The Clausius Inequality

Imagine a steam engine idling in a cold factory. Its firebox roars, but its efficiency is capped by nature, not bad engineering. Every engine ever built, however cleverly designed, obeys a strict accounting rule. That rule links the heat it exchanges to the temperature at which the exchange happens. That rule is the Clausius inequality. The inequality matters because it is the hinge on which the entire Second Law swings. It converts Carnot's efficiency limit into a completely general statement. This statement applies to any cyclic process, reversible or not. From it, in one short step, falls the existence of entropy as a state function. This article builds the inequality from Carnot's theorem. It states the inequality in its general and differential forms. It walks through the derivation line by line. It applies the inequality to two worked cycles. You will see where it comes from. You will see what it forbids. You will see how it opens the door to entropy.

Key takeaways

The Clausius inequality states that the cyclic integral of đq/T is never positive, for any cycle.
Equality holds only for a reversible cycle; strict inequality signals irreversibility.
It generalizes Carnot's efficiency bound from a single reversible engine to an arbitrary cycle, reversible or not.
The proof uses a chain of tiny Carnot engines coupled to one reservoir, plus the Kelvin statement of the Second Law.
It is the direct ancestor of the state function entropy, defined via dS = đq_rev/T.
Key idea: No cyclic process can extract net positive "heat divided by temperature" from its surroundings; this deficit, zero only in the reversible limit, is the seed of entropy.

The core insight

The Clausius inequality says that for any closed cycle, the sum (integral) of heat exchanged divided by the temperature at which each exchange occurs can never exceed zero, with equality reserved exclusively for reversible cycles — turning Carnot's single-engine efficiency bound into a universal law about heat and temperature that defines entropy as a state function.

The physical picture

Picture a working substance—gas in a cylinder, say—carried through any cyclic path on a $p$ – $V$ diagram, returning exactly to its starting state. Along the way it exchanges tiny packets of heat $đq$ with a sequence of reservoirs at various temperatures $T$ . Clausius's insight was to weight each packet by the temperature at which it was absorbed, forming the quantity $đq/T$ , and then sum this weighted heat around the entire loop. For a reversible cycle, this weighted sum returns exactly to zero: what flows in as $đq/T$ at one stage is exactly balanced by what flows out at another. For any real, irreversible cycle, the sum comes back negative. It can never come back positive. This asymmetry is not a matter of engineering skill; it is baked into the structure of heat flow itself, as Carnot's theorem already hints. The Clausius inequality makes that hint into a sharp, general statement covering every possible cycle, not just idealized reversible ones. The physical root of the asymmetry is irreversibility itself: friction, unrestrained expansion, finite-rate heat conduction across a temperature gap. Each of these degrades the "quality" of heat transfer relative to the reversible ideal, and the inequality tracks that degradation quantitatively.

Figure

A generic cycle exchanging heat with a continuum of reservoirs at temperatures T_i

The second figure shows why the reversible case is special: a reversible cycle can be approximated arbitrarily well by a zig-zag chain of infinitesimal Carnot engines, each exchanging heat isothermally at a well-defined local temperature.

Figure

A reversible cycle decomposed into a chain of infinitesimal Carnot strips

The core relations

\oint \frac{đq}{T} \leq 0

where: the integral runs once around any closed cycle, $đq$ is heat absorbed by the system, and $T$ is the temperature of the reservoir supplying it at that instant.

\oint \frac{đq_{\text{rev}}}{T} = 0

where: equality holds specifically when every step of the cycle is reversible.

\frac{đq}{T} \leq dS

where: $dS$ is the entropy change of the system for an infinitesimal step, with equality only in the reversible limit — the differential form that leads directly to the entropy state function.

\eta_{\text{rev}} = 1 - \frac{T_{\text{c}}}{T_{\text{h}}} \geq \eta_{\text{irrev}}

where: $\eta$ is engine efficiency, $T_h$ and $T_c$ are hot and cold reservoir temperatures — the Carnot bound from which the inequality is built.

Derivation AFrom Carnot's theorem to the Clausius inequality

Goal. Show that for an arbitrary cyclic process, exchanging heat with any number of reservoirs at different temperatures, the cyclic integral $\oint đq/T$ cannot exceed zero, with equality only for a reversible cycle.

