Chapter 18 of 18

100%

JEE Main Physics★ ~1 question (4 marks) in JEE Main; now a notably smaller chapter than it used to be, since the entire transistor section was removed in 2023, which makes what remains very high yield per page

Electronic Devices (Semiconductors)

Doping is the whole subject — one impurity atom per million changes conductivity by orders of magnitude and lets you choose the sign of the carriers. Every device here is a junction between two doped regions.

⏱ 23 min readMedium difficulty✓ Free · exam-pattern aligned

Timed practiceMCQs from this chapter with instant scoring

FlashcardsFormulas & traps on spaced repetition

JEE Main PYQsSolved previous-year questions

By the end of this chapter you'll be able to…

1Classify materials by band gap and explain why conduction rises so steeply with temperature
2Distinguish n-type from p-type, and justify why both remain electrically neutral
3Apply the mass action law and compute conductivity from carrier densities and mobilities
4Describe depletion region formation and explain why no external voltage is measurable across it
5Analyse diode circuits, rectifier output frequency and capacitor ripple
6Distinguish LED, photodiode, solar cell and Zener by bias condition and function, and evaluate gate truth tables

💡

Why this chapter matters in JEE Main

The chapter reads like a list of devices to memorise and is really one mechanism applied repeatedly. Doping is the entire subject: a pure semiconductor is a poor conductor and a poor insulator, useful for almost nothing, while roughly one impurity atom per million raises the conductivity by orders of magnitude and — far more importantly — lets you choose whether the charge carriers are negative or positive. Every device here is then a junction between two differently doped regions, where forward bias narrows the depletion region and reverse bias widens it, and that single asymmetry produces rectification, light emission, light detection, photovoltaic generation and voltage regulation. JEE Main returns to neutrality of doped material, the mass action law, rectifier output frequency, Zener regulator arithmetic, LED band gaps and gate truth tables.

Before you start — revise these

🔗

Resistivity and its temperature dependence, from Current Electricity

🔗

Photon energy and the relation

h c = 1240

eV nm, from Dual Nature

🔗

Electric field and potential difference, from Electrostatics

🔗

Capacitor charging and discharging

Electronic Devices (Semiconductors)

A silicon crystal is doped with phosphorus, giving n-type material. Does it carry a net negative charge?

Most say yes — that is what the "n" seems to mean.

It is exactly neutral. Each phosphorus atom donates one mobile electron and is itself left as a fixed positive ion of equal magnitude. The "n" names the sign of the majority carriers, not a charge on the material.

The chapter has a reputation as a list of devices to memorise. It is one mechanism applied repeatedly:

Doping is the entire subject. Pure semiconductor is a poor conductor and a poor insulator, useful for almost nothing. One impurity atom per million raises conductivity by orders of magnitude and — far more importantly — lets you choose the sign of the carriers.
Every device here is a junction between two differently doped regions. Forward bias narrows the depletion region, reverse bias widens it, and that one asymmetry produces rectification, light emission, light detection, photovoltaic generation and voltage regulation.

Scope note. Junction transistors, transistor action, and the transistor as amplifier, oscillator and switch were all removed from JEE Main in the 2023 revision and stay out for 2026. Diodes and logic gates are fully in, and this chapter is scoped accordingly.

1. Energy Bands

In an isolated atom electrons occupy discrete levels. In a solid the vast number of interacting atoms broadens those levels into bands. The valence band is the highest filled or partly filled band; the conduction band is the next one up. The gap between them decides everything.

Material	Band gap	Behaviour
Conductor	Zero or overlapping	Conducts at all temperatures
Semiconductor	0.7 to 1.1 eV	Insulating cold, conducting warm
Insulator	Above ~3 eV	Effectively no conduction

Germanium 0.7 eV, silicon 1.1 eV, diamond about 5.5 eV — which is why diamond is transparent and insulating while silicon is neither.

Trap. Thermal energy at room temperature is only 0.026 eV, far below either semiconductor gap. Conduction happens because a small fraction of electrons sit in the tail of the thermal distribution. That is why the carrier population rises so steeply with temperature.

Illustration 1

Intrinsic carrier concentration goes as $n_{i} \propto e^{- E_{g} /2 k T}$ . Compare germanium ( $E_{g} = 0.7$ eV) with silicon (1.1 eV) at 300 K, where $k T = 0.026$ eV.

