Helmholtz Energy and Maximum Work

Squeeze a gas isothermally in a cylinder and it can turn a crank, charge a spring, or run a tiny motor. How much useful work can you ever extract? The Helmholtz energy answers this question exactly. It is not just an abstract combination of $U$ and $S$ ; it is a work-accounting function. This matters because chemists constantly ask how much work a reaction, a battery, or an expanding gas can deliver at fixed temperature. The answer is never "all of the internal energy released." Some energy must be paid to the surroundings as heat to keep entropy bookkeeping honest, and the Helmholtz energy tells you exactly how much is left over. This article builds the maximum-work theorem from the Clausius inequality. You will see the proof in full. Then you will see why $-\Delta A$ is the ceiling on total work, expansion plus everything else. Finally two worked numerical examples pin the idea down, before pointing toward the Gibbs energy and non-expansion work.

Key takeaways

The Helmholtz energy is defined as $A = U - TS$ , building on the Helmholtz energy concept already introduced.
At constant temperature, the total work done on a system in a reversible process equals $\Delta A$ .
For any real (possibly irreversible) process at constant $T$ , the work done by the system can never exceed $-\Delta A$ .
$-\Delta A$ is the maximum total work, including expansion work, obtainable from a process — this is why $A$ is called the "work function" (Arbeit).
Irreversibility always wastes potential work as unrecoverable heat dumped to the surroundings, never violating the Second Law.
Key idea: at constant temperature, $-\Delta A$ sets an unbeatable ceiling on total work extractable from a process, reversibility alone achieves it.

The core insight

At constant temperature, the decrease in Helmholtz energy, $-\Delta A$ , equals the maximum total work (expansion plus non-expansion) that a process can possibly deliver, with equality only in the reversible limit.

The physical picture

Picture an isothermal gas expansion inside a cylinder connected to a large heat reservoir at temperature $T$ . As the gas expands it can push a piston (expansion work) and, in principle, also drive a shaft connected to a generator or a chemical cell doing electrical work (non-expansion work). The question "how much total work can we extract?" is really a question about how efficiently we can convert the energy released into work rather than heat. The First Law says the internal energy change $\Delta U$ is fixed by the initial and final states — it doesn't care about the path. But work and heat individually do depend on the path: a reversible, quasi-static expansion delivers more work and absorbs more heat from the reservoir than a fast, irreversible one delivering the same $\Delta U$ . The Second Law, through the Clausius inequality, puts a hard ceiling on how much heat can be absorbed (and hence how much work can be produced) at a given temperature. That ceiling is expressed most cleanly through $A = U - TS$ . Think of $A$ as the portion of $U$ that is "free" to be converted into work at constant $T$ ; the remaining part, $TS$ , is energy that must stay behind, effectively rented from the surroundings as heat to keep the entropy books balanced. Only a perfectly reversible process collects every last joule of that free portion as work. Any real, dissipative process — friction, an internal chemical reaction running away from equilibrium, an electric current pushed against resistance — throws part of that potential work away as extra heat, satisfying $dS_{\text{univ}} \geq 0$ but delivering less to you.

Figure

Reversible vs irreversible isothermal expansion: same ΔU, different work delivered

Figure

Energy bookkeeping at constant T: U splits into A (free to do work) and TS (retained)

The core relations

A \equiv U - TS

where: $A$ is the Helmholtz energy, $U$ the internal energy, $T$ the absolute temperature, and $S$ the entropy.

dw \leq dA \quad (\text{constant } T)

where: $dw$ is the work done on the system in an infinitesimal step; equality holds only for a reversible change.

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

where: $w_{\text{by system, max}}$ is the maximum total work (expansion + non-expansion) obtainable, achieved only reversibly.

\Delta A = \Delta U - T\Delta S \quad (\text{constant } T)

where: this is the working formula for computing $-\Delta A$ from tabulated or calculated $\Delta U$ and $\Delta S$ .

Derivation AMaximum work theorem from the Clausius inequality

Goal. Show that at constant temperature, the work done on a system in any process satisfies $dw \geq dA$ when work is done on the system (equivalently, the work done by the system can never exceed $-dA$ ), with equality only for a reversible path.

Step 1. Start from the First Law. For any process, reversible or not,

dU = dq + dw

where $dq$ is heat absorbed by the system and $dw$ is work done on it, using the convention that both add to $U$ .

Step 2. Invoke the Clausius inequality. The Second Law states that for any process,

dS \geq \frac{dq}{T}

with equality only for a reversible process, because a reversible path generates zero additional entropy in the universe. Rearranging,

dq \leq T\,dS.

