In Chapter 14 we established that the maximum work obtainable from a chemical reaction at constant temperature and pressure is given by Gibbs’s equation:
Maximum Work = −ΔGrxn = −(ΔHrxn − TΔSrxn) constant T, P
in which ΔHrxn, as discussed in Chapter 13, is measured directly in a reaction calorimeter.
Gibbs’s equation tells us that ΔHrxn is not itself the maximum work available, as scientists once thought. A portion of it — TΔSrxn — is unavailable for useful purposes. What, physically, is this portion? What is actually happening to the atoms that makes this energy unavailable?
The theory that TΔS broke
To appreciate what TΔS represents, it helps to understand what happened when it was ignored.
In the mid-1800s, Thomsen and Berthelot proposed that the heat released in a calorimeter — ΔHrxn — represents the maximum work a reaction can deliver and the sole criterion for spontaneity. An exothermic reaction, they argued, releases energy and is therefore spontaneous; an endothermic reaction absorbs energy and therefore cannot be. This thermal theory of affinity worked well for many reactions. But it failed for the spontaneous endothermic reaction — one that absorbs energy from its surroundings and yet proceeds on its own. According to Thomsen and Berthelot, it wasn’t supposed to happen. And yet it did.
The resolution required recognizing something the calorimeter could not reveal: that ΔHrxn contains two physically distinct contributions, and that it is only one of them (ΔGrxn) — not the total — that governs spontaneity and maximum work. TΔSrxn is the other contribution, the one the thermal theory of affinity wasn’t aware of. An endothermic reaction can be spontaneous when TΔSrxn is positive and large enough to overcome a positive ΔHrxn. The spontaneous endothermic reaction was not a violation of thermodynamics. It was the first clear evidence that ΔHrxn alone was insufficient, and that something else was needed to complete the picture.
The puzzle Gibbs solved
The answer to what TΔS physically represents arrived not from theory but from a piece of laboratory apparatus: the electrochemical cell.
In the late 1800s, a group of researchers — Pierre Antoine Favre and Johann Silbermann, François-Marie Raoult, and Hans Max Jahn among them — set out to compare two ways of running the same chemical reaction. [1] In the first, the reaction proceeds inside a calorimeter, generating a measurable heat effect: ΔHrxn. In the second, the same reaction runs inside an electrochemical cell, generating an electrical current instead. That current can be passed through a large resistor and converted entirely back into heat, allowing a direct comparison.
The two heat measurements didn’t match. The energy balance didn’t close. Something was missing.
The resolution came when these researchers placed the entire electrochemical cell into a constant-temperature bath calorimeter. They discovered an additional heat effect — one occurring inside the cell itself, exchanged with the surrounding bath in order to keep the cell at constant temperature as the reaction proceeded. Once this exchange was included, the energy balance closed exactly:
Energy from the reaction (ΔHrxn) = electrical work delivered externally + heat exchanged with the constant-temperature bath
This third term — the heat exchanged with the bath to maintain constant temperature — is precisely TΔSrxn. Gibbs, reviewing this body of experimental work, concluded that the voltage generated by an electrochemical cell is directly proportional to the maximum external work, −ΔGrxn, while the heating or cooling required to hold the cell at constant temperature is TΔSrxn. [2]
It continues to amaze me that the terms in Gibbs’s equation line up exactly with the cell voltage and the heating/cooling requirements of the bath — two independently measurable quantities, each corresponding to one term in the equation.
This is the physical reality of TΔS. It is not an abstract mathematical remainder. It is a real, measurable heat effect — the energy a reaction must exchange with its surroundings, in addition to whatever work it performs, simply to keep going at constant temperature.
TΔS as the change in structural energy
Recall from Chapter 12 that absolute entropy is built by integrating heat capacity from absolute zero: S = ∫(Cp/T)dT, plus contributions from any phase transitions crossed along the way. As Chapter 5 established, this integral accounts for two things: the thermal energy needed to set atoms into motion, and the energy needed to drive them apart from one another into the volume they occupy. Absolute entropy, arrived at by this calorimetric path, equals the entropy arrived at by Boltzmann’s microstate counting — a fact first demonstrated by Tetrode and Sackur, who independently showed that S = kBln W and dS = δQrev/T lead to the same value for a monatomic ideal gas. (What fascinates me about this demonstration is that Boltzmann’s entropy doesn’t history while Clausius’s entropy does.)
TΔSrxn therefore quantifies the difference between the energies required to establish the reactant system and the product system — where “system” means the full atomic arrangement, both the momentum and the spatial configuration of the atoms involved. This energy is bound up in the structure of the system itself, present in how the atoms are arranged and how they move. I call it structural energy.
The sign carries direct physical meaning. When TΔS is negative, the products require less structural energy than the reactants — the difference is released as heat and must be removed by the constant-temperature bath to hold conditions steady; entropy decreases. When TΔS is positive, the products require more — the bath must supply the difference; entropy increases. Either way, this energy is exchanged with the surroundings separately from the maximum work the reaction performs.
