Chemical thermodynamics: free energies and equilibrium

Intuition [Beginner]

Will a reaction happen? That is the central question of chemical thermodynamics. The answer depends on the Gibbs free energy $G$, defined as $G=H-TS$ where $H$ is enthalpy (heat content), $T$ is temperature, and $S$ is entropy (disorder).

At constant temperature and pressure, the change in Gibbs free energy $\Delta G$ tells you everything. If $\Delta G<0$, the reaction is **spontaneous** -- it will proceed on its own. If $\Delta G>0$, the reaction is non-spontaneous. If $\Delta G=0$, the system is at equilibrium -- the forward and reverse reactions occur at the same rate, and the composition stops changing.

The equation $\Delta G=\Delta H-T\Delta S$ shows that spontaneity is a tug-of-war between enthalpy and entropy. A reaction that releases heat ($\Delta H<0$) and increases disorder ($\Delta S>0$) is spontaneous at all temperatures. A reaction that absorbs heat ($\Delta H>0$) and decreases disorder ($\Delta S<0$) is never spontaneous. The interesting cases are the mixed ones, where temperature determines the outcome.

Visual [Beginner]

Picture four scenarios in a $2\times 2$ grid. The rows are $\Delta H<0$ (exothermic) and $\Delta H>0$ (endothermic). The columns are $\Delta S>0$ (more disorder) and $\Delta S<0$ (less disorder).

	$\Delta S>0$	$\Delta S<0$
$\Delta H<0$	Spontaneous at all $T$	Spontaneous at low $T$
$\Delta H>0$	Spontaneous at high $T$	Never spontaneous

The boundary between spontaneous and non-spontaneous occurs at $T=\Delta H/\Delta S$, where $\Delta G=0$.

Worked example [Beginner]

The Haber process: ${\text{N}}_2(g)+3{\text{H}}_2(g)\rightleftharpoons 2{\text{NH}}_3(g)$. Compute $\Delta G^{\circ }$ at 298 K and find $K$.

Step 1. Look up standard Gibbs energies of formation (all in kJ/mol): $\Delta G_f^{\circ }({\text{N}}_2)=0$, $\Delta G_f^{\circ }({\text{H}}_2)=0$, $\Delta G_f^{\circ }({\text{NH}}_3)=-16.45$.

$$\Delta G^{\circ }=2(-16.45)-1(0)-3(0)=-32.90\text{kJ/mol}$$

Step 2. Compute $K$ from $\Delta G^{\circ }=-RT\ln K$:

$$\ln K=-\Delta G^{\circ }/(RT)=32900/(8.314\times 298.15)=13.28$$ $$K=e^{13.28}=5.85\times 10^5$$

$K$ is large, meaning the equilibrium strongly favours ${\text{NH}}_3$ at 298 K.

Step 3. Why does the industrial process run at high temperature ($\approx 700$ K) and high pressure? The reaction is exothermic ($\Delta H^{\circ }=-92.4$ kJ/mol), so raising the temperature shifts equilibrium toward reactants (Le Chatelier). But at 298 K the reaction is too slow. High temperature is needed for kinetics, not thermodynamics. High pressure favours ${\text{NH}}_3$ because the reaction reduces the number of gas molecules (4 mol reactant gas $\to$ 2 mol product gas), and Le Chatelier's principle says increasing pressure shifts equilibrium toward fewer gas moles.

Check your understanding [Beginner]

Formal definition [Intermediate+]

The Gibbs free energy is a thermodynamic potential defined by

$$G:=H-TS=U+PV-TS.$$

Its natural variables are temperature $T$, pressure $P$, and composition (moles $n_i$ of each species). The differential of $G$ for a closed system with composition change is

$$dG=-SdT+VdP+\sum _i{\mu }_{i}d{n}_i,$$

where ${\mu }_i=(\partial G/\partial n_i{)}_{T,P,n_{j\ne i}}$ is the chemical potential of species $i$.

At constant $T$ and $P$, $dG={\sum }_i{\mu }_{i}d{n}_i$. The system minimises $G$ at equilibrium: any spontaneous process at constant $T$, $P$ satisfies $dG\le 0$, with equality only at equilibrium.

For a chemical reaction ${\sum }_i{\nu }_i{A}_i=0$ (products positive, reactants negative), the Gibbs energy of reaction is

$${\Delta }_{r}G=\sum _i{\nu }_i{\mu }_i.$$

At equilibrium, ${\Delta }_{r}G=0$. Under standard-state conditions (1 bar reference pressure, unit activity for all species), the standard reaction Gibbs energy is ${\Delta }_r{G}^{\circ }={\sum }_i{\nu }_i{\mu }_i^{\circ }$.

The equilibrium constant. The reaction quotient is

$$Q=\prod _i{a}_i^{{\nu }_i},$$

where $a_i$ is the activity of species $i$ (approximately the partial pressure for gases, concentration for solutes). The relation between $Q$ and ${\Delta }_{r}G$ is

$${\Delta }_{r}G={\Delta }_r{G}^{\circ }+RT\ln Q.$$

At equilibrium, $Q=K$ and ${\Delta }_{r}G=0$, giving

$$\boxed{{\Delta }_r{G}^{\circ }=-RT\ln K.}$$

Le Chatelier's principle. If a system at equilibrium is disturbed, it responds to partially counteract the disturbance. Formally:

• Temperature change: The van't Hoff equation gives $\frac{d\ln K}{dT}=\frac{{\Delta }_r{H}^{\circ }}{R{T}^2}$. For exothermic reactions ($\Delta H^{\circ }<0$), $K$ decreases with $T$ (equilibrium shifts toward reactants). For endothermic reactions ($\Delta H^{\circ }>0$), $K$ increases with $T$.

• Pressure change: For gas-phase reactions with a change in total moles ($\Delta n_{\text{gas}}\ne 0$), increasing total pressure shifts equilibrium toward the side with fewer gas moles. For the Haber process, $\Delta n_{\text{gas}}=2-4=-2$, so high pressure favours ${\text{NH}}_3$.

• Concentration change: Adding a reactant increases $Q<K$, driving the reaction forward until $Q$ returns to $K$.

Standard-state conventions

Standard-state properties (denoted by the ${}^{\circ }$ superscript) refer to: pure substance at 1 bar pressure, or unit activity in solution, at a specified temperature (usually 298.15 K unless stated otherwise). Standard Gibbs energies of formation ${\Delta }_f{G}^{\circ }$ are tabulated for most compounds; ${\Delta }_f{G}^{\circ }=0$ for elements in their reference form (${\text{N}}_2(g)$, ${\text{H}}_2(g)$, ${\text{O}}_2(g)$, C(graphite), etc.).

Key theorem with proof [Intermediate+]

Theorem (Equilibrium condition from Gibbs energy minimisation). For a closed system at constant $T$ and $P$ undergoing a single chemical reaction ${\sum }_i{\nu }_i{A}_i=0$, the equilibrium composition satisfies

$${\Delta }_{r}G={\Delta }_r{G}^{\circ }+RT\ln K=0,$$

where $K={\prod }_i(a_i^{\text{eq}}{)}^{{\nu }_i}$ is the equilibrium constant.

Proof. The Gibbs energy of the system is $G={\sum }_i{n}_i{\mu }_i$. As the reaction proceeds by an extent $d\xi$, the mole numbers change by $d{n}_i={\nu }_{i}d\xi$, and

$$dG=\sum _i{\mu }_{i}d{n}_i=\left(\sum _i{\nu }_i{\mu }_i\right)d\xi ={\Delta }_{r}Gd\xi .$$

At equilibrium, $G$ is minimised, so $dG/d\xi =0$, which requires ${\Delta }_{r}G=0$.

Now express each chemical potential as ${\mu }_i={\mu }_i^{\circ }+RT\ln a_i$ (the fundamental relation between chemical potential and activity). Then

$${\Delta }_{r}G=\sum _i{\nu }_i{\mu }_i^{\circ }+RT\sum _i{\nu }_i\ln a_i={\Delta }_r{G}^{\circ }+RT\ln \prod _i{a}_i^{{\nu }_i}={\Delta }_r{G}^{\circ }+RT\ln Q.$$

Setting ${\Delta }_{r}G=0$ at equilibrium gives $Q=K$ and ${\Delta }_r{G}^{\circ }=-RT\ln K$. $\square$

Corollary (van't Hoff equation). Differentiating $\ln K=-{\Delta }_r{G}^{\circ }/(RT)$ with respect to $T$ and using the Gibbs-Helmholtz equation $d(\Delta G/T)/dT=-\Delta H/T^2$ gives

$$\frac{d\ln K}{dT}=\frac{{\Delta }_r{H}^{\circ }}{R{T}^2}.$$

For exothermic reactions (${\Delta }_r{H}^{\circ }<0$), $K$ decreases with increasing $T$. For endothermic reactions (${\Delta }_r{H}^{\circ }>0$), $K$ increases with $T$.

