Gibbs free energy in three intuitions

Gibbs free energy is the single number that says whether a reaction goes forward at constant temperature and pressure. If ΔG < 0, the reaction is spontaneous; if ΔG = 0, it is at equilibrium; if ΔG > 0, it is spontaneous in the reverse direction. That description is correct, portable, and — for most students — deeply unsatisfying. This guide gives three intuitions for what G is actually doing, each of which is useful in a different situation. Once all three click, you can look at a reaction, guess the sign of ΔG, and be right almost every time.

Intuition 1: G is enthalpy corrected for the universe’s entropy tax

The defining relation, first written this way by Gibbs (1878):

G = H − TS

At constant T and P,

ΔG = ΔH − T ΔS

Read that literally. ΔH is the heat released or absorbed by the reaction — from the reaction’s point of view. ΔS is the entropy change of the reaction system alone. The T ΔS term takes both into account. The correction has a purpose: it converts everything into what matters to the universe.

The second law says a spontaneous process must increase the total entropy of the universe (system + surroundings). Since the surroundings’ entropy change is essentially −ΔH_system / T (heat leaving the system enters the surroundings, so the surroundings gain entropy proportional to enthalpy released divided by temperature),

ΔS_universe = ΔS_system − ΔH_system / T

Multiply that by −T:

−T ΔS_universe = ΔH_system − T ΔS_system = ΔG_system

So ΔG < 0 is exactly ΔS_universe > 0, at constant T and P. ΔG is not a system property with a mysterious extra term — it is a proxy for the universe’s entropy change that you can compute from the system’s own data. That is the entire theoretical justification for the free-energy criterion.

This intuition is best when a student is stuck asking “why does the T ΔS term exist” — the answer is “because entropy of the surroundings is real, and this converts it into system quantities you can measure or tabulate.”

Intuition 2: G is the maximum non-expansion work the reaction can deliver

Rewrite the first and second laws for a closed system at constant T:

dG = dH − T dS = dU + P dV − T dS = δq − δw_pv − δw_other + P dV − T dS

At constant T and P for a reversible process, δq_rev = T dS and δw_pv = P dV. That collapses to

dG = − δw_other

or, integrating,

ΔG = −w_max,non-PV

ΔG for a chemical reaction is the maximum useful work you can extract from it while running it reversibly at constant T and P. The “non-PV” qualifier matters: expansion work (pushing back the atmosphere) is not counted, because the reaction has no choice about paying that. What is counted is the electrical work you can extract from a battery, the mechanical work a muscle can do from ATP hydrolysis, the shaft work a fuel cell can deliver.

For a reaction with ΔG = −237 kJ/mol (hydrogen combustion at 25 °C, CODATA 2018 tables), that is the largest possible amount of electrical work per mole of H₂ you can get out of a hydrogen fuel cell. Real fuel cells hit ~60% of this because they run irreversibly (fast reactions do not stay on the equilibrium track). But no fuel cell can beat 237 kJ/mol — that ceiling is thermodynamic, not engineering.

This intuition is the best one when a student asks why they should care about ΔG in the first place. The answer is: it is the payout on the reaction as a work engine.

Intuition 3: G is the driving force away from equilibrium

The third framing is the most operationally useful and the one that hooks into every equilibrium-constant problem you will ever solve. Under non-standard conditions,

ΔG = ΔG° + RT ln Q

where Q is the reaction quotient (same form as K, but computed from whatever the concentrations or partial pressures actually are, not their equilibrium values). At equilibrium Q = K and ΔG = 0, so

ΔG° = −RT ln K

That equation is the whole bridge between thermodynamics and equilibrium. A reaction with a strongly negative ΔG° has a large K — products dominate at equilibrium. A reaction with ΔG° ≈ 0 sits at K ≈ 1 — no strong preference. A reaction with a strongly positive ΔG° has K ≪ 1 — reactants dominate.

The RT ln Q term is what makes non-standard-state reactions runnable. Even if ΔG° > 0, you can drive the reaction forward by keeping Q low (removing products, adding reactants). This is Le Chatelier in one equation — the ΔG° tells you the standard-state preference, and RT ln Q tells you how far you can move the goalposts.

At 25 °C, RT = 8.314 × 298.15 J/mol = 2478 J/mol, or 2.478 kJ/mol. Multiplied by ln 10 ≈ 2.303, that gives 5.706 kJ/mol per decade of Q. Every factor of 10 shift in Q moves ΔG by ~5.7 kJ/mol at room temperature. That is a useful sanity-check ruler for how much a concentration change can bend a thermodynamic preference.

This intuition is best when a student needs to relate ΔG to something they can see — an equilibrium constant, a concentration ratio, a fuel-cell voltage.

Sign combinations of ΔH and ΔS

Because ΔG = ΔH − T ΔS depends on T, there are four regimes:

• ΔH < 0, ΔS > 0 — spontaneous at every T. Combustion is the canonical example: exothermic (energy out, ΔH < 0) and entropy-increasing (fuel + O₂ → many gas molecules, ΔS > 0). You cannot un-burn wood by cooling it.
• ΔH > 0, ΔS < 0 — non-spontaneous at every T. The reverse of combustion. No temperature makes it go.
• ΔH < 0, ΔS < 0 — spontaneous at low T. Freezing of water at 1 atm below 273.15 K: releases enthalpy (crystallization is exothermic) but decreases entropy (ordered lattice). Cold enough, the enthalpy term wins.
• ΔH > 0, ΔS > 0 — spontaneous at high T. Melting of ice above 273.15 K, dissolution of NH₄NO₃ in water (endothermic but entropy-driven), most cracking reactions of large hydrocarbons.

The temperature at which ΔG changes sign in the mixed cases is exactly T = ΔH / ΔS. For water freezing, ΔH_fus = 6.01 kJ/mol and ΔS_fus = 22.0 J K⁻¹ mol⁻¹, so T_change = 6010 / 22.0 = 273.2 K — the standard freezing point (Atkins 11e Table 3.3). The whole first-year formulation of “why does ice melt at 0 °C” is the temperature at which ΔG_fus crosses zero.

A canonical worked example

The Haber-Bosch case is the classic illustration that thermodynamics is only half the story: a favourable ΔG says the reaction can happen, not that it will on any human timescale. Everything downstream (catalysis, activation energy, transition-state theory — see the half-life calculator for first-order rate analysis) is the other half.

Where does the Nernst equation come in?

The Nernst equation is intuition 2 and intuition 3 combined for an electrochemical cell:

ΔG = −nFE and ΔG = ΔG° + RT ln Q

Substituting,

−nFE = −nFE° + RT ln Q E = E° − (RT / nF) ln Q

The cell voltage is a direct measurement of ΔG per unit charge. Every millivolt on a battery is 96.485 J of free energy per mole of electrons (F = 96 485.332 12 C/mol, CODATA 2018). If your battery reads 0 V, that reaction is at equilibrium — no more work extractable. The Nernst equation calculator is the shortest bridge from ΔG° to a lab-measurable voltage.

Practice problems

Where to go next

• The Nernst equation calculator is the fastest way to see intuition 2 turn into a voltage.
• The ideal gas law calculator covers the T, P, V corrections you need when a reaction shifts moles of gas.
• The Ka/Kb cheatsheet has enough values to run ΔG° = −RT ln K on nearly any weak-acid/base equilibrium you meet in solution.

Once you can flip fluently between the three framings — universe’s entropy proxy, max non-PV work, distance from equilibrium — Gibbs free energy stops feeling like an abstract accounting device and starts being the number you reach for first when someone asks “will this reaction go?”