Ideal Solutions and the Chemical Potential of Liquids

Pour benzene and toluene together and something quietly beautiful happens: molecules of each barely notice the swap. Their intermolecular attractions are so similar that mixing costs almost nothing. The only cost is the entropy of shuffling molecules randomly. This is the ideal solution, physical chemistry's cleanest model for liquid mixtures. The idea matters because it gives every component a chemical potential formula. This formula is simple enough to solve by hand. Yet it is rich enough to predict boiling points, vapor pressures, and phase diagrams. It is the liquid-phase analogue of the ideal gas law. It is also the reference state against which every real, imperfect solution is measured. This article builds the chemical potential of a liquid in an ideal solution from Raoult's law. You will see where the logarithmic mole-fraction term comes from. Then you will see why it defines ideality rather than merely describing it. Finally you will use it to compute real vapor pressures and mixing free energies.

Key takeaways

An ideal solution is defined by Raoult's law: every component obeys $p_i = x_i p_i^*$ across the whole composition range.
The chemical potential of a liquid component becomes $\mu_i = \mu_i^* + RT\ln x_i$ , mirroring the ideal-gas mixing formula.
This form follows from equating liquid and vapor chemical potentials at equilibrium, using Raoult's law for the vapor pressure.
Because $x_i \le 1$ , $\ln x_i \le 0$ , so mixing always lowers each component's chemical potential — mixing is spontaneous.
Ideal solutions have zero enthalpy of mixing and zero volume of mixing; the mixing free energy is purely entropic.
Key idea: Raoult's law is not a consequence of ideality — it is the definition, and $\mu_i = \mu_i^* + RT\ln x_i$ is its thermodynamic translation.

The core insight

An ideal liquid solution is one in which every component's vapor pressure is directly proportional to its mole fraction over the entire composition range, $p_i = x_i p_i^*$ ; this experimental statement (Raoult's law) is thermodynamically equivalent to the chemical potential expression $\mu_i = \mu_i^* + RT\ln x_i$ , because equating liquid and vapor chemical potentials at equilibrium and substituting Raoult's law for the partial pressure produces exactly that logarithmic form.

The physical picture

Picture a closed vessel containing a liquid mixture of components A and B in equilibrium with its vapor. At equilibrium, the chemical potential of A in the liquid must equal the chemical potential of A in the vapor above it — otherwise A would spontaneously flow from the phase of high potential to the phase of low potential. The same holds for B. This equality is the thermodynamic anchor for everything that follows. Now ask: how does the vapor pressure of A depend on how much A is actually present in the liquid? Raoult's law answers empirically for many mixtures of chemically similar molecules (benzene/toluene, hexane/heptane): the partial vapor pressure of A is simply the mole fraction of A times the vapor pressure of pure A,

p_A = x_A p_A^*

This is a strong statement — it says the escaping tendency of A molecules from the liquid is diluted in direct proportion to how much of the liquid is A, independent of what B is (provided A and B are similar enough). When every component of a mixture obeys this law across all compositions, the mixture is called an ideal solution. Real mixtures approach this only when A–A, B–B, and A–B intermolecular forces are essentially indistinguishable in strength and character. Because ideality is defined this way, translating it into a statement about chemical potential is a short algebraic step (carried out in the derivation below) rather than a new physical assumption. The result is the liquid-phase mixing law that plays the same role in solution thermodynamics that $pV = nRT$ plays for gases.

Figure

Raoult's law: partial and total vapor pressure versus mole fraction

Figure

Chemical potential of a component as its mole fraction falls below 1

The core relations

p_i = x_i\, p_i^{*}

where: $p_i$ is the partial vapor pressure of component $i$ over the solution, $x_i$ its liquid mole fraction, and $p_i^{*}$ the vapor pressure of the pure liquid $i$ — this equation is Raoult's law and is the operational definition of an ideal solution.

