Skip to content

Semiconductor Devices: Junctions, Diodes, BJTs, FETs and Heterojunctions

CSIR NET electronics: p-n junction physics, diode equation, special diodes, BJT and FET operation, small-signal gain, heterojunctions, with practice.

  • 23 min read
  • Updated
On this page · 15 sections
  1. Prerequisites
  2. The Hook: One Junction, Every Device
  3. Carriers in Semiconductors
  4. The p-n Junction in Equilibrium
  5. The Diode Equation
  6. Bipolar Junction Transistors
  7. Field-Effect Transistors
  8. Heterojunctions
  9. Memory Tricks
  10. Don’t Confuse
  11. Common Mistakes
  12. Practice
  13. Quick Revision Table
  14. Key Takeaways
  15. What’s Next

Prerequisites

The Hook: One Junction, Every Device

Carriers in Semiconductors

In a non-degenerate semiconductor in equilibrium,

$$n = N_c\,e^{-(E_c - E_F)/kT}, \qquad p = N_v\,e^{-(E_F - E_v)/kT}, \qquad \boxed{np = n_i^2 = N_cN_v\,e^{-E_g/kT}}$$

The product \( np \) depends only on temperature and band gap (the mass-action law). With full ionisation, an n-type sample has \( n \approx N_D \), \( p = n_i^2/N_D \); p-type has \( p \approx N_A \). The Fermi level moves from the intrinsic level by

$$E_F - E_i = kT\ln\frac{n}{n_i} \quad (\text{n-type}), \qquad E_i - E_F = kT\ln\frac{p}{n_i} \quad (\text{p-type})$$

Current has two parts: drift \( J = q(n\mu_n + p\mu_p)\mathcal E \) and diffusion \( J_n = qD_n\,dn/dx \), \( J_p = -qD_p\,dp/dx \). They are tied by the Einstein relation \( D/\mu = kT/q \).

The thermal voltage \( V_T = kT/q \) is 25.9 mV at 300 K. Every exponential on this page is written in units of it.

Material (300 K)\( E_g \) (eV)Gap type
Ge0.66indirect
Si1.12indirect
GaAs1.42direct
GaN3.4direct

The p-n Junction in Equilibrium

When p and n regions meet, electrons diffuse into the p side and holes into the n side, leaving behind uncovered ionised acceptors (negative) and donors (positive). This depletion region holds an electric field that opposes further diffusion. In equilibrium the Fermi level is flat, and the bands bend by the built-in potential:

$$\boxed{V_{bi} = \frac{kT}{q}\ln\frac{N_AN_D}{n_i^2}}$$

Solving Poisson’s equation with the depletion approximation (no free carriers in the region, abrupt edges):

$$\boxed{W = \sqrt{\frac{2\varepsilon(V_{bi} - V)}{q}\left(\frac{1}{N_A} + \frac{1}{N_D}\right)}}, \qquad N_Ax_p = N_Dx_n, \qquad \mathcal E_{\max} = \frac{qN_Dx_n}{\varepsilon}$$

where \( V \) is the applied forward bias (negative for reverse bias). Three consequences carry most Part B questions:

  1. Charge neutrality \( N_Ax_p = N_Dx_n \): the depletion region extends mostly into the lightly doped side. In a \( p^+n \) junction, \( W \approx x_n \).
  2. The field is triangular; the potential is piecewise parabolic.
  3. \( W \propto \sqrt{V_{bi} - V} \): reverse bias widens the region.

Junction capacitance

The depletion region is a parallel-plate capacitor of width \( W \):

$$C_j = \frac{\varepsilon A}{W} \propto (V_{bi} - V)^{-1/2} \ \ (\text{abrupt}), \qquad \propto (V_{bi} - V)^{-1/3} \ \ (\text{linearly graded})$$

For an abrupt one-sided junction,

$$\boxed{\frac{1}{C_j^2} = \frac{2(V_{bi} + V_R)}{q\varepsilon N_DA^2}}$$

so a plot of \( 1/C^2 \) against reverse bias \( V_R \) is a straight line: the slope gives the doping of the lightly doped side and the intercept on the voltage axis is \( -V_{bi} \). This C–V profiling is a standard laboratory measurement and a natural Part C data question. A varactor diode exploits the voltage-dependent \( C_j \) for electronic tuning.

Under forward bias a second capacitance dominates: the diffusion capacitance \( C_d = \tau I/(\eta V_T) \), from the stored minority charge, where \( \tau \) is the minority-carrier lifetime.

