Chemical potential μ = (∂G - ∂N)_{T,P}: Visual walkthrough

2.4.5 · D2 · Physics › Thermodynamics & Statistical Mechanics (Advanced) › Chemical potential μ = (∂G - ∂N)_{T,P}

Step 0 — Teen cheezein jo hum actually change kar sakte hain

Definition Hamare box ke knobs

Ek sealed box of gas ki picture banao. Sirf teen independent cheezein hain jo tum control kar sakte ho:

Garam karo — tum heat andar ya bahar push karte ho temperature $T$ badal ke (yeh woh dial hai jo tumhara haath actually ghumata hai, jaise ek water bath). Heat andar aane se ek bookkeeping quantity jise entropy $S$ kehte hain, badhti hai. Entropy count karti hai, ek logarithm ke through, ki particles ke kitne microscopic arrangements $Ω$ possible hain: $S = k_{B} ln Ω$ , jahan $k_{B}$ Boltzmann's constant hai. $ln$ zaruri hai — $S$ raw count $Ω$ nahi hai, yeh us count ka (scaled) logarithm hai, aur yahi cheez $S$ ko additive banati hai jab tum do boxes join karte ho.
Dabao — tum volume $V$ change karte ho (woh space jisme particles rehte hain).
Particles daalo ya nikalo — tum particle number $N$ change karte ho (literally, andar kitne molecules hain).

Figure — Chemical potential μ = (∂G - ∂N)_{T,P}

Figure mein box ke teen controls hain: ek thermometer (tum $T$ set karte ho; heat flow phir entropy $S$ move karta hai), ek piston (volume $V$ control karta hai), aur ek hatch (particle number $N$ control karta hai). Note karo yeh honest wiring: tum $T$ directly ghumao, aur $S$ respond karta hai — tum entropy ko haath se pakad nahi sakte. Is page ki har cheez iske baare mein hai ki jab hum hatch — woh $N$ dial jo purana first law ignore karta tha — ko nudge karte hain toh kya hota hai.

Step 1 — Box mein stored energy, aur uske "response slopes"

KYA. Hum kehte hain $U$ teen state variables par depend karta hai: $U = U (S, V, N)$ .

KYUN. Kyunki yahi sirf cheezein hain jo change ho sakti hain, $U$ tabhi change ho sakta hai jab inme se koi change ho. Toh $U$ teen variables ke upar ek surface hai.

PICTURE. Figure dekho: $U$ ek curved sheet ki tarah draw ki gayi hai. Agar tum uske upar chalte ho, toh uski har direction mein steepness ek slope hai. Har slope ka ek naam hai:

$d U = slope in S T d S - slope in V P d V + slope in N μ d N$

Term by term, theek jahan yeh baithte hain:

$d U$ — box ki total energy mein tiny change.
$T d S$ — woh reversible heat jo andar flow karti hai jab entropy $d S$ badlti hai; energy $T$ (temperature) slope ke saath respond karti hai. Woh slope hi hai jiska matlab temperature hai.
$- P d V$ — volume ko $d V$ nudge karo; energy $- P$ slope ke saath respond karti hai. Minus sign kehta hai: box expand hone do ( $d V > 0$ ) aur woh energy khota hai duniya ko push karne mein.
$+ μ d N$ — hatch ko $d N$ particles se nudge karo; energy $μ$ slope ke saath respond karti hai.

Yeh kyun honest definition hai: humne koi extra assumption nahi ki: humne sirf notice kiya ki $U$ ka $N$ direction mein slope hai aur use ek symbol diya. Baaki sab bookkeeping hai.

Step 2 — Lab ke liye $U$ galat potential kyun hai

KYA. Hum ek naya energy bookkeeper chahte hain jiski natural variables $T, P, N$ hon.

KYUN. Taaki "freeze $T$ aur $P$ , wiggle $N$ " kuch aisa ho jo tum bench par actually kar sako.

PICTURE. Figure mein hum "hard" variables $S, V$ ko "easy" variables $T, P$ se trade karte hain. Is trade ka ek naam aur ek recipe hai.

Step 3 — Gibbs free energy $G$ build karna

KYA. Humne ek naya energy $G$ define kiya jo $U$ mein do correction terms ke saath hai.

