Phase Space Concepts and Liouville's Theorem
In classical mechanics, the traditional focus is often on the trajectory of a single particle under the influence of forces. However, when dealing with complex systems comprising $N$ particles—such as an ideal gas or a liquid—tracking the precise position and momentum of every individual particle becomes computationally intractable. To address this, statistical mechanics introduces the geometric concept of phase space and utilizes Liouville's Theorem to reveal the fundamental laws governing the evolution of such systems.
The Geometry of Phase Space
In the framework of classical mechanics, the complete state of a system is determined by the collective positions and momenta of all its constituent particles.
Microstates and Coordinates
Consider a system consisting of $N$ particles, each moving in three-dimensional space. To fully describe the system's state at a specific instant, one must specify:
- The position coordinates of each particle: $\mathbf{q}_1, \mathbf{q}_2, \dots, \mathbf{q}_N$ (totaling $3N$ components).
- The momentum coordinates of each particle: $\mathbf{p}_1, \mathbf{p}_2, \dots, \mathbf{p}_N$ (totaling $3N$ components).
Combining these $6N$ variables defines a single point in a high-dimensional space. This point represents the system's microstate.
Definition and Dimensionality
Phase space is the mathematical space encompassing all possible microstates. For a system of $N$ particles, phase space is a $6N$-dimensional manifold.
- Dimensionality: $6N$ dimensions.
- Points: Each point $\Gamma = (\mathbf{q}_1, \dots, \mathbf{q}_N, \mathbf{p}_1, \dots, \mathbf{p}_N)$ corresponds uniquely to an instantaneous state of the system.
- Trajectories: As time evolves, the system follows Hamiltonian dynamics, causing the point in phase space to move along a specific path known as a phase trajectory.
By transforming complex particle dynamics into the motion of points within this high-dimensional space, we can employ geometric tools to analyze statistical properties.
Dynamics in Phase Space
The evolution of the system is strictly governed by Hamilton's equations:
$$\dot{q}_i = \frac{\partial H}{\partial p_i}, \quad \dot{p}_i = -\frac{\partial H}{\partial q_i}$$
where $H(\mathbf{q}, \mathbf{p}, t)$ is the Hamiltonian, typically representing the total energy of the system.
In phase space, these equations define a velocity field. Every point $\Gamma$ possesses a unique velocity vector $\mathbf{v} = (\dot{\mathbf{q}}, \dot{\mathbf{p}})$. Consequently, if we consider an ensemble of systems with varying initial conditions, their evolution resembles a fluid flowing through phase space.
Liouville's Theorem
Liouville's Theorem is one of the cornerstones of statistical mechanics, describing how probability density evolves within phase space.
Probability Density Function
Since we cannot know the exact state of a macroscopic system, we use a probability density function $\rho(\mathbf{q}, \mathbf{p}, t)$ to describe the likelihood of the system occupying a specific region of phase space. Physically, $\rho d\Gamma$ represents the probability of finding the system within the phase space volume element $d\Gamma$.
Statement of the Theorem
Liouville's Theorem asserts that the probability density $\rho$ remains constant along the phase trajectories as the system evolves.
Mathematically, if $\rho$ is viewed as a scalar field varying with time, its total derivative (or material derivative) is zero:
$$\frac{d\rho}{dt} = \frac{\partial \rho}{\partial t} + \sum_{i=1}^{3N} \left( \frac{\partial \rho}{\partial q_i} \dot{q}_i + \frac{\partial \rho}{\partial p_i} \dot{p}_i \right) = 0$$
Physical Intuition: Incompressibility
A more intuitive interpretation is that the "probability fluid" in phase space is incompressible.
Imagine a region in phase space defined by a set of initial states. As time progresses, this region may undergo significant deformation, stretching, or twisting. However, the volume it occupies in phase space remains strictly constant.
We can prove this using the continuity equation. Probability conservation requires:
$$\frac{\partial \rho}{\partial t} + \nabla \cdot (\rho \mathbf{v}) = 0$$
where $\nabla \cdot$ denotes the divergence in phase space. Expanding the divergence term:
$$\frac{\partial \rho}{\partial t} + \mathbf{v} \cdot \nabla \rho + \rho (\nabla \cdot \mathbf{v}) = 0$$
Noting that $\frac{\partial \rho}{\partial t} + \mathbf{v} \cdot \nabla \rho = \frac{d\rho}{dt}$, the theorem holds if we can show that the divergence of the phase space velocity field is zero ($\nabla \cdot \mathbf{v} = 0$).
Using Hamilton's equations, we calculate the divergence:
$$\nabla \cdot \mathbf{v} = \sum_{i=1}^{3N} \left( \frac{\partial \dot{q}_i}{\partial q_i} + \frac{\partial \dot{p}i}{\partial p_i} \right) = \sum{i=1}^{3N} \left( \frac{\partial^2 H}{\partial q_i \partial p_i} - \frac{\partial^2 H}{\partial p_i \partial q_i} \right) = 0$$
Since mixed partial derivatives commute (assuming $H$ is sufficiently smooth), the divergence is identically zero. This confirms the incompressibility of the phase space flow.
Examples and Implications
Example: The Harmonic Oscillator
Consider a simple harmonic oscillator (a pendulum with small angles). Let $q$ be the angular displacement and $p$ the momentum.
- The Hamiltonian is $H = \frac{p^2}{2m} + \frac{1}{2}kq^2$.
- In the 2D phase space $(q, p)$, lines of constant energy form ellipses.
- A point representing the system moves along these elliptical trajectories.
- If you draw a small circle in phase space, it will move along the ellipse. While nonlinear effects might stretch it into a thin filament over time, its area (the 2D phase space volume) remains invariant.
Significance in Statistical Mechanics
Liouville's Theorem provides the foundation for Ensemble Theory:
- Definition of Equilibrium: If a system's probability distribution $\rho$ is time-independent ($\frac{\partial \rho}{\partial t} = 0$), Liouville's Theorem implies that the distribution must satisfy specific constraints. This is essential for deriving equilibrium distributions, such as the Boltzmann distribution in the canonical ensemble.
- Conservation of Information: The invariance of phase space volume implies that the microscopic evolution is reversible. Information encoded in the initial microstate is neither created nor destroyed during the dynamics.
- The Paradox of Entropy: Although phase space volume is conserved, macroscopic systems appear to evolve toward equilibrium, suggesting an increase in entropy. This leads to the concept of coarse-graining. Macroscopically, the probability distribution becomes increasingly complex and filamented. While the total volume is preserved, the distribution spreads out so finely that macroscopic observables can no longer resolve the fine structure, resulting in an apparent increase in statistical entropy.