Statistical Mechanics — Boltzmann Distribution & Ensembles

Statistical mechanics explains how the huge number of possible microscopic states in a system lead to predictable macroscopic quantities such as energy, entropy, and pressure. The main idea is simple: identify the allowed microstates, assign probabilities that match the physical setup, and average over them.

For many students, the subject starts to click once two ideas are clear. The Boltzmann distribution tells you how probability depends on energy in thermal equilibrium at fixed temperature. Ensembles tell you which probability model matches the constraints of the system.

What Statistical Mechanics Means

A microstate is one complete microscopic configuration of the system. A macrostate is a coarse description such as fixed energy, temperature, volume, or particle number.

Many different microstates can produce the same macrostate. That is why counting states and weighting them correctly matters. Statistical mechanics does not replace mechanics. It gives a workable way to predict systems with far too many particles to track one by one.

When The Boltzmann Distribution Applies

If a system is in thermal equilibrium with a heat reservoir at temperature $T$, the canonical ensemble says that a microstate with energy $E_i$ has probability

$$P_i = \frac{e^{-E_i/(k_B T)}}{Z}$$

where the normalization constant is

$$Z = \sum_j e^{-E_j/(k_B T)}.$$

This is the Boltzmann distribution for discrete microstates. The quantity $Z$, called the partition function, normalizes the probabilities so they add to $1$. Lower-energy states get larger weight, but higher-energy states are still possible.

The condition matters. This formula is not a universal rule for every statistical-mechanics problem. It applies when the system is in equilibrium and can exchange energy with a reservoir, so temperature is fixed.

Which Ensemble Matches The Physical Setup

An ensemble is a probability model for a physical setup. The three standard cases are:

Microcanonical ensemble: fixed energy

Use this for an isolated system with fixed energy, fixed particle number, and fixed volume. In equilibrium, the accessible microstates are taken to be equally likely.

Canonical ensemble: fixed temperature

Use this when the system can exchange energy with a heat bath, so the temperature is fixed but the system energy can fluctuate. This is where the Boltzmann distribution appears.

Grand canonical ensemble: fixed temperature and chemical potential

Use this when the system can exchange both energy and particles with a reservoir. Temperature and chemical potential are fixed, while the particle number can fluctuate.

The point is simple: ensembles are not interchangeable labels. They encode different physical constraints.

Worked Example: Boltzmann Factor Versus Degeneracy

Suppose a system is in the canonical ensemble at temperature $T$. It has four microstates:

• one ground microstate with energy $0$
• three excited microstates, each with energy $\Delta$

Take $\Delta = 2k_B T$. Then each excited microstate gets Boltzmann weight

$$e^{-\Delta/(k_B T)} = e^{-2} \approx 0.135.$$

The ground microstate has weight $1$, so the partition function is

$$Z = 1 + 3e^{-2} \approx 1 + 3(0.135) \approx 1.406.$$

Now the probabilities are easy to read off from the weights divided by $Z$.

The ground microstate has probability

$$P_{\text{ground}} = \frac{1}{1 + 3e^{-2}} \approx 0.711.$$

Each excited microstate has probability

$$P_{\text{one excited microstate}} = \frac{e^{-2}}{1 + 3e^{-2}} \approx 0.096.$$

But the probability of the excited energy level is the sum over all three excited microstates:

$$P_{\text{excited level}} = \frac{3e^{-2}}{1 + 3e^{-2}} \approx 0.289.$$

This example shows the core competition clearly. Energy pushes probability downward, but multiplicity pushes probability upward. A higher-energy level can still matter if many microstates share it.

The Main Intuition To Keep

The Boltzmann factor rewards low energy. State counting rewards multiplicity. Equilibrium behavior comes from both.

That is why statistical mechanics explains familiar macroscopic patterns. Heat capacities, magnetization, ideal-gas behavior, and phase transitions all depend on how energy and multiplicity compete under the constraints of the system.

Common Mistakes In Statistical Mechanics

Using the Boltzmann distribution without checking the setup

The Boltzmann distribution is for canonical equilibrium. If the system is isolated, driven, or out of equilibrium, you need to stop and check the assumptions.

Confusing an energy level with a microstate

If several microstates have the same energy, you must add their probabilities to get the probability of that energy level. Ignoring degeneracy can give the wrong physical conclusion.

Treating all ensembles as the same idea with different names

The ensemble is part of the problem statement. Fixed energy and fixed temperature are not the same physical condition.

Using Celsius in the exponent

The quantity $k_B T$ uses absolute temperature, so $T$ must be in kelvin.

Where Statistical Mechanics Is Used

Statistical mechanics is used whenever microscopic randomness still leads to reliable large-scale behavior. That includes gases, solids, magnetism, chemical equilibrium, radiation, semiconductors, and many-body quantum systems.

In practice, the subject is often the bridge between thermodynamics and microscopic physics. Thermodynamics tells you what must happen macroscopically. Statistical mechanics helps explain why.

Try A Similar Problem

Keep the same four-state example, but change the gap from $\Delta = 2k_B T$ to $\Delta = k_B T$ or $\Delta = 4k_B T$. Recompute $Z$ and the excited-level probability. That one exercise builds good intuition for when energy dominates and when multiplicity still matters.