$\frac{n _{i} ( Ge )}{n _{i} ( Si )} = exp (\frac{1.1 - 0.7}{2 ( 0.026 )}) = e^{7.7} \approx 2200$

Germanium carries about two thousand times more intrinsic carriers. That is exactly why a germanium diode leaks microamperes in reverse where silicon leaks nanoamperes, and why silicon displaced germanium once low leakage mattered more than a low knee voltage.

2. Intrinsic Semiconductors

A pure semiconductor conducts by thermal generation of electron-hole pairs. Promoting an electron to the conduction band leaves a vacancy in the valence band, and that vacancy behaves as a mobile positive charge — a hole.

$n_{e} = n_{h} = n_{i} (intrinsic)$

necessarily, since every carrier of one type is created alongside one of the other.

A hole is not a physical particle. It is the absence of an electron, appearing to move when a neighbouring electron shifts into it — like the empty square in a sliding tile puzzle.

Illustration 2

Pure silicon has $n_{i} = 1.5 \times 1 0^{16}$ m⁻³ at 300 K. It is doped with donors until $n_{e} = 5 \times 1 0^{22}$ m⁻³. Find the hole concentration.

The mass action law holds whatever the doping:

$n_{e} n_{h} = n_{i}^{2} \Rightarrow n_{h} = \frac{( 1.5 \times 1 0 ^{16} ) ^{2}}{5 \times 1 0 ^{22}} = \frac{2.25 \times 1 0 ^{32}}{5 \times 1 0 ^{22}} = 4.5 \times 1 0^{9} m^{- 3}$

Notice what doping did to the minority carriers. Before doping there were $1.5 \times 1 0^{16}$ holes per cubic metre; afterwards there are $4.5 \times 1 0^{9}$ , about three million times fewer.

Adding carriers of one type removes carriers of the other. The flood of electrons raises the recombination rate until the hole population is driven down far enough to keep the product fixed. The material stays electrically neutral throughout, because the fixed donor ions supply the balancing positive charge.

Intrinsic conductivity is far too small and far too temperature-sensitive for any device, which is precisely why doping exists.

3. Doping

Type	Dopant group	Majority	Minority	Impurity called
n-type	15, pentavalent (P, As)	Electrons	Holes	Donor
p-type	13, trivalent (B, In)	Holes	Electrons	Acceptor

Phosphorus has five valence electrons: four bond, one is loosely held and easily promoted. Boron has three: one bond is left incomplete, creating a hole.

Trap. Both types remain electrically neutral overall — the chapter's most misunderstood point. A donor contributes a mobile electron but is left as a fixed positive ion, and the totals balance exactly.

$n_{e} n_{h} = n_{i}^{2} (mass action law, at fixed T)$

So raising one population suppresses the other. Heavy n-doping does not merely add electrons; it actively removes holes.

Conductivity and temperature

$σ = e (n_{e} μ_{e} + n_{h} μ_{h})$

Electron mobility beats hole mobility in silicon by roughly three times, because a hole moves only through successive electron hops while an electron moves directly.

Semiconductors have a negative temperature coefficient of resistance — the opposite of metals — and it follows straight from $ρ = m / n e^{2} τ$ met in Current Electricity:

	$n$	$τ$	Result
Metal	fixed	falls on heating	$ρ$ rises
Semiconductor	rises exponentially	falls	$ρ$ falls

Doped semiconductors are far less temperature-sensitive than intrinsic ones, because the carrier population is set by the impurity concentration rather than by thermal generation.

Illustration 3

Silicon is doped to $n_{e} = 5 \times 1 0^{22}$ m⁻³ with $μ_{e} = 0.135$ m² V⁻¹ s⁻¹. Find its conductivity and resistivity, neglecting the hole contribution.

$σ = e n_{e} μ_{e} = (1.6 \times 1 0^{- 19}) (5 \times 1 0^{22}) (0.135) \Rightarrow σ = 1080 S m^{- 1}$

$ρ = \frac{1}{σ} = 9.3 \times 1 0^{- 4} Ω m$

Neglecting holes is safe here: by the mass action law they number only $4.5 \times 1 0^{9}$ m⁻³, thirteen orders below the electrons. Yet even so, this doped silicon is still about 50,000 times more resistive than copper — a semiconductor is never trying to be a conductor.

4. The p-n Junction

Join p-type and n-type and carriers diffuse across, driven purely by the concentration difference — electrons from n to p, holes from p to n.

Each departing carrier leaves behind a fixed ionised impurity, so the n side acquires a positive layer and the p side a negative one. That region, stripped of mobile carriers, is the depletion region, typically a fraction of a micrometre wide.

