Gibbs Energy of Mixing - Physical Chemistry 1

SECTION 1

Historical Context & Motivation

The question of why certain substances mix spontaneously while others resist blending has occupied chemists and physicists for over two centuries. Early natural philosophers observed that gases invariably intermingle when brought into contact, yet many liquids remain stubbornly immiscible. The development of a rigorous thermodynamic framework for mixing required contributions from several intellectual traditions, spanning the formulation of the free energy concept itself to the statistical–mechanical interpretation of entropy. The Gibbs energy of mixing (Δ_mixG) ultimately provided the single thermodynamic criterion that determines whether a mixture forms spontaneously at constant temperature and pressure: mixing is favored when Δ_mixG < 0.

1824 • Carnot's Foundation

Sadi Carnot publishes Réflexions sur la puissance motrice du feu, laying the groundwork for the second law of thermodynamics and the notion that spontaneous processes are governed by energetic considerations beyond simple energy conservation.

1876 • Gibbs Free Energy Defined

J. Willard Gibbs publishes On the Equilibrium of Heterogeneous Substances, introducing the thermodynamic potential G = H − TS. This quantity provides the definitive spontaneity criterion at constant T and P and underpins the modern treatment of mixing.

1886 • Raoult's Law

François-Marie Raoult demonstrates that the partial vapor pressure of each component in an ideal solution is proportional to its mole fraction, providing an empirical foundation for the concept of an ideal mixture.

1929 • Regular Solution Theory

Joel Henry Hildebrand introduces the regular solution model, which retains ideal entropy of mixing but allows for a nonzero enthalpy of mixing through an interaction parameter. This framework bridges the gap between ideal and fully non-ideal solutions.

1941 • Flory–Huggins Theory

Paul Flory and Maurice Huggins independently develop a lattice-based model for polymer–solvent mixing, extending the Gibbs energy of mixing to systems where molecules differ greatly in size and introducing the Flory–Huggins interaction parameter χ.

The central question that the Gibbs energy of mixing answers is deceptively simple: given two or more pure substances at a specified temperature and pressure, will they form a single homogeneous phase spontaneously? If so, what composition is most stable, and how does the thermodynamic driving force depend on temperature? Answering these questions quantitatively requires decomposing Δ_mixG into its enthalpic and entropic contributions — a decomposition that reveals fundamentally different behavior for ideal solutions, regular solutions, and real solutions.

SECTION 2

Core Principles & Definitions

The Gibbs energy of mixing is defined as the change in Gibbs energy when pure components are combined to form a mixture at constant temperature and pressure. Formally, Δ_mixG = G_mixture − Σ n_iG_i^*, where G_i^* denotes the molar Gibbs energy of pure component i. Because Gibbs energy governs spontaneity at constant T and P, this single quantity encapsulates whether mixing is thermodynamically favorable. To fully appreciate the framework, several foundational ideas must be established.

Gibbs Energy & Spontaneity

At constant temperature and pressure, a process is spontaneous if and only if ΔG < 0. The Gibbs energy combines the system's enthalpy and entropy through G = H − TS, capturing the competition between energetic stability and dispersal of energy.

Ideal vs. Non-Ideal Mixing

An ideal solution is one in which all intermolecular interactions (A–A, B–B, A–B) are energetically equivalent. In such a system Δ_mixH = 0, and mixing is driven entirely by entropy. Non-ideal solutions have Δ_mixH ≠ 0.

Entropy of Mixing

For an ideal mixture, the entropy of mixing Δ_mixS = −nR Σ x_i ln x_i is always positive because mole fractions lie between 0 and 1, making each logarithm negative. This guarantees an entropic driving force toward mixing.

Chemical Potential & Composition

The chemical potential μ_i = (∂G/∂n_i)_T,P,n_j≠i connects Δ_mixG to measurable quantities such as vapor pressure and activity coefficients. Equilibrium demands equal chemical potentials across phases.

✦

KEY TAKEAWAY

Think of mixing like shuffling two decks of differently colored cards together. Even if no card 'prefers' being next to any other (Δ_mixH = 0), there are astronomically more ways to arrange a shuffled deck than two separated stacks. That enormous increase in configurational arrangements is the entropy of mixing, and it is the fundamental reason ideal gases and ideal solutions always mix spontaneously: the system accesses a vastly larger number of microstates.