\mu_i = \mu_i^{*} + RT\ln x_i

where: $\mu_i$ is the chemical potential of component $i$ in the ideal liquid solution, $\mu_i^{*}$ is the chemical potential of pure liquid $i$ at the same temperature and pressure, $R$ is the gas constant, $T$ is the absolute temperature.

\Delta_{\text{mix}}G = nRT\left(x_A\ln x_A + x_B\ln x_B\right)

where: $n$ is the total number of moles mixed, $x_A, x_B$ the mole fractions of the two components — always negative for $0<x_i<1$ , so mixing is spontaneous.

\Delta_{\text{mix}}H = 0 \qquad \Delta_{\text{mix}}V = 0

where: these vanish identically for an ideal solution — mixing is driven purely by entropy, with no heat absorbed or released and no volume change.

Derivation AFrom Raoult's law to

\mu_i = \mu_i^{*} + RT\ln x_i

Goal. Starting only from the equilibrium condition between liquid and vapor and from Raoult's law, derive the chemical potential of a liquid component in an ideal solution.

Step 1 — Equilibrium condition. At a given temperature and total pressure, a solution component $i$ in the liquid phase is in equilibrium with the same substance in the vapor phase above it. Equilibrium requires equal chemical potential in both phases:

\mu_i(\text{liquid}) = \mu_i(\text{vapor})

Step 2 — Chemical potential of the vapor. Treating the vapor as an ideal gas at its partial pressure $p_i$ , the chemical potential is

\mu_i(\text{vapor}) = \mu_i^{\ominus}(g) + RT\ln\!\left(\frac{p_i}{p^{\ominus}}\right)

where $\mu_i^{\ominus}(g)$ is the standard chemical potential of gaseous $i$ at the standard pressure $p^{\ominus}$ (usually 1 bar).

Step 3 — Reference to the pure liquid. Apply the same reasoning to pure liquid $i$ in equilibrium with its own vapor at pressure $p_i^{*}$ :

\mu_i^{*}(\text{liquid}) = \mu_i^{\ominus}(g) + RT\ln\!\left(\frac{p_i^{*}}{p^{\ominus}}\right)

This holds because the pure liquid is simply the special case $x_i = 1$ of the same equilibrium.

Step 4 — Subtract to eliminate the gas-phase standard state. Subtracting the Step 3 equation from the combination of Steps 1 and 2 removes the awkward $\mu_i^{\ominus}(g)$ term entirely:

\mu_i(\text{liquid}) - \mu_i^{*}(\text{liquid}) = RT\ln\!\left(\frac{p_i}{p^{\ominus}}\right) - RT\ln\!\left(\frac{p_i^{*}}{p^{\ominus}}\right) = RT\ln\!\left(\frac{p_i}{p_i^{*}}\right)

Step 5 — Insert Raoult's law. Ideality means $p_i = x_i p_i^{*}$ for every $x_i$ . Substituting,

\mu_i(\text{liquid}) - \mu_i^{*}(\text{liquid}) = RT\ln\!\left(\frac{x_i p_i^{*}}{p_i^{*}}\right) = RT\ln x_i

Step 6 — Result. Renaming $\mu_i(\text{liquid})$ as $\mu_i$ for brevity gives the boxed relation:

\boxed{\mu_i = \mu_i^{*} + RT\ln x_i}

Since $0 < x_i \le 1$ implies $\ln x_i \le 0$ , every component's chemical potential in solution is lower than in the pure liquid — dilution always favors mixing thermodynamically, consistent with the negative $\Delta_{\text{mix}}G$ found from summing this expression over both components weighted by moles.