The charged layers set up a field opposing further diffusion, and equilibrium arrives when drift exactly balances diffusion. The resulting step is the barrier potential: about 0.3 V for germanium, 0.7 V for silicon.

Trap. No current flows at equilibrium and no external voltage appears across an unbiased junction. Attaching leads creates metal-to-semiconductor contacts whose own potentials cancel the barrier exactly around the loop. If any net voltage survived, a junction lying on a bench could drive a current forever.

Illustration 4

A silicon junction has a 0.7 V barrier across a depletion region 0.5 µm wide. Find the average electric field there.

$E = \frac{V}{d} = \frac{0.7}{0.5 \times 1 0 ^{- 6}} \Rightarrow E = 1.4 \times 1 0^{6} V m^{- 1}$

Nearly half the field that breaks down air, sitting inside an unpowered component on the shelf. The depletion region is thin enough that a fraction of a volt produces a colossal field — which is also why a modest reverse voltage can trigger breakdown.

5. Biasing and the I-V Characteristic

	Forward bias	Reverse bias
Positive terminal to	p side	n side
Depletion region	Narrows	Widens
Barrier	Lowered	Raised
Resistance	Low, a few ohms	Very high
Current	Rises steeply past the knee	Tiny saturation current only

Reverse saturation current depends on temperature but scarcely on the applied voltage, because it is limited by how fast minority carriers are generated, not by the field pushing them.

Beyond a critical reverse voltage the junction breaks down. Ordinary diodes are destroyed by this; Zener diodes are built to survive and exploit it.

Illustration 5

A silicon diode with a 0.7 V knee is in series with a 100 Ω resistor across a 5 V supply. Find the current and the power dissipated in the diode.

Treat the diode as a fixed 0.7 V drop once conducting:

$I = \frac{5 - 0.7}{100} = 43 mA \Rightarrow P_{d i o d e} = (0.7) (0.043) = 30 mW$

The resistor takes the other 185 mW. This is exactly how every LED in every appliance is driven — the diode fixes the voltage, the resistor fixes the current, and forgetting the resistor destroys the diode.

Illustration 6

A 10 V supply feeds a 30 Ω resistor in series with two parallel branches: branch A is an ideal diode connected forward with 10 Ω, branch B an ideal diode connected reverse with 20 Ω. Find the current drawn from the supply.

Check each diode before touching any arithmetic.

Branch B's diode is reverse biased, so it carries nothing and the 20 Ω might as well not be in the diagram.

Branch A conducts, so the circuit is simply 30 Ω in series with 10 Ω:

$I = \frac{10}{30 + 10} = 0.25 A$

The whole question is whether each diode conducts. Once that is settled the circuit is ordinary series-parallel, and a branch whose diode blocks is deleted outright — never averaged in, and never given some share of the current.

6. The Diode as a Rectifier

Property	Half-wave	Full-wave
Diodes	1	2 (centre-tap) or 4 (bridge)
Output frequency	Same as input	Twice the input
Efficiency	~40.6 %	~81.2 %
Ripple	Large	Smaller

The frequency doubling is a favourite exam point, and the fastest way to identify a full-wave circuit from a trace without following the wiring.

Rectified output is unidirectional but far from steady, so a capacitor filter goes across the load: it charges at each peak and discharges slowly between them.

Illustration 7

A full-wave rectifier at 50 Hz feeds a 1 kΩ load through a 100 µF capacitor, at 10 V peak. Estimate the ripple voltage.

The capacitor must supply the load between pulses, which arrive at $f_{r} = 100$ Hz:

$Δ V \approx \frac{I}{f _{r} C} = \frac{10/1000}{( 100 ) ( 100 \times 1 0 ^{- 6} )} \Rightarrow Δ V = 1 V$

Ten per cent ripple. A half-wave circuit would have twice as long to discharge, giving 2 V — which is the real reason full-wave is preferred, more than the efficiency figure. Bigger capacitor, smaller ripple, in inverse proportion.

Illustration 8

A transformer secondary supplies 20 V peak from 50 Hz mains. Find the output frequency and the peak inverse voltage for a half-wave rectifier, a centre-tapped full-wave rectifier and a bridge rectifier.

Output frequency first. Half-wave passes one hump per input cycle, so 50 Hz; both full-wave circuits pass two, so 100 Hz. That doubling is why the filter capacitor has half as long to discharge.