Worked example at intermediate level

For the Haber process at 700 K: compute $K$ given $\Delta H^{\circ }=-92.4$ kJ/mol and $\Delta S^{\circ }=-198.8$ J/(mol K).

$$\Delta G_{700}^{\circ }=\Delta H^{\circ }-T\Delta S^{\circ }=-92400-700(-198.8)=-92400+139160=+46760\text{J/mol}$$

$$\ln K_{700}=-46760/(8.314\times 700)=-8.035$$ $$K_{700}=e^{-8.035}=3.24\times 10^{-4}$$

At 298 K, $K=5.85\times 10^5$. At 700 K, $K=3.24\times 10^{-4}$. The equilibrium has shifted enormously toward reactants, confirming Le Chatelier's prediction for this exothermic reaction. The industrial process compensates by using high pressure (150-300 atm) to push the equilibrium back toward ${\text{NH}}_3$.

Exercises [Intermediate+]

The four laws, thermodynamic potentials, and Maxwell relations [Master]

Chemical thermodynamics rests on four laws, each a constraint on the energy and entropy bookkeeping of macroscopic systems. The zeroth law establishes temperature as a well-defined equivalence-class label on states in mutual thermal equilibrium: if $A$ and $C$ are each in equilibrium with $B$, they are in equilibrium with each other. This sounds bookkeeping but is the foundational reason that thermometers work and that "temperature" is a property of a single body rather than a relation between two. The first law writes the differential energy balance for a closed system as $dU=\delta q+\delta w$, where $\delta q$ is heat added and $\delta w$ is work performed on the system; the inexact differentials $\delta q$ and $\delta w$ are path-dependent, but their sum $dU$ is an exact differential because $U$ is a state function. For a simple compressible system the reversible work is $\delta w_{\text{rev}}=-PdV$ and the reversible heat is $\delta q_{\text{rev}}=TdS$, so $dU=TdS-PdV$ in the absence of composition change.

The second law asserts the existence of an extensive state function $S$ (entropy) satisfying $dS\ge \delta q/T$ with equality for reversible processes; equivalently, the entropy of an isolated system never decreases. Clausius's 1865 formulation in terms of the integral $\oint \delta q/T\le 0$ around any cyclic process gives entropy as the path-independent integral $S(B)-S(A)={\int }_A^B\delta q_{\text{rev}}/T$. The third law (Nernst-Planck) fixes the entropy zero: ${\lim }_{T\to 0}S=0$ for a perfect crystalline solid in its ground state. Without the third law, entropy is defined only up to an additive constant; with it, $S(T)={\int }_0^T{C}_P(T^{\prime })/T^{\prime }d{T}^{\prime }$ is computable from heat-capacity measurements, and tabulated standard entropies $S^{\circ }$ are physical quantities.

For a single-component open system with composition change, the Gibbs equation writes the fundamental relation as $$ dU = T,dS - P,dV + \mu,dN, $$ where the chemical potential $\mu =(\partial U/\partial N{)}_{S,V}$ measures the energy cost of adding one particle at fixed entropy and volume. For a multicomponent mixture ${\mu }_i=(\partial U/\partial N_i{)}_{S,V,N_{j\ne i}}$ is one chemical potential per species, and ${\mu }_{i}d{N}_i$ generalises to a sum. The natural variables of $U$ are $(S,V,N_i)$ — these are the variables in which $U$ is a characteristic function in Gibbs's sense, meaning every thermodynamic property of the system can be derived from $U$ as a function of $(S,V,N_i)$ by differentiation, with no additional information needed.

The other three potentials are obtained from $U$ by Legendre transforms, the standard mathematical operation for swapping a natural variable from an extensive coordinate to its conjugate intensive variable while preserving the characteristic-function property. For a convex function $f(x)$, the Legendre transform $f^{\ast }(p)={\sup }_x(px-f(x))$ has the property that if $p=f^{\prime }(x)$, then $f^{\ast }(p)=px-f(x)$; the conjugate variables $x$ and $p$ trade places. Applied to $U(S,V,N)$ we get: $$ \begin{aligned} H(S, P, N) &= U + PV = \text{enthalpy} \quad \text{(Legendre-transform } V \leftrightarrow P), \ F(T, V, N) &= U - TS = \text{Helmholtz free energy} \quad (S \leftrightarrow T), \ G(T, P, N) &= U - TS + PV = \text{Gibbs free energy} \quad (S \leftrightarrow T, V \leftrightarrow P). \end{aligned} $$

Each new potential has its own fundamental differential, obtained by direct manipulation of $dU=TdS-PdV+\mu dN$: $$ \begin{aligned} dH &= T,dS + V,dP + \mu,dN, \ dF &= -S,dT - P,dV + \mu,dN, \ dG &= -S,dT + V,dP + \mu,dN. \end{aligned} $$ Each potential is minimised at equilibrium when its natural variables are held constant: $U$ at constant $(S,V,N)$, $H$ at constant $(S,P,N)$, $F$ at constant $(T,V,N)$, and $G$ at constant $(T,P,N)$. The Gibbs energy $G$ is the central object of chemical thermodynamics because most laboratory and biological reactions occur at constant temperature and pressure — open to the atmosphere, in a thermostatted vessel or a cellular cytoplasm.

Maxwell relations as exterior-derivative consistency. Each fundamental differential is a closed one-form on the thermodynamic state-space manifold. For $G(T,P,N)$ at fixed composition, $dG=-SdT+VdP$ is a one-form $\omega =-SdT+VdP$ on the two-dimensional submanifold parameterised by $(T,P)$. Because $G$ is a state function, the one-form $dG$ is exact: $dG=d(\text{function})$. The exterior derivative of an exact form vanishes ($d^2=0$), so $d(dG)=0$. Expanding: $$ d(dG) = d(-S,dT + V,dP) = -\left(\frac{\partial S}{\partial P}\right)_T dP \wedge dT + \left(\frac{\partial V}{\partial T}\right)_P dT \wedge dP = 0, $$ which gives the Maxwell relation $$ \boxed{\left(\frac{\partial S}{\partial P}\right)_T = -\left(\frac{\partial V}{\partial T}\right)_P.} $$

The same exterior-derivative argument applied to $U,H,F$ produces three more Maxwell relations: $$ \begin{aligned} \left(\frac{\partial T}{\partial V}\right)_S &= -\left(\frac{\partial P}{\partial S}\right)_V \quad &\text{(from } dU\text{)}, \ \left(\frac{\partial T}{\partial P}\right)_S &= +\left(\frac{\partial V}{\partial S}\right)_P \quad &\text{(from } dH\text{)}, \ \left(\frac{\partial S}{\partial V}\right)_T &= +\left(\frac{\partial P}{\partial T}\right)_V \quad &\text{(from } dF\text{)}. \end{aligned} $$ Each Maxwell relation is the closure condition for the corresponding potential's fundamental one-form, and each lets experimentally hard-to-measure quantities (like $(\partial S/\partial P{)}_T$) be replaced by experimentally easy ones (like $(\partial V/\partial T{)}_P$, which is the thermal expansion coefficient times the volume). The chemistry-physics literature treats Maxwell relations as a coordinate-algebra identity; the differential-geometric content is that they are the components of the structure equation $d^2=0$ on the thermodynamic-state-space manifold. Hermann Weyl's 1949 Philosophy of Mathematics and Natural Science and Souriau's 1970 Structure des systèmes dynamiques gave the modern geometric formulations; the chemistry presentation hides the exterior-algebra machinery under the coordinate-partial-derivative notation but the content is identical.

The Gibbs-Duhem equation. Extensivity of $G$ in the variables $(N_1,\dots ,N_n)$ at fixed $(T,P)$ forces, by the Euler theorem for homogeneous functions of degree one, $$ G(T, P, N_1, \ldots, N_n) = \sum_i N_i, \mu_i(T, P, N_1, \ldots, N_n). $$ Differentiating both sides and comparing with $dG=-SdT+VdP+{\sum }_i{\mu }_{i}d{N}_i$ produces $$ S,dT - V,dP + \sum_i N_i,d\mu_i = 0, $$ the Gibbs-Duhem equation. At constant $T$ and $P$ it reduces to ${\sum }_i{N}_{i}d{\mu }_i=0$: in a multicomponent system at equilibrium, the chemical potentials cannot all vary independently. For a binary mixture, this couples $d{\mu }_1$ and $d{\mu }_2$ through the mole fractions, which is the origin of the Duhem-Margules equation relating activity coefficients in binary phase equilibria. The chemistry 11.01.02 pending thermodynamic-potentials cross-link develops the same machinery from the physics side; the present treatment privileges $G$ because chemistry uses constant-$(T,P)$ reaction conditions while general thermodynamics gives equal weight to all four potentials.

