To understand the parent topic, you must be fluent in every symbol it throws around. Below we build each one from absolutely nothing — plain words, then a picture, then why the topic can't live without it. Read top to bottom: every idea leans on the one above it, and no symbol appears in a formula before it has its own section.
The whole point of chemical potential is "what happens when N changes?" — a molecule evaporates, diffuses through a wall, or reacts. So N is the star variable. Everything else measures how the system reacts to nudging N up or down by one.
Before any formula, you must read these three symbols correctly. They all mean "change," but at different zoom levels.
Figure s01 (below) shows a smooth curve of "some quantity" versus a variable X. The dotted black steps mark a big finite change ΔX; the red straight line is the tangent, showing how the curve looks under infinite zoom — the world of dX.
Why the topic needs ∂ (partial derivative): the energy budget we'll build (call it G) depends on temperature, pressure, AND particle number all together. When we later write (∂N∂G)T,P we mean: "nudge N a tiny bit, watch how G responds, but keep the other two pinned." The little subscript is a promise about what we held still. Skip that promise and the number means something else entirely.
Figure s02 (below) plots a rising curve of G against N. The red tangent line touches it at one point; its rise over a run of one particle is the slope — and that slope is what we will name μ.
The tool we're using here — the derivative — is the right tool because the question is literally "rate of change of budget with count." No other operation answers "per one more particle." Integration would add things up; algebra can't capture a rate; only a derivative gives a local slope. (And thanks to the thermodynamic limit of §1, that slope is well defined even though N counts whole particles.)
Figure s03 (below) shows two identical boxes. On the left the particles are clustered in a corner (few arrangements, low S); on the right the red particles are spread across the whole box (many arrangements, high S). The arrow reminds you which way "spread-out-ness" grows.
P and V form the second conjugate pair: PdV is the mechanical work done as the walls move. We will later hold P fixed (not V) in the definition of μ because in a real lab you control the pressure — you leave the beaker open to the atmosphere — while volume is free to adjust.
Now that T,S,P,V each have a meaning, we can safely write the energy law that ties them together.
The First Law says energy is conserved. In its differential form for a system whose particle count is fixed:
Each piece is a "driving force × the thing it changes." Once we allow N to vary, we simply add the third conjugate pair(μ,N) — the very knob promised in §1:
Force (a "potential")
Times
The change it drives
Temperature T
×
entropy change dS
Pressure P
×
volume change dV
Chemical potential μ
×
particle change dN
So the generalized first law reads
dU=TdS−PdV+μdN,
and this equation is literally where μ is born: it is the coefficient sitting in front of dN.
Now every symbol in this formula is earned: ∂/∂N is the slope (§3, valid because of the thermodynamic limit of §1), the subscript T,P is the "held fixed" promise (§2), and G with its differential is the fixed-T,P budget (§7). μ is simply the third conjugate force, partner of N, in the same family as T (partner of S) and P (partner of V).
Every arrow means "you need the tail before the head makes sense." Notice all roads pass through G and its differential into μ — that's the spine of the topic.