Reduced Density Matrices and Phase-Space Distributions in Thermofield Dynamics

This paper derives formal expressions for reduced 1-particle density matrices and Wigner distributions within the inverse Bogoliubov transformation variant of Thermofield Dynamics by leveraging correlations encoded in the reduced 2-particle density matrix, demonstrating the approach for both harmonic and anharmonic oscillators.

The Gist

The Big Picture: Simulating Heat Without the Heat

Imagine you are a chef trying to bake a perfect cake. To do this, you need to know exactly how the ingredients behave at a specific temperature. In the quantum world (the world of atoms and molecules), things get messy when they get hot. Atoms vibrate, jitter, and interact in complex ways.

Usually, to simulate a hot quantum system, scientists have to run thousands of simulations with different random starting points (like baking 1,000 cakes with slightly different temperatures) and then average the results. This is slow, expensive, and computationally heavy.

Thermofield Dynamics (TFD) is a clever trick invented to solve this. Instead of baking 1,000 cakes, TFD says: "Let's bake just one cake, but imagine it has a magical twin."

The "Magic Twin" Trick (The Duplicated Space)

In this paper, the authors are working with a specific version of TFD called iBT-TFD (Inverse Bogoliubov Transformation). Here is how the "Magic Twin" works:

1. The Real World: You have your actual molecule (let's call it Bob).
2. The Fictitious World: You create a ghost twin for Bob (let's call him Shadow Bob). Shadow Bob doesn't exist in reality; he is a mathematical tool.
3. The Entanglement: You link Bob and Shadow Bob together with a special "thermal rope." This rope is tied in a specific knot that represents the temperature.
   • If it's 0 Kelvin (absolute zero), the rope is slack; they are just normal.
   • If it's Hot, the rope is pulled tight and twisted. This twist encodes all the thermal energy.

The Innovation:
In the old way of doing TFD, you had to start with the rope already tied in a complex knot (the "thermal vacuum"). This made the math for the starting point very hard.

The iBT-TFD method (the focus of this paper) flips the script. It starts with Bob and Shadow Bob standing apart, holding a slack rope (the "bare vacuum"). Then, as the simulation runs, the "knot" (the temperature effect) is moved from the rope into the rules of the game (the Hamiltonian).

• Analogy: Instead of starting with a tangled ball of yarn, you start with a straight string and change the laws of physics so the string naturally twists as you pull it. This makes the simulation much faster and easier to run.

The Problem: Seeing the Real World Through a Foggy Mirror

Here is the catch. Because we moved the "heat knot" into the rules of the game, the math describing Bob and Shadow Bob is now a weird mix of the two. They are entangled.

If you want to know what Bob (the real molecule) is actually doing—like where he is standing or how fast he is moving—you can't just look at him directly. The "Shadow Bob" is interfering with the view.

• The Analogy: Imagine you are trying to take a photo of a dancer (Bob) in a room full of mirrors (Shadow Bob). The mirrors are reflecting the dancer in confusing ways. If you just look at the room, you see a mess of reflections. To get a clear photo of the dancer, you have to mathematically "undo" the reflections.

The paper addresses a major headache: How do we get a clear picture of the real molecule (Bob) when our simulation is a messy mix of Bob and his Shadow?

The Solutions: Three Ways to Clear the Fog

The authors propose three ways to extract the "real" picture from this messy simulation:

1. The "Exact" Method (The Master Key)

They derived a precise mathematical formula to "undo" the mirror effect.

• How it works: You take the messy data from the simulation (which includes both Bob and Shadow Bob) and apply a specific mathematical transformation (the "Bogoliubov back-transformation").
• The Catch: This is like trying to solve a 10,000-piece puzzle. It works perfectly, but it is incredibly slow and requires massive computer power. It's only feasible for small systems.

2. The "Ignore the Twin" Method (The Shortcut)

This is an approximation where you pretend Bob and Shadow Bob aren't talking to each other.

• The Analogy: Imagine you are looking at the dancer through the mirrors, but you just assume the mirrors are clean and ignore the reflections.
• Pros: It's super fast.
• Cons: It works great at the very beginning of the dance, but as time goes on, the "thermal rope" twists tighter, and the reflections become important. Ignoring them leads to a blurry, inaccurate picture later on.