Stability conditions from $d^{2}G>0$. Equilibrium requires not only $dG=0$ but $d^{2}G>0$ in every accessible direction (a minimum, not a saddle or maximum). The Hessian of $G$ with respect to its natural extensive variables yields the standard stability criteria: $C_P>0$ (positive heat capacity — the system resists temperature fluctuations); ${\kappa }_T=-(1/V)(\partial V/\partial P{)}_T>0$ (positive isothermal compressibility — the system resists volume fluctuations); $(\partial {\mu }_i/\partial N_i{)}_{T,P,N_{j\ne i}}>0$ (chemical stability — adding a component raises its own chemical potential). Violations of these criteria are not mathematical curiosities: they signal real instabilities. A region of negative compressibility is the spinodal regime inside a phase coexistence loop, where the homogeneous mixture is unstable to spontaneous phase separation. A region of $(\partial {\mu }_1/\partial N_1)<0$ is the chemical analog: the mixture lowers its Gibbs energy by demixing into two distinct phases of different composition. These instability conditions are derived in standard treatments via convexity arguments on $G$; the geometric content is that the equilibrium states form the lower convex envelope of the raw fundamental relation, and points above the envelope are accessible only as metastable states that eventually relax to the envelope through nucleation or spinodal decomposition.

Chemical equilibrium, the reaction quotient, and the van't Hoff equation [Master]

A chemical reaction ${\sum }_i{\nu }_i{A}_i=0$ — with stoichiometric coefficients ${\nu }_i$ positive for products and negative for reactants — proceeds by an extent variable $\xi$ such that $N_i(\xi )=N_i^0+{\nu }_i\xi$. The instantaneous Gibbs energy as a function of $\xi$ at fixed $(T,P)$ is

$$G(\xi )=\sum _i{N}_i(\xi ){\mu }_i=\sum _i(N_i^0+{\nu }_i\xi ){\mu }_i,$$

and the reaction Gibbs energy is its derivative:

$${\Delta }_{r}G\equiv {\left(\frac{\partial G}{\partial \xi }\right)}_{T,P}=\sum _i{\nu }_i{\mu }_i.$$

Equilibrium occurs when ${\Delta }_{r}G=0$ — the Gibbs energy is stationary with respect to reaction progress. The chemical-potential expansion ${\mu }_i={\mu }_i^{\circ }+RT\ln a_i$, where $a_i$ is the activity of species $i$ (approximately the partial pressure for gases at low density, the molar concentration for dilute solutes, or unity for pure solids and pure liquids), yields the central result of reaction thermodynamics:

$${\Delta }_{r}G={\Delta }_r{G}^{\circ }+RT\ln Q,\text{where }Q=\prod _i{a}_i^{{\nu }_i}.$$

$Q$ is the reaction quotient — the product of activities raised to stoichiometric powers, evaluated at whatever composition the system currently has. Setting ${\Delta }_{r}G=0$ at equilibrium with the equilibrium activities $a_i^{\text{eq}}$ gives

$$\boxed{{\Delta }_r{G}^{\circ }=-RT\ln K,K=\prod _i(a_i^{\text{eq}}{)}^{{\nu }_i}=\exp \left(-\frac{{\Delta }_r{G}^{\circ }}{RT}\right).}$$

This is the central insight of chemical equilibrium: the equilibrium constant $K$ is a Boltzmann-factor function of the standard Gibbs energy of reaction. $K$ depends only on temperature (through ${\Delta }_r{G}^{\circ }(T)$); the equilibrium composition that realises $K$ depends on the total amounts and the choice of standard state. Putting these together, the reaction quotient $Q$ serves as a thermodynamic compass: if $Q<K$ the system is reactant-rich relative to equilibrium and ${\Delta }_{r}G<0$, so the reaction proceeds forward; if $Q>K$ the system is product-rich and ${\Delta }_{r}G>0$, so the reaction reverses; at $Q=K$ the system is at equilibrium and ${\Delta }_{r}G=0$.

Le Chatelier's principle as a derivative inequality. The qualitative statement that "a system at equilibrium responds to a perturbation in the direction that partially counteracts the perturbation" is rigorously derived from the second-derivative stability conditions on $G$. Take temperature as the perturbation. The temperature dependence of $\ln K$ follows from differentiating $\ln K=-{\Delta }_r{G}^{\circ }/(RT)$ with respect to $T$ and applying the Gibbs-Helmholtz identity $[\partial (\Delta G/T)/\partial T{]}_P=-\Delta H/T^2$: $$ \boxed{\frac{d \ln K}{dT} = \frac{\Delta_r H^\circ}{RT^2} \quad \text{(van't Hoff equation, 1884)}.} $$ For exothermic reactions (${\Delta }_r{H}^{\circ }<0$), $K$ decreases with increasing $T$ — raising the temperature shifts equilibrium toward reactants. For endothermic reactions (${\Delta }_r{H}^{\circ }>0$), $K$ increases with $T$ — equilibrium shifts toward products. The foundational reason is heat-capacity bookkeeping: treating heat as a stoichiometric participant ($\text{reactants}\rightleftharpoons \text{products}+\text{heat}$ for exothermic), Le Chatelier's qualitative response identifies with the van't Hoff equation's quantitative direction. Pressure perturbations on gas-phase reactions follow the analogous derivative $(d\ln K_x/dP{)}_T=-\Delta n_{\text{gas}}/P$ when $K_x$ is expressed in mole fractions; high pressure favours the side with fewer moles of gas, as the Haber process exploits.

Integrated van't Hoff and the temperature dependence of $K$. If ${\Delta }_r{H}^{\circ }$ is approximately temperature-independent over the interval $[T_1,T_2]$, the van't Hoff equation integrates to $$ \ln!\left(\frac{K(T_2)}{K(T_1)}\right) = -\frac{\Delta_r H^\circ}{R}!\left(\frac{1}{T_2} - \frac{1}{T_1}\right). $$ A plot of $\ln K$ versus $1/T$ is a straight line of slope $-{\Delta }_r{H}^{\circ }/R$ and intercept ${\Delta }_r{S}^{\circ }/R$ (from $\Delta G^{\circ }=\Delta H^{\circ }-T\Delta S^{\circ }$). This is the basis of the experimental determination of $\Delta H^{\circ }$ from equilibrium measurements at multiple temperatures, complementing direct calorimetric measurement and providing the cross-check on tabulated standard enthalpies of reaction.

For a real example, consider the protein folding equilibrium $\text{U}\rightleftharpoons \text{F}$ where U is the unfolded denatured state and F is the native folded state. At physiological temperature, ${\Delta }_r{G}^{\circ }$ is on the order of $-30$ kJ/mol — modestly negative, so K is moderately large but not astronomical. ${\Delta }_r{H}^{\circ }$ is typically around $-100$ to $-300$ kJ/mol (hydrogen bonds and van der Waals contacts in the folded core), while ${\Delta }_r{S}^{\circ }$ is around $-400$ to $-1000$ J/(mol K) (loss of chain conformational entropy upon folding). The two terms nearly cancel — this is the entropy-enthalpy compensation that gives proteins their characteristic "marginal stability" near body temperature. The compensation is not coincidental: a folded protein with much larger $|{\Delta }_r{G}^{\circ }|$ would be too stable to permit the conformational dynamics required for catalysis and signalling, and the cell would not be able to degrade it readily. Evolution selects for $|{\Delta }_r{G}^{\circ }|\sim 30$ kJ/mol precisely because that is the regime where folding is reliable but not irreversible.

Entropy-enthalpy tradeoff in chemistry. The compensation pattern recurs across many reactions: hydrophobic effect in aqueous solutions (positive entropy of water release upon hydrophobic association compensates against unfavourable enthalpy or vice versa, depending on temperature); ligand binding to receptors (enthalpic contacts versus entropic loss of translational and rotational freedom of the bound ligand); ATP hydrolysis (small $|\Delta H|$ but large positive $\Delta S$ from the release of inorganic phosphate). The compensation can be misleading: comparing two ligands on the basis of $\Delta H$ alone is unreliable because $\Delta H$ and $\Delta S$ are correlated through a common physical mechanism. The thermodynamic gold standard is $\Delta G$, which is what actually determines binding affinity. Modern isothermal titration calorimetry (ITC) measures $\Delta H$ directly and computes $\Delta S$ from the simultaneous $K$ measurement; both are reported, but only $\Delta G$ matters for the equilibrium ratio.

