Spontaneity — second law; entropy ΔS: Visual walkthrough

Step 1 — Draw the "arrangements" of a tiny world

WHAT. Before any formula, picture the smallest possible world: 4 boxes and 2 identical energy coins. A "microstate" is one specific way of dropping the coins into boxes. Count them.

WHY. Entropy is a counting idea. If we cannot count arrangements by hand first, the symbol $W$ later means nothing. We start this small so the whole set fits on one figure.

PICTURE. Figure s01 lays out every arrangement. Placing 2 identical coins in 4 boxes gives $W = 6$ distinct microstates (the red grid). Now spread the coins over 6 boxes and the count jumps to $W = 15$ (the violet grid) — more room means more ways.

Figure — Spontaneity — second law; entropy ΔS

$W = 6 add boxes ⟶ W = 15$

$W$ — the raw count of arrangements. Bigger $W$ = more ways for the world to "be" that state.

Step 2 — Why we take the logarithm: $S = k_{B} ln W$

WHAT. Glue two independent little worlds side by side. The combined count is the product $W_{1} \times W_{2}$ (every arrangement of the left pairs with every arrangement of the right). But we want a bookkeeping number that adds. The logarithm is the one tool that turns "multiply" into "add".

WHY this tool and not another? We demand entropy be additive: two separate systems should have total entropy $S_{1} + S_{2}$ . Yet their microstates multiply. The only continuous function $f$ with $f (W_{1} W_{2}) = f (W_{1}) + f (W_{2})$ is the logarithm. So the log is not a guess — it is forced by the additivity requirement.

PICTURE. Figure s02: left world ( $W_{1} = 3$ ) and right world ( $W_{2} = 4$ ) combine to $W = 12$ microstates (the grid of pairings), yet $ln 12 = ln 3 + ln 4$ — heights of the bars literally stack.

See Boltzmann distribution & microstates for how $W$ is computed for real systems.

Step 3 — The heat definition: $d S = \frac{d q _{r e v}}{T}$

WHAT. Counting microstates works for toy boxes but is hopeless for a beaker of water. We need a way to measure a change in entropy from lab quantities: heat and temperature. The link is $d S = \frac{d q _{r e v}}{T}$ .

WHY divide heat by $T$ ? The same dollop of heat scrambles a cold, quiet system far more than a hot, already-jittery one — a whisper in a silent room versus a shout in a stadium. So the entropy gained is heat relative to how energetic the surroundings already are: hence $\div T$ . And we insist on $q_{r e v}$ — the heat along the reversible (gentlest, infinitely slow) path — because only that unique path makes $S$ a state function (depends on start & end alone).

PICTURE. Figure s03: the same heat arrow $d q$ dropped into a cold system (big entropy splash, magenta) versus a hot system (small splash, orange). The splash size is $d q / T$ .

Step 4 — Isothermal shortcut: pull $T$ out

WHAT. If temperature is held constant (isothermal), $T$ is the same for every slice, so it slides out of the sum.

WHY. A constant multiplier can leave an integral. This collapses the messy integral into one clean division — the workhorse formula for melting, boiling, and dissolving (all happen at fixed $T$ ).

PICTURE. Figure s04: many equal-width heat slices, each divided by the same $T$ , adding up to one bar of height $q_{r e v} / T$ .

$Δ S = \int \frac{d q _{r e v}}{T} = \frac{1}{T} \int d q_{r e v} = \frac{q _{r e v}}{T}$

$\frac{1}{T}$ — pulled out because it never changes.
$\int d q_{r e v} = q_{r e v}$ — total reversible heat for the whole step.

Step 5 — Split the world: system + surroundings

WHAT. The second law speaks about the universe, so we cut the world into two pieces that share a wall: the system (our beaker) and the surroundings (everything else, a giant reservoir at fixed $T$ ).

WHY split? A system alone can get more ordered (water freezing → neat lattice, $Δ S_{sy s} < 0$ ). That looks like a second-law violation — until you notice the heat it dumped into the surroundings. To judge spontaneity honestly we must add both books.

PICTURE. Figure s05: a box (system) inside a large ring (surroundings). Heat $q_{sy s}$ crosses the wall. Whatever leaves the system, $- q_{sy s}$ , enters the surroundings.