3. The "Moment Expansion" Method (The Sketch Artist)

This is the paper's most creative and practical solution. Instead of trying to reconstruct the whole messy picture, they ask: "What are the most important numbers?"

• The Analogy: Instead of drawing every single hair on the dancer's head, you just measure their height, weight, and how fast they are spinning. You use these few "moments" (statistics) to draw a very good sketch of the dancer.
• How it works: The simulation easily calculates the average position and speed of the mixed-up system. The authors use a mathematical trick (using something called Hermite polynomials, which are like building blocks for shapes) to reconstruct the full shape of the molecule's distribution based only on these few numbers.
• Result: It's fast, accurate, and gives a clear picture of the real molecule without needing to solve the impossible 10,000-piece puzzle.

Why Does This Matter?

This research is a bridge between high-level theoretical physics and practical chemistry.

• For Chemists: It allows them to simulate how molecules behave in hot environments (like inside a living cell or a chemical reactor) with high precision.
• For Technology: Understanding these thermal effects is crucial for designing better solar cells, more efficient batteries, and new materials.

Summary in One Sentence

The authors figured out how to take a fast, easy-to-run quantum simulation that uses a "ghost twin" to represent heat, and they invented a clever mathematical way to filter out the ghost so we can see exactly what the real molecule is doing, even when it's hot and chaotic.

Technical Summary

1. Problem Statement

Thermofield Dynamics (TFD) is a powerful framework for simulating finite-temperature quantum systems by purifying mixed states into a pure state within a doubled Hilbert space ($H \otimes \tilde{H}$). A specific variant, the Inverse Bogoliubov Transformation (iBT-TFD), has gained traction in chemistry and quantum dynamics. In this variant, the thermal state is initialized as a simple vacuum, and the complex thermal effects are transferred into the Hamiltonian (propagator) via an inverse Bogoliubov transformation.

The Core Challenge:
While the iBT-TFD approach simplifies the propagation of the wavefunction (allowing the use of efficient tensor network methods like MCTDH), it complicates the analysis of physical observables.

• In standard TFD, the thermal vacuum is a correlated state, but physical observables are straightforward to extract.
• In iBT-TFD, the wavefunction evolves as a simple vacuum, but the physical density matrix requires a "back-transformation" (Bogoliubov transformation) to recover.
• Specific Issue: Calculating reduced 1-particle distributions (e.g., reduced density matrices or Wigner functions) from the iBT wavefunction is non-trivial. It requires knowledge of the reduced 2-particle density matrix (2-RDM) involving both physical ($z$) and fictitious tilde ($\tilde{z}$) modes. The necessary transformation involves non-linear coordinate mixing and high-dimensional tensor contractions, making exact calculations computationally prohibitive for high-dimensional systems.

2. Methodology

The authors develop a formalism to bridge the gap between the iBT wavefunction and physical observables, proposing both exact and approximate schemes.

A. Theoretical Framework

• Formal Derivation: The paper derives exact expressions for the reduced 1-particle density matrix (1-RDM) and Wigner distributions in the iBT representation.
• The Transformation: The physical 1-RDM, $\rho_z$, is obtained by tracing out the tilde degrees of freedom from a transformed 2-RDM:
  $$\rho_{t|z} = \text{Tr}_{\tilde{z}} \left[ e^{-iG_z(\beta)} \gamma^{\beta}_{t|z\tilde{z}} e^{+iG_z(\beta)} \right]$$
  where $\gamma^{\beta}_{t|z\tilde{z}}$ is the 2-RDM in the iBT picture, and $G_z(\beta)$ is the generator of the thermal Bogoliubov transformation.
• Harmonic Oscillator (HO) Case: For a harmonic oscillator (shifted or unshifted), the authors derive analytical expressions showing that the transformation involves a linear mixing of coordinates ($z$ and $\tilde{z}$) via hyperbolic functions ($\cosh\theta, \sinh\theta$) and displacement parameters. This results in a "two-mode squeezing" effect on the 2-RDM.

B. Numerical Implementation (MCTDH)

The authors apply these formalisms to Multi-Configurational Time-Dependent Hartree (MCTDH) wavefunctions.