The connection to mass-action kinetics. Chemical kinetics 14.08.01 writes forward and reverse rates as $r_{\text{f}}=k_{\text{f}}{\prod }_{\text{reactants}}{c}_i^{|{\nu }_i|}$ and $r_{\text{r}}=k_{\text{r}}{\prod }_{\text{products}}{c}_i^{|{\nu }_i|}$. At equilibrium $r_{\text{f}}=r_{\text{r}}$, so $$ K_c = \frac{k_\text{f}}{k_\text{r}} = \prod_i c_i^{\nu_i}\big|\text{eq}. $$ The thermodynamic equilibrium constant (derivable from free energies) equals the ratio of forward and reverse rate constants — a beautiful synthesis of thermodynamics and kinetics. Importantly, this identifies a thermodynamic ratio with a microscopic-rate ratio without claiming that thermodynamics determines the absolute rate. Catalysts lower the activation energy of both forward and reverse reactions equally, leaving $K$ unchanged but accelerating approach to the same equilibrium. The bridge is the Arrhenius form: $k\text{f} = A_\text{f} e^{-E_\text{a,f}/RT}$,$k_\text{r} = A_\text{r} e^{-E_\text{a,r}/RT}$,with$E_\text{a,f} - E_\text{a,r} = \Delta_r U^\circ \approx \Delta_r H^\circ$. The Arrhenius-Eyring transition-state perspective recovers the van't Hoff equation through the rate-constant ratio, and the central insight is that thermodynamics determines the destination while kinetics determines the path and pace.

Phase equilibria, chemical potential, and the Clausius-Clapeyron equation [Master]

Phase equilibrium is the simplest instance of chemical equilibrium: the "reaction" is a phase transition $\alpha \rightleftharpoons \beta$ of a pure substance, and the equilibrium condition is equality of chemical potentials between coexisting phases. For a single-component system with phases labelled $\alpha$ and $\beta$, the equilibrium condition at temperature $T$ and pressure $P$ is $$ \mu_\alpha(T, P) = \mu_\beta(T, P). $$ At any $(T,P)$ off the phase boundary, one phase has a lower chemical potential and is the stable phase; the other is metastable or absent. The boundary in the $(T,P)$ plane is the locus of points where the two potentials are equal; it is a one-dimensional curve (or, when three phases coexist, a single point — the triple point). The complete picture is encoded in the Gibbs phase rule: $F=C-P+2$, where $C$ is the number of independent components, $P$ the number of phases, and $F$ the number of intensive degrees of freedom. For a one-component, two-phase system $F=1-2+2=1$: the coexistence curve has one degree of freedom, which we conventionally take as $T$.

Derivation of the Clausius-Clapeyron equation. Suppose we move infinitesimally along the coexistence curve from $(T,P)$ to $(T+dT,P+dP)$. The chemical potentials of the two phases must remain equal: $d{\mu }_{\alpha }=d{\mu }_{\beta }$. Using $d\mu =-S_{m}dT+V_{m}dP$ (the molar form of $dG$, valid for a pure substance because $\mu =G_m$): $$ -S_{m,\alpha},dT + V_{m,\alpha},dP = -S_{m,\beta},dT + V_{m,\beta},dP. $$ Solving for the slope of the coexistence curve: $$ \boxed{\frac{dP}{dT}\bigg|\text{coex} = \frac{S{m,\beta} - S_{m,\alpha}}{V_{m,\beta} - V_{m,\alpha}} = \frac{\Delta S_\text{trans}}{\Delta V_\text{trans}} = \frac{\Delta H_\text{trans}}{T,\Delta V_\text{trans}},} $$ where the last step uses $\Delta S_{\text{trans}}=\Delta H_{\text{trans}}/T$ at the reversible phase transition (since $\Delta G_{\text{trans}}=0$ on the coexistence curve). This is the Clapeyron equation — fully exact for any phase transition of a pure substance. For a transition involving a gas with $V_{m,\text{gas}}\gg V_{m,\text{condensed}}$ and ideal-gas behaviour $V_{m,\text{gas}}=RT/P$, we get the Clausius-Clapeyron equation: $$ \frac{d \ln P}{dT} = \frac{\Delta H_\text{vap}}{RT^2}, $$ which integrates (assuming temperature-independent $\Delta H_{\text{vap}}$) to $$ \ln!\left(\frac{P_2}{P_1}\right) = -\frac{\Delta H_\text{vap}}{R}!\left(\frac{1}{T_2} - \frac{1}{T_1}\right). $$ The structural parallel to the van't Hoff equation is exact: both arise from the same Gibbs-Helmholtz identity, both express how an equilibrium "constant" (a vapor pressure or an equilibrium constant $K$) depends on temperature through the corresponding enthalpy change. The bridge is that a phase transition is just a chemical reaction in which the "products" and "reactants" are the same substance in different phases.

Worked numeric example. For water at $25^{\circ }$C ($T_1=298$ K), the vapour pressure is $P_1=3.17$ kPa and $\Delta H_{\text{vap}}=44.0$ kJ/mol. Predict the vapor pressure at $T_2=100^{\circ }$C ($373$ K) — the normal boiling point, where $P_2=101.3$ kPa. Apply Clausius-Clapeyron: $$ \ln!\left(\frac{P_2}{P_1}\right) = -\frac{44,000}{8.314}!\left(\frac{1}{373} - \frac{1}{298}\right) = -5294 \times (-6.74 \times 10^{-4}) = 3.57, $$ giving $P_2/P_1=e^{3.57}=35.5$, so $P_2=3.17\times 35.5=112.5$ kPa. The measured value is 101.3 kPa — a 10% overprediction, attributable to the temperature variation of $\Delta H_{\text{vap}}$ (which decreases from 44.0 kJ/mol at 25${}^{\circ }$C to 40.7 kJ/mol at 100${}^{\circ }$C). Using the average $\Delta H_{\text{vap}}=42.4$ kJ/mol gives $P_2=102.6$ kPa — within 1% of measured. The lesson: Clausius-Clapeyron is exact within its assumptions (ideal-gas vapour, condensed-phase volume negligible, $\Delta H_{\text{vap}}$ constant); the approximations dominate the small residual discrepancies, but the equation captures the dominant exponential dependence of vapour pressure on temperature.

Binodal and spinodal curves in mixtures. For binary mixtures the chemical-potential picture extends naturally. Plot $G_m$ as a function of mole fraction $x$ at fixed $(T,P)$. A homogeneous single-phase mixture is stable when $G_m(x)$ is convex; a region of $G_m$ that is locally concave is unstable to phase separation. The binodal is the common-tangent construction: two compositions $x_{\alpha }$ and $x_{\beta }$ such that the tangent line to $G_m$ at $x_{\alpha }$ also touches $G_m$ at $x_{\beta }$ — these are the equilibrium compositions of the two coexisting phases, with the common tangent being the locus of the mixed-phase Gibbs energy as the relative amounts of $\alpha$ and $\beta$ vary. The spinodal is the inflection-point locus $({\partial }^2{G}_m/\partial x^2{)}_{T,P}=0$ — the boundary between locally stable (positive curvature) and locally unstable (negative curvature) homogeneous mixtures. Between the binodal and the spinodal, the homogeneous mixture is metastable: it survives small perturbations but decomposes once a critical fluctuation nucleates a new phase. Inside the spinodal, it is unstable to any perturbation and decomposes spontaneously — spinodal decomposition — producing a characteristic mesoscopic pattern of intertwined phases that the Cahn-Hilliard equation (1958) models quantitatively as a fourth-order diffusion equation for the composition field. Spinodal decomposition is the chemistry-side analog of the unstable-mode growth in early-universe cosmology and in Turing-pattern morphogenesis.

Eutectic phase diagrams. A eutectic point is the lowest-melting composition in a binary solid-liquid system. Below the eutectic temperature $T_E$ the mixture is fully solid (two coexisting crystalline phases); above $T_E$ but below the pure-component melting points, a single liquid phase coexists with one of the two pure solids. At the eutectic composition $x_E$, the system passes directly from a single liquid to two-phase solid at $T_E$ without an intermediate two-phase liquid+solid region — the lowest-temperature liquid in the diagram. The Gibbs phase rule constrains $T_E$ and $x_E$: with two components and three phases (liquid plus two solids), $F=2-3+2=1$, but at fixed $P$ that becomes $F=0$, so the eutectic point is a single point in the $(T,x)$ diagram. Eutectics underlie the metallurgy of solder (Pb-Sn eutectic at 183${}^{\circ }$C, $x_{\text{Sn}}=0.62$, well below either pure component's melting point), antifreeze (water-ethylene glycol eutectic at $-50^{\circ }$C), and many ice-cream-making and food-preservation applications.