The peak inverse voltage is the largest reverse voltage any single diode has to survive:

Half-wave: the idle diode sees the full secondary peak, so $P I V = 20$ V.
Centre-tapped: with 20 V peak on each half winding, the idle diode sees its own half plus the conducting half, so $P I V = 40$ V.
Bridge: the two idle diodes share the reverse voltage, so $P I V = 20$ V.

The bridge needs four diodes where the centre-tapped circuit needs two, but halves the voltage rating each must carry and needs no centre tap at all. That is the trade which decides between them in practice, and the centre-tapped circuit's doubled PIV is the part most often forgotten.

7. Special Diodes

Zener diode. Heavily doped, so the depletion region is very thin and breakdown is reached at a low reverse voltage. Always operated in reverse breakdown, where the voltage across it stays almost constant however the current varies. A series resistor absorbs the excess, and any change in supply or load is taken up by a change in Zener current rather than in output voltage.

LED. Forward biased; light is produced when electrons recombine with holes. Photon energy is roughly the band gap, so the colour is fixed by the material, never by the drive voltage — which is why silicon at 1.1 eV emits invisible infrared and gallium compounds are needed for visible light.

Photodiode. Reverse biased; incident light generates electron-hole pairs and raises the reverse current. Reverse bias is used because the dark current is already tiny, so a small photocurrent is a large fractional change and far easier to measure.

Solar cell. A photodiode with no external bias, generating its own voltage from light, made with a large junction area to collect as much of it as possible.

The three light-related diodes, stated plainly: LED forward biased and emits, photodiode reverse biased and detects, solar cell unbiased and generates.

Illustration 9

A silicon solar cell of area 100 cm² under full sunlight (1000 W m⁻²) delivers 0.5 V at 3 A. Find its efficiency.

$P_{in} = (1000) (0.01) = 10 W, P_{o u t} = (0.5) (3) = 1.5 W$

$η = \frac{1.5}{10} = 15%$

The other 85 per cent is lost mostly to the band gap itself: photons below 1.1 eV pass straight through unabsorbed, and photons well above it waste their surplus as heat. That trade-off — a wider gap catches fewer photons but extracts more from each — is what sets the theoretical ceiling on any single-junction cell.

Illustration 10

Design a 5 V regulated supply for a 20 mA load from 230 V, 50 Hz mains, using a transformer, bridge rectifier, capacitor filter and Zener.

Transformer. Aim for roughly 9 V of usable DC. A bridge loses two diode drops, so target a secondary near 8 V RMS:

$V_{p e ak} = 82 = 11.3 V \Rightarrow V_{D C} \approx 11.3 - 1.4 = 9.9 V, \frac{N _{p}}{N _{s}} = \frac{230}{8} \approx 29$

Filter. With a 470 µF capacitor and $f_{r} = 100$ Hz:

$Δ V = \frac{I}{f _{r} C} = \frac{0.02}{( 100 ) ( 470 \times 1 0 ^{- 6} )} = 0.43 V$

so the rail sits between about 9.5 and 9.9 V.

Regulator. Take a 5.1 V Zener needing at least 5 mA. The worst case is the lowest rail voltage, where least current is available:

$R \leq \frac{9.5 - 5.1}{0.025} = 176 Ω \Rightarrow use 150 Ω$

Check the other extreme. At 9.9 V the resistor passes $(9.9 - 5.1) /150 = 32$ mA, the load takes 20 mA, and the Zener absorbs the surplus 12 mA — comfortably in breakdown, so the output holds at 5.1 V across the whole range.

Every stage is one device from this chapter doing exactly one job: the transformer scales, the bridge rectifies, the capacitor smooths, the Zener regulates.

8. Logic Gates

Gate	Logic	Output is 1 when
OR	$A + B$	Either input is 1
AND	$A \cdot B$	Both inputs are 1
NOT	$\overset{ˉ}{A}$	The single input is 0
NAND	$\overline{A \cdot B}$	Not both inputs are 1
NOR	$\overline{A + B}$	Neither input is 1

NAND and NOR are universal: any logic function whatsoever can be built from copies of either alone. That matters industrially, not just theoretically — a fabrication line that makes one gate type reliably can make anything, which is why real integrated circuits are dominated by NAND structures.

Reading a truth table is usually faster than manipulating Boolean algebra for the two-input cases JEE asks about, and much harder to get wrong.

Illustration 11

Inputs $A$ and $B$ feed both a NAND gate and an OR gate, and the two outputs feed an AND gate. Identify the result.

