Canonical Ensemble

Describes systems in thermal equilibrium with a reservoir at fixed temperature.

Conceptual context

The canonical ensemble describes a system that can exchange energy with a much larger reservoir while keeping its volume and particle number fixed. The reservoir imposes the $T$, but it does not fix the system’s instantaneous energy. The system therefore fluctuates among microstates. The central question is not which energy the system always has, but how its possible energies are distributed and what its $U$ becomes. Each $E_i$ receives a $p_i$ controlled by the thermal scale set by $k_B$ and $T$. The $Z$ organizes those weights and connects microscopic possibilities with thermodynamic observables.

🟢 Essential level

In the canonical ensemble, the system does not keep exactly the same energy because it exchanges small amounts with the reservoir.

The combined situation is nevertheless in thermal equilibrium. The $T$ determines how strongly high-energy microstates are suppressed. At low $T$, most population gathers in states with small $E_i$. As $T$ rises, energetic states gain $p_i$, the distribution broadens, and the $U$ usually increases. The $Z$ acts as a thermal counter: it combines all allowed weights and guarantees normalized probabilities. The $\beta$ carries the same thermal information in inverse form. Fluctuations measured by $\sigma_E^2$ are not accidental noise but a physical consequence of thermal contact. Their size determines the $C_V$ and links microscopic randomness with macroscopic thermal response. This redistribution is the central thermal mechanism.

Josiah Willard Gibbs formalized the statistical ensemble concept in 1902, in his book "Elementary Principles in Statistical Mechanics," providing the general framework that unified decades of prior work by Boltzmann and Maxwell on gases. The name "canonical" is his: he chose that word precisely because he wanted a standard reference distribution, applicable to any system in thermal contact.

🔵 Formal level

Basic formulas

The thermal scale is introduced through:

Inverse temperature

Each microstate receives a normalized canonical weight:

Canonical probability of a microstate

Normalization is built by summing all weights:

Canonical partition function

These three relations fix the thermal scale, the relative weight of each microstate, and the overall normalization connecting them.

Real student questions

Is the system energy constant? No. It fluctuates because of exchange with the reservoir, although the $U$ is stable.

Why do high-energy states not disappear completely? At finite $T$, they retain nonzero $p_i$.

Is the $Z$ just the number of states? No. It is a weighted sum, so low-energy states count more.

Can $U$ be negative? Yes, depending on the chosen energy zero. Differences and trends are what matter.

Why is $C_V$ related to fluctuations? A broader energy distribution responds more strongly to thermal changes.

When should another ensemble be used? Use the microcanonical ensemble for strictly fixed energy and the grand canonical ensemble when particles are exchanged.

Final synthesis

The $T$ determines $U$ by reorganizing microstate probabilities. The $Z$ summarizes that reorganization, $\sigma_E^2$ measures its fluctuations, $C_V$ expresses the response, and $F$ selects equilibrium. The canonical ensemble turns a microscopic spectrum into coherent thermodynamic predictions.