The Diode Equation

Forward bias lowers the barrier by \( V \), so minority-carrier densities at the depletion edges rise by \( e^{V/V_T} \). Diffusion of these excess carriers gives the Shockley equation:

$$\boxed{I = I_s\left(e^{V/\eta V_T} - 1\right)}, \qquad I_s = qAn_i^2\left(\frac{D_p}{L_pN_D} + \frac{D_n}{L_nN_A}\right)$$

with ideality factor \( \eta \) between 1 (diffusion current) and 2 (recombination in the depletion region dominates). \( L = \sqrt{D\tau} \) is the diffusion length.

Results that follow directly:

QuantityResult (300 K, \( \eta = 1 \))
Voltage per decade of current\( \eta V_T\ln10 = 59.5 \) mV
Dynamic resistance\( r_d = \eta V_T/I \): 25.9 Ω at 1 mA
Reverse current\( -I_s \), independent of \( V \) until breakdown
\( I_s \) temperature dependence\( \propto n_i^2 \propto T^3e^{-E_g/kT} \): rises steeply with \( T \)
Forward voltage at fixed currentfalls by roughly 2 mV per °C for silicon

The ratio \( I_s(\text{Ge})/I_s(\text{Si}) \) is enormous because \( n_i^2 \) carries \( e^{-E_g/kT} \): this is why germanium diodes conduct at about 0.2–0.3 V and silicon at about 0.6–0.7 V.

Breakdown and special diodes

Reverse breakdown happens by two mechanisms:

  • Zener (tunnelling) breakdown: heavily doped junctions, very thin depletion region, high field; electrons tunnel from valence to conduction band. Typically below about 5 V. Breakdown voltage falls with temperature (negative coefficient).
  • Avalanche breakdown: lighter doping, wider region; carriers gain enough energy between collisions for impact ionisation. Typically above about 6 V. Breakdown voltage rises with temperature (positive coefficient).

Between them the two effects roughly cancel, which is why reference diodes near 5–6 V have small temperature coefficients.

DeviceConstructionKey propertyUse
Zener diodeheavily doped p-nsharp, stable reverse breakdownvoltage regulation, reference
Varactorreverse-biased junction\( C_j(V) \)tuning, frequency multipliers
Tunnel diodeboth sides degeneratenegative differential resistance in forward biasmicrowave oscillators
Schottky diodemetal–semiconductormajority-carrier device, low forward drop, no minority storagefast switching, clamping
PIN diodeintrinsic layer between p and nlow capacitance, wide depletionRF switches, photodetectors
LED / photodiodedirect-gap junctionlight emission / absorptionsee Opto-electronic Devices

Rectifiers in one table

With peak secondary voltage \( V_m \) and ideal diodes:

CircuitDC outputRipple factorEfficiency (max)PIV per diodeRipple frequency
Half-wave\( V_m/\pi \)1.2140.6%\( V_m \)\( f \)
Full-wave, centre tap\( 2V_m/\pi \)0.48281.2%\( 2V_m \)\( 2f \)
Bridge\( 2V_m/\pi \)0.48281.2%\( V_m \)\( 2f \)

With a capacitor filter, the peak-to-peak ripple is about \( V_m/(fRC) \) (half-wave) or \( V_m/(2fRC) \) (full-wave), for load \( R \).

Bipolar Junction Transistors

A BJT is two junctions sharing a thin, lightly doped base. In the active region the emitter–base junction is forward biased and the collector–base junction reverse biased. The emitter injects carriers into the base; because the base is thin compared with the diffusion length, almost all of them diffuse across and are swept into the collector.

$$I_E = I_B + I_C, \qquad \alpha = \frac{I_C}{I_E}, \qquad \beta = \frac{I_C}{I_B}, \qquad \boxed{\beta = \frac{\alpha}{1 - \alpha}, \quad \alpha = \frac{\beta}{1 + \beta}}$$

Including leakage, \( I_C = \beta I_B + (1 + \beta)I_{CBO} \): the collector–base leakage is multiplied by \( 1 + \beta \), which is why fixed-bias circuits drift with temperature.

RegionE–B junctionC–B junctionBehaviour
Activeforwardreverseamplifier, \( I_C \approx \beta I_B \)
Saturationforwardforwardclosed switch, \( V_{CE} \approx 0.2 \) V
Cut-offreversereverseopen switch
Inverse activereverseforwardpoor gain (roles swapped)

Early effect: increasing \( V_{CE} \) widens the C–B depletion region, narrows the effective base and raises \( I_C \). The output characteristics, extended backwards, meet near \( -V_A \) (the Early voltage); the output resistance is \( r_o = V_A/I_C \).