$Y = (A + B) \cdot \overline{A \cdot B}$

$A$	$B$	$A + B$	$\overline{A \cdot B}$	$Y$
0	0	0	1	0
0	1	1	1	1
1	0	1	1	1
1	1	1	0	0

Output is 1 exactly when the inputs differ — this is XOR. Read in words the circuit says "at least one, but not both", which is what exclusive-or means. Building the table beat guessing at the algebra.

Illustration 12

Show how NOT, AND and OR are each built from NAND gates alone.

NOT. Tie both inputs of a single NAND together:

$\overline{A \cdot A} = \overset{ˉ}{A}$

AND. A NAND is an AND followed by an inversion, so undo the inversion with a second NAND wired as a NOT:

$\overline{\overline{A \cdot B}} = A \cdot B$

OR. Invert both inputs first and NAND the results, and De Morgan's law finishes the job:

$\overline{\overset{ˉ}{A} \cdot \overset{ˉ}{B}} = A + B$

Three NAND gates for OR, two for AND, one for NOT. Every Boolean function can be written using only NOT, AND and OR, so every Boolean function can be built from NAND gates alone. That is what "universal" means, and it is why a fabrication line able to make one gate type reliably can make anything at all.

Summary

Doping is the whole subject — it changes conductivity by orders of magnitude and lets you choose the sign of the carriers.
Band gap separates the classes: zero for conductors, 0.7 to 1.1 eV for semiconductors, above 3 eV for insulators.
$k T$ at room temperature is only 0.026 eV, so conduction lives in the tail of the distribution and climbs steeply with $T$ .
Intrinsic: $n_{e} = n_{h} = n_{i}$ , since carriers are created only in pairs. A hole is an absence, not a particle.
Pentavalent gives n-type, trivalent gives p-type — and both are electrically neutral.
$n_{e} n_{h} = n_{i}^{2}$ : raising one population suppresses the other.
$σ = e (n_{e} μ_{e} + n_{h} μ_{h})$ ; semiconductors have a negative temperature coefficient because $n$ rises faster than $τ$ falls.
Depletion region and barrier potential: 0.3 V germanium, 0.7 V silicon — and no measurable external voltage.
Forward bias narrows and conducts; reverse bias widens and blocks. That asymmetry is what a diode is for.
Reverse saturation current depends on temperature, hardly on voltage.
Full-wave gives twice the input frequency and about twice the efficiency; ripple goes as $I / f_{r} C$ .
Zener: heavily doped, reverse breakdown, holds voltage while absorbing surplus current.
LED forward emits, photodiode reverse detects, solar cell unbiased generates.
LED colour is set by the band gap, never by the drive voltage.
NAND and NOR are universal — any function can be built from either alone.
Transistors are off the JEE Main syllabus since 2023; effort belongs on diodes and gates.

Key formulas & results

Everything to memorise for the exam hall, in one card. Screenshot this for revision.

Band gap classification

conductor \approx 0; semiconductor 0.7 - 1.1 eV; insulator > 3 eV

Germanium 0.7 eV, silicon 1.1 eV, diamond about 5.5 eV. Thermal energy at room temperature is only 0.026 eV, so conduction depends on the tail of the distribution — which is why the carrier count climbs so steeply with temperature.

Intrinsic carriers

n_{e} = n_{h} = n_{i}, n_{i} \propto e^{- E_{g} /2 k T}

Carriers are created only in pairs, so the two populations are necessarily equal. The exponential is why germanium carries about two thousand times more intrinsic carriers than silicon at 300 K, and hence leaks far more in reverse.

Mass action law

n_{e} n_{h} = n_{i}^{2}

Holds at a given temperature whatever the doping. Raising one population therefore suppresses the other: heavy n-doping does not merely add electrons, it actively removes holes.

Conductivity

σ = e (n_{e} μ_{e} + n_{h} μ_{h})

Electron mobility beats hole mobility in silicon by roughly three times, because a hole moves only through successive electron hops. Even heavily doped silicon remains tens of thousands of times more resistive than copper.

Temperature coefficient

ρ = \frac{m}{n e ^{2} τ} : metal n fixed, semiconductor n rises

The same relation as Current Electricity, with the two materials differing only in which factor dominates. Heating a semiconductor promotes exponentially more electrons across the gap, overwhelming the fall in $\tau$ and giving a negative temperature coefficient.

Barrier potential

V_{b} \approx 0.3 V (Ge), 0.7 V (Si)

Set up by fixed ionised impurities left behind as carriers diffuse across. Equilibrium is reached when drift balances diffusion, and no external voltage is measurable — the metal contacts' own potentials cancel it exactly around the loop.