Step 3. Substitute the bound on heat into the First Law. Since $dq \leq T\,dS$ ,

dw = dU - dq \geq dU - T\,dS.

This says the work you must put in is at least $dU - T\,dS$ ; equivalently, the work you can get out is at most $T\,dS - dU$ .

Step 4. Introduce Helmholtz energy at constant $T$ . By definition $A = U - TS$ , so at constant temperature,

dA = dU - T\,dS.

Substituting into Step 3,

dw \geq dA.

Step 5. Integrate over a finite change at constant $T$ . Integrating both sides between the initial and final states,

w \geq \Delta A,

where $w$ is the total work done on the system (mechanical/expansion plus any non-expansion contribution, such as electrical work). Equality holds only when every step of the path is reversible, since only then does $dq = T\,dS$ exactly.

Step 6. Flip sign to express work done by the system. Writing $w_{\text{by}} = -w$ ,

w_{\text{by}} \leq -\Delta A.

\boxed{w_{\text{by, max}} = -\Delta A \ \ (\text{constant } T, \text{ achieved only reversibly})}

This is the maximum-work theorem: at fixed temperature, $-\Delta A$ is an unreachable-in-practice but mathematically exact ceiling on total work output, and any irreversibility (friction, finite driving forces, spontaneous reaction) strictly lowers the work delivered below this ceiling while keeping $dS_{\text{universe}} \geq 0$ .

Worked examples

Example 1 — Isothermal reversible expansion of an ideal gas (foundational). Consider 1.00 mol of an ideal gas expanding reversibly and isothermally at $T = 298$ K from $V_1 = 10.0$ L to $V_2 = 20.0$ L. For an ideal gas at constant $T$ , $\Delta U = 0$ (internal energy depends only on $T$ ). The entropy change is

\Delta S = nR\ln\frac{V_2}{V_1} = (1.00\ \text{mol})(8.314\ \text{J K}^{-1}\text{mol}^{-1})\ln(2.00) = 5.763\ \text{J K}^{-1}.

\Delta A = \Delta U - T\Delta S = 0 - (298\ \text{K})(5.763\ \text{J K}^{-1}) = -1717\ \text{J}.

The maximum work obtainable is

w_{\text{max}} = -\Delta A = +1717\ \text{J} \approx 1.72\ \text{kJ}.

Check against the direct formula for reversible isothermal expansion of an ideal gas,

w_{\text{by}} = nRT\ln\frac{V_2}{V_1} = (1.00)(8.314)(298)\ln(2.00) = 1717\ \text{J}.

The two routes agree exactly, as they must, since this path is reversible and this is a purely expansion-work process — the theorem's equality condition is satisfied. Sanity check: the gas absorbs heat $q = T\Delta S = 1717$ J from the reservoir (consistent with $\Delta U = q + w = 1717 - 1717 = 0$ ), and every joule of that heat is converted to work — the maximum possible, since expansion of an ideal gas against vacuum-adjacent conditions reversibly is the textbook case of 100% "free energy" utilization for this specific $\Delta S$ -driven process.

Example 2 — Irreversible expansion delivers less work; a chemical cell delivers non-expansion work (applied). (a) Irreversible free expansion, same states. Now let the same 1.00 mol of gas undergo the same net change of state (10.0 L → 20.0 L at 298 K) but by sudden expansion against a constant external pressure equal to the final pressure, $p_{\text{ext}} = p_2$ . Since $U$ is a state function, $\Delta U = 0$ again, and since $S$ is a state function, $\Delta S = 5.763$ J K⁻¹ again, so $\Delta A = -1717$ J still — the theorem's ceiling is unchanged, because it depends only on initial and final states, not on path. Compute the actual work done irreversibly. With

p_1 = \frac{nRT}{V_1} = \frac{(1.00)(8.314)(298)}{0.0100\ \text{m}^3} = 2.478\times10^5\ \text{Pa},

p_2 = \frac{nRT}{V_2} = \frac{(1.00)(8.314)(298)}{0.0200\ \text{m}^3} = 1.239\times10^5\ \text{Pa}.

Work done by the gas against constant external pressure $p_2$ :

w_{\text{by}} = p_{\text{ext}}\Delta V = (1.239\times10^5\ \text{Pa})(0.0100\ \text{m}^3) = 1239\ \text{J}.