Small-signal model

In the active region the collector current follows the emitter-junction exponential, \( I_C = I_Se^{V_{BE}/V_T} \), so

$$\boxed{g_m = \frac{I_C}{V_T}}, \qquad r_\pi = \frac{\beta}{g_m}, \qquad r_e = \frac{V_T}{I_E} \approx \frac{1}{g_m}$$

In h-parameter language, \( h_{fe} \approx \beta \) and \( h_{ie} \approx r_\pi \). For a common-emitter stage with collector load \( R_C \):

$$A_v = -g_mR_C \quad (\text{emitter bypassed}), \qquad A_v \approx -\frac{R_C}{R_E} \quad (\text{unbypassed } R_E,\ \beta R_E \gg r_\pi)$$

The unbypassed emitter resistor trades gain for stability: the gain no longer depends on \( \beta \) or temperature. This is negative feedback, the same idea that makes operational amplifier circuits predictable.

ConfigurationVoltage gainCurrent gainInput impedanceOutput impedancePhase
Common emitterhigh\( \beta \) (high)mediummedium–high180°
Common basehigh\( \alpha < 1 \)lowhigh0°
Common collector (emitter follower)\( \approx 1 \)\( 1 + \beta \)highlow0°

Frequency response

At high frequency the junction capacitances short out the signal. In the hybrid-\( \pi \) model the forward-biased base–emitter junction contributes \( C_\pi \) (mostly diffusion capacitance) and the reverse-biased collector–base junction \( C_\mu \) (depletion capacitance). Two results matter:

  • Transition frequency, where the short-circuit current gain falls to 1: \( f_T = \dfrac{g_m}{2\pi(C_\pi + C_\mu)} \). The current gain behaves as \( \beta(f) \approx \beta_0/(1 + jf/f_\beta) \) with \( f_\beta = f_T/\beta_0 \).
  • Miller effect: \( C_\mu \) connects input to an inverted, amplified output, so it appears at the input of a common-emitter stage multiplied: \( C_{in} = C_\pi + C_\mu(1 + g_mR_C) \). This sets the upper cut-off of a CE amplifier. Common-base and emitter-follower stages avoid the multiplication and are faster.

At the low-frequency end, coupling and bypass capacitors set the lower cut-off; the band between is the amplifier’s mid-band, where the formulas above hold.

Biasing

The voltage-divider bias (resistors \( R_1 \), \( R_2 \) from \( V_{CC} \) to ground at the base, \( R_E \) at the emitter) is the standard stable circuit. Replace the divider by its Thevenin equivalent \( V_{Th} = V_{CC}R_2/(R_1 + R_2) \), \( R_{Th} = R_1 \parallel R_2 \); then

$$I_B = \frac{V_{Th} - V_{BE}}{R_{Th} + (\beta + 1)R_E}, \qquad V_{CE} = V_{CC} - I_CR_C - I_ER_E$$

If \( (\beta + 1)R_E \gg R_{Th} \), \( I_C \approx (V_{Th} - V_{BE})/R_E \), independent of \( \beta \). Always check \( V_{CE} > V_{CE,\text{sat}} \) to confirm the active region before using small-signal results.

Field-Effect Transistors

A FET controls the conductance of a channel with an electric field from the gate. Only one carrier type carries current (unipolar), and the gate draws essentially no current, so the input impedance is very high.

JFET

In an n-channel JFET, reverse-biased p-type gate regions deplete the channel. At \( V_{GS} = V_P \) (the pinch-off voltage, negative for n-channel) the channel closes. In saturation (\( V_{DS} \geq V_{GS} - V_P \)) the drain current follows Shockley’s equation:

$$\boxed{I_D = I_{DSS}\left(1 - \frac{V_{GS}}{V_P}\right)^2}, \qquad g_m = \frac{2I_{DSS}}{\lvert V_P\rvert}\left(1 - \frac{V_{GS}}{V_P}\right) = g_{m0}\left(1 - \frac{V_{GS}}{V_P}\right)$$

\( I_{DSS} \) is the drain current at \( V_{GS} = 0 \), the maximum for a JFET (forward-biasing the gate junction is not allowed).

MOSFET

A metal (or polysilicon) gate sits on a thin oxide over the semