Gibbs Energy and Maximum Non-Expansion Work

A charged lithium-ion cell sits on your desk. As it discharges, the chemical reaction inside does not merely warm the room — it lights a screen. That useful electrical output is bounded by a thermodynamic quantity: the Gibbs energy change of the reaction. Chemists need to know the absolute ceiling on "useful" work — work other than pushing back the atmosphere — that a process can deliver. Batteries, fuel cells, and muscle contraction all convert chemical change into non-expansion work at essentially constant temperature and pressure. The Gibbs energy sets the exact limit. This article derives why $-\Delta G$ equals the maximum non-expansion work at constant $T$ and $p$ . You will see the derivation from the combined first and second laws. Then you will apply it to a real electrochemical cell. Finally, you will see how real, irreversible operation always falls short of this limit.

Key takeaways

At constant $T$ and $p$ , $-\Delta G$ is the maximum non-expansion (e.g., electrical, mechanical-non-pV) work obtainable from a process.
The bound is only reached in the reversible limit; real processes always deliver less useful work.
Non-expansion work includes electrical work, work of extension, and work against gravity or magnetic fields — anything besides $p\Delta V$ .
The relation follows directly from combining the Clausius inequality with the first law and the definition $G = H - TS$ .
For electrochemical cells, $-\Delta G = w_{e,\max} = -nFE_{\text{cell}}$ , linking thermodynamics to measurable cell voltage.
Key idea: Gibbs energy is a "work function" for non-expansion work under the everyday constraints of fixed temperature and pressure.

The core insight

Under constant temperature and pressure, the total work a process can extract splits into unavoidable expansion work against the surroundings and everything else; the Gibbs energy change is precisely the negative of the maximum value of that "everything else," achieved only when the process runs reversibly.

The physical picture

Consider a system that can exchange heat with surroundings at temperature $T$ , expand or contract against a constant external pressure $p$ , and also deliver some other, more useful form of work — electrical current through a wire, a rotating shaft, a stretched fiber. Call this second category non-expansion work, $w_{\text{add}}$ (also written $w_e$ or $w'$ ). Total work is $w = w_{pV} + w_{\text{add}}$ , where $w_{pV} = -p\Delta V$ is the ordinary expansion work exchanged with the atmosphere. This portion is thermodynamically "wasted" in the sense that it just displaces air; it cannot run a motor or charge a capacitor. The question that Gibbs energy answers is: how much of the remaining work can we extract, at most? The answer depends on the path. Only a reversible path — one passing through a continuous sequence of equilibrium states, with no dissipation via friction, resistive heating, or finite-rate driving forces — extracts the maximum possible work of any kind. Any real, spontaneous, finite-rate process dissipates some potential work as heat internally, so it delivers less. The Gibbs energy change quantifies the reversible, maximal case: $-\Delta G$ is achieved only in the limit of a perfectly reversible transformation at constant $T$ and $p$ . This is why a real battery under load, or a muscle doing mechanical work, always operates below the thermodynamic ceiling — some chemical potential energy is inevitably lost as heat to resistive or viscous dissipation, exactly as friction reduces the work extracted from a falling weight.

Figure

Energy bookkeeping for a spontaneous process at constant T and p

Figure

Real vs reversible discharge: work delivered as a fraction of the thermodynamic limit

The core relations

dG = dH - T\,dS \quad (\text{at constant } T)

where: $H$ is enthalpy, $S$ is entropy, $T$ is absolute temperature.

w_{\text{add,max}} = \Delta G \quad \text{(sign convention: work done } \textit{on}\text{ the system)}

where: $w_{\text{add,max}}$ is the maximum non-expansion work done on the system, achieved reversibly at constant $T$ , $p$ .

w_{\text{add,max, by system}} = -\Delta G

where: this is the maximum useful work the system can deliver to the surroundings, the quantity of practical interest for engines, batteries, and cells.

\Delta G = \Delta H - T\Delta S \quad (\text{constant } T)

where: $\Delta H$ is the enthalpy change, $\Delta S$ is the total entropy change of the system, both at the temperature $T$ .

\Delta G_{\text{cell}} = -nFE_{\text{cell}}

where: $n$ is moles of electrons transferred, $F = 96{,}485\ \text{C mol}^{-1}$ is the Faraday constant, $E_{\text{cell}}$ is the cell potential.

The core derivation

Derivation A−ΔG as the maximum non-expansion work at constant T, p

Goal. Show that for a process at constant temperature $T$ and pressure $p$ , the non-expansion work done by the system on the surroundings can never exceed $-\Delta G$ , with equality only for a reversible path.

Step 1 — Start from the Clausius inequality. For any process, real or reversible, the entropy change of the system obeys

dS \geq \frac{dq}{T}

with equality only for a reversible path. Rearranged for heat,

dq \leq T\,dS.

Step 2 — Bring in the first law with two kinds of work. Write the work done on the system in two pieces: ordinary expansion work $dw_{pV} = -p\,dV$ (against a constant external pressure) and additional (non-expansion) work $dw_{\text{add}}$ . The first law gives

dU = dq + dw_{pV} + dw_{\text{add}}.

