Maximum Work and Reversibility

Push a piston out too fast and you waste part of the gas's potential to do work as useless turbulence and heat. Push it out infinitely slowly, always balanced against the surroundings, and you extract every last joule the process can give. This gap between sloppy and careful expansion is not a minor engineering detail. It is a fundamental limit written into the structure of thermodynamics itself. This idea matters because it sets the theoretical ceiling for engines, batteries, and any device that converts disorganized molecular motion into organized mechanical work. Chemists use it to define exergy, to bound cell potentials, and to justify why free energy functions exist at all. This article proves that reversible work is the maximum work obtainable between two fixed states. You will see the mechanical argument first. Then you will see the general first-law derivation. Finally you will connect it to gas expansion, foreshadowing the isothermal ideal-gas case.

Key takeaways

For a given pair of initial and final states, no process delivers more work to the surroundings than a reversible one.
Irreversible expansion against a fixed external pressure always does less work than reversible expansion against a continuously adjusted pressure.
The proof rests on the first law: internal energy change is fixed by the endpoints, so more work out means more heat must flow in.
The second law caps how much heat can flow in reversibly, which caps the work.
This result underlies the definitions of Gibbs and Helmholtz free energy as "maximum non-expansion work" functions.
Key idea: because $\Delta U$ depends only on the endpoints, any irreversible loss of work reappears as heat that a reversible path could have converted into more work instead.

The core insight

Reversible work is maximum work because the internal energy drop between two fixed states is a constant budget; a reversible path converts the largest possible share of that budget into work by never squandering potential through unbalanced, dissipative driving forces.

The physical picture

Imagine a cylinder of gas expanding from volume $V_1$ to $V_2$ at constant temperature, prerequisite background: [Reversible vs Irreversible Processes]. There are two extreme ways to run this expansion. In the irreversible case, you suddenly drop the external pressure to some fixed low value $p_{\text{ext}}$ and let the gas rush out against it. The gas does work $p_{\text{ext}}(V_2-V_1)$ , and because $p_{\text{ext}}$ is smaller than the gas's own pressure for almost the whole expansion, this work is small. The mismatch between internal and external pressure is dissipated as unrecoverable turbulence and eventually appears as extra heat that must be dumped to the surroundings; no work is captured from it. In the reversible case, you decrease $p_{\text{ext}}$ in infinitesimal steps, always keeping it equal to the gas pressure $p$ (or infinitesimally below it). At every stage the piston is in near-equilibrium. Work is done in the maximally efficient way, with no internal pressure gradients wasted as friction-like dissipation. Integrating $p\,dV$ using the actual equation of state gives a larger total than any single-step external pressure can match, because $p \ge p_{\text{ext}}$ at every point along an expansion. The key geometric fact: on a pressure-volume diagram, the work done by the gas is the area under the path from $V_1$ to $V_2$ . The reversible path traces the true equation-of-state curve $p(V)$ , which lies above any constant $p_{\text{ext}}$ used in a single irreversible step. More area under the curve means more work extracted.

Figure

Work as area under p–V curves: reversible path lies above and captures more area

A second picture makes the energy bookkeeping explicit. Both paths start and end at the same $U_1$ and $U_2$ , so $\Delta U$ is identical. Whatever work is not extracted must instead leave (or fail to leave) as heat, since $\Delta U = q + w$ is fixed.

Figure

Same ΔU, different q–w split: irreversible path sacrifices work for wasted heat

The core relations

\Delta U = q + w

where: $\Delta U$ is fixed by the initial and final states only; $q$ is heat absorbed by the system and $w$ is work done on the system (IUPAC sign convention).

w_{\text{by system}} = -w = \int_{V_1}^{V_2} p_{\text{ext}}\, dV

where: for expansion ( $V_2 > V_1$ ) this integral is the work delivered to the surroundings.

w_{\text{rev, by system}} = \int_{V_1}^{V_2} p\, dV \ \ge\ \int_{V_1}^{V_2} p_{\text{ext}}\, dV = w_{\text{irr, by system}}

where: $p$ is the system's own (equilibrium) pressure at each instant; the inequality holds for any expansion because $p_{\text{ext}} \le p$ is required for the gas to expand at all.

w_{\text{by system}}^{\max} = w_{\text{rev, by system}}

where: the maximum work obtainable in expansion between fixed endpoints is the reversible work.

