Formation of a p-n junction

No factory takes a finished block of p-type silicon and a finished block of n-type silicon and glues them together to make a diode. If you tried, the join would be a mess of misaligned atoms, oxide skin, and stray defects — and none of the delicate electronic behaviour a p-n junction is famous for would ever appear. The real trick is quieter: take one single crystal, unbroken from one end to the other, and change which impurity atoms are diffused into it as you cross an imaginary line partway through. The lattice never notices; the electrons and holes very much do. What happens next, once that line is drawn, is a small drama that plays out in nanoseconds and then freezes into a permanent, self-maintaining boundary. Holes crowd on one side, electrons crowd on the other, and the sheer imbalance sets both marching across the line — not because any field is pushing them yet, but because concentration itself is a kind of pressure. As they cross and vanish into each other, they leave behind something that cannot move: the charged skeletons of the dopant atoms themselves. Those fixed charges build an invisible wall of electric field exactly at the place where the crystal seemed, moments ago, perfectly uniform. This article follows that story end to end — from why the crystal must be continuous, through diffusion and recombination, to the depletion region and its barrier potential settling into equilibrium — the essential background for everything a p-n junction diode later does under forward and reverse bias.

Key takeaways

A p-n junction is never assembled from two separate crystals; it must be one continuous single crystal with donor impurities diffused into one region and acceptor impurities into the adjoining region.
The metallurgical junction is the internal plane where the dopant type changes; the crystal lattice itself runs unbroken across it.
Diffusion starts the process: holes flood from the p-side into the n-side and electrons flood the other way, driven purely by the concentration difference, with no field involved yet.
Recombination near the junction consumes these wandering carriers and uncovers the fixed, immobile dopant ions underneath — negative acceptor ions on the p-side, positive donor ions on the n-side.
This ion-only, carrier-free zone is the depletion region, and the exposed ions create an internal electric field pointing from n to p that opposes further diffusion.
Growth stops at equilibrium, when the diffusion current is exactly balanced by an opposite drift current, leaving a fixed barrier potential (about $0.7\ \text{V}$ for silicon, $0.3\ \text{V}$ for germanium) with zero net current in the unbiased junction.
Key idea: a p-n junction forms only inside one continuous crystal, where diffusion of carriers across the doping boundary, followed by recombination, exposes fixed dopant ions whose electric field grows until it exactly balances further diffusion — freezing in a depletion region and a barrier potential at equilibrium.

The core insight

A p-n junction is built as a single continuous crystal in which diffusion of majority carriers across the doping boundary, followed by recombination, exposes fixed dopant ions whose field grows until it balances further diffusion, leaving a stable depletion region and barrier potential at equilibrium.

Where it fits in the syllabus

Within Unit IX, "Semiconductor Electronics: Materials, Devices and Simple Circuits," the CBSE 2025-26 prescription explicitly keeps "p-n junction (formation, depletion region — qualitative)" in scope, alongside n-type and p-type extrinsic semiconductors as prerequisites and the semiconductor diode's forward/reverse I-V behaviour as the topic this one feeds into. The treatment is deliberately qualitative — you are expected to describe and reason about diffusion, recombination, the depletion region, and the barrier potential, but no exponential diode-current formula or band-bending calculus is required or examined. This article stays strictly within that qualitative frame; anything resembling a derived current-voltage law belongs to the diode's characteristics, not to formation.

The physical picture

Picture a single crystal of intrinsic semiconductor in which one half has been doped, during growth or by diffusion, to become n-type, and the other half to become p-type, meeting at a well-defined internal plane. Right at that plane, immediately after doping, the two sides are still locally out of balance: the p-side has a very high hole concentration and very few free electrons, while the n-side has a very high electron concentration and very few free holes. Away from the junction, each region individually remains electrically neutral, since every dopant ion still has its compensating mobile carrier nearby. That sharp concentration gradient right at the boundary is itself a driving force, exactly as a drop of ink pushes outward into still water. Holes diffuse from the p-side into the n-side, and electrons diffuse from the n-side into the p-side, purely because there are many more of them on one side than the other — no electric field is needed to start this, and initially none exists.