Substituting the Clausius bound on $dq$ ,

dU \leq T\,dS - p\,dV + dw_{\text{add}}.

Step 3 — Isolate the non-expansion work. Rearranging,

dw_{\text{add}} \geq dU - T\,dS + p\,dV.

This says the work that must be supplied to the system is at least this combination; equivalently, the work the system can deliver is at most the negative of it. To connect to $G$ , we now impose the constant- $T$ , constant- $p$ conditions explicitly.

Step 4 — Introduce $G = H - TS = U + pV - TS$ . At constant $T$ and constant $p$ , differentiating,

dG = dU + p\,dV - T\,dS.

Compare this with the right-hand side of the Step 3 inequality: they are identical. Therefore

dw_{\text{add}} \geq dG \quad (\text{constant } T, p).

Step 5 — Interpret the sign convention and integrate. $dw_{\text{add}}$ here is work done on the system (positive when energy flows in). The inequality states that at least $dG$ of work must be supplied to force the change if $dG > 0$ ; if $dG < 0$ , the system can instead deliver work outward. Flip perspective to work done by the system, $dw_{\text{add,by}} = -dw_{\text{add}}$ :

dw_{\text{add,by}} \leq -dG.

Integrating from initial to final states at constant $T$ , $p$ ,

w_{\text{add,by}} \leq -\Delta G.

Step 6 — Identify the equality condition. Equality holds precisely when the Clausius inequality in Step 1 is an equality — that is, when the process is carried out reversibly at every stage, with the system passing through a continuous sequence of equilibrium states. Any irreversibility (finite-rate reaction, resistive dissipation, friction, an unbalanced driving force) makes the inequality strict, and less work is obtained.

\boxed{w_{\text{add,by, max}} = -\Delta G \quad \text{(constant } T, p\text{, reversible limit)}}

Worked examples

Example 1 — The hydrogen–oxygen fuel cell. The overall cell reaction is

\mathrm{H_2(g) + \tfrac{1}{2}O_2(g) \rightarrow H_2O(l)}

at 298 K and 1 bar. Standard thermodynamic data give $\Delta H^\circ = -285.8\ \text{kJ mol}^{-1}$ and $\Delta S^\circ = -163.3\ \text{J K}^{-1}\text{mol}^{-1}$ (using $S^\circ_{\text{H}_2\text{O(l)}} = 69.9$ , $S^\circ_{\text{H}_2} = 130.7$ , $S^\circ_{\text{O}_2} = 205.2\ \text{J K}^{-1}\text{mol}^{-1}$ , all per mole). Check the entropy value:

\Delta S^\circ = 69.9 - \left(130.7 + \tfrac{1}{2}(205.2)\right) = 69.9 - 233.3 = -163.4\ \text{J K}^{-1}\text{mol}^{-1}.

Compute $\Delta G^\circ$ :

\Delta G^\circ = \Delta H^\circ - T\Delta S^\circ = -285{,}800\ \text{J mol}^{-1} - (298\ \text{K})(-163.4\ \text{J K}^{-1}\text{mol}^{-1})

= -285{,}800 + 48{,}693 = -237{,}100\ \text{J mol}^{-1} = -237.1\ \text{kJ mol}^{-1}.

By the boxed result, the maximum non-expansion (electrical) work available per mole of $\text{H}_2$ consumed is 237.1 kJ, obtainable only if the cell is discharged infinitely slowly (reversibly). This is the value tabulated as $\Delta_f G^\circ(\text{H}_2\text{O, l}) = -237.1\ \text{kJ mol}^{-1}$ , consistent with standard tables. Convert to a cell potential. With $n = 2$ mol electrons transferred per mole of reaction ( $\text{H}_2 \to 2\text{H}^+ + 2e^-$ ),

E^\circ_{\text{cell}} = \frac{-\Delta G^\circ}{nF} = \frac{237{,}100\ \text{J mol}^{-1}}{(2)(96{,}485\ \text{C mol}^{-1})} = \frac{237{,}100}{192{,}970}\ \text{V} = 1.229\ \text{V}.

This matches the well-known standard potential of the hydrogen–oxygen fuel cell, 1.23 V. Sanity check: the sign is positive, consistent with a spontaneous, work-producing reaction ( $\Delta G^\circ < 0$ ).

Example 2 — Real (irreversible) discharge and the efficiency gap. Suppose the same fuel cell, discharged at a practical current, delivers a working terminal voltage of only $0.70\ \text{V}$ per cell instead of the reversible $1.229\ \text{V}$ , due to overpotentials (activation losses, ohmic resistance, concentration polarization). The actual electrical work delivered per mole of $\text{H}_2$ is

w_{\text{add, actual}} = nFE_{\text{actual}} = (2)(96{,}485\ \text{C mol}^{-1})(0.70\ \text{V}) = 135{,}079\ \text{J mol}^{-1} = 135.1\ \text{kJ mol}^{-1}.