KYUN. Har correction ek "hard" variable ko cancel karke hume ek "easy" wala deta hai. Yeh standard Legendre swap hai — ek variable ko uske slope se trade karo. Deeper machinery ke liye Euler Relation and Gibbs-Duhem dekho.

PICTURE. Figure mein tall bar $U$ ke upar se ek chunk $T S$ shave off kiya gaya hai aur ek chunk $P V$ stack kiya gaya hai; bacha hua height $G$ hai.

Step 4 — $G$ differentiate karo (sirf product rule)

KYA. $G = U - T S + P V$ ka tiny change lo.

KYUN. Hum dekhna chahte hain ki $G$ naturally kaunse variables par respond karta hai. Uske slopes find karne ke liye hum differentiate karte hain.

$- T S$ par product rule use karo (yeh $- T d S - S d T$ banta hai) aur $+ P V$ par (yeh $+ P d V + V d P$ banta hai):

$d G = from Step 1 d U product rule on - T S - T d S - S d T product rule on + P V + P d V + V d P$

PICTURE. Figure mein woh paanch chote rectangles dikhaye gaye hain jo product rule create karta hai — har product term se do do, plus $d U$ block jo replace hone ka intezaar kar raha hai.

Abhi tak kuch physical nahi hua — yeh pure calculus hai. Magic agli substitution mein hai.

Step 5 — First law substitute karo: cancellation

KYA. $d U$ ko Step 1 se replace karo, $d U = T d S - P d V + μ d N$ .

KYUN. Kyunki $d U$ woh akela term hai jo abhi bhi hard variables $S, V$ carry kar raha hai. Jaise hi hum ise plug karte hain, dekho kya khatam hota hai.

$d G = (A T d S B - P d V + μ d N) - A - T d S - S d T - B + P d V + V d P$

First law ka $T d S$ product rule ke $- T d S$ ko cancel karta hai ( $A$ aur $- A$ ). $- P d V$ $+ P d V$ ko cancel karta hai ( $B$ aur $- B$ ). Yeh cancellation hi poori baat hai — yeh wajah hai ki $G$ ko exactly un do correction terms ke saath build kiya gaya tha.

PICTURE. Figure mein, matching amber blocks pairs mein annihilate hote hain; sirf teen bachte hain.

Jo bachta hai:

Step 6 — Do dials freeze karke μ read off karo

KYA. $T$ fixed rakho ( $d T = 0$ ) aur $P$ fixed rakho ( $d P = 0$ ).

KYUN. Yahi water bath aur khula atmosphere tumhare liye automatically karte hain. Un do variables ko freeze karna master differential ke pehle do terms ko khatam kar deta hai.

$d G = - S d T^{0} + V d P^{0} + μ d N = μ d N$

$d N$ se divide karo:

PICTURE. Figure mein $G$ surface $(N)$ axis ke upar dikhaya gaya hai jisme $T, P$ floor level rakha gaya hai; tangent line ki steepness $μ$ hai.

Recall

Subscript $T, P$ kyun matter karta hai? ::: Kyunki usi surface ka alag slope (jaise $S, V$ hold karke) ek alag physical quantity hai; sirf $G$ ka $T, P$ -frozen slope $μ$ ke barabar hota hai.

Step 7 — Edge case: pure substance shortcut μ = G/N

KYA. Ek single pure substance ke liye, stuff ki amount ko kisi bhi factor $λ$ se scale karna (jaise $λ = 2$ matlab "do guna", $λ = \frac{1}{2}$ matlab "aadha") $G$ ko bhi same $λ$ se scale karta hai, fixed $T, P$ par. Yahan $λ$ sirf ek dimensionless scale factor hai, ek plain positive number jisse tum amount multiply karte ho. Symbols mein:

$G (T, P, λ N) = λ G (T, P, N) .$

KYUN yeh ek per-particle number mein collapse hota hai. Agar $G$ fixed $T, P$ par strictly $N$ ke proportional hai, toh slope $\partial G / \partial N$ average $G / N$ ke barabar hoga — origin se guzarne wali straight line ka slope = height ÷ base. Euler's theorem (Euler Relation and Gibbs-Duhem) yeh cleanly state karta hai:

$G = N (\frac{\partial G}{\partial N})_{T, P} = N μ ⟹ μ = \frac{G}{N}$

PICTURE. Figure mein, $G$ vs $N$ origin se ek straight ray hai. Slope $μ$ aur ratio $G / N$ same dashed line hain — woh equality sirf pure substance ki straight ray ke liye true hai. Mixture ke liye curve bend hoti hai aur tumhe partial molar quantities $μ_{i} = (\partial G / \partial N_{i})_{T, P, N_{j \neq = i}}$ use karni padti hain.