Chemical potential as the driver of mass transport. Diffusion and osmosis are driven by gradients in chemical potential, not by gradients in concentration. The thermodynamic flux $J_i=-L_i\nabla {\mu }_i/T$ (linear-response Onsager relation) generalises Fick's law $J_i=-D_i\nabla c_i$, which is the dilute-solution limit where ${\mu }_i={\mu }_i^{\circ }+RT\ln c_i$ and $\nabla {\mu }_i\approx (RT/c_i)\nabla c_i$. In nonideal solutions, or in the presence of an external field (gravity, electric field, osmotic pressure), diffusion can run "uphill" in concentration if downhill in chemical potential. Membrane biology exploits this constantly: an ion can move from low concentration to high concentration through a passive channel if the membrane potential creates an electrochemical potential gradient in the favourable direction, with the appropriate chemical potential being ${\tilde{\mu }}_i={\mu }_i+z_{i}F\varphi$ (the electrochemical potential, with $z_{i}F$ the molar charge and $\varphi$ the electric potential).

The statistical-mechanical bridge: partition functions and the molecular origin of free energy [Master]

The macroscopic thermodynamics developed above is empirically grounded but conceptually opaque about why the laws hold and what physical quantities they encode. The statistical-mechanical bridge, due to Boltzmann (1877) and developed by Gibbs (1902 Elementary Principles in Statistical Mechanics), is that macroscopic thermodynamic potentials are logarithms of microscopic partition functions, and equilibrium is the maximum-entropy distribution subject to macroscopic constraints. This identifies thermodynamics as the macroscopic shadow of a deeper microscopic theory.

The canonical ensemble. A system at fixed $(T,V,N)$ in thermal contact with a heat bath has its microstates $i$ occupied with probabilities $$ p_i = \frac{e^{-\beta E_i}}{Z}, \qquad Z(T, V, N) = \sum_i e^{-\beta E_i}, \qquad \beta = \frac{1}{k_B T}, $$ where $E_i$ is the energy of microstate $i$ and the canonical partition function $Z$ normalises the distribution. The cross-link to the physics-side derivation is 11.04.01 pending canonical ensemble (pending), where the maximum-entropy principle subject to fixed $⟨E⟩=U$ delivers the Boltzmann factor by Lagrange-multiplier construction with $\beta$ as the multiplier conjugate to energy. The Helmholtz free energy emerges as $$ \boxed{F(T, V, N) = -k_B T \ln Z(T, V, N).} $$ This is the foundational reason the partition function controls all thermodynamics: every macroscopic equation of state, every heat capacity, every chemical equilibrium constant is derivable from $Z$ by differentiation. The mean energy is $U=-\partial \ln Z/\partial \beta =k_B{T}^2(\partial \ln Z/\partial T{)}_V$; the entropy is $S=-(\partial F/\partial T{)}_V=k_B(\ln Z+\beta U)$; the pressure is $P=-(\partial F/\partial V{)}_T=k_{B}T(\partial \ln Z/\partial V{)}_T$. The mass-action equilibrium constant for a chemical reaction emerges from the appropriate ratios of partition functions, as developed below.

Translational, rotational, vibrational, and electronic partition functions. For a dilute gas of distinguishable non-interacting molecules, the system partition function factors as $Z=(q^N/N!)$ where $q$ is the single-molecule partition function (the $1/N!$ corrects for indistinguishability of identical molecules in classical statistics, the Gibbs paradox resolution). The single-molecule $q$ factors further as $$ q = q_\text{trans} \cdot q_\text{rot} \cdot q_\text{vib} \cdot q_\text{el}, $$ because the molecular Hamiltonian decouples into independent translational, rotational, vibrational, and electronic pieces to a good approximation (Born-Oppenheimer separation, plus rigid-rotor and harmonic-oscillator approximations for the nuclear motion). Each factor is computable in closed form:

The translational factor for a free particle in a volume $V$ is $q_{\text{trans}}=V/{\Lambda }^3$, where $\Lambda =h/\sqrt{2\pi m{k}_{B}T}$ is the thermal de Broglie wavelength. For ${\text{N}}_2$ at 298 K, $\Lambda =1.91\times 10^{-11}$ m, and $q_{\text{trans}}/V\approx 1.4\times 10^{32}$ m${}^{-3}$ — astronomically large, reflecting the enormous number of translational microstates available to a gas molecule at room temperature.

The rotational factor for a diatomic with moment of inertia $I$ in the rigid-rotor approximation is $q_{\text{rot}}=T/(\sigma {\theta }_{\text{rot}})$ in the high-temperature limit $T\gg {\theta }_{\text{rot}}={\hslash }^2/(2I{k}_B)$, where $\sigma$ is the symmetry number (2 for homonuclear, 1 for heteronuclear). For ${\text{N}}_2$, ${\theta }_{\text{rot}}=2.88$ K, so at 298 K we are deep in the classical limit and $q_{\text{rot}}\approx 51.6$.

The vibrational factor for a harmonic oscillator of angular frequency $\omega$ is $q_{\text{vib}}=1/(1-e^{-\hslash \omega /k_{B}T})$, choosing the zero of energy at the lowest vibrational level. For ${\text{N}}_2$ with $\hslash \omega /k_B=3374$ K (a very stiff bond), at 298 K we have $q_{\text{vib}}\approx 1.0003$ — essentially no vibrational excitation, which is why ${\text{N}}_2$ at room temperature has heat capacity $\frac{5}{2}R$ (translational $\frac{3}{2}R$ + rotational $R$) rather than the $\frac{7}{2}R$ that would include vibrational degrees of freedom.

The electronic factor is $q_{\text{el}}={\sum }_n{g}_n{e}^{-(E_n-E_0)/k_{B}T}$, the Boltzmann sum over electronic energy levels with degeneracies $g_n$. For most molecules at room temperature, only the ground electronic state contributes and $q_{\text{el}}\approx g_0$ (the ground-state degeneracy, typically 1 for singlets, 2 for doublets, 3 for triplets like ${\text{O}}_2$).

The molecular partition function and the chemical equilibrium constant. The chemical potential of an ideal gas in terms of its molecular partition function is $$ \mu_i = -k_B T \ln!\left(\frac{q_i}{N_i}\right) = \mu_i^\circ(T) + k_B T \ln!\left(\frac{N_i}{q_i^\circ}\right), $$ where $q_i^{\circ }$ is the partition function per unit standard-state volume or pressure. For a chemical reaction ${\sum }_i{\nu }_i{A}_i=0$ in an ideal gas mixture, the equilibrium condition ${\sum }_i{\nu }_i{\mu }_i=0$ combined with the partition-function expression for ${\mu }_i$ yields $$ \boxed{K = \prod_i !\left(\frac{q_i / V}{N_A / V_\text{ref}}\right)^{\nu_i} !!\times, \exp!\left(-\frac{\Delta E_0}{R T}\right),} $$ where $\Delta E_0={\sum }_i{\nu }_i{E}_{0,i}$ is the difference of molecular ground-state energies (zero-point inclusive) between products and reactants, and the prefactor depends only on the partition-function ratios. This is the statistical-mechanical derivation of the mass-action law. The bridge between quantum chemistry and chemical thermodynamics is laid here: $E_{0,i}$ comes from solving the molecular Schrödinger equation (or its computational approximations — Hartree-Fock, DFT, coupled cluster, as developed in 14.05.02 pending for diatomics and generalised to polyatomics); $q_{\text{rot}}$ and $q_{\text{vib}}$ come from the spectroscopic constants ${\theta }_{\text{rot}}$ and $\hslash \omega$ measured by infrared and microwave spectroscopy; $q_{\text{trans}}$ is universal. From these molecular inputs, $K(T)$ is computable from first principles for any gas-phase reaction.

A worked microscopic example: H${}_2$ dissociation. Consider ${\text{H}}_2(g)\rightleftharpoons 2\text{H}(g)$. The bond dissociation energy is $D_0=432$ kJ/mol, giving $\Delta E_0=+432$ kJ/mol (endothermic). At 2000 K, the partition-function ratio (using the molecular constants of ${\text{H}}_2$: ${\theta }_{\text{rot}}=87.6$ K, $\hslash \omega /k_B=6325$ K) gives a prefactor of order $10^5$ (driven mainly by the $V/{\Lambda }^3$ translational factor for the dissociated H atoms gaining translational entropy). The exponential factor is $\exp (-432000/(8.314\times 2000))=\exp (-25.96)=5.3\times 10^{-12}$. Multiplying, $K\sim 5\times 10^{-7}$ — strongly favouring intact ${\text{H}}_2$ at 2000 K, consistent with experiment. At 5000 K the exponential factor becomes $\exp (-10.4)=3\times 10^{-5}$ and the prefactor grows, giving $K\sim 10$ — dissociation now favoured, again consistent with experiment. The bridge is exact: chemical thermodynamics built from $\Delta G^{\circ }=-RT\ln K$ recovers what statistical mechanics predicts from partition functions, and both recover what spectroscopy measures.