Diode in a circuit

I = \frac{V _{s u ppl y} - V _{k n ee}}{R}

Treat a conducting diode as a fixed voltage drop. This is how every LED is driven — the diode fixes the voltage and the series resistor fixes the current. A reverse-biased branch carries nothing and is deleted from the circuit entirely.

Rectifier output

f_{o u t} = f_{in} (half-wave), f_{o u t} = 2 f_{in} (full-wave)

Full-wave inverts the negative half rather than discarding it, giving two pulses per input cycle and about 81 per cent efficiency against 41 per cent. The frequency doubling is the fastest way to identify a full-wave circuit from a trace.

Capacitor filter ripple

Δ V \approx \frac{I}{f _{r} C}

The capacitor supplies the load between pulses. A half-wave circuit has twice as long to discharge and therefore twice the ripple, which is the practical reason full-wave is preferred rather than the efficiency figure.

Light-related diodes and gates

LED forward emits; photodiode reverse detects; solar cell unbiased generates

LED photon energy is roughly the band gap, so colour is fixed by material and never by drive voltage. NAND and NOR are universal gates — any logic function can be built from copies of either alone, which is why real chips are dominated by NAND.

Peak inverse voltage

P I V = V_{m} (half-wave, bridge), P I V = 2 V_{m} (centre-tapped full-wave)

The centre-tapped circuit's idle diode sees its own half winding **plus** the conducting half, which is the doubling most often forgotten. The bridge needs four diodes rather than two but halves the rating each must carry, and needs no centre tap.

Zener regulator series resistor

R = \frac{V _{in} - V _{Z}}{I _{Z} + I _{L}}, size it at the lowest V_{in}

The worst case is the bottom of the ripple, where least current is available to share between Zener and load. Then check the top of the ripple to confirm the Zener current stays within its power rating; the Zener absorbs whatever the load does not take.

NAND and NOR universality

\overline{A \cdot A} = \overset{ˉ}{A}, \overline{\overline{A \cdot B}} = A \cdot B, \overline{\overset{ˉ}{A} \cdot \overset{ˉ}{B}} = A + B

One NAND gives NOT, two give AND, three give OR, and De Morgan's law supplies the last step. Since every Boolean function is expressible in NOT, AND and OR, NAND alone suffices for all of them — and NOR does the same by the dual construction.

JEE Main test series

Ready to be tested on this?

Electronic Devices (Semiconductors) shows up in every JEE Main paper — sit a timed one and find out where you actually stand.

~1Est. Qs

Mediumpriority

Full JEE Main mocks, real clock
Physics-only sectionals & chapter tests
All-India rank + topic-level analysis

Free forever · solver-verified questions

⚠️

Traps JEE Main sets — and how to dodge them

These are the exact option-traps and misreads that cost marks under negative marking.

WATCH OUT

✗ Thinking n-type material carries a net negative charge

✓ The material is exactly neutral. Each donor releases a mobile electron and is left as a fixed positive ion of equal magnitude. The n refers only to the sign of the majority carriers.

Why it happens: The letter n is read as "negative", and the dopant does visibly supply an electron.

WATCH OUT

✗ Believing doping only adds carriers without affecting the other type

✓

n_{e} n_{h} = n_{i}^{2}

must hold at a given temperature, so raising the electron count by a million forces the hole count down by the same factor. The extra electrons recombine with holes until the product is restored.

Why it happens: Adding donors obviously adds electrons, and nothing visibly removes holes.

WATCH OUT

✗ Giving the same output frequency for half-wave and full-wave rectifiers

✓ Half-wave discards the negative half and gives one pulse per cycle at the input frequency. Full-wave inverts it and gives two pulses per cycle, so the output is at twice the input frequency.

Why it happens: Both are fed from the same supply, so the output is assumed to inherit its frequency.

WATCH OUT

✗ Forward biasing a Zener diode or a photodiode

✓ A Zener is operated in reverse breakdown, which is the only place its voltage is constant. A photodiode is reverse biased because the dark current is tiny there, making a small photocurrent a large fractional change. Only the LED is forward biased.

Why it happens: Forward bias is where a diode "works", so it is applied to every diode by habit.

WATCH OUT

✗ Expecting semiconductor resistance to rise with temperature as a metal's does

✓ In a metal the carrier count is fixed, so more scattering means more resistance. In a semiconductor heating promotes exponentially more carriers across the gap, and that increase in

n

swamps the scattering entirely, giving a negative temperature coefficient.

Why it happens: Heating increases lattice vibration in both, so the same conclusion is drawn for both.

