Foundations — Free energy from partition function

Before you can read the parent note Free energy from the partition function, you must own — with a picture — every symbol it throws at you. Nothing below is assumed. We build each piece from the one before it.

1. The stage: microstates

Picture a single coin: it has exactly two microstates, heads or tails. A system of three coins has $2\times 2\times 2=8$ microstates. A gas has an astronomically huge number, but the idea is identical — a microstate is a single "snapshot" from the pile of all possible snapshots.

Why the topic needs this: $Z$ is going to be a sum over microstates. If you don't picture the pile of snapshots first, the summation symbol later is meaningless. See Boltzmann distribution and microstates for the deeper story.

2. Energy of a microstate: $E_i$

In the coin picture, imagine a coin that costs nothing to sit heads-up ($E=0$) but costs a fixed amount $\epsilon$ (Greek letter "epsilon", a small energy) to flip tails-up. Then the tails microstate has $E_i=\epsilon$ and the heads microstate has $E_i=0$.

Why the topic needs it: energy is the currency that decides which microstates are cheap (common) and which are expensive (rare). Everything that follows weighs microstates by their $E_i$.

3. Temperature and the "coldness" $\beta$

Why introduce $\beta$ at all — why not just use $T$? Because every Boltzmann weight (next section) is an exponential of $-\beta E_i$. The energies always show up multiplied by $\beta$, never by $T$ directly. Writing $\beta$ makes those formulas clean and — crucially — lets us differentiate with respect to $\beta$ later to pull energies out of $Z$.

4. The Boltzmann weight $e^{-\beta E_i}$

Here we meet the exponential. Why this tool and not a simpler one?

We want a number that is large for cheap (low-$E$) microstates and small for expensive (high-$E$) ones, and that smoothly interpolates as temperature changes. The function $e^{-x}$ does exactly this: it equals $1$ at $x=0$, shrinks toward $0$ as $x$ grows, and never goes negative (weights must be positive). Feeding it $x=\beta E_i$ gives the Boltzmann weight $e^{-\beta E_i}$.

Why the topic needs it: this single factor encodes the entire preference rule "cold prefers low energy." Sum these factors and you get $Z$.

5. The partition function $Z$: the sum symbol

Why "partition"? Because $Z$ tells you how the total probability gets partitioned (divided up) among the microstates. It is a weighted count: at high temperature $Z$ approaches the raw number of microstates; at low temperature it collapses toward just counting the ground states.

Why the topic needs it: $Z$ is the master quantity. Once you have it, the parent note shows every thermodynamic property drops out by differentiation — no more looping over microstates by hand.

6. Probability $p_i$ and the log $\ln$

Now the logarithm. Why does $\ln$ appear everywhere?

The logarithm answers one question: "to what power must I raise $e$ to get this number?" Its magic property is $\ln (a\times b)=\ln a+\ln b.$ It turns multiplication into addition. This is the whole reason free energy uses $\ln Z$ and not $Z$: when you join two independent systems their partition functions multiply ($Z=Z_1{Z}_2$), but energies and free energies should add. Taking $\ln$ converts the product into a sum, making $F$ extensive (proportional to system size).

Why the topic needs it: $\ln$ is the exact tool that makes $Z$ (multiplicative) into $F$ (additive). Without it, $F$ would double when you'd expect it to double — but $Z$ squares instead. See Legendre transforms in thermodynamics for how $F$ then relates to other potentials.

7. The thermodynamic cast: $U$, $S$, $F$

Why the topic needs all three: the parent note assembles $F=U-TS$ and watches $U$ cancel, leaving the clean result $F=-k_{B}T\ln Z$. You cannot follow that cancellation without knowing what each letter is.

8. The derivative $\partial /\partial \beta$: the slope tool

Why this tool? Look at a single Boltzmann weight: $\frac{\partial }{\partial \beta }{e}^{-\beta E_i}=-E_i{e}^{-\beta E_i}.$ Differentiating pulls the energy $E_i$ down in front. That is precisely the trick the parent note uses to get internal energy: differentiating $\ln Z$ with respect to $\beta$ conjures up the average energy without a separate sum. The derivative is the machine that extracts physics from $Z$.

Prerequisite map

Each arrow says "you need this before that." The whole page flows left-to-right into the boxed result $F=-k_{B}T\ln Z$.

Equipment checklist

Cover the right side; can you answer each before moving on?