SECTION 3

Visual Explanation — Δ_mix G for a Binary Ideal Solution

The following diagram illustrates how the Gibbs energy of mixing for an ideal binary solution varies with composition. The horizontal axis represents the mole fraction of component B (x_B), running from 0 (pure A) to 1 (pure B). The vertical axis shows Δ_mixG normalized per mole of mixture. Because the entropy of mixing is always positive and the enthalpy of mixing vanishes for an ideal solution, the curve is everywhere negative and symmetric about x_B = 0.5, where it reaches its most negative value of −RT ln 2 ≈ −1.72 kJ mol⁻¹ at 298 K.

Figure 1: The Gibbs energy of mixing for an ideal binary solution plotted against mole fraction x_B at T = 298 K. The curve is always negative and symmetric, with a minimum at x_B = 0.5 corresponding to −RT ln 2. The shaded region under the curve represents the thermodynamic driving force for mixing; the deeper the curve, the stronger the tendency to form a homogeneous mixture.

Several key features deserve attention. First, the curve touches zero at both endpoints (x_B = 0 and x_B = 1), because a pure substance mixed with nothing of course experiences no change in Gibbs energy. Second, the slope of the curve is infinite at both endpoints — a mathematical consequence of the logarithmic terms — meaning that even a trace of a second component produces a disproportionately large drop in Gibbs energy. Third, because the curve is concave everywhere for an ideal solution, the mixture is stable at all compositions; there is no composition range where phase separation would lower G. This last feature is precisely what changes for non-ideal systems with positive Δ_mixH.

SECTION 4

Mathematical Framework

The derivation of the Gibbs energy of mixing begins with the general thermodynamic identity Δ_mixG = Δ_mixH − TΔ_mixS. For an ideal solution, Δ_mixH = 0 and Δ_mixV = 0, so the entire thermodynamic driving force originates in entropy. We derive the entropy of mixing from a statistical-mechanical argument, then extend the treatment to non-ideal systems.

GIBBS ENERGY OF MIXING (GENERAL)

Δ_mix G = Δ_mix H − T Δ_mix S

This master equation decomposes the free energy change on mixing into an enthalpic contribution (Δ_mixH, arising from changes in intermolecular interactions) and an entropic contribution (−TΔ_mixS, reflecting the increase in molecular disorder).

IDEAL ENTROPY OF MIXING (BINARY)

Δ_mix S = −nR (x_A ln x_A + x_B ln x_B)

Here n is the total moles of mixture, R = 8.314 J mol⁻¹ K⁻¹ is the gas constant, and x_A, x_B are mole fractions. Because 0 < x_i < 1, each ln x_i < 0, making Δ_mixS > 0 for all compositions.

IDEAL GIBBS ENERGY OF MIXING (BINARY)

Δ_mix G^ideal = nRT (x_A ln x_A + x_B ln x_B)

Setting Δ_mixH = 0 and substituting the ideal entropy of mixing yields this key result. Because each logarithmic term is negative, Δ_mixG^ideal < 0 at every composition — ideal solutions always mix spontaneously.

REGULAR SOLUTION MODEL (BINARY)

Δ_mix G = nRT (x_A ln x_A + x_B ln x_B) + n w x_A x_B

The regular solution model adds an interaction parameter w (sometimes written as Ω) that captures the excess enthalpy: Δ_mixH = n w x_A x_B. When w > 0 (endothermic mixing), the enthalpy opposes entropy; when w < 0 (exothermic mixing), both terms favor the mixed state. For w > 2RT, the Δ_mixG curve develops a double-well shape, signaling a miscibility gap.

ℹ

Derivation Insight

The ideal entropy of mixing can be derived rigorously from statistical mechanics. For an ideal gas mixture, Δ_mixS equals the sum of the individual isothermal expansions of each gas from its initial volume into the total volume: Δ_mixS = −Σ n_iR ln(V_i/V_total). Using V_i/V_total = x_i (for ideal gases at equal T and P) recovers the standard formula. The same expression holds for ideal liquid solutions by analogy, assuming random mixing on a lattice.

SECTION 5

Ideal vs. Non-Ideal Mixing — Effect of the Interaction Parameter

The behavior of Δ_mixG changes dramatically depending on the sign and magnitude of the interaction parameter w in the regular solution model. When w = 0, we recover the ideal solution. When w is moderately positive, the enthalpy of mixing partially opposes the entropy, reducing the magnitude of Δ_mixG but not enough to make it positive anywhere — the system still forms a single phase. However, when w exceeds a critical value of 2RT, the Gibbs energy curve develops a double-well structure with two minima and a central maximum, indicating a miscibility gap where the system separates into two phases.