Indeed $1239\ \text{J} < 1717\ \text{J} = -\Delta A$ , confirming the inequality $w_{\text{by}} \leq -\Delta A$ strictly for this irreversible path — about 478 J of potential work (28%) was squandered because the piston was pushed against a driving force smaller than it could have sustained. (b) A chemical cell delivering non-expansion (electrical) work. Consider a reaction with $\Delta_r U = -285.8$ kJ mol⁻¹ and $\Delta_r S = -163.0$ J K⁻¹mol⁻¹ at $T = 298$ K (values close to the hydrogen combustion reaction run electrochemically at constant volume). Then

\Delta_r A = \Delta_r U - T\Delta_r S = -285{,}800\ \text{J mol}^{-1} - (298\ \text{K})(-163.0\ \text{J K}^{-1}\text{mol}^{-1})

= -285{,}800 + 48{,}574 = -237{,}226\ \text{J mol}^{-1} \approx -237.2\ \text{kJ mol}^{-1}.

The maximum total work (here almost entirely electrical, since a fuel cell at constant volume does negligible expansion work) is

w_{\text{max}} = -\Delta_r A = 237.2\ \text{kJ mol}^{-1}.

A real fuel cell running with internal resistance and finite current delivers less electrical work per mole than 237.2 kJ, the shortfall reappearing as waste heat — exactly the pattern established in Example 2(a), now applied to useful electrochemical work rather than gas expansion.

Limiting cases & common traps

Situation	What happens to the theorem
Fully reversible path	$w_{\text{by}} = -\Delta A$ exactly, the ceiling is achieved
Any irreversible path, same endpoints	$w_{\text{by}} < -\Delta A$ strictly; deficit appears as extra heat released
Ideal gas, purely expansion work	$\Delta A = \Delta U - T\Delta S$ ; if $\Delta U=0$ , reduces to $-T\Delta S = nRT\ln(V_2/V_1)$
Non-expansion work present (electrical, etc.)	$-\Delta A$ still bounds the total work, expansion plus non-expansion combined
Constant $V$ , no non-expansion work	$\Delta A = q_V$ for a reversible process; useful for calorimetry at constant volume

The chemist behind it

Hermann von Helmholtz, a German physicist and physiologist, introduced the function $A = U - TS$ in 1882 while formalizing the thermodynamics of chemical affinity and electrochemical cells. He called it the "free energy" (freie Energie), reasoning that only this portion of a system's internal energy is thermodynamically "free" to appear as work at constant temperature — the rest is inseparably bound to the entropy of the system via the term $TS$ . His derivation predates the Gibbs energy's widespread chemical use and directly inspired Gibbs' own treatment of constant-pressure systems, cementing the "maximum work" language that persists in every physical chemistry course today.

Origins · 1 minHermann von Helmholtz and how this was discovered

Hermann von Helmholtz

1821–1894 · German physicist who linked free energy to maximum obtainable work in electrochemical and thermal systems

Ask the tutor: How does the maximum-work theorem change when pressure, not volume, is held constant — and why does that lead to the Gibbs energy?

Is −ΔA the same as the heat released in a reaction?

No. Heat released depends on the path, while $-\Delta A$ is a state-function-based ceiling on total work at constant $T$ . Only for a fully reversible process does the heat absorbed equal $T\Delta S$ , and even then that heat is separate from the work; $-\Delta A$ specifically bounds work, not heat. Conflating the two is a common exam mistake — always track $\Delta U$ , $q$ , and $w$ separately.

Why is it called "free" energy?

Helmholtz used "free" to mean the portion of internal energy that is thermodynamically free to be converted into work at constant temperature, as opposed to the $TS$ portion which is tied up maintaining the system's entropy at that temperature via heat exchange with surroundings. It's a statement about availability for work, not about being free of cost or free of constraints in any other sense.

Does the maximum-work theorem apply at constant pressure too?

The Helmholtz version applies strictly at constant temperature and volume (or more generally whenever volume work is separated out explicitly). At constant temperature and pressure, the analogous and more chemically common bound uses the Gibbs energy, where $-\Delta G$ gives the maximum non-expansion work, since the $p\Delta V$ term is automatically "spent" on pushing back the atmosphere. That relation is developed next.

Can a real process ever exceed −ΔA in work output?

No — this would violate the Second Law by requiring $dS_{\text{universe}} < 0$ . The inequality $w_{\text{by}} \leq -\Delta A$ is as fundamental as the Clausius inequality it derives from. Any apparent exception signals either a measurement error, an unaccounted heat exchange, or a system that isn't truly at constant temperature throughout the process.

Why does reversibility matter so much if it's never achieved in practice?

Reversible processes are idealized, infinitely slow limits, but they define the theoretical maximum efficiency against which all real devices are judged. Engineers benchmark real engines, cells, and separations against this reversible ceiling, and the gap between actual and reversible performance quantifies exactly how much room for improvement exists.