This is the sense in which TΔS is, as Helmholtz described it in 1882, “bound” energy — present in the structural arrangement of the system, exchangeable only as heat, never convertible to useful work. [3]
Testing this against real data
A hypothesis of this kind deserves a numerical check. Consider the reaction:
Hg(l) + ½Cl₂(g) → ½Hg₂Cl₂(s)
Bertram Dodge’s 1944 thermodynamics textbook reported the following measurements at 25°C: [4]
ΔHrxn = −31,300 cal/mol, measured in a constant T,P calorimeter
Wext = 25,140 cal/mol, electrical work measured from the cell’s electromotive force when operating reversibly
Using Gibbs’s equation, the heat exchanged with the constant-temperature bath:
25,140 = 31,300 + TΔSrxn
TΔSrxn = −6,160 cal/mol
ΔSrxn = −6,160 / 298 = −20.7 cal/(mol·K)
This value was obtained entirely from electrochemical and calorimetric measurements — no absolute entropy data involved. It can now be checked independently. NIST tabulates absolute entropy values for each species: mercury liquid, chlorine gas, and Hg₂Cl₂ solid. [5] These absolute values were not available to Gibbs — the low-temperature heat capacity measurements required to calculate them would not exist until the early 1900s. That the two approaches agree at all is itself a confirmation of the theoretical framework.
Combining the NIST values on a per-mole-of-HgCl basis:
ΔSrxn = ½S°(Hg₂Cl₂) − S°(Hg) − ½S°(Cl₂) = ½(46.02) − 18.14 − ½(53.32) = 23.01 − 18.14 − 26.66 = −21.79 cal/(mol·K)
The two values — −20.7 from the electrochemical cell and −21.8 from absolute entropies — agree to within roughly 5 percent. This is a genuine, quantitative confirmation that the heat exchanged with the bath to maintain constant temperature in the electrochemical cell is the same physical quantity as the entropy change calculated from Clausius’s thermodynamic definition.
What this means
Gibbs’s equation can be read in two directions:
Maximum Work = ΔGrxn = ΔHrxn − TΔSrxn
or equivalently and perhaps more informatively:
ΔHrxn = ΔGrxn + TΔSrxn
As Chapter 13 established, the heat measured by a calorimeter combines all contributions — orbital electron rearrangement, changes in intramolecular potential energies, changes in intermolecular potential energies, and any PV work at constant pressure. The calorimeter measures their sum as ΔHrxn. Gibbs’s equation then partitions that sum into two terms: ΔGrxn, which sets the maximum work available, and TΔSrxn, which is the structural energy exchanged with the surroundings to maintain constant temperature.
How exactly the individual physical contributions from Chapter 13 map onto ΔG versus TΔS is not completely resolved. Certain contributions — particularly those involving orbital electron rearrangement — are more closely associated with ΔG; others — particularly those involving intermolecular forces and configurational freedom — are more closely associated with TΔS. But the exact parsing is subtle, and the honest statement is that Gibbs’s equation identifies what must be exchanged as heat and what is available for work, without providing a complete atomic-level account of why each contribution falls where it does. Chapter 15 marks a starting point for that deeper analysis, not its conclusion.
Final comments
TΔS is not a correction factor invented to make Gibbs’s equation balance. It is a physically real heat effect — first identified empirically by thermochemists puzzled by an energy balance that wouldn’t close, given theoretical structure by Gibbs and Helmholtz, confirmed numerically by comparing electrochemical measurements against absolute entropy values built independently from heat capacity data, and now understood as the change in structural energy bound up in the rearrangement of a system’s atoms — their motion and their spatial arrangement — as reactants become products. It is the energy that must be exchanged with the surroundings to hold temperature constant, and for that reason it can never be converted to useful work.
The thermal theory of affinity failed because it saw only ΔH. Gibbs’s framework succeeded because it separated what ΔH contains.
References
[1] Ostwald, W., Electrochemistry: History and Theory, Amerind Publishing Co., New Delhi, 1980, pp. 741–1016.
[2] Gibbs, J.W., The Scientific Papers of J. Willard Gibbs, Volume One: Thermodynamics, Ox Bow Press, 1993, pp. 331–349.
[3] von Helmholtz, H., “On the Thermodynamics of Chemical Processes,” Physical Memoirs Selected and Translated from Foreign Sources, 1, 43–97, 1882.
[4] Dodge, B.F., Chemical Engineering Thermodynamics, McGraw-Hill, New York, 1944, p. 71. Dodge’s original formula HgCl is understood to correspond to mercury(I) chloride, Hg₂Cl₂. NIST lists a standard enthalpy of formation for Hg₂Cl₂ of −265 kJ/mol, equivalent to −31,668 cal per mole of HgCl — within 1.2% of Dodge’s −31,300, supporting the identification.
[5] NIST Chemistry WebBook, SRD 69: mercury, chlorine, mercury(I) chloride.
END