Step 1: Set up the auxiliary reservoir. Let the system undergo an arbitrary cycle, absorbing heat $đq_i$ from a reservoir at temperature $T_i$ at each stage $i$ . Introduce one additional very large reservoir at a fixed reference temperature $T_0$ , and for each stage insert a small reversible Carnot engine operating between $T_0$ and $T_i$ . This Carnot engine absorbs $đq_i^0$ from the $T_0$ reservoir and delivers exactly $đq_i$ to the system's local reservoir at $T_i$ , matching what the system needs.

Step 2: Apply Carnot's efficiency relation to each auxiliary engine. Since each auxiliary engine is reversible, running between $T_0$ and $T_i$ ,

\frac{đq_i^0}{T_0} = \frac{đq_i}{T_i}

so that $đq_i^0 = T_0 \left( đq_i / T_i \right)$ . This step uses Carnot's theorem directly: a reversible engine's heat ratio equals its temperature ratio, regardless of the working substance.

Step 3: Sum over the whole cycle. Summing (integrating) over one complete cycle of the system, the total heat withdrawn from the single reservoir at $T_0$ is

q_0 = \oint đq^0 = T_0 \oint \frac{đq}{T}

Meanwhile, the composite device — system plus all the little Carnot engines — undergoes a full cycle itself and returns to its initial state, exchanging heat only with the single reservoir at $T_0$ .

Step 4: Invoke the Kelvin statement of the Second Law. A cyclic device exchanging heat with a single reservoir cannot produce net positive work output; equivalently it cannot absorb net positive heat from that one reservoir, since by the First Law over a full cycle $w = -q_0$ and any positive $q_0$ converted entirely to work would violate Kelvin's statement. Therefore

q_0 \leq 0

Step 5: Translate back. Combining Steps 3 and 4, and noting $T_0 > 0$ ,

T_0 \oint \frac{đq}{T} \leq 0 \quad \Longrightarrow \quad \oint \frac{đq}{T} \leq 0

Step 6: Identify the equality condition. If the original cycle itself is reversible, it can be run in reverse. Applying the same argument to the reversed cycle gives $\oint đq/T \geq 0$ for the reverse path, which is $-\oint đq/T \geq 0$ for the forward path, i.e. $\oint đq/T \leq 0$ again but now with the opposite sign forced too. Both inequalities together force

\boxed{\oint \frac{đq_{\text{rev}}}{T} = 0}

while any irreversible cycle, lacking this reversal symmetry, retains strict inequality:

\boxed{\oint \frac{đq}{T} < 0 \ \ (\text{irreversible cycle})}

Worked examples

Example 1 — Carnot cycle sanity check. A reversible Carnot engine operates between $T_h = 500\ \text{K}$ and $T_c = 300\ \text{K}$ , absorbing $q_h = 1000\ \text{J}$ from the hot reservoir per cycle. Step 1: Carnot efficiency.

\eta = 1 - \frac{T_c}{T_h} = 1 - \frac{300}{500} = 0.400

Step 2: Work and rejected heat.

w = \eta q_h = 0.400 \times 1000\ \text{J} = 400\ \text{J}

q_c = q_h - w = 1000\ \text{J} - 400\ \text{J} = 600\ \text{J} \ \ (\text{heat rejected, so } đq = -600\ \text{J at } T_c)

Step 3: Evaluate the cyclic integral, which for a two-reservoir cycle collapses to a two-term sum.

\oint \frac{đq}{T} = \frac{q_h}{T_h} + \frac{-q_c}{T_c} = \frac{1000}{500} + \frac{-600}{300}

= 2.00\ \text{J K}^{-1} - 2.00\ \text{J K}^{-1} = 0

Sanity check: the result is exactly zero, exactly as required for a reversible cycle. This confirms the entropy generated by the universe over one cycle is zero, consistent with reversibility.

Example 2 — Irreversible engine violating the reversible bound. Now suppose a real engine operates between the same two reservoirs, $T_h = 500\ \text{K}$ and $T_c = 300\ \text{K}$ , but with internal friction and irreversible heat leaks. It still absorbs $q_h = 1000\ \text{J}$ from the hot reservoir, but because of losses it can only convert $w = 300\ \text{J}$ into work (below the reversible maximum of 400 J found above). Step 1: Heat rejected to the cold reservoir, from the First Law over one cycle ( $\Delta U = 0$ ).

q_c = q_h - w = 1000\ \text{J} - 300\ \text{J} = 700\ \text{J}

Step 2: Evaluate the cyclic integral.