Figure 2: Comparison of Δ_mixG versus x_B for a regular binary solution at four values of the interaction parameter w. For w = 0 (ideal, cyan), the curve is everywhere negative. At w = 2RT (amber), the curve flattens to a critical point. For w = 3RT (red), a double-well structure emerges, signaling a thermodynamically unstable composition range and a miscibility gap.

Regime	w value	Δ_mix H	Δ_mix G curve shape	Phase behavior
Ideal	0	0	Single minimum at x = 0.5, concave everywhere	Complete miscibility
Endothermic (moderate)	0 < w < 2RT	> 0	Single minimum, shallower than ideal	Complete miscibility
Critical	w = 2RT	> 0	Flat inflection at x = 0.5	Onset of immiscibility (upper critical solution T)
Strongly endothermic	w > 2RT	≫ 0	Double-well with two minima and a central maximum	Miscibility gap; two coexisting phases
Exothermic	w < 0	< 0	Deeper minimum than ideal	Strong mixing tendency; may form ordered compounds

Table 1: Summary of mixing behavior as a function of the regular-solution interaction parameter w.

SECTION 6

Worked Example

Gibbs Energy of Mixing for an Ideal Binary Solution

1
Step 1 — Identify Given Values

Consider mixing 2.00 mol of liquid benzene (A) with 3.00 mol of liquid toluene (B) at T = 298 K. The benzene–toluene system is a classic example of a nearly ideal solution because both molecules are aromatic hydrocarbons of similar size and polarity. We treat the mixture as ideal, so Δ_mixH = 0.

n_A = 2.00 mol, n_B = 3.00 mol, T = 298 K
2
Step 2 — Calculate Mole Fractions

n_total = 2.00 + 3.00 = 5.00 mol. Therefore x_A = 2.00/5.00 = 0.400 and x_B = 3.00/5.00 = 0.600.

x_A = 0.400, x_B = 0.600
3
Step 3 — Evaluate Logarithmic Terms

ln(0.400) = −0.9163 and ln(0.600) = −0.5108. The sum x_A ln x_A + x_B ln x_B = (0.400)(−0.9163) + (0.600)(−0.5108) = −0.3665 − 0.3065 = −0.6730.

Σ x_i ln x_i = −0.6730
4
Step 4 — Compute Δ_mix G

Δ_mixG = nRT(Σ x_i ln x_i) = (5.00 mol)(8.314 J mol⁻¹ K⁻¹)(298 K)(−0.6730) = (5.00)(2477.6 J mol⁻¹)(−0.6730) = −8336 J = −8.34 kJ.

Δ_mixG = −8.34 kJ
5
Step 5 — Interpret the Result

The large negative value of Δ_mixG confirms that benzene and toluene mix spontaneously at 298 K. On a per-mole-of-mixture basis, Δ_mixG/n = −8.34 kJ / 5.00 mol = −1.67 kJ mol⁻¹, close to but slightly less negative than the symmetric minimum of −RT ln 2 = −1.72 kJ mol⁻¹ because the mixture is not equimolar.

Δ_mixG per mole = −1.67 kJ mol⁻¹

SECTION 7

Strengths & Limitations of the Ideal and Regular Solution Models

The ideal solution model provides an elegant and analytically tractable baseline, but real systems frequently deviate from its predictions. Understanding the strengths and limitations of both the ideal and regular solution treatments is essential for choosing the appropriate model in research and engineering contexts.

Feature	Ideal Solution Model	Regular Solution Model
Δ_mix H	Exactly zero by assumption	Non-zero: Δ_mix H = w $x_A$ $x_B$
Δ_mix S	Ideal (random mixing entropy)	Assumed equal to ideal (no excess entropy)
Excess Gibbs energy	$G^E$ = 0	$G^E$ = w $x_A$ $x_B$ (symmetric, one-parameter)
Predicts phase separation?	No — always predicts complete miscibility	Yes, when w > 2RT
Best applied to	Mixtures of chemically similar species (benzene/toluene, isotopic mixtures)	Mixtures with moderate intermolecular interaction differences (some metal alloys, simple organic pairs)
Key limitation	Cannot account for any enthalpy effects or asymmetric behavior	Assumes symmetric excess properties and no excess entropy; fails for strongly non-ideal or asymmetric systems

Table 2: Comparison of the ideal and regular solution models for Δ_mix G.

