A theory of what is possible
Undergraduate thermodynamics can feel like a grab-bag: Boyle's law here, a Carnot engine there, a formula for entropy you half-believe. The graduate view is different and far more powerful. Thermodynamics is a small set of universal constraints on macroscopic matter. It never tells you the microscopic mechanism, but it tells you what any system, whatever it is made of, is allowed to do — which processes can happen, which cannot, and how much work you can get.
A thermodynamic system in equilibrium is pinned down by a handful of state variables — temperature T, pressure P, volume V, particle number N. The central distinction of the whole subject is between state functions, which depend only on where you are, and path functions, which depend on how you got there. Internal energy U and entropy S are state functions; heat and work are not.
State functions and exact differentials
The mathematical spine of the whole subject is one idea: a state function has an exact differential. Because U depends only on the state, its change around any closed cycle is zero, \oint dU = 0. In coordinates, a differential df = M\,dx + N\,dy is exact precisely when its cross-derivatives agree.
The exactness (equality-of-mixed-partials) test — the engine behind every result to come.
This innocent-looking test is what will later generate the Maxwell relations for free. Because energy and entropy are state functions, their mixed second partial derivatives must commute — and each such identity is a nontrivial statement about real measurements.
The fundamental relation
Now combine the first law dU = \delta Q - \delta W with the second law along a reversible path, where \delta Q = T\,dS and \delta W = P\,dV. Everything path-dependent cancels and you are left with a relation between state functions alone.
The fundamental thermodynamic relation for a closed system — the first and second laws fused into one exact differential.
Read it as U=U(S,V): entropy and volume are the natural variables of the energy. Then T and P are not new inputs — they are just the slopes of U. This is the fundamental thermodynamic relation, and although we derived it along a reversible path, it relates state functions only, so it holds for every equilibrium state.
Temperature and pressure are the conjugate slopes of the energy in its natural variables.
If particles can flow in and out — an open system — add one more conjugate pair. Each extra particle costs an energy \mu, the chemical potential, which will dominate the later story of phase and chemical equilibrium.
The fundamental relation for an open system, with the chemical potential \mu conjugate to particle number.
A preview of the payoff
Look at the everyday pressure–temperature phase diagram of a pure substance: regions of solid, liquid and gas, meeting along coexistence lines at a single triple point, with the liquid–gas line ending abruptly at a critical point. Remarkably, thermodynamics alone — with no atomic model whatsoever — constrains this picture: why phases meet along lines, why the triple point is a point, and why the liquid–gas line must terminate.
The road ahead: build the four thermodynamic potentials and the Legendre transform that links them (Guide 2), extract the Maxwell relations and put them to work (Guide 3), derive phase coexistence and the Clausius–Clapeyron slope (Guide 4), and finally reach the universal physics of the critical point (Guide 5).