Step 8 — Degenerate limits: kya μ abhi bhi make sense karta hai?

Worked example Teen boundary cases, check kiye gaye

$N \to 0$ (almost empty box). Ideal gas ke liye $μ = μ^{\circ} (T) + k_{B} T ln (P / P^{\circ})$ , aur pressure $\to 0$ jaata hai jaise box khali hota hai, toh $ln (P / P^{\circ}) \to - \infty$ : $μ \to - \infty$ . Picture: ek empty room mein ghusna infinitely "sasta" hai — particles rush in karte hain. Yahi wajah hai ki vacuum suck karta hai.
$d N = 0$ (sealed hatch). $μ d N$ term vanish ho jata hai aur $G$ sirf $T, P$ ke through change hota hai. $μ$ ek slope ki tarah abhi bhi exist karta hai; bas exercise nahi hota. Purana first law recover ho jaata hai.
Do boxes, $μ_{A} = μ_{B}$ (equilibrium). Koi slope difference nahi matlab koi drive nahi: $d G = (μ_{A} - μ_{B}) d N_{A} = 0$ kisi bhi $d N_{A}$ ke liye. Kuch flow nahi karta. Yeh diffusive (aur chemical) equilibrium hai: jab bhi particles connected systems ke beech move kar sakte hain, woh tabhi ruke jab chemical potentials equal hon. (Related lekin distinct sawaal ki $P$ – $T$ plane mein ek phase boundary line kaise tilt hoti hai, woh Phase Equilibrium and Clausius-Clapeyron ka subject hai.)

Figure teeno dikhata hai: $N \to 0$ par diverging $μ$ (left), ek sealed hatch (middle), aur do connected boxes ke beech equilibrium mein ek level "μ-landscape" (right, flat = no flow).

Ek-picture summary

Yeh single figure poora page compress karta hai: $d U = T d S - P d V + μ d N$ se shuru karo (top), hard variables ko easy ones se swap karo $- T S + P V$ add karke $G$ paane ke liye (middle), matching terms cancel karo, aur surviving $N$ -slope ko $μ = (\partial G / \partial N)_{T, P}$ ki tarah read off karo (bottom). Dashed amber arrow particle flow ko hamesha $μ$ mein downhill trace karta hai.

(Diagram mein, "chem-pot" Greek letter $μ$ ka plain-text stand-in hai jo is page par baaki jagah use hota hai — Mermaid labels plain text rehne chahiye.)

Recall Feynman: walkthrough plain words mein

Humne ek box se shuru kiya jiske teen controls hain — ek temperature knob, ek piston, aur particles add karne ke liye ek hatch. Temperature knob ghumane se heat flow hoti hai, aur woh heat quietly ek bookkeeping quantity ko move karti hai jise entropy kehte hain (roughly, un ways ka logarithm jisme particles arrange ho sakte hain). Purana energy rulebook sirf heat aur squeeze track karta tha, toh humne hatch ko uska apna slope diya aur use $μ$ naam diya: woh energy jo ek naye particle ke saath aati hai. Lekin woh slope un quantities mein likha tha jo koi lab mein fixed nahi rakh sakta (entropy, volume). Toh humne ek naya energy scorecard, $G$ , banaya — heat-energy $T S$ shave karke aur push-energy $P V$ add karke — ek bookkeeping trick jo awkward variables ko easy ones se magically swap karti hai, temperature aur pressure. Jab humne $G$ ke slopes recompute kiye, heat aur squeeze terms perfectly cancel ho gayi (woh cancellation hi wajah hai ki humne $- T S + P V$ choose kiya), teen clean terms bachte hue. Temperature aur pressure freeze karo — jaise lab mein baithe ho — aur sirf particle term bachti hai: $μ$ $G$ ka slope hai jab tum particles add karte ho. Ek pure ingredient ke liye, woh slope sirf $G$ divided by $N$ hai, kyunki stuff double karne par $G$ straight line mein double hota hai. Aur jab do connected boxes same $μ$ reach karte hain, slopes match ho jaate hain, kuch flow nahi karta — yahi equilibrium hai.