✦

KEY TAKEAWAY

The ideal solution model is analogous to the ideal gas law — it works well for 'well-behaved' systems and provides the reference state against which all real behavior is measured. Just as we define departure functions from ideal gas behavior, we define excess Gibbs energy G^E = Δ_mixG − Δ_mixG^ideal to quantify deviations from ideal mixing. More sophisticated models like Margules, van Laar, Wilson, NRTL, and UNIQUAC equations progressively relax the assumptions of the regular solution model to handle asymmetric and multi-component systems.

SECTION 8

Connection to Phase Diagrams & Advanced Theory

The Gibbs energy of mixing is not merely an abstract thermodynamic quantity; it is the computational engine behind binary phase diagrams. The common tangent construction on a Δ_mixG versus composition curve identifies the compositions of coexisting phases at equilibrium. The condition that a tangent line simultaneously touches the Δ_mixG curve at two points is equivalent to requiring that the chemical potential of each component be equal in both phases. By repeating the common tangent construction at many temperatures, one maps out the binodal (coexistence) curve on a T–x phase diagram.

Concept	This Lesson	Advanced Treatment
Stability criterion	Δ_mix G < 0 for spontaneous mixing	(∂²G/∂x²)_T,P > 0 for local stability; spinodal decomposition when this inequality fails
Excess properties	$G^E$ = w $x_A$ $x_B$ (regular solution)	$G^E$ modeled by Margules, Wilson, NRTL, UNIQUAC equations with multiple adjustable parameters
Number of components	Binary (two-component) systems	Multi-component systems requiring Gibbs–Duhem consistency and computational methods (CALPHAD)
Molecular model	Mean-field lattice model (random mixing)	Molecular simulation (Monte Carlo, molecular dynamics), COSMO-RS, group contribution methods (UNIFAC)
Phase equilibria	Qualitative identification of miscibility gap	Quantitative VLE, LLE, SLE calculations via activity coefficients and fugacities

Table 3: This lesson's treatment versus advanced approaches.

The spinodal and binodal curves, the upper and lower critical solution temperatures (UCST and LCST), and the detailed topology of ternary and higher-order phase diagrams all emerge naturally from extensions of the Gibbs energy of mixing framework. In materials science, the CALPHAD method (Computer Coupling of Phase Diagrams and Thermochemistry) constructs multi-component Gibbs energy surfaces from assessed thermodynamic data, enabling predictive alloy and materials design. Mastering the binary ideal and regular solution cases covered in this lesson is the essential first step toward these powerful techniques.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

Explain why the Gibbs energy of mixing for an ideal solution is always negative at any composition between 0 and 1, despite the enthalpy of mixing being zero. What is the microscopic origin of this driving force?

Show answer

For an ideal solution, Δ_mixG = nRT(Σ x_i ln x_i). Since each mole fraction x_i is between 0 and 1, each ln x_i is negative, making the entire sum negative and therefore Δ_mixG < 0. Microscopically, this arises from the entropy of mixing: the number of distinguishable arrangements (microstates) of molecules increases enormously when two pure substances are combined, increasing S and lowering G = H − TS. The driving force is purely entropic for an ideal solution.

Question 2: BASIC CALCULATION

Calculate Δ_mixG for an equimolar ideal binary mixture (1.00 mol A + 1.00 mol B) at T = 350 K. Express your answer in kJ.

Show answer

With x_A = x_B = 0.500 and n = 2.00 mol: Δ_mixG = nRT(x_A ln x_A + x_B ln x_B) = (2.00)(8.314)(350)(0.500 × (−0.6931) + 0.500 × (−0.6931)) = (2.00)(2910)(−0.6931) = −4033 J = −4.03 kJ. At equimolar composition, Δ_mixG per mole of mixture equals −RT ln 2 = −(8.314)(350)(0.6931) = −2016 J mol⁻¹ = −2.02 kJ mol⁻¹.

Question 3: INTERMEDIATE

A binary regular solution has an interaction parameter w = 4.00 kJ mol⁻¹ at 298 K. Determine whether the system exhibits a miscibility gap at this temperature by comparing w to the critical value 2RT. Then calculate Δ_mixG (per mole of mixture) at x_B = 0.500.