Worked examples

Example 1 — Vapor pressure and chemical potential lowering in an ideal mixture. Liquid benzene (A) and toluene (B) form a solution close to ideal at 298 K. Pure vapor pressures are $p_A^{*} = 12.8\ \text{kPa}$ (benzene) and $p_B^{*} = 3.85\ \text{kPa}$ (toluene). Prepare a solution with $x_A = 0.40$ , $x_B = 0.60$ . Partial pressures (Raoult's law):

p_A = x_A p_A^{*} = (0.40)(12.8\ \text{kPa}) = 5.12\ \text{kPa}

p_B = x_B p_B^{*} = (0.60)(3.85\ \text{kPa}) = 2.31\ \text{kPa}

Total pressure:

p_{\text{total}} = p_A + p_B = 5.12 + 2.31 = 7.43\ \text{kPa}

Chemical potential change of benzene relative to the pure liquid, at $T = 298\ \text{K}$ , $R = 8.314\ \text{J K}^{-1}\text{mol}^{-1}$ :

\mu_A - \mu_A^{*} = RT\ln x_A = (8.314)(298)\ln(0.40)

= (2477.8\ \text{J/mol})(-0.9163) = -2270\ \text{J/mol} = -2.27\ \text{kJ/mol}

Chemical potential change of toluene:

\mu_B - \mu_B^{*} = RT\ln(0.60) = (2477.8)(-0.5108) = -1266\ \text{J/mol} = -1.27\ \text{kJ/mol}

Sanity check: both values are negative, as required since $x_i<1$ always lowers $\mu_i$ below the pure-liquid value. The more dilute component (benzene, $x_A=0.40$ ) shows the larger drop in chemical potential, consistent with $\ln x_i$ becoming more negative as $x_i$ decreases.

Example 2 — Gibbs energy of mixing and checking self-consistency with the composition of the vapor. Mix $n_A = 2.0\ \text{mol}$ benzene with $n_B = 3.0\ \text{mol}$ toluene at 298 K (same system as above, so $x_A = 0.40$ , $x_B = 0.60$ , $n = 5.0\ \text{mol}$ total). Step 1 — Gibbs energy of mixing:

\Delta_{\text{mix}}G = nRT(x_A\ln x_A + x_B\ln x_B)

= (5.0\ \text{mol})(8.314\ \text{J K}^{-1}\text{mol}^{-1})(298\ \text{K})\big[(0.40)(-0.9163)+(0.60)(-0.5108)\big]

= (12{,}389\ \text{J})\big[-0.3665 - 0.3065\big] = (12{,}389\ \text{J})(-0.6730) = -8338\ \text{J} = -8.34\ \text{kJ}

This is negative, confirming mixing is spontaneous at constant $T,p$ , exactly as anticipated because $\ln x_i<0$ for both components. Step 2 — Cross-check using entropy of mixing. Since $\Delta_{\text{mix}}H=0$ for an ideal solution, $\Delta_{\text{mix}}G = -T\Delta_{\text{mix}}S$ , so

\Delta_{\text{mix}}S = -\frac{\Delta_{\text{mix}}G}{T} = -\frac{-8338\ \text{J}}{298\ \text{K}} = 27.98\ \text{J/K}

Independently, $\Delta_{\text{mix}}S = -nR(x_A\ln x_A + x_B\ln x_B) = -(5.0)(8.314)(-0.6730) = 27.98\ \text{J/K}$ — the two routes agree exactly, as they must. Step 3 — Vapor mole fraction check. Using the partial pressures from Example 1, the vapor is richer in the more volatile component:

y_A = \frac{p_A}{p_{\text{total}}} = \frac{5.12}{7.43} = 0.689

Since $y_A = 0.689 > x_A = 0.40$ , benzene (the higher-vapor-pressure component) is indeed enriched in the vapor, matching physical expectation and the shape of a benzene–toluene $p$ – $x$ – $y$ diagram.

