting quantum fields in curved spacetime, specifically focusing on cosmological particle production during the expansion of the universe.

• Theoretical Gap: Perturbative methods, standard in flat-space QFT, fail in curved spacetime due to the non-uniqueness of vacua, lack of Poincaré invariance, and the breakdown of effective field theories in high-curvature regimes.
• Limitations of Existing Methods: Euclidean lattice field theory (e.g., Lattice QCD) cannot handle real-time evolution due to the "sign problem." Quantum simulation is promising but currently limited by hardware noise and qubit counts.
• Specific Focus: The authors aim to study interacting scalar and gauge theories in a dynamically expanding background (1+1 dimensions), a regime largely unexplored compared to free-field theories. They specifically target the $\lambda\phi^4$ scalar theory and the Schwinger model (1+1D QED).

2. Methodology

The authors employ classical tensor network methods, specifically Matrix Product States (MPS) and Matrix Product Operators (MPO), to simulate the real-time evolution of these systems.

A. Theoretical Framework

• Models:
  1. $\lambda\phi^4$ Theory: A real scalar field with quartic self-interaction.
  2. Schwinger Model: A Dirac fermion coupled to a $U(1)$ gauge field. Crucially, via bosonization, this is mapped to a massive scalar field with a cosine self-interaction potential ($\cos(2\sqrt{\pi}\phi)$).
• Background Geometry: The simulations are performed in a 1+1 dimensional Friedmann-Lemaitre-Robertson-Walker (FLRW) spacetime using conformal coordinates ($g_{\mu\nu} = \Omega(t)^2 \eta_{\mu\nu}$).
• Expansion Profile: A "quench" model is used where the scale factor evolves as $\Omega^2(t) = A + B \tanh[\rho(t - t_0)]$. This allows for well-defined "in" (asymptotic past) and "out" (asymptotic future) vacuum states.
• Lattice Formulation:
  • $\lambda\phi^4$: Discretized on a lattice with open boundary conditions (OBC). The local Hilbert space is truncated to dimension $K=10$.
  • Schwinger Model: Formulated using staggered fermions (Kogut-Susskind). The gauge field is integrated out using Gauss's law, reducing the system to a spin-chain-like Hamiltonian with local constraints.
  • Curved Space Adaptation: The authors derive that in conformal coordinates, the expansion effectively rescales the mass and coupling constants by powers of the scale factor $\Omega(t)$. For the Schwinger model, this is the first derivation of a Hamiltonian lattice gauge theory in a general gravitational background.

B. Numerical Techniques

• Ground State Preparation: The system is initialized in the ground state of the Hamiltonian at $t \to -\infty$ using the Density Matrix Renormalization Group (DMRG) algorithm.
• Real-Time Evolution: Time evolution is performed using the Time-Dependent Variational Principle (TDVP) applied to the MPS.
• Validation: Results are benchmarked against known analytical solutions for free fields in curved spacetime (using Bogoliubov transformations and hypergeometric functions).
• Supplementary Methods: For the Schwinger model, Exact Diagonalization (ED) is used on small systems ($N \le 20$) to extract the low-lying excitation spectrum and validate the MPS results.

3. Key Contributions

1. First Non-Perturbative Lattice Gauge Theory in Curved Spacetime: The authors provide a complete first-principles derivation of the Hamiltonian lattice formulation for the Schwinger model in an expanding universe, demonstrating that the bosonization dictionary holds even in time-dependent gravitational backgrounds.
2. Non-Perturbative Treatment of Interactions: Unlike previous studies focusing on free fields, this work fully incorporates self-interactions ($\lambda\phi^4$ and the cosine potential in the bosonized Schwinger model) without relying on perturbation theory.
3. Discovery of Interaction-Induced Suppression: The central physical finding is that self-interactions suppress gravitational particle production compared to the free-field case.
4. Entanglement Dynamics: The study analyzes the generation of real-space entanglement entropy, revealing a complex interplay between suppressed particle production and enhanced inter-particle correlations in the presence of interactions.

4. Key Results

A. Validation of Bosonization in Curved Spacetime

• In the free limit ($\lambda=0$ for scalars, $m_f=0$ for Schwinger), the numerical results for two-point correlation functions and particle spectra match analytical predictions with high precision.
• This confirms that the equivalence between the massless fermionic Schwinger model and a free massive scalar field persists in a dynamical gravitational background.

B. Suppression of Particle Production

• $\lambda\phi^4$ Theory: As the coupling $\lambda$ increases, the survival probability (probability of remaining in the instantaneous ground state) increases, and the excitation probability decreases. The occupation number per mode $n(k)$ is significantly reduced compared to the free case.
• Schwinger Model: Increasing the fermion mass $m_f$ (which controls the strength of the bosonic cosine interaction) similarly suppresses particle production.
• Mechanism: The self-interaction creates a "stiffness" in the field that resists the excitations caused by the expansion. The ground state is less disturbed by the quench when interactions are strong.

C. Entanglement Dynamics

• $\lambda\phi^4$: Stronger interactions lead to reduced entanglement growth. This aligns with the suppression of particle production; fewer particles mean fewer correlations generated across the bipartition.
• Schwinger Model: The behavior is more nuanced. While particle production is suppressed, the entanglement entropy does not decrease monotonically with interaction strength. At large fermion masses ($m_f/g_f \approx 1/2$), the entanglement production rate is actually higher than at intermediate couplings.
• Interpretation: In the Schwinger model, interactions not only suppress the number of produced particles but also enhance the correlations between the produced excitations. At strong coupling, this enhancement of inter-particle correlations outweighs the suppression of particle number, leading to complex entanglement behavior.

D. Spectral Analysis

• Using excited-state tomography (via ED), the authors confirmed that the produced states in the Schwinger model follow the expected dispersion relations for pseudoscalar and scalar mesons.
• The probability of creating specific multi-particle states matches analytical free-theory predictions at low momentum but deviates at higher momenta, likely due to lattice artifacts and the breakdown of the simple two-particle interpretation at high energies.

5. Significance and Outlook

• Theoretical Impact: This work establishes a robust framework for studying non-perturbative quantum dynamics in curved spacetime, bridging the gap between cosmological particle production and strongly coupled gauge theories.
• Cosmological Relevance: The suppression of particle production by interactions suggests that pre-heating scenarios in the early universe (where inflaton fields decay) might be less efficient than free-field estimates if strong self-interactions are present.
• Future Directions:
  • Higher Dimensions: While 1+1D is a testbed, the authors note that 3+1D simulations will require quantum simulators due to the limitations of classical tensor networks in higher dimensions.
  • Backreaction: The framework can be extended to solve effective Friedmann equations sourced by the quantum expectation value of the energy-momentum tensor (EMT), allowing for the study of semi-classical backreaction.
  • De Sitter Space: Extending the formalism to genuinely expanding geometries (like de Sitter space) is a natural next step for cosmological applications.

In summary, the paper demonstrates that interactions fundamentally alter the quantum response of fields to gravitational expansion, suppressing particle creation but potentially enhancing quantum correlations, a result that can only be captured through non-perturbative, real-time lattice simulations.