Show answer

First, 2RT = 2 × 8.314 × 298 = 4955 J mol⁻¹ ≈ 4.96 kJ mol⁻¹. Since w = 4.00 kJ mol⁻¹ < 4.96 kJ mol⁻¹, the system does not exhibit a miscibility gap at 298 K — it remains fully miscible, though with significant positive deviation from ideality. For Δ_mixG per mole at x_B = 0.500: ideal part = RT(0.5 ln 0.5 + 0.5 ln 0.5) = (2478)(−0.6931) = −1717 J mol⁻¹; excess part = w × x_A × x_B = 4000 × 0.25 = 1000 J mol⁻¹. Therefore Δ_mixG = −1717 + 1000 = −717 J mol⁻¹ = −0.72 kJ mol⁻¹. The mixing is still spontaneous but much less favorable than the ideal case.

Question 4: APPLIED

In polymer processing, a polymer–solvent system can be modeled with a Flory–Huggins-type interaction parameter χ. Suppose a certain polymer–solvent pair has an effective w = 6.00 kJ mol⁻¹ at 300 K. (a) Determine the critical value of w at this temperature. (b) Based on your result, predict whether the polymer will dissolve or remain phase-separated. (c) Estimate the critical temperature T_c above which the system would become fully miscible.

Show answer

(a) 2RT = 2 × 8.314 × 300 = 4988 J mol⁻¹ ≈ 4.99 kJ mol⁻¹. (b) Since w = 6.00 kJ mol⁻¹ > 4.99 kJ mol⁻¹, the Δ_mixG curve has a double-well shape, and the system exhibits a miscibility gap. The polymer will not fully dissolve at 300 K — certain intermediate compositions will phase-separate. (c) Setting w = 2RT_c gives T_c = w/(2R) = 6000/(2 × 8.314) = 361 K. Above 361 K (assuming w is temperature-independent), the entropic term dominates at all compositions and the system becomes fully miscible. This is the upper critical solution temperature (UCST).

Question 5: CRITICAL THINKING

The regular solution model assumes that the excess entropy of mixing is zero (S^E = 0), meaning deviations from ideality are entirely enthalpic. Provide a physical argument for why this assumption might fail for (a) a mixture of molecules with very different sizes (e.g., a polymer in a small-molecule solvent) and (b) a mixture with strong, directional intermolecular interactions (e.g., hydrogen bonding). In each case, suggest how the Δ_mixG expression would need to be modified.

Show answer

(a) When molecules differ greatly in size, the assumption of random mixing on a lattice breaks down. A large polymer chain occupies many lattice sites and constrains the configurations available to neighboring solvent molecules, reducing the actual entropy of mixing below the ideal prediction. The Flory–Huggins model addresses this by replacing mole fractions in the entropy term with volume fractions: Δ_mixS = −R(n₁ ln φ₁ + n₂ ln φ₂), yielding a smaller (less positive) entropy of mixing. (b) Strong directional interactions like hydrogen bonds create local ordering in the mixture — molecules don't arrange randomly but instead form preferred configurations. This introduces a negative excess entropy (S^E < 0) because the system is more ordered than a random mixture. Models like the Wilson equation and NRTL incorporate local composition effects through non-random factors, effectively adding composition-dependent excess entropy terms to Δ_mixG. Both cases illustrate that the regular solution model's neat separation of ideal entropy and excess enthalpy is an approximation that must be relaxed for strongly asymmetric or associating systems.

SUMMARY

Summary — Gibbs Energy of Mixing

Summary

The Gibbs energy of mixing (Δ_mixG) is the central thermodynamic quantity governing whether substances form a homogeneous mixture at constant T and P. It decomposes into an enthalpic contribution (Δ_mixH) reflecting changes in intermolecular interactions and an entropic contribution (−TΔ_mixS) reflecting the increase in molecular disorder. For an ideal solution (Δ_mixH = 0), the expression simplifies to Δ_mixG = nRT Σ x_i ln x_i, which is always negative, guaranteeing spontaneous mixing at every composition.

The regular solution model extends this framework by introducing an interaction parameter w that captures the excess enthalpy: Δ_mixH = w x_A x_B. When w > 2RT, the Gibbs energy curve develops a double-well structure, indicating a miscibility gap and phase separation. The common tangent construction on the Δ_mixG curve identifies coexisting phase compositions and links directly to binary phase diagrams. Mastery of these concepts provides the foundation for advanced activity-coefficient models, excess property analysis, and computational thermodynamics (CALPHAD).