WATCH OUT

✗ Thinking LED colour can be changed by varying the applied voltage

✓ Photon energy is roughly the band gap of the material, so the colour is fixed at manufacture. Raising the drive current makes the same colour brighter. This is exactly why silicon can only emit infrared and blue LEDs required new materials.

Why it happens: Brightness does change with drive current, so colour seems likely to follow.

Exam-pattern practice

PYQ-style questions with full solutions. Work through them as a readiness check — mark yourself honestly and get your gap report at the end.

Readiness check

Are you exam-ready for Electronic Devices (Semiconductors)?

12 problems from this chapter. Try each one, reveal the worked solution, mark yourself honestly — get your gap report at the end.

12 questions~8 min worth ~4 marks in JEE Main exams

5-minute revision

The whole chapter, distilled. Read this the night before the exam.

•Doping is the whole subject — it sets both the size and the sign of the conductivity
•Gaps: zero for conductors, 0.7 eV Ge and 1.1 eV Si, above 3 eV for insulators; $k T$ is only 0.026 eV
•Intrinsic $n_{e} = n_{h} = n_{i}$ ; a hole is an absence, not a particle
•Pentavalent gives n-type, trivalent gives p-type, and both are electrically neutral
• $n_{e} n_{h} = n_{i}^{2}$ , so raising one population suppresses the other
•Semiconductors have a negative temperature coefficient because $n$ rises faster than $τ$ falls
•Barrier potential 0.3 V Ge, 0.7 V Si, with no measurable external voltage
•Forward bias narrows and conducts; reverse bias widens and blocks, leaving only a temperature-limited saturation current
•Full-wave gives twice the input frequency and about twice the efficiency; ripple goes as $I / f_{r} C$
•LED forward emits, photodiode reverse detects, solar cell unbiased generates; NAND and NOR are universal
•PIV is $V_{m}$ for half-wave and bridge but $2 V_{m}$ for the centre-tapped full-wave — the doubling is the part most often missed
•Doping drives the minority carriers down: $n_{e} n_{h} = n_{i}^{2}$ holds whatever the doping, so more electrons means fewer holes

JEE Main question blueprint

How this topic is asked, tier by tier — so you can prep to the pattern.

Typical weightage: ~1 question (4 marks) of the 100-mark Physics section

Question style	Marks each	Typical count	What it tests
Diodes, rectifiers and special diodes	2	1	Deciding which diodes conduct before doing any arithmetic, rectifier output frequency and peak inverse voltage, capacitor filter ripple, and Zener regulator design sized at the lowest rail voltage
Semiconductors and doping	1	1	Band gaps and the exponential intrinsic carrier count, the mass action law driving minority carriers down, $\sigma = e(n_e\mu_e+n_h\mu_h)$, and why heating lowers a semiconductor's resistivity but raises a metal's
Logic gates	1	1	Truth tables for combinations of two-input gates, identifying the resulting function, and building any gate from NAND or NOR alone using De Morgan's law

Exam-hall strategy

Battle-tested tips from mentors and toppers for this topic under the sectional clock.

In any diode network, decide first which diodes conduct. A reverse-biased branch carries nothing and is deleted from the diagram entirely, after which the circuit is ordinary series-parallel.
For rectifier questions, check half-wave against full-wave before computing anything. The output frequency, efficiency and ripple all differ by a factor of about two, and which one is being asked for is usually the whole question.
In Zener problems, work the worst case first — the minimum supply voltage, where the least current is available — then verify the maximum does not over-drive the diode.
Convert LED wavelength to band gap with $h c = 1240$ eV nm rather than through $h$ and $c$ separately. It is one step and keeps everything in electronvolts.
For gate questions, build the truth table rather than manipulating Boolean algebra. For the two-input cases JEE asks about it is faster and far harder to get wrong.

Beyond the exam

Where this skill shows up in the job you're competing for — and in life.

Every integrated circuit in every computer and phone is b…

Every integrated circuit in every computer and phone is built from doped silicon junctions, and the whole industry rests on controlling impurity concentrations to parts per billion

Mains adapters use a bridge rectifier

Mains adapters use a bridge rectifier, capacitor filter and regulator in exactly the chain worked through at the end of this chapter, converting 230 V AC to a few volts of steady DC

Solar photovoltaic panels are large-area p-n junctions wh…

Solar photovoltaic panels are large-area p-n junctions whose efficiency ceiling is set directly by the band gap — too wide and most photons pass through, too narrow and each wastes its surplus as heat

Where else this topic is tested

Prepare once, score in every exam that asks it.