The Boltzmann factor's predictive power. The exponential $e^{-E/k_{B}T}$ is the most consequential function in chemistry. It controls reaction rates (Arrhenius factor $e^{-E_a/k_{B}T}$), conformational populations of biomolecules ($p_{\text{folded}}/p_{\text{unfolded}}=e^{-\Delta G/k_{B}T}$), spectroscopic line intensities (population of upper level in a transition), the population inversion threshold for lasers, the magnetic susceptibility of paramagnets through Curie's law, the dissociation of weak electrolytes, the fraction of high-energy molecules in a gas (Maxwell-Boltzmann velocity distribution), the rate of nuclear fusion in stellar cores (Gamow penetration through Coulomb barriers), the temperature dependence of semiconductor conductivity, the entropy of mixing in alloys and polymers, the Eyring transition-state-theory rate constants, the Marcus theory of electron transfer, and the kinetics of protein unfolding. The Boltzmann factor identifies the macroscopic with the microscopic: macroscopic concentration ratios are microscopic Boltzmann-weighted population ratios, and the temperature is the universal exchange rate between energy and entropy that the second law enforces.

Chemistry emergent from molecular ensembles. The macroscopic concept of a chemical species — "an N${}_2$ molecule" with definite energy levels and partition functions — itself depends on the timescale of observation. The Born-Oppenheimer separation between electronic and nuclear motion that lets us define "the bond energy" $D_e$ as a well-defined quantity rests on the large mass ratio between nuclei and electrons and the consequent timescale separation. Slower timescales (rotational state populations, conformational equilibria, isomerisation) appear in the ensemble as additional partition-function factors; faster timescales (electronic transitions induced by light) appear as time-dependent perturbations not captured by the equilibrium ensemble. The thermodynamic limit — that macroscopic intensive variables become well-defined as $N\to \infty$ with $N/V$ fixed — guarantees that the macroscopic concentrations are sharp (no thermal fluctuations of order unity), while small systems (single molecules, nanoscale aggregates, individual cells) exhibit measurable thermodynamic fluctuations governed by the same statistical mechanics.

Synthesis. Putting these together, chemical thermodynamics is the macroscopic average of a microscopic statistical theory, with the partition function $Z$ as the central object; the Gibbs energy is $-k_{B}T\ln Z$ in the appropriate ensemble; the equilibrium constant is a ratio of molecular partition functions weighted by a Boltzmann factor of the energy difference. The central insight is that this bridge identifies macroscopic equilibrium with microscopic maximum entropy, and the bridge is exact in the thermodynamic limit. The pattern recurs across every domain where statistical mechanics applies: builds toward the canonical-ensemble framework 11.04.01 pending, generalises to the Maxwell-Boltzmann velocity distribution 11.02.01 pending, and is dual to the grand-canonical ensemble used for open systems where particle number fluctuates. Appears again in the connection between the molecular partition function and quantum chemistry 14.04.01 pending, where the molecular energy levels $E_{0,i}$ that feed into $\Delta E_0$ are computed by solving the electronic Schrödinger equation.

Full proof set [Master]

Proposition 1 (Equilibrium condition from Gibbs energy minimisation). For a closed system at constant $T$ and $P$ undergoing a single chemical reaction ${\sum }_i{\nu }_i{A}_i=0$, the equilibrium composition satisfies ${\Delta }_r{G}^{\circ }=-RT\ln K$, where $K={\prod }_i(a_i^{\text{eq}}{)}^{{\nu }_i}$ is the equilibrium constant.

Proof. The Gibbs energy of the system at composition $\{N_i\}$ is $G={\sum }_i{N}_i{\mu }_i$. As the reaction proceeds by extent $d\xi$, the mole numbers change by $d{N}_i={\nu }_{i}d\xi$, so $$ dG = \sum_i \mu_i, dN_i = \left(\sum_i \nu_i \mu_i\right) d\xi = \Delta_r G, d\xi. $$ At equilibrium at constant $(T,P)$, $G$ is minimised with respect to $\xi$, so $(\partial G/\partial \xi {)}_{T,P}=0$ requires ${\Delta }_{r}G=0$. Substituting ${\mu }_i={\mu }_i^{\circ }+RT\ln a_i$: $$ \Delta_r G = \sum_i \nu_i \mu_i^\circ + RT \sum_i \nu_i \ln a_i = \Delta_r G^\circ + RT \ln Q, $$ where $Q={\prod }_i{a}_i^{{\nu }_i}$. Setting ${\Delta }_{r}G=0$ at equilibrium with $Q=K$: $$ \Delta_r G^\circ + RT \ln K = 0 \implies \Delta_r G^\circ = -RT \ln K. \quad \square $$

Proposition 2 (van't Hoff equation). The temperature dependence of the equilibrium constant satisfies $d\ln K/dT={\Delta }_r{H}^{\circ }/(R{T}^2)$.

Proof. From Proposition 1, $\ln K=-{\Delta }_r{G}^{\circ }/(RT)$. Differentiate with respect to $T$: $$ \frac{d \ln K}{dT} = -\frac{1}{R}, \frac{d}{dT}!\left(\frac{\Delta_r G^\circ}{T}\right). $$ The Gibbs-Helmholtz equation $[\partial (G/T)/\partial T{]}_P=-H/T^2$ (proved by direct calculation from $G=H-TS$ and $(\partial G/\partial T{)}_P=-S$) applied to ${\Delta }_r{G}^{\circ }$ gives $d({\Delta }_r{G}^{\circ }/T)/dT=-{\Delta }_r{H}^{\circ }/T^2$. Substituting: $$ \frac{d \ln K}{dT} = -\frac{1}{R} \cdot \left(-\frac{\Delta_r H^\circ}{T^2}\right) = \frac{\Delta_r H^\circ}{RT^2}. \quad \square $$

Proposition 3 (Clausius-Clapeyron equation). Along the liquid-vapour coexistence curve of a pure substance, assuming the vapour behaves as an ideal gas and the condensed-phase molar volume is negligible compared to that of the vapour, $d\ln P/dT=\Delta H_{\text{vap}}/(R{T}^2)$.

Proof. Equality of chemical potentials across the coexistence curve: ${\mu }_{\text{liq}}(T,P)={\mu }_{\text{vap}}(T,P)$. Differentiating along the curve, $d{\mu }_{\text{liq}}=d{\mu }_{\text{vap}}$. Substituting the molar Gibbs differential $d\mu =-S_{m}dT+V_{m}dP$: $$ -S_{m,\text{liq}} dT + V_{m,\text{liq}} dP = -S_{m,\text{vap}} dT + V_{m,\text{vap}} dP. $$ Rearranging: $$ \frac{dP}{dT} = \frac{S_{m,\text{vap}} - S_{m,\text{liq}}}{V_{m,\text{vap}} - V_{m,\text{liq}}} = \frac{\Delta S_\text{vap}}{\Delta V_\text{vap}}. $$ At the reversible phase transition $\Delta G_{\text{vap}}=0$, so $\Delta H_{\text{vap}}=T\Delta S_{\text{vap}}$, giving the Clapeyron equation $dP/dT=\Delta H_{\text{vap}}/(T\Delta V_{\text{vap}})$. Assuming $V_{m,\text{liq}}\ll V_{m,\text{vap}}$, $\Delta V_{\text{vap}}\approx V_{m,\text{vap}}$. Assuming ideal-gas behaviour for the vapour, $V_{m,\text{vap}}=RT/P$, so $$ \frac{dP}{dT} = \frac{P \Delta H_\text{vap}}{R T^2} \implies \frac{d \ln P}{dT} = \frac{\Delta H_\text{vap}}{RT^2}. \quad \square $$

Proposition 4 (Helmholtz free energy from the canonical partition function). For a system in the canonical ensemble at temperature $T$, volume $V$, and particle number $N$, the Helmholtz free energy is $F(T,V,N)=-k_{B}T\ln Z(T,V,N)$, where $Z={\sum }_i{e}^{-E_i/(k_{B}T)}$ is the canonical partition function.