\oint \frac{đq}{T} = \frac{q_h}{T_h} + \frac{-q_c}{T_c} = \frac{1000}{500} + \frac{-700}{300}

= 2.000\ \text{J K}^{-1} - 2.333\ \text{J K}^{-1} = -0.333\ \text{J K}^{-1}

Sanity check: the sum is negative, exactly as the Clausius inequality demands for an irreversible cycle. Notice also that this engine's actual efficiency is

\eta_{\text{irrev}} = \frac{w}{q_h} = \frac{300}{1000} = 0.300 < 0.400 = \eta_{\text{rev}}

consistent with Carnot's theorem: no irreversible engine beats the reversible bound between the same two reservoirs. The magnitude $0.333\ \text{J K}^{-1}$ is, as later shown, the entropy generated in the surroundings and is the direct numerical seed of the entropy state function introduced next.

Limiting cases & common traps

Situation	What happens to $\oint đq/T$	Common trap
Fully reversible cycle	exactly $0$	Forgetting that equality is a special, idealized case, not the generic result
Any real (irreversible) cycle	strictly $< 0$	Assuming a positive result is merely "small error"; it is thermodynamically forbidden
Adiabatic reversible step within a cycle	contributes $0$ to the integral since $đq = 0$	Confusing adiabatic with isothermal steps in a Carnot cycle
Using $T$ = system temperature instead of reservoir temperature for irreversible steps	gives wrong (often positive) sign	The $T$ in $đq/T$ for irreversible heat transfer must be the temperature of the reservoir supplying the heat, not necessarily the system's own instantaneous temperature

The chemist behind it

Rudolf Clausius spent the 1850s wrestling with a puzzle: Carnot's caloric-theory reasoning gave the right efficiency formula, but its underlying picture of heat as a conserved fluid was wrong, as Joule's mechanical equivalent of heat had shown. Clausius rebuilt Carnot's result on the correct foundation—heat and work as interconvertible forms of energy—and in doing so isolated the quantity $đq/T$ as the true invariant. His 1865 paper introduced not just the inequality bearing his name but the word "entropy" itself, from the Greek for "transformation," deliberately chosen to echo "energy." He also gave the Second Law its most quoted slogan: the energy of the universe is constant, its entropy tends to a maximum.

Origins · 1 minRudolf Clausius and how this was discovered

Rudolf Clausius

1822-1888 · German physicist who reconciled Carnot's efficiency limit with energy conservation and coined the term entropy

Ask the tutor: How does the differential form of the Clausius inequality lead to the formal thermodynamic definition of entropy as a state function?

Why is the inequality written with heat divided by temperature, rather than just heat?

Raw heat $q$ is not a state function and depends heavily on the path and on the temperature at which transfer occurs. Dividing by $T$ produces a quantity that, when summed reversibly around any closed cycle, always returns to zero — the hallmark of a state function. Clausius recognized that this ratio, not heat itself, is the thermodynamically fundamental bookkeeping quantity, and it is exactly this recognition that lets him define entropy as $dS = đq_{\text{rev}}/T$ .

Does the inequality apply to open, non-cyclic processes too?

The cyclic integral form applies strictly to closed cycles, where the system returns to its initial state. For a single non-cyclic step, the differential form $đq/T \leq dS$ applies instead, which is how the inequality generalizes beyond engines to arbitrary chemical and physical transformations, including reactions and phase changes.

How does the Clausius inequality connect to Carnot's theorem specifically?

Carnot's theorem states that no engine operating between two fixed reservoirs can exceed the reversible (Carnot) efficiency, and all reversible engines between the same two reservoirs share that same efficiency. The Clausius inequality is the generalization of this idea to cycles involving arbitrarily many reservoirs at arbitrarily many temperatures, proved using a chain of auxiliary Carnot engines exactly as Carnot's theorem describes for a single pair.

What physically goes wrong in an irreversible cycle to make the integral negative?

Irreversible features — friction, finite-rate heat conduction across a nonzero temperature difference, unrestrained expansion — all degrade useful work output relative to the reversible ideal. Each such feature effectively "wastes" some heat-carrying capacity at a favorable temperature and instead dumps equivalent heat at a less favorable (often colder) temperature, which numerically pulls the weighted sum $\oint đq/T$ below zero.

Why does the sign convention matter so much here?

The inequality relies on the standard convention that $đq$ is positive when heat flows into the system. Flipping the sign convention, or forgetting to flip the sign on rejected heat (heat leaving the system), is the single most common source of an incorrectly "positive" or zero result for what should be a strictly negative irreversible cycle sum, as seen explicitly in Worked Example 2.