Limiting cases & common traps

Situation	What happens	Common error
$x_i \to 1$ (nearly pure)	$\mu_i \to \mu_i^{}$ , Raoult's law holds exactly for the major component even in real* solutions	Assuming Raoult's law only applies to ideal mixtures — it is always exact in the $x_i\to 1$ limit (this is the basis of Henry's law behavior for the solute)
$x_i \to 0$ (trace component)	$RT\ln x_i \to -\infty$ ; the model still gives a defined but very negative $\mu_i-\mu_i^{*}$	Forgetting that real dilute solutes obey Henry's law, $p_i = x_iK_i$ , not Raoult's law — ideal-solution formulas do not describe trace solutes in real mixtures
Mixing similar liquids (benzene/toluene, hexane/heptane)	Excellent approximation to ideality	Assuming any two miscible liquids mix ideally — polar/nonpolar or H-bonding pairs deviate strongly
Computing $\Delta_{\text{mix}}H$ or $\Delta_{\text{mix}}V$	Both are exactly zero for a truly ideal solution	Confusing this with real mixtures, where nonzero $\Delta_{\text{mix}}H$ signals non-ideality (see excess functions)

The chemist behind it

François-Marie Raoult spent decades in Grenoble painstakingly measuring vapor pressures of solutions, motivated by a practical puzzle: why does dissolving a solute always depress a solvent's vapor pressure, freezing point, and raise its boiling point, and by how much? His systematic 1880s measurements across many organic solvent–solute pairs revealed the simple proportionality between vapor pressure and mole fraction that today carries his name. He could not have derived the chemical potential form shown above — Gibbs's chemical potential and the modern thermodynamic apparatus came slightly later and were fused with Raoult's empirical law by subsequent workers, particularly through the statistical framework that Gilbert Lewis and others made rigorous. The result is a rare case where a purely empirical law, discovered from careful bench measurements, turned out to be thermodynamically exact for an entire idealized class of mixtures.

Origins · 1 minFrançois-Marie Raoult and how this was discovered

François-Marie Raoult

1830–1901 · French chemist whose vapor-pressure measurements on solutions established the law that now defines solution ideality

Ask the tutor: How does replacing $x_i$ with the activity $a_i = \gamma_i x_i$ let this same formula describe real, non-ideal solutions?

Is Raoult's law an assumption or a derived result?

Neither exactly — it is the empirical definition of solution ideality. Raoult observed the proportionality $p_i = x_ip_i^{*}$ experimentally for chemically similar liquid pairs. Once you accept that a mixture obeys this law for every component at every composition, the chemical potential expression $\mu_i = \mu_i^{*}+RT\ln x_i$ follows rigorously from equilibrium thermodynamics, as shown in the derivation above. So the law is empirical input; the chemical potential formula is its logical, derived consequence.

Why does the chemical potential use mole fraction and not concentration?

Mole fraction is the natural variable because Raoult's law itself is written in mole fraction: $p_i = x_ip_i^{*}$ . The derivation substitutes this ratio directly into $RT\ln(p_i/p_i^{*})$ , so the $p_i^{*}$ cancels and only $x_i$ survives. Concentration units like molarity would introduce solvent density and temperature dependence that clutter this otherwise clean, dimensionless result.

Do any real mixtures actually behave ideally?

Approximately, yes. Mixtures of structurally similar nonpolar molecules — benzene and toluene, or hexane and heptane — come very close because their intermolecular forces (dispersion forces of similar strength) barely change on mixing. No real mixture is perfectly ideal, but these near-ideal systems are the standard textbook and laboratory benchmarks for testing Raoult's law.

How does this connect to activities and real solutions?

The ideal-solution chemical potential is the reference state for real solutions. Chemists write $\mu_i = \mu_i^{*}+RT\ln a_i$ , where the activity $a_i = \gamma_ix_i$ absorbs all non-ideal behavior into the activity coefficient $\gamma_i$ . When $\gamma_i=1$ , this collapses back to the ideal formula derived here. Excess Gibbs energies and activity coefficients, covered next, measure exactly how far real mixtures deviate from this baseline.

Why is the chemical potential lowering always negative and never zero for a mixture?

Because $x_i<1$ strictly whenever more than one component is present, so $\ln x_i<0$ strictly. Only in the limit of a pure substance, $x_i=1$ , does $\ln x_i=0$ and $\mu_i=\mu_i^{*}$ exactly. Any actual mixing, however slight, produces a nonzero (negative) shift — this is precisely why mixing of ideal liquids is always thermodynamically spontaneous.