• They analyze how to compute the 1-RDM when the physical and tilde modes are either combined in a single subspace or separated.
• They demonstrate that exact evaluation requires contracting 2-dimensional Discrete Variable Representation (DVR) grids, which scales poorly with system size.

C. Approximate Schemes

To overcome the computational cost of exact 2-RDM evaluation, two approximation strategies are introduced:

1. Neglecting Real-Tilde Correlations: Assumes the iBT 2-RDM is a product of independent 1-RDMs for the real and tilde modes. This is exact for harmonic systems at $t=0$ but fails as anharmonicity induces correlations over time.
2. Moment Expansion (Hermite Expansion):
   • Instead of reconstructing the full density matrix, the method reconstructs the probability density (or Wigner function) using a limited set of moments ($\langle z^n p^m \rangle$).
   • Moments in the physical frame are calculated from the iBT wavefunction using a linear transformation of moments (Eq. 40).
   • The distribution is expanded in a basis of weighted Hermite polynomials. The expansion coefficients are determined recursively from the calculated moments.
   • An optimization algorithm adjusts the basis width ($\sigma$) and shift ($\mu$) to minimize negative probability values, ensuring a physically valid distribution.

3. Key Contributions

1. Formal Derivation of 1-RDM in iBT-TFD: The paper provides the first rigorous derivation of how to extract reduced 1-particle densities and Wigner functions from the iBT-TFD wavefunction, explicitly linking them to the 2-RDM of the doubled space.
2. Identification of "Two-Mode Squeezing": The authors clarify that the iBT transformation induces a specific geometric distortion (squeezing) on the phase-space distribution of the real and tilde modes, which must be accounted for to retrieve physical observables.
3. Development of Efficient Approximations:
   • The Moment Expansion method is proposed as a robust, low-cost alternative to exact 2-RDM calculations. It allows for the reconstruction of thermalized distributions using only low-order moments, which are easily calculable in the iBT framework.
   • The Uncorrelated Approximation is validated as a valid short-time approximation for weakly anharmonic systems.
4. Algorithmic Implementation: Detailed algorithms are provided for reconstructing 1D and 2D (Wigner) distributions from moments, including optimization of hyperparameters to ensure positivity.

4. Results

The methods were tested on a thermalized anharmonic oscillator (quartic potential) using ML-MCTDH simulations.

• Exact vs. Approximate:
  • Exact Calculation: Using the full 2-RDM (Eq. 36) is feasible for small systems but computationally expensive.
  • Uncorrelated Approximation: Accurate at very short times ($t < 50$ fs) but deviates significantly as thermal correlations build up, leading to an over-broadened wavepacket.
  • Moment Expansion: Successfully reconstructs the physical density with high fidelity. Even with thermal correlations neglected in the input, the moment expansion captures the correct wavepacket width (variance) much better than the uncorrelated approximation.
• Phase-Space Dynamics:
  • The study visualizes the "two-mode squeezing" effect on the diagonal of the 2-RDM, showing elongation along the $45^\circ$ axis in phase space.
  • The moment expansion accurately reproduces the time evolution of the first moment (position) and the second moment (variance) of the physical mode.
• Temperature Effects: The simulations clearly show the broadening of the reduced 1-particle density as temperature increases from 0 K to 300 K, validating the method's ability to capture thermal effects.

5. Significance

This work is significant for the field of quantum chemistry and quantum dynamics for several reasons:

• Enabling High-Temperature Simulations: It resolves a major bottleneck in applying iBT-TFD to chemical systems. By providing a way to extract physical observables without the prohibitive cost of full 2-RDM reconstruction, it makes high-dimensional thermal dynamics simulations (e.g., for electron transfer, exciton dynamics, or conical intersections) practically feasible.
• Bridging Theory and Practice: It connects the abstract formalism of Thermofield Dynamics with practical numerical tools (MCTDH, Tensor Networks), offering a pathway to study non-adiabatic dynamics at finite temperatures.
• Generalizability: The moment expansion approach is not limited to 1D systems; the authors outline how it can be extended to higher-dimensional phase-space distributions and other quantum dynamics contexts, provided sufficient moments can be computed.
• Validation of iBT-TFD: The results confirm that the iBT-TFD variant, despite its initial complexity in extracting observables, is a viable and efficient alternative to standard TFD for studying complex, anharmonic, thermalized quantum systems.