JEE Main

JEE Advanced

NEET UG

BITSAT

CBSE Class 12 Physics

📝 My notes

Saved on this device — sign in to sync · how this works

0/600

Questions aspirants ask

Pulled from the Q&A community and mentor sessions.

Because connecting the voltmeter creates two more junctions — metal to p-type at one end and metal to n-type at the other — and each of those has its own contact potential. Going right round the loop, they cancel the barrier exactly. That cancellation is not a coincidence but a requirement of thermodynamics: if any net voltage survived, a diode lying in a drawer could drive a current through a resistor forever. The barrier is entirely real; it simply cannot be tapped without supplying energy.

The same way the gap moves in a sliding tile puzzle. No physical object travels — an electron from a neighbouring bond hops into the vacancy, which leaves a vacancy where that electron came from, and repeating this shifts the vacancy steadily across the crystal. Treating the hole as a positive particle is a bookkeeping device that happens to work extremely well, and it even gets the mobility right, because holes move more slowly than electrons precisely on account of needing those successive hops.

Because they are optimised for different jobs. A photodiode is a sensor: it needs to report light level fast and linearly, and reverse bias widens the depletion region, speeds the response and keeps the dark current tiny so a small photocurrent is a large fractional change. A solar cell is a source: applying an external bias would consume power, defeating the point. It is run unbiased in the fourth quadrant of its characteristic, where it delivers rather than absorbs, and it is built with a large area to collect as much light as possible.

Because the photon energy is roughly the band gap, so blue light at around 450 nm needs a gap near 2.8 eV — far wider than red needs. The difficulty was not knowing this but finding a material with that gap that could also be doped both n-type and p-type and grown without ruinous defects. Gallium nitride eventually worked, after decades in which most researchers had given up on it. The 2014 Nobel Prize in Physics was awarded for exactly this, which gives a sense of how hard the materials problem was.

Because 0.026 eV is the average, not a ceiling. Thermal energies follow a distribution with a long exponential tail, and at any instant a small fraction of electrons has far more than the average — enough to cross a 1.1 eV gap. The fraction is tiny, which is why intrinsic silicon is such a poor conductor, but it is not zero. It also explains the steep temperature dependence: raising $T$ slightly fattens that tail enormously, so the carrier count climbs exponentially rather than linearly.

More Physics chapters

KinematicsIdentify what is held constant and the equations pick themselves. The three famous equatio…

Laws of MotionChoose the body, draw every force on it, add as vectors. Almost every wrong answer here is…

Work, Energy and PowerEnergy methods bypass the path. A problem that is intractable with forces collapses to one…

Rotational MotionTranslational mechanics with a dictionary applied — except that the new mass depends on wh…

Sources and How This Chapter Was CheckedSyllabus scope, what was derived rather than quoted, and how every answer here was checked.

Scope follows the NTA JEE Main syllabus (Unit 20, Electronic Devices): energy bands in solids with a qualitative treatment, conductors, insulators and semiconductors, semiconductor diodes, and I-V characteristics in forward and reverse bias.

It also covers the diode as a rectifier, the I-V characteristics of LED, photodiode and solar cell, the Zener diode and its use as a voltage regulator, and logic gates including OR, AND, NOT, NAND and NOR.

Junction transistors, transistor action and the transistor as amplifier, oscillator and switch were all removed in the 2023 revision and are not covered. The chapter says so explicitly rather than leaving a candidate to revise material that cannot be examined.

Results were derived rather than quoted: the germanium-to-silicon carrier ratio from the exponential dependence on band gap; the junction field from the barrier potential and depletion width; the ripple from the capacitor supplying the load between pulses; and the XOR identity by building the truth table rather than manipulating Boolean algebra.

Every illustration was checked. The conductivity calculation was verified against the mass action law to confirm that neglecting holes is safe by thirteen orders of magnitude, and against copper's resistivity to keep the scale honest. The supply design was checked at both extremes of the rail voltage to confirm the Zener stays in breakdown at the minimum and is not over-driven at the maximum.

The minority carrier concentration was computed from the mass action law and checked against the intrinsic value, confirming that doping drives the hole population about three million times below where it started. The three NAND constructions were each verified against the corresponding truth table. The peak inverse voltages were compared across the three rectifier circuits to confirm that only the centre-tapped one doubles, which is the figure most often missed.

The illustrations are teaching problems written for this chapter, not previous-year questions, and are not labelled as such.

Take the next step