J_{\text{diff}} \propto -\frac{dn}{dx}, \qquad J_{\text{diff}} \propto -\frac{dp}{dx}

where: $J_{\text{diff}}$ is the diffusion current density carried by each type of carrier, and $dn/dx$ , $dp/dx$ are the concentration gradients of electrons and holes across the junction. The physical reading is only qualitative here — the steeper the concentration changes near the boundary, the stronger the initial tendency to diffuse; CBSE does not require evaluating this expression, only the idea that diffusion is driven by a concentration gradient, not by a field. As holes cross into the n-side they meet a sea of free electrons and recombine; as electrons cross into the p-side they meet a sea of free holes and recombine. This recombination consumes carriers on both sides of the boundary and, crucially, uncovers something that was always there but invisible while neutralised: the dopant atoms themselves, fixed in the crystal lattice, unable to move. On the p-side these are negatively charged acceptor ions; on the n-side, positively charged donor ions. A region straddling the junction, now stripped of free carriers and populated only by these fixed charges, is the depletion region (also called the space-charge region).

Figure

A single crystal shown as a horizontal bar with a p-type region on the left and an n-type region on the right meeting at a central depletion region, showing negative acceptor ions on the left edge of the depletion zone, positive donor ions on the right edge, an internal electric field arrow pointing from the n-side toward the p-side across the depletion zone, and small diffusion arrows for holes moving right and electrons moving left just outside it

The exposed ions do not stay silent. Negative acceptor ions on the p-side and positive donor ions on the n-side set up an internal electric field spanning the depletion region, pointing from the n-side toward the p-side. This field opposes any further diffusion of majority carriers across the junction — it pushes holes back toward the p-side and electrons back toward the n-side — while simultaneously sweeping the few minority carriers (stray electrons wandering into the p-side, stray holes wandering into the n-side) across the junction as a small drift current in the opposite sense to diffusion. As the depletion region widens, the field strengthens, diffusion slows, and drift grows, until the two exactly cancel:

I_{\text{net}} = I_{\text{diffusion}} - I_{\text{drift}} = 0

where: $I_{\text{diffusion}}$ is the net current of majority carriers still managing to cross against the growing field, and $I_{\text{drift}}$ is the current of minority carriers being swept across by that same field. The physical reading is that this is a dynamic, not a static, equilibrium — carriers continue crossing in both directions at every instant, but their rates are equal, so the observable net current through an unbiased junction is zero. The stabilised field corresponds to a fixed potential difference across the depletion region, the barrier potential $V_0$ , roughly $0.7\ \text{V}$ for silicon and $0.3\ \text{V}$ for germanium at room temperature — small, fixed numbers worth memorising exactly as given, since CBSE does not require deriving them.

Formation as an energy-balance argument

Derivation AWhy diffusion and drift settle into a fixed equilibrium

The goal is to explain, without calculus and without the diode's exponential current law, why the depletion region does not grow forever and instead locks into a fixed width with a fixed barrier potential — using nothing beyond the qualitative link between concentration gradients, exposed charge, and the field that charge produces. This is exactly the level of argument CBSE expects for "formation, depletion region — qualitative," and it is the reasoning examiners look for in a full-mark answer.

Step 1. Start from the trigger. The moment p-type and n-type regions coexist in one crystal, the p-side has far more holes and far fewer electrons than the n-side, and vice versa. This concentration mismatch, by itself, causes diffusion — holes flow p to n, electrons flow n to p — with no field required to start it, because diffusion is driven purely by the gradient in carrier concentration.

Step 2. Track what diffusion leaves behind. Every hole that diffuses into the n-side and recombines there removes one mobile positive charge from circulation near the junction and leaves behind the acceptor ion it had been neutralising, now exposed and immobile, on the p-side. The mirror process on the n-side exposes donor ions. Because recombination happens close to the junction, the exposed ions accumulate in a narrow region straddling it — the depletion region — while the far bulk of each side remains as neutral as before.

Step 3. Connect exposed charge to a field. A layer of negative ions next to a layer of positive ions is precisely the arrangement that produces an internal electric field, directed from the positive (n-side) layer toward the negative (p-side) layer. Wider depletion region means more exposed ions means a stronger field — the field's strength is tied directly to how far diffusion has already proceeded, which is the key coupling that eventually shuts the process down.

Step 4. Let the field push back. This field acts on the very carriers that are trying to diffuse: it decelerates further holes attempting to leave the p-side and further electrons attempting to leave the n-side, while accelerating the rare minority carriers (electrons in the p-side, holes in the n-side) across the junction in the drift direction, opposite to diffusion.