$Δ S_{s u r r} = \frac{q _{s u r r}}{T} = \frac{- q _{sy s}}{T} = \frac{- Δ H _{sy s}}{T} (const P)$

$q_{s u r r} = - q_{sy s}$ — the wall conserves heat: what one loses, the other gains.
$q_{sy s} = Δ H_{sy s}$ at constant pressure (heat at fixed $P$ is enthalpy).
The minus sign is the whole story: an exothermic system ( $Δ H_{sy s} < 0$ ) makes $Δ S_{s u r r} > 0$ .

Step 6 — Add the books: the second law appears

WHAT. Sum the two entropy changes. The reversible (ideal) path gives exactly zero; any real path leaks extra disorder, so the total can only be $\geq 0$ .

WHY $\geq$ and not $=$ ? Real processes take shortcuts (friction, finite temperature gaps, rushing). Every shortcut generates entropy inside the universe that never comes back. So $Δ S_{u ni v} = 0$ only for the perfect reversible limit; for everything real, $> 0$ .

PICTURE. Figure s06: two stacked bars, $Δ S_{sy s}$ and $Δ S_{s u r r}$ , combining. Green case (sum climbs = spontaneous), grey case (sum flat = equilibrium), red case (sum drops = forbidden).

Step 7 — Every case on one dial

WHAT. Walk all sign combinations of $Δ H$ and $Δ S_{sy s}$ so no scenario surprises you.

WHY. The contract: cover every quadrant. Spontaneity flips depending on which term wins, and temperature $T$ is the referee.

PICTURE. Figure s07: a 2×2 grid of $(Δ H, Δ S_{sy s})$ with each cell's verdict.

$Δ H$	$Δ S_{sy s}$	$Δ S_{u ni v} = Δ S_{sy s} - Δ H / T$	Spontaneous?
$< 0$ (exo)	$> 0$	both push up ⇒ $> 0$	Always
$> 0$ (endo)	$< 0$	both pull down ⇒ $< 0$	Never
$< 0$ (exo)	$< 0$	fight; $- Δ H / T$ wins at low $T$	Only low $T$
$> 0$ (endo)	$> 0$	fight; $Δ S_{sy s}$ wins at high $T$	Only high $T$

Step 8 — The degenerate cases

WHAT. Check the edges: $T \to 0$ , exactly reversible ( $Δ S_{u ni v} = 0$ ), and a perfect crystal.

WHY. Formulas must survive their limits, or they hide a bug.

PICTURE. Figure s08: a number line of $T$ . As $T \to 0^{+}$ , the term $- Δ H / T$ blows up in magnitude, so enthalpy utterly dominates near absolute zero. As $Δ S_{u ni v} \to 0$ we sit at equilibrium (ice ⇌ water at $273 K$ ). At $T = 0$ a perfect crystal has $W = 1$ , so $S = k_{B} ln 1 = 0$ .

$W = 1 \Rightarrow S = k_{B} ln 1 = 0$

One arrangement only ⇒ zero entropy — this is the Third Law of Thermodynamics.
Near $T = 0$ , $- \frac{Δ H}{T} \to \infty$ ⇒ the surroundings term overwhelms everything.

The one-picture summary

Figure s09 compresses the whole chain: count → log → heat/T → split world → add → verdict. Read it left to right and you have re-derived the second law.

Recall Feynman retelling of the walkthrough

Start with a toy world of a few boxes and a couple of energy coins, and just count the ways to arrange them — that count is $W$ . Glue two toy worlds together and the counts multiply, but we want a score that simply adds, so we take the logarithm: that gives $S = k_{B} ln W$ . For a real beaker we can't count coins, so we measure entropy by heat instead: pour heat in gently ( $q_{r e v}$ ) and divide by how hot it already is ( $T$ ) — a whisper in a quiet room changes things more than a shout in a stadium. Now cut the universe into the beaker (system) and everything else (surroundings) sharing a wall; heat leaving one enters the other, which is why the surroundings' entropy is $- Δ H / T$ . Add the two books together and you get $Δ S_{u ni v} = Δ S_{sy s} - Δ H / T$ . If that total climbs, the process happens by itself; if it's flat, you're at equilibrium; if it would drop, it's forbidden. Temperature is the referee that decides which term wins — and that referee is exactly what Gibbs Free Energy ΔG packages up.

Connections

Parent: 2.5.12 Spontaneity — second law; entropy ΔS (Hinglish)
Boltzmann distribution & microstates · Gibbs Free Energy ΔG · Enthalpy ΔH
First Law of Thermodynamics · Reversible vs Irreversible Processes · Third Law of Thermodynamics