Compare to the maximum:

\text{efficiency} = \frac{w_{\text{add,actual}}}{-\Delta G^\circ} = \frac{135.1\ \text{kJ mol}^{-1}}{237.1\ \text{kJ mol}^{-1}} = 0.570, \quad \text{i.e. } 57.0\%.

The "lost" work, $237.1 - 135.1 = 102.0\ \text{kJ mol}^{-1}$ , is not destroyed — energy is conserved — but it is degraded: it appears as extra heat released to the surroundings beyond the reversible heat exchange $T\Delta S$ , because the internal irreversibility (Step 1 inequality, strict) generates additional entropy. This illustrates precisely why $-\Delta G$ is a ceiling, not a promise: real fuel cells and batteries always operate at some voltage below $E^\circ$ , and the shortfall is the thermodynamic cost of running at a finite, useful rate.

Limiting cases & common traps

Situation	What −ΔG tells you	Common pitfall
Reversible, constant $T,p$	$-\Delta G$ exactly equals maximum non-expansion work	Assuming any real device reaches this — it never quite does
Irreversible, constant $T,p$	$-\Delta G$ is an unreachable upper bound; actual work is smaller	Forgetting that $\Delta G$ itself is a state function value, unaffected by the path's reversibility
No non-expansion work possible (e.g. simple gas expansion)	$-\Delta G$ still computable, but irrelevant to $w_{pV}$	Confusing $-\Delta G$ with $-\Delta A$ (which bounds total work, including $pV$ work, at constant $T$ )
Constant $T,V$ instead of $T,p$	Use Helmholtz energy $A$ instead: $w_{\max} = -\Delta A$	Using $\Delta G$ when pressure is not constant
$\Delta G = 0$	System at equilibrium; no net non-expansion work extractable	Thinking $\Delta G=0$ means "nothing happens" rather than "no net driving force"

The chemist behind it

Josiah Willard Gibbs developed the entire framework of chemical thermodynamics — including the free energy function that bears his name — in a series of extraordinarily dense papers in the 1870s, working largely in isolation at Yale. He recognized that combining energy and entropy into a single function evaluated at fixed temperature and pressure would predict spontaneity and quantify available work in one stroke, unifying mechanics, heat, and chemical change decades before the electron or the modern battery existed.

Origins · 1 minJosiah Willard Gibbs and how this was discovered

Josiah Willard Gibbs

1839–1903 · American mathematical physicist who founded chemical thermodynamics and defined the free energy that governs spontaneity and available work

Ask the tutor: How does coupling a highly exergonic reaction to an unfavorable one let cells "spend" −ΔG to drive biosynthesis uphill?

Why does the bound require both constant T and constant p, not just one

Both constraints enter the derivation directly. Constant $T$ lets the Clausius inequality term $T\,dS$ be integrated cleanly using a single temperature. Constant $p$ makes $dG = dU + p\,dV - T\,dS$ match exactly the rearranged first-law expression from Step 3 of the derivation. If pressure varies, $G$ is still defined, but $\Delta G$ no longer equals the extractable non-expansion work; you would need to track the $p\,dV$ term explicitly instead of letting it cancel.

Is −ΔG the work delivered, or just an upper limit

It is strictly an upper limit, achieved only in the reversible limit — an idealization of infinitely slow, frictionless, dissipation-free operation. Every real process, run at a finite rate, delivers strictly less non-expansion work than $-\Delta G$ . The gap represents entropy generated internally by irreversibility, which converts potential work into extra low-grade heat rather than useful output.

How does this relate to the maximum work concept from Helmholtz energy

At constant temperature and volume, an analogous derivation shows $-\Delta A$ bounds the total maximum work, including ordinary expansion work. Gibbs energy specializes this idea to the more common laboratory and biological condition of constant pressure, separating out the unavoidable $p\Delta V$ contribution so that $-\Delta G$ isolates only the "useful" work — the quantity relevant to electrochemical cells, muscles, and industrial engines operating open to the atmosphere.

Does a negative ΔG guarantee the reaction actually happens quickly

No. $\Delta G < 0$ only guarantees that the reaction is thermodynamically spontaneous and that work could in principle be extracted; it says nothing about rate. Kinetically hindered reactions (like diamond converting to graphite, or hydrogen and oxygen mixing without a spark) can have very negative $\Delta G$ yet proceed immeasurably slowly without a catalyst or ignition source. Thermodynamics sets the destination and the ceiling on work; kinetics sets the timescale.

How does this connect to coupled reactions in biochemistry

Living cells routinely use a reaction with very negative $\Delta G$ , such as ATP hydrolysis, to drive an otherwise unfavorable reaction with positive $\Delta G$ , by mechanically or chemically coupling the two so that the combined $\Delta G$ is negative. The "non-expansion work" extracted from the favorable half-reaction is spent directly on the unfavorable half — a preview of coupled-reaction thermodynamics built entirely on the maximum-work interpretation developed here.