Step 5. Reach equilibrium. As the depletion region widens, the field intensifies, so diffusion current keeps falling while drift current keeps rising — until they become equal in magnitude, giving zero net current. Beyond that point there is no further net flow to expose more ions, so the field stops growing, the depletion width stops growing, and the barrier potential locks at its equilibrium value ( $\approx 0.7\ \text{V}$ silicon, $\approx 0.3\ \text{V}$ germanium). Nothing external has been applied; this is the junction's own self-limiting response to its own initial imbalance. Verified: the logical chain (concentration gradient causes diffusion, diffusion causes recombination, recombination exposes ions, exposed ions create a field, the field opposes diffusion and drives drift, equality of diffusion and drift currents is the equilibrium condition) matches the qualitative CBSE treatment of p-n junction formation with no formula outside the syllabus ceiling invoked.

Worked examples

Board level - explaining why the depletion region has a definite, finite width

Question: A freshly doped p-n junction is observed to stop growing after its depletion region reaches a certain width, with a stable barrier potential thereafter. Explain, step by step, why the growth halts rather than continuing indefinitely.

Step 1. State what starts and sustains growth. The depletion region grows because holes and electrons diffuse across the junction and recombine, each recombination event exposing one more fixed ion — negative on the p-side, positive on the n-side.

Step 2. State what the exposed ions produce. As more ions are exposed, the internal electric field across the depletion region strengthens; the field always points from the n-side toward the p-side.

Step 3. State the field's effect on further diffusion. This strengthening field opposes the diffusion of majority carriers, making it progressively harder for holes to leave the p-side and electrons to leave the n-side, while it simultaneously assists a small drift current of minority carriers in the opposite direction.

Step 4. Identify the stopping condition. Growth halts precisely when the diffusion current has fallen, and the drift current has risen, until the two become equal in magnitude — the net current is then zero, no further ions are exposed, and the depletion width and barrier potential freeze at their equilibrium values. Answer: the depletion region stops growing because the field it creates progressively opposes further diffusion until diffusion current exactly equals drift current, at which point net carrier flow across the junction is zero.

Competitive level - comparing silicon and germanium junctions by their barrier potential

Question: Two otherwise similar freshly formed junctions are made, one in silicon (barrier potential $\approx 0.7\ \text{V}$ ) and one in germanium (barrier potential $\approx 0.3\ \text{V}$ ). Using only the equilibrium argument for formation, reason out which junction is likely to have exposed more fixed ions — that is, a comparatively wider depletion region — at equilibrium, and justify your reasoning.

Step 1. Recall the coupling established in the derivation: a wider depletion region means more exposed ions, and more exposed ions produce a stronger internal field, which corresponds to a larger barrier potential $V_0$ across the junction.

Step 2. Read the comparison the other way. Since silicon's equilibrium barrier potential ( $0.7\ \text{V}$ ) is more than double germanium's ( $0.3\ \text{V}$ ), and barrier potential tracks the strength of the field built from exposed ions, the silicon junction must have needed to expose more fixed ionic charge before diffusion and drift could balance.

Step 3. State the qualitative conclusion. For comparable doping levels, the silicon junction settles into equilibrium only after a comparatively wider depletion region has formed than the germanium junction, precisely because it takes a stronger internal field — built from more exposed ions across a wider zone — to bring silicon's higher diffusion tendency into balance with an equal drift current.

Step 4. Sanity-check against the physical trend the syllabus expects. This is consistent with the general fact (used again later for the diode's forward-bias behaviour) that silicon junctions require a larger externally applied forward voltage than germanium junctions before appreciable current flows — the higher $0.7\ \text{V}$ built-in barrier for silicon must first be overcome. Answer: the silicon junction, with its larger $0.7\ \text{V}$ barrier potential, has the wider depletion region at equilibrium compared with germanium's $0.3\ \text{V}$ barrier, for similar doping. Timing note: under ninety seconds. The whole comparison rides on one coupling learned in the derivation — larger barrier potential implies more exposed ion charge implies a wider depletion region — so there is no arithmetic to perform, only the correct direction of that one qualitative link. The trap is reversing the logic and assuming a smaller barrier potential means a wider region; always check which junction needs a stronger field to reach balance.

