At thermal equilibrium the law of mass action holds:
np=ni2
When we inject excess carriers (Δn, Δp) the product np>ni2. Nature abhors this excess, so net recombination kicks in to drag np back to ni2. If instead we deplete carriers (np<ni2), net generation dominates. So:
The whole subject is: what is U as a function of the excess?
An electron in the conduction band and a hole in the valence band must annihilate — but energy ≈Eg and momentum must go somewhere. Where it goes defines the mechanism.
HOW we derive its rate. The recombination rate should be proportional to the chance of an electron meeting a hole — so proportional to the product np:
R=Bnp
Generation at equilibrium must cancel recombination at equilibrium:
G0=R0=Bn0p0=Bni2
Assuming the (thermal) generation rate stays fixed at G0 (it depends only on temperature, not on injected carriers):
Urad=R−G=B(np−ni2)
Simplify for low injection in n-type (n0≫p0, Δn=Δp). Write n=n0+Δp, p=p0+Δp:
np−ni2=(n0+Δp)(p0+Δp)−n0p0=(n0+p0)Δp+Δp2
Low injection ⇒ Δp≪n0 ⇒ drop Δp2 and p0:
Urad≈Bn0Δp=τpΔp,τp=Bn01
HOW (result of detailed balance of four capture/emission events):USRH=τp(n+n1)+τn(p+p1)np−ni2
where n1=nie(Et−Ei)/kT, p1=nie(Ei−Et)/kT, and τn,p=1/(σn,pvthNt).
Key readings of this formula (the 80/20 insight):
Numerator np−ni2 → same "distance from equilibrium" driver as before.
Traps are most effective at mid-gap (Et=Ei): then n1=p1=ni, minimizing the denominator ⇒ maximum U. Mid-gap traps are "lifetime killers."
Low injection, n-type: n≈n0 huge, so U≈Δp/τp. Lifetime τp=1/(σpvthNt) — inversely proportional to trap density.
Two carriers of one type + one of the other:
RAuger=Cnn2p+Cpnp2
Net rate:
UAuger=(Cnn+Cpp)(np−ni2)
n-type low injection: τAuger=1/(Cnn02) — falls as 1/ND2, so it kills lifetime in heavily doped regions.
What single quantity drives all three net rates? → np−ni2.
Where must a trap sit to be most effective? → mid-gap, Et=Ei.
Why can't Si make good LEDs? → indirect gap, radiative path weak.
Which mechanism dominates at heavy doping? → Auger (∝n02).
How do lifetimes combine? → 1/τ=∑1/τi.
Recall Feynman: explain to a 12-year-old
Imagine a two-floor building. Kids upstairs (electrons in conduction band) sometimes slide down to the ground floor (valence band) where there's an empty spot (hole), and the two "disappear" together — that's recombination. Three ways to slide down: (1) shout out a flash of light (radiative — this is a glow-in-the-dark toy / LED), (2) use a mid-floor landing as a stepping stone (SRH trap — easiest in dirty buildings/silicon), or (3) bump into a friend and give them a shove of speed (Auger — happens in crowded rooms). Meanwhile heat keeps kicking other kids back upstairs — that's generation. When the crowd is normal, up = down and nothing changes. Shine a light and suddenly there are too many pairs, so the "sliding down" speeds up to clean up the extra. How fast it cleans up is the lifetime.
Net recombination rate definition
U=R−G; positive above equilibrium, zero at equilibrium.
Common driving factor in all recombination rate expressions
np−ni2 (distance from mass-action equilibrium).
Radiative net rate
Urad=B(np−ni2).
Radiative low-injection lifetime (n-type)
τ=1/(Bn0).
SRH net recombination rate
U=τp(n+n1)+τn(p+p1)np−ni2.
Where is an SRH trap most effective
at mid-gap, Et=Ei (minimizes denominator, so max U).
n1 and p1 definitions
n1=nie(Et−Ei)/kT, p1=nie(Ei−Et)/kT, with n1p1=ni2.
SRH lifetime formula
τ=1/(σvthNt), inversely proportional to trap density.
Auger net rate
U=(Cnn+Cpp)(np−ni2); Auger lifetime ∝1/ND2.
Which mechanism dominates in direct-gap GaAs
radiative (band-to-band) — basis of LEDs/lasers.
Which dominates in silicon
SRH (trap-assisted), because Si is indirect-gap.
Which dominates at very high doping/injection
Auger recombination.
How do the three lifetimes combine
rates add in parallel: 1/τ=1/τrad+1/τSRH+1/τAuger.
Excess-carrier decay law
Δp(t)=Δp(0)e−t/τ from dΔp/dt=−Δp/τ.
Why can't Si be an efficient LED
indirect gap needs a phonon for radiative transitions, so SRH (heat) wins.
Dekho, semiconductor equilibrium par ek "traffic jam jaisa balance" hai: kuch electrons upar conduction band mein jaate rehte (generation) aur kuch wapas gir ke holes ke saath recombine hote rehte (recombination), dono barabar. Jab hum light daalte hain ya current inject karte hain, to extra electron-hole pairs ban jaate hain aur np>ni2 ho jaata hai. Ab crystal ye extra pasand nahi karta, isliye net recombinationU=R−G start hoti hai jo system ko wapas np=ni2 tak le aati hai. Poore chapter ka asli sawaal bas yahi hai: U kitna hai — aur hamesha driver hota hai np−ni2.
Teen raaste hain energy release karne ke. Radiative: electron seedha girta hai aur photon nikalta hai — ye direct-gap materials (GaAs) mein strong hota hai, isliye LED aur laser inhi se bante hain. SRH (trap-assisted): mid-gap defect ek "seedhi ki step" ki tarah kaam karta hai, do chhoti chhalaang ek badi jump se easy hoti hai — Silicon indirect-gap hai isliye yahi dominate karta hai aur energy heat (phonon) ban jaati hai. Auger: energy ek teesre carrier ko mil jaati hai — heavy doping ya high injection par matter karta hai.
Lifetime τ ka concept simple hai: low injection mein excess exponentially decay hota hai, Δp(t)=Δp(0)e−t/τ. n-type ke liye radiative τ=1/(Bn0), SRH τ=1/(σvthNt), Auger τ∝1/ND2. Teeno parallel channels hain, isliye rates add hote hain: 1/τ=∑1/τi — sabse fast raasta (smallest τ) hi jeetata hai. Yaad rakho: mid-gap trap sabse khatarnak "lifetime killer" hai, aur U hamesha net rate hai, sirf R nahi. Yeh cheezein device speed, LED brightness aur solar cell efficiency sab decide karti hain.