Proof. The canonical probability of microstate $i$ is $p_i=e^{-\beta E_i}/Z$ with $\beta =1/(k_{B}T)$. The Gibbs entropy is $$ S = -k_B \sum_i p_i \ln p_i = -k_B \sum_i p_i (-\beta E_i - \ln Z) = k_B \beta \langle E \rangle + k_B \ln Z = \frac{U}{T} + k_B \ln Z. $$ Solving: $TS=U+k_{B}T\ln Z$, hence $U-TS=-k_{B}T\ln Z$. The left side is the Helmholtz free energy $F$ by definition, so $F=-k_{B}T\ln Z$. $\square$

Corollary (Boltzmann's entropy formula in the microcanonical limit). For an isolated system with $\Omega$ accessible microstates at fixed energy, $S=k_B\ln \Omega$ (Boltzmann's entropy formula). Inside the microcanonical ensemble, all $\Omega$ microstates are equiprobable, so the Gibbs entropy collapses to $S=-k_B{\sum }_i(1/\Omega )\ln (1/\Omega )=k_B\ln \Omega$.

Connections [Master]

• Thermodynamic potentials and Legendre transforms 11.01.02 pending. The physics-side treatment of the four thermodynamic potentials $U,H,F,G$ develops the Legendre-transform machinery in detail and gives the geometric framing of Maxwell relations as closed-form identities on the thermodynamic state-space manifold. The chemistry-side treatment here builds on that foundation by specialising to constant-$(T,P)$ chemical-reaction conditions where $G$ is the natural object. The cross-link is reciprocal: the physics unit develops the abstract structure; this unit develops the chemical-reaction specialisation.

• Canonical ensemble 11.04.01 pending. The statistical-mechanical bridge developed in §4 of this unit (partition function $Z$, Helmholtz free energy $F=-k_{B}T\ln Z$) is grounded in the canonical ensemble. The peer unit derives the Boltzmann distribution from the maximum-entropy principle with energy constraint; this unit applies that derivation to molecular partition functions and reaction equilibria. The cross-link is the foundational reason chemical equilibrium constants emerge from microscopic dynamics: $K$ is a ratio of partition-function factors weighted by a Boltzmann factor of the energy-level difference.

• Maxwell-Boltzmann distribution 11.02.01 pending. The Boltzmann factor that controls equilibrium populations of energy levels in chemical equilibrium is the same factor that controls the velocity distribution of a classical gas. The bridge is that both are special cases of the canonical-ensemble probability $p_i\propto e^{-E_i/(k_{B}T)}$, with $E_i$ being electronic+vibrational+rotational energy for chemical equilibrium or kinetic energy for the velocity distribution. The peer unit develops the Maxwell-Boltzmann case in detail; this unit's partition-function machinery generalises to all degrees of freedom.

• Hydrogen-atom quantum chemistry 14.04.01 pending. The molecular ground-state energy $E_0$ that enters the partition-function expression for the equilibrium constant is computed by solving the electronic Schrödinger equation. For chemistry beyond hydrogen, the methods are Hartree-Fock, density-functional theory, and post-HF correlation methods (CI, CC, MP perturbation), developed in the MO theory unit 14.05.02 pending. The bridge from quantum chemistry to chemical thermodynamics is: molecular electronic-structure calculations produce $E_0$ and the spectroscopic constants (${\theta }_{\text{rot}}$, $\hslash \omega$); these feed into the molecular partition function $q$; the equilibrium constant $K$ is then computable from first principles.

• Molecular orbital theory 14.05.02 pending. The bond dissociation energies and zero-point energies that enter $\Delta E_0$ in the partition-function derivation of $K$ are direct outputs of molecular orbital calculations. The MO unit gives the methodology for computing $E_0$ for diatomics and polyatomics; this unit gives the framework in which $E_0$ controls the macroscopic equilibrium constant. The cross-link is essential for ab initio prediction of equilibrium constants without empirical fitting.

• Stoichiometry and gas laws 14.03.01 provide the balanced equations, mole-fraction algebra, and ideal-gas law that underpin standard-state calculations throughout this unit. The stoichiometric coefficients ${\nu }_i$ in ${\Delta }_{r}G={\sum }_i{\nu }_i{\mu }_i$ are the same as those in the balanced reaction equation, and the ideal-gas law underlies the relation ${\mu }_i={\mu }_i^{\circ }+RT\ln (P_i/P^{\circ })$ for gas-phase species.

• Chemical kinetics 14.08.01. Thermodynamics predicts where equilibrium lies (the destination); kinetics predicts how long it takes to get there (the path and pace). The rate constants $k_{\text{f}}$ and $k_{\text{r}}$ of the forward and reverse reactions satisfy $K=k_{\text{f}}/k_{\text{r}}$ at equilibrium — a beautiful identification of a thermodynamic ratio with a microscopic-rate ratio. The Arrhenius factor $e^{-E_a/(k_{B}T)}$ in the rate constants is the same Boltzmann factor that governs the partition-function ratios.

• Acid-base equilibrium 14.10.01 is a direct application of the equilibrium constant to proton-transfer reactions. The acid dissociation constant $K_a$, the base ionisation constant $K_b$, and the water self-ionisation constant $K_w$ are all instances of the general $K=\exp (-\Delta G^{\circ }/(RT))$ framework developed here, with pH being a logarithm of a chemical potential.

• Electrochemistry 14.11.01 applies $\Delta G=-nFE$ to redox reactions, identifying the cell potential $E$ with the Gibbs energy of the redox reaction. The Nernst equation $E=E^{\circ }-(RT/(nF))\ln Q$ is the electrochemical analog of ${\Delta }_{r}G={\Delta }_r{G}^{\circ }+RT\ln Q$, with the bridge being $\Delta G=-nFE$.

• Metabolic thermodynamics 17.04.01. Living cells run reactions at constant body temperature and constant cytoplasmic pressure — exactly the conditions where Gibbs energy is the natural variable. ATP hydrolysis ($\Delta G^{\circ }\approx -30.5$ kJ/mol at pH 7) drives coupled reactions that would otherwise be non-spontaneous, exploiting the additivity of $\Delta G$ across coupled reactions. The cellular maintenance of metabolic concentration ratios far from chemical equilibrium is the macroscopic signature of life as a non-equilibrium thermodynamic system.

• Origin of life and prebiotic thermodynamics 19.15.01 pending. The free-energy landscape of prebiotic chemistry — the reducing conditions of the early Earth that made organic synthesis thermodynamically favourable, the disequilibrium environments (hydrothermal vents, atmosphere-ocean interfaces) that powered the first metabolisms — is governed by the same $\Delta G$ framework developed here. The bridge between geochemistry and biochemistry runs through chemical thermodynamics.

• Biomolecules in cells overview 17.01.01. The condensation and hydrolysis reactions that assemble and disassemble biological macromolecules — peptide bonds, phosphodiester bonds, glycosidic bonds — are governed by the same $\Delta G=\Delta H-T\Delta S$ framework developed here. ATP coupling, the central thermodynamic mechanism in living systems, is introduced in the biomolecules unit as the strategy by which cells drive thermodynamically unfavourable polymerisation reactions, and the additivity of $\Delta G$ across coupled reactions explains how catabolic exergonicity powers anabolic endergonicity.

• Membrane transport — passive and active 17.02.02. The Nernst equation derived here for electrochemical cells is the same equation that governs ion electrochemical potentials across biological membranes. The free-energy framework $\Delta G=RT\ln (C_{\text{out}}/C_{\text{in}})+zF{V}_m$ used to quantify passive and active transport in 17.02.02 is a direct application of the chemical-potential and equilibrium thermodynamics developed in this unit. Active transport (Na+/K+-ATPase, V-type ATPase) consumes ATP's free energy to move ions against their electrochemical gradients, illustrating coupled-reaction thermodynamics at the membrane scale.

Historical & philosophical context [Master]

The concept of free energy emerged from the problem of reconciling the first and second laws of thermodynamics for chemical processes. Hermann von Helmholtz ^{[Helmholtz1882]} introduced the function $A=U-TS$ (now Helmholtz free energy) in 1882 in his treatise Die Thermodynamik chemischer Vorgänge, identifying it as the maximum work obtainable from a closed system at constant temperature. Josiah Willard Gibbs ^[Gibbs1878] developed the complete framework of chemical thermodynamics in his 1876-78 monograph On the Equilibrium of Heterogeneous Substances, published in the obscure Transactions of the Connecticut Academy. Gibbs introduced the chemical potential ${\mu }_i$, the phase rule $F=C-P+2$, the conditions for chemical equilibrium ${\sum }_i{\nu }_i{\mu }_i=0$, and the geometric construction of phase diagrams from the common-tangent method on the Gibbs energy surface. His work was so mathematically dense that it took decades to be fully appreciated in Europe; Wilhelm Ostwald's 1892 German translation and Pierre Duhem's 1893 French commentary were instrumental in its dissemination, and Le Châtelier's 1899 French translation made Gibbs accessible to the chemical-engineering community.