Common traps

Imagining the junction is made by pressing two separately grown crystals together — a real p-n junction must be one continuous, unbroken lattice for a clean depletion region to form.
Believing an electric field is present from the first instant of doping — the field is a consequence of diffusion and recombination exposing ions; it does not pre-exist and start the process.
Treating the depletion region as forming instantly and staying frozen forever — it is a dynamic equilibrium of continuously balanced diffusion and drift, and its width changes further once an external bias is applied.
Assuming the entire p-type or n-type block carries net charge once the junction forms — only the thin depletion region has exposed, uncompensated ions; the bulk of each side stays neutral.
Mixing up which fixed ions sit where — acceptor ions (negative) are exposed on the p-side, donor ions (positive) on the n-side; reversing this also reverses the field direction.
Quoting the wrong barrier potential values — roughly $0.7\ \text{V}$ for silicon and $0.3\ \text{V}$ for germanium are fixed, recalled numbers, not values derived from a formula.

Common trapThe junction is not two crystals glued together, and the field is not there from the start

The single most costly misconception is picturing p-n junction formation as a mechanical assembly — take a p-type chip, take an n-type chip, clamp them together, and call it a junction. Real junctions must be fabricated as one continuous single crystal, with the dopant type changing across an internal metallurgical junction while the lattice itself stays unbroken; any real mechanical interface between separately grown pieces has surface defects and misaligned planes that make smooth diffusion, recombination, and a clean depletion region impossible. The second half of the same mistake is assuming the internal electric field exists from the moment the two doping regions are created, somehow "causing" carriers to move. It is the reverse: diffusion, driven only by the concentration gradient, comes first; recombination near the junction then exposes fixed ions; only those exposed ions create the field, which afterwards opposes further diffusion. Get the order right — diffusion, then recombination, then exposed ions, then field, then equilibrium — and both traps disappear together.

You should be able to…

Explain why a p-n junction must be fabricated as one continuous single crystal rather than two separately grown blocks joined together.
State, in correct order, the sequence of physical events during formation: diffusion, recombination, exposed fixed ions, growing internal field, equilibrium.
Identify which fixed ions (acceptor or donor) are exposed on each side of the depletion region, and the direction of the resulting internal field.
State the equilibrium condition (diffusion current equals drift current, net current zero) for an unbiased, freshly formed junction.
Quote the approximate barrier potential for silicon ( $0.7\ \text{V}$ ) and germanium ( $0.3\ \text{V}$ ) and reason qualitatively about how a larger barrier potential relates to a wider depletion region.
Distinguish a dynamic equilibrium (continuous, balanced crossing in both directions) from a static, frozen boundary.

Ask the tutor: Want to see how forward and reverse bias each disturb this diffusion-drift balance to produce the diode's I-V characteristics — ask your AI tutor to walk through what changes once an external voltage is applied to this same junction.

Why can't we just glue a p-type block and an n-type block together to make a junction faster?

Because the join between two separately grown crystals is never atomically perfect — there will be surface oxide, misaligned lattice planes, and defect states right at the interface, even if you clamp them as tightly as possible. The whole formation story depends on electrons and holes diffusing smoothly through a continuous, regular lattice and recombining cleanly near the junction; a mechanical seam disrupts that diffusion with scattering and trapping, so no well-defined depletion region or barrier potential ever forms properly. Real junctions are grown or diffused as one crystal specifically to avoid this.

Does the depletion region have any free electrons or holes in it at all?

Essentially none, once equilibrium is reached — that is exactly why it is called the depletion region: free carriers have been depleted from it through recombination, leaving only the fixed, immobile dopant ions behind (negative acceptor ions on the p-side, positive donor ions on the n-side). A tiny trickle of minority carriers does drift through it continuously as part of the balanced drift current, but their number is small compared with the carrier densities in the neutral bulk regions on either side.

If diffusion and drift are equal at equilibrium, doesn't that mean the junction is doing nothing?

It means the net current is zero, not that nothing is happening. Carriers are still crossing the junction in both directions continuously — holes and electrons still diffuse, minority carriers still drift — but the two flows exactly cancel at every instant, which is what a dynamic equilibrium means. This is very different from a truly static, frozen boundary; the moment an external voltage is applied (forward or reverse bias), that balance is disturbed and a measurable net current appears, which is the entire basis of diode action covered next.

Why is silicon's barrier potential bigger than germanium's, roughly 0.7 V versus 0.3 V?

At this syllabus level, treat these as recalled, fixed values rather than something you derive — they arise from differences in the two materials' band structure and intrinsic carrier concentration, which are beyond the qualitative CBSE treatment. What you should be able to do is reason with the numbers once given: a larger barrier potential corresponds to a stronger equilibrium internal field, which corresponds to more exposed ionic charge, which corresponds to a comparatively wider depletion region for similar doping, as used in the competitive worked example above.