The core relations

Derivation AReversible work is maximum work

Goal. Show that for a system moving between the same initial state 1 and final state 2, no irreversible path can deliver more work to the surroundings than the reversible path connecting the same two states.

Step 1: fix the energy budget. Internal energy $U$ is a state function. For any process, reversible or not, connecting the same state 1 to the same state 2,

\Delta U = U_2 - U_1

is exactly the same number regardless of path. This is the entire leverage of the argument: the total energy released or absorbed by the system is path-independent, even though how it is split between heat and work is not.

Step 2: write the first law for an arbitrary infinitesimal step. For expansion work only,

dU = \delta q - p_{\text{ext}}\, dV

where $p_{\text{ext}}$ is whatever external pressure is actually resisting the piston at that instant — it need not equal the gas's own pressure $p$ if the process is irreversible.

Step 3: compare $p_{\text{ext}}$ to $p$ . For the gas to expand at all, the surroundings must be pushing back with $p_{\text{ext}} \le p$ ; otherwise the piston would not move outward. Equality holds only in the reversible limit, where the process proceeds through a continuous sequence of equilibrium states with $p_{\text{ext}} = p - dp \approx p$ . Any finite mismatch $p_{\text{ext}} < p$ defines an irreversible process.

Step 4: integrate the work. The work done by the system on the surroundings over the whole path is

w_{\text{by system}} = \int_{V_1}^{V_2} p_{\text{ext}}\, dV.

Since $p_{\text{ext}} \le p$ pointwise, and $p$ is the same equilibrium function of $V$ regardless of how fast we run the process (for a fixed $T$ -path or fixed intermediate states),

\int_{V_1}^{V_2} p_{\text{ext}}\, dV \;\le\; \int_{V_1}^{V_2} p\, dV.

The right-hand side is attained only when $p_{\text{ext}} = p$ throughout, i.e. only for the reversible path.

Step 5: translate into heat. Combine with Step 1. Since $\Delta U$ is fixed,

q = \Delta U + w_{\text{by system}}.

If $w_{\text{by system}}$ is smaller (irreversible case), then to keep $\Delta U$ fixed, less heat need be supplied to the system — or equivalently, the "lost" work reappears as extra heat that must be rejected or is simply never converted. Either way, the reversible path is the unique path that maximizes the fraction of $|\Delta U|$ that appears as usable work rather than as heat exchanged with a single reservoir.

Step 6: generalize beyond gas expansion. The same argument holds for electrical work, e.g. a battery driving a current against an external voltage $E_{\text{ext}} \le E_{\text{cell}}$ . Reversible discharge (near-zero current, back-EMF nearly balancing $E_{\text{cell}}$ ) extracts the maximum electrical work,

w_{\text{elec}}^{\max} = -nFE_{\text{cell}},

while any finite-current (irreversible) discharge delivers less, the difference appearing as resistive (Joule) heating inside the cell.

Boxed result.

\boxed{w_{\text{by system}} \le w_{\text{rev, by system}} = \int_{V_1}^{V_2} p\, dV \quad \text{(maximum work theorem)}}

Worked examples

Example 1 — Foundational: single-step vs reversible isothermal expansion of an ideal gas. One mole of an ideal gas at $T = 298\ \text{K}$ expands from $V_1 = 10.0\ \text{L}$ to $V_2 = 30.0\ \text{L}$ . (a) Irreversible, single-step expansion against constant $p_{\text{ext}} = 1.00\ \text{bar}$ . Work done by the gas:

w_{\text{by gas}} = p_{\text{ext}}(V_2 - V_1) = (1.00\times10^5\ \text{Pa})(30.0 - 10.0)\times10^{-3}\ \text{m}^3

w_{\text{by gas}} = (1.00\times10^5)(0.0200) = 2.00\times10^{3}\ \text{J} = 2.00\ \text{kJ}.

(b) Reversible isothermal expansion.

w_{\text{by gas, rev}} = \int_{V_1}^{V_2} \frac{nRT}{V}\, dV = nRT \ln\!\left(\frac{V_2}{V_1}\right)

= (1.00\ \text{mol})(8.314\ \text{J K}^{-1}\text{mol}^{-1})(298\ \text{K}) \ln\!\left(\frac{30.0}{10.0}\right)

= (2477.8\ \text{J})(\ln 3.00) = (2477.8\ \text{J})(1.0986) = 2.722\times10^{3}\ \text{J} = 2.72\ \text{kJ}.

Check: $2.72\ \text{kJ} > 2.00\ \text{kJ}$ , confirming the reversible path extracts more work for the identical endpoints, exactly as the theorem requires. Since $\Delta U = 0$ for isothermal ideal-gas expansion, $q_{\text{rev}} = w_{\text{by gas,rev}} = 2.72\ \text{kJ}$ flows in, while $q_{\text{irr}} = 2.00\ \text{kJ}$ flows in — the reversible path pulls in more heat from the surroundings and converts all of it to work, whereas the irreversible path pulls in less heat and still wastes potential relative to what was available.

Example 2 — Applied: multi-step irreversible expansion approaching the reversible limit, and electrochemical cross-check. Take the same gas and endpoints, but now split the expansion into two irreversible steps to see convergence toward the reversible value. Step (i): expand from $10.0\ \text{L}$ to $20.0\ \text{L}$ against constant $p_{\text{ext},1}$ equal to the gas pressure at $V = 20.0\ \text{L}$ :

p_{\text{ext},1} = \frac{nRT}{V} = \frac{(1.00)(8.314)(298)}{20.0\times10^{-3}\ \text{m}^3} = \frac{2477.8}{0.0200} = 1.239\times10^{5}\ \text{Pa}.

Work in this step:

w_1 = p_{\text{ext},1}(V' - V_1) = (1.239\times10^5)(0.0100) = 1.239\times10^{3}\ \text{J}.

Step (ii): expand from $20.0\ \text{L}$ to $30.0\ \text{L}$ against constant $p_{\text{ext},2}$ equal to the gas pressure at $V = 30.0\ \text{L}$ :

p_{\text{ext},2} = \frac{(1.00)(8.314)(298)}{30.0\times10^{-3}} = \frac{2477.8}{0.0300} = 0.8259\times10^{5}\ \text{Pa}.

w_2 = p_{\text{ext},2}(V_2 - V') = (0.8259\times10^5)(0.0100) = 0.826\times10^{3}\ \text{J}.

Total two-step work:

w_{\text{total}} = w_1 + w_2 = 1.239\times10^3 + 0.826\times10^3 = 2.065\times10^{3}\ \text{J} = 2.07\ \text{kJ}.

Check against the theorem: $2.00\ \text{kJ}$ (one step) $< 2.07\ \text{kJ}$ (two steps) $< 2.72\ \text{kJ}$ (infinite steps, reversible). Adding more steps monotonically approaches — but never exceeds — the reversible value, exactly the trend the maximum work theorem predicts. Electrochemical cross-check. A galvanic cell with $E_{\text{cell}} = 1.10\ \text{V}$ (a Daniell-type cell) transfers $n = 2$ mol electrons. Reversible (zero-current) electrical work:

w_{\text{elec}}^{\max} = -nFE_{\text{cell}} = -(2)(96485\ \text{C mol}^{-1})(1.10\ \text{V}) = -2.123\times10^{5}\ \text{J} = -212.3\ \text{kJ}.

If instead the cell is discharged rapidly through a low resistor so that the working voltage sags to $E_{\text{working}} = 0.90\ \text{V}$ (internal resistance dissipates the rest as heat),

w_{\text{elec,irr}} = -nFE_{\text{working}} = -(2)(96485)(0.90) = -1.737\times10^{5}\ \text{J} = -173.7\ \text{kJ},

which in magnitude ( $173.7\ \text{kJ}$ ) is again smaller than the reversible maximum ( $212.3\ \text{kJ}$ ) — the same theorem, a different kind of work.

Limiting cases & common traps

Situation	What happens to work	Note
Reversible path, $p_{\text{ext}} = p$ at every point	$w$ is maximum for expansion, minimum in magnitude for compression	Requires infinitely slow, quasi-static driving
Single-step irreversible expansion	$w < w_{\text{rev}}$	Loss grows with the size of the pressure mismatch
Free expansion, $p_{\text{ext}} = 0$	$w = 0$	The absolute floor; no work at all is recovered
Compression (reversible vs irreversible)	Reversible work is the minimum work you must do	Sign flips: "maximum work theorem" becomes "minimum work of compression"
Increasing the number of irreversible steps	$w$ increases monotonically toward $w_{\text{rev}}$	Demonstrated numerically above (2.00 to 2.07 to 2.72 kJ)

The chemist behind it

The maximum work theorem is inseparable from the work of Sadi Carnot, who first showed that an idealized, frictionless, quasi-static heat engine sets an upper bound on convertible work between two thermal reservoirs. Rudolf Clausius and later Josiah Willard Gibbs translated this mechanical insight into the chemical thermodynamics used today, with Gibbs explicitly defining what we now call the Gibbs energy as the maximum non-expansion work obtainable at constant temperature and pressure — a direct descendant of the argument given here.

Origins · 1 minJosiah Willard Gibbs and how this was discovered

Josiah Willard Gibbs

1839–1903 · American theoretical physicist and chemist who formalized free energy as maximum available work in chemical systems

Ask the tutor: Want to see how this same maximum-work argument leads directly to the formula for entropy change in a reversible isothermal expansion?

Why is reversible work called "maximum" for expansion but "minimum" for compression?

The sign of $w$ flips with the direction of the process while the inequality $p_{\text{ext}} \le p$ for expansion becomes $p_{\text{ext}} \ge p$ for compression. In expansion you want to extract as much work as possible, and the reversible path gives the largest magnitude. In compression you must supply work, and the reversible path requires you to supply the least. Both statements are the same underlying inequality viewed from different physical goals.

Does the maximum work theorem require the process to be isothermal?

No. The proof in the derivation only assumes that $p_{\text{ext}} \le p$ pointwise during expansion; it does not fix temperature. Isothermal expansion is simply the most common textbook case because it gives a clean closed-form integral, $nRT\ln(V_2/V_1)$ . The theorem applies equally to adiabatic, non-isothermal, or electrochemical processes, as shown by the cell example above.

How does this connect to entropy and the second law?

Reversibility is defined by zero net entropy production in the universe. The maximum work theorem is a mechanical restatement of that same idea: any dissipative (entropy-generating) step converts potentially useful work into heat that cannot be fully recovered. The link becomes quantitative once you define Gibbs and Helmholtz energies, which express maximum work directly in terms of entropy and enthalpy or internal energy.

Is reversible work actually achievable in a real experiment?

Strictly, no — true reversibility requires infinitely slow driving with zero net force imbalance, which takes infinite time. Real processes always dissipate some potential as heat. The reversible value is a theoretical limit, useful the same way "frictionless" is useful in mechanics: it defines the best possible performance against which real, finite-rate processes are judged.

What comes next after this result?

The natural next step is to compute the reversible work explicitly for an ideal gas undergoing isothermal expansion, since that is the cleanest case where the integral $\int p\,dV$ has a simple closed form. That calculation, and its link to $q_{\text{rev}}$ and entropy, is developed in the discussion of isothermal expansion of an ideal gas.

Maximum Work and Reversibility

The physical picture

The core relations

The core relations

Worked examples

Limiting cases & common traps

The chemist behind it

Before this

Where this leads