Le Chatelier ^{[LeChatelier1884]} published his principle in 1884, and Karl Ferdinand Braun independently stated a similar result in 1887. The combined Le Chatelier-Braun principle is a qualitative statement of what the van't Hoff equation (also 1884) quantifies: the temperature dependence of $K$ follows from the sign of $\Delta H^{\circ }$. The formal proof that Le Chatelier's principle follows from thermodynamic stability conditions ($d^{2}G>0$) was given by Prigogine and Defay in their 1954 Chemical Thermodynamics. Jacobus Henricus van't Hoff ^{[vantHoff1884]} derived the equation $d\ln K/dT=\Delta H^{\circ }/(R{T}^2)$ in his 1884 Études de Dynamique Chimique, building on his earlier 1874 La Chimie dans l'Espace (which introduced stereochemistry) and on his identification of osmotic pressure with gas pressure through the law $\Pi V=nRT$ — a deep insight that won him the inaugural 1901 Nobel Prize in Chemistry.

The Clausius-Clapeyron equation has a more complicated history. Émile Clapeyron published the equation $dP/dT=\Delta H/(T\Delta V)$ for general phase transitions in 1834, working within the caloric theory of heat that predated the first law; Clausius reformulated it in modern thermodynamic language in 1850 and applied it to the liquid-vapour equilibrium with the ideal-gas approximation that gives the modern Clausius-Clapeyron form. The phase-diagram methodology of binary mixtures was developed by H. W. Bakhuis Roozeboom in the 1890s-1910s in his five-volume Die heterogenen Gleichgewichte vom Standpunkte der Phasenlehre, applying Gibbs's phase rule to metallurgy, geology, and chemistry.

The statistical-mechanical bridge that identifies thermodynamic potentials with logarithms of partition functions is due to Ludwig Boltzmann (1877) ^{[Boltzmann1877]}, who derived the entropy formula $S=k_B\ln \Omega$ from the combinatorial counting of microstates, and Josiah Willard Gibbs again, in his 1902 Elementary Principles in Statistical Mechanics. Gibbs introduced the modern ensemble formalism (microcanonical, canonical, grand canonical) and proved that the canonical-ensemble Helmholtz energy equals $-k_{B}T\ln Z$. The Boltzmann-Gibbs framework was challenged by the ergodicity question — does the time average over a trajectory equal the ensemble average over phase space? — which was settled in important special cases by Birkhoff (1931) and von Neumann (1932) ^{[vonNeumann1932]} in the mathematical ergodic theorem; the general case remains open for most realistic chemical systems but the framework's empirical predictions are unambiguous.

The distinction between thermodynamic spontaneity ($\Delta G<0$) and kinetic feasibility (reasonable reaction rate) is one of the most important conceptual points in chemistry. The hydrogen-oxygen reaction to form water has $\Delta G^{\circ }=-237$ kJ/mol at 298 K, yet a mixture of ${\text{H}}_2$ and ${\text{O}}_2$ is kinetically stable at room temperature because the activation barrier is about 400 kJ/mol. A spark provides the activation energy, and the reaction proceeds explosively. This distinction motivates the entire field of catalysis: lowering the activation barrier without changing $\Delta G$ is what makes industrially viable conversion of thermodynamically favourable reactions possible.

The Haber-Bosch process is a case study in the thermodynamics-kinetics tension. Fritz Haber and Carl Bosch developed the high-pressure, moderate-temperature process between 1908 and 1913 ^[Haber1909], using an iron catalyst to achieve reasonable rates at temperatures where equilibrium is still moderately favourable. Bosch's engineering of the high-pressure equipment was as significant as Haber's chemistry; the process earned Haber the 1918 Nobel Prize and Bosch the 1931 Nobel Prize. The Haber-Bosch process now produces around 200 million tonnes of ammonia per year and supports an estimated half of the world's food production via synthetic nitrogen fertilisers — perhaps the single most consequential application of chemical thermodynamics in human history.

Bibliography [Master]

@article{Gibbs1878,
  author = {Gibbs, J. W.},
  title = {On the Equilibrium of Heterogeneous Substances},
  journal = {Transactions of the Connecticut Academy of Arts and Sciences},
  volume = {3},
  year = {1876-1878},
  pages = {108-248, 343-524}
}

@article{Helmholtz1882,
  author = {Helmholtz, H. von},
  title = {Die Thermodynamik chemischer Vorgänge},
  journal = {Sitzungsberichte der Königlich Preussischen Akademie der Wissenschaften zu Berlin},
  year = {1882},
  pages = {22-39}
}

@article{LeChatelier1884,
  author = {Le Chatelier, H.},
  title = {Sur un énoncé général des lois des équilibres chimiques},
  journal = {Comptes Rendus de l'Académie des Sciences},
  volume = {99},
  year = {1884},
  pages = {786-789}
}

@book{vantHoff1884,
  author = {van't Hoff, J. H.},
  title = {Études de Dynamique Chimique},
  publisher = {Frederik Muller and Co., Amsterdam},
  year = {1884}
}

@article{Boltzmann1877,
  author = {Boltzmann, L.},
  title = {Über die Beziehung zwischen dem zweiten Hauptsatze der mechanischen Wärmetheorie und der Wahrscheinlichkeitsrechnung respektive den Sätzen über das Wärmegleichgewicht},
  journal = {Wiener Berichte},
  volume = {76},
  year = {1877},
  pages = {373-435}
}

@book{GibbsStatMech1902,
  author = {Gibbs, J. W.},
  title = {Elementary Principles in Statistical Mechanics},
  publisher = {Yale University Press, New Haven},
  year = {1902}
}

@article{Haber1909,
  author = {Haber, F.},
  title = {Über die Darstellung des Ammoniaks aus Stickstoff und Wasserstoff},
  journal = {Zeitschrift für Elektrochemie und angewandte physikalische Chemie},
  volume = {16},
  year = {1910},
  pages = {244-246}
}

@book{vonNeumann1932,
  author = {von Neumann, J.},
  title = {Mathematische Grundlagen der Quantenmechanik},
  publisher = {Springer, Berlin},
  year = {1932}
}

@article{Clapeyron1834,
  author = {Clapeyron, B. P. É.},
  title = {Mémoire sur la puissance motrice de la chaleur},
  journal = {Journal de l'École Polytechnique},
  volume = {14},
  year = {1834},
  pages = {153-190}
}

@article{Clausius1850,
  author = {Clausius, R.},
  title = {Über die bewegende Kraft der Wärme und die Gesetze, welche sich daraus für die Wärmelehre selbst ableiten lassen},
  journal = {Annalen der Physik und Chemie},
  volume = {79},
  year = {1850},
  pages = {368-397, 500-524}
}

@article{CahnHilliard1958,
  author = {Cahn, J. W. and Hilliard, J. E.},
  title = {Free energy of a nonuniform system. I. Interfacial free energy},
  journal = {Journal of Chemical Physics},
  volume = {28},
  year = {1958},
  pages = {258-267}
}

@book{Callen1985,
  author = {Callen, H. B.},
  title = {Thermodynamics and an Introduction to Thermostatistics},
  edition = {2},
  publisher = {Wiley, New York},
  year = {1985}
}

@book{McQuarrie2000,
  author = {McQuarrie, D. A.},
  title = {Statistical Mechanics},
  publisher = {University Science Books, Sausalito},
  year = {2000}
}

@book{AtkinsPaula2023,
  author = {Atkins, P. and de Paula, J.},
  title = {Physical Chemistry},
  edition = {12},
  publisher = {Oxford University Press},
  year = {2023}
}

@book{Tro2023,
  author = {Tro, N. J.},
  title = {Chemistry: A Molecular Approach},
  edition = {6},
  publisher = {Pearson},
  year = {2023}
}

@book{PrigogineDefay1954,
  author = {Prigogine, I. and Defay, R.},
  title = {Chemical Thermodynamics},
  publisher = {Longmans, London},
  year = {1954}
}

---

Cycle 6 Track B deepening of the Wave 3 chemistry unit (originally produced by claude-glm-5.1 in Wave 3). Master tier expanded from a single sub-section (1000 words) to four named H2 sub-sections (9000 words combined) per the Cycle 4 style-parity contract. Status promoted to shipped. Prerequisites trimmed to the one shipped/pending-registered ID (11.04.01 canonical ensemble); other foundation cross-links moved to Connections. Lean status remains none with an expanded mathlib_gap that names the differential-forms hierarchy and the Legendre-transform library as the specific Mathlib modules whose gluing to chemistry-specific concepts would be required for a meaningful formalisation. All hooks_out targets remain proposed; promotion to confirmed awaits cross-domain reviewer audit per CHEMISTRY_PLAN §6.