Phase Space, Microstates, and Macrostates

1. From Determinism to Statistics: Why We Need a New Approach

Your work with classical mechanics and simulations has shown you that, in principle, the evolution of a system of N particles is completely determined by Hamilton's equations, given a set of initial conditions. However, for a macroscopic system where $N\sim 10^{23}$, this deterministic approach becomes both practically impossible and conceptually unenlightening.

Professor Leonard Susskind opens his statistical mechanics course with a discussion of this very point.

The issue is not just the sheer number of particles. Classical many-body systems are often chaotic. An infinitesimal uncertainty in the initial state can lead to exponentially divergent trajectories over time—the "butterfly effect." A deterministic prediction becomes meaningless after just a few collision times.

The following reading provides a striking quantitative illustration of this sensitivity.

The conclusion is inescapable: for macroscopic systems, a deterministic, trajectory-by-trajectory description is a dead end. We must instead adopt a statistical approach, asking not "What is the precise state of the system?" but rather "What is the probability of finding the system in a certain state?" To formalize this, we first need to define the space of all possible states.

2. Phase Space: The Arena of Classical Dynamics

For a classical system, the complete instantaneous state is specified by the generalized coordinates and their conjugate momenta for all particles. Your study of advanced classical mechanics provides the necessary framework here.

For a system of $N$ particles moving in three dimensions, we have:

• $3N$ position coordinates $q_i$
• $3N$ momentum coordinates $p_i$

The phase space, often denoted by $\Gamma$, is the $6N$-dimensional space spanned by these coordinates. A single point in this space, $\mathbf{x}=(q_1,\dots ,q_{3N},p_1,\dots ,p_{3N})$, represents the complete mechanical state of the entire system at a single instant in time. As the system evolves according to Hamilton's equations, this point traces a unique trajectory through phase space.

This $6N$-dimensional space is the stage upon which all of classical statistical mechanics unfolds. Every possible configuration and motion of the particles in the system is represented by a single point within it.

3. Microstates and Macrostates

With the concept of phase space established, we can now define the two most crucial terms in our statistical vocabulary.

• A microstate is a single point in phase space. It is the most complete, detailed description possible for a classical system, specifying the position and momentum of every single particle.

• A macrostate is a description of the system using a small number of macroscopic, measurable variables, such as total energy (E), volume (V), number of particles (N), pressure (P), and temperature (T).

The central idea of statistical mechanics is that a single macrostate corresponds to an immense number of different microstates. When we measure the pressure of a gas, we are not measuring the property of one specific microstate. Instead, we are observing a property that is an average over the countless microstates the system rapidly transitions through, all of which are consistent with the observed macrostate (e.g., all having the same total energy).

The relationship is illustrated below. Different microscopic configurations (microstates) of a polymer chain can have the same total energy, thereby belonging to the same macroscopic energy level (macrostate).

4. A Concrete Example: Counting States

To make these abstract definitions tangible, let's analyze a simple "toy model" where we can explicitly count the microstates for each macrostate. The following video provides an excellent, step-by-step walkthrough of this counting process.

Let's summarize the key results from the video for the system of 4 distinguishable particles (A,B,C,D) in 2 compartments:

The number of microstates corresponding to a given macrostate, $\Omega$, is a fundamentally important quantity. It is often called the thermodynamic probability or statistical weight of that macrostate. As you saw in the video, this can be calculated using combinatorics. For a macrostate ($N_1,N_2,\dots ,N_k$) where $N$ distinguishable particles are distributed among $k$ compartments, the number of microstates is given by the multinomial coefficient:

For our two-compartment example, this simplifies to the binomial coefficient $\Omega (n_1,n_2)=\frac{4!}{n_1!n_2!}$. For the (2,2) macrostate, this gives $\frac{4!}{2!2!}=\frac{24}{4}=6$, matching our explicit count.

Conclusion

In this lesson, we have laid the conceptual groundwork for equilibrium statistical mechanics. We have established that a deterministic approach is untenable for macroscopic systems and have defined the essential vocabulary to build a statistical framework.

Key Takeaways:

• Due to extreme sensitivity to initial conditions, a statistical description of many-particle systems is not just a convenience but a necessity.
• Phase space is the $6N$-dimensional space of positions and momenta that contains every possible mechanical state of a classical N-particle system.
• A microstate is a single point in phase space, representing a complete microscopic specification of the system.
• A macrostate is specified by a few macroscopic observables (like E, V, N).
• The core principle connecting these is that a single macrostate corresponds to a vast number of accessible microstates. The quantity $\Omega$, the number of microstates for a given macrostate, is called the thermodynamic probability.

We have so far treated microstates as static configurations. However, a real system is dynamic, constantly evolving from one microstate to another. In our next lesson, we will investigate how an entire collection of microstates, represented as a fluid-like distribution in phase space, evolves in time. This will lead us to Liouville's theorem, a crucial result that describes the conservation of phase-space volume and serves as the mechanical underpinning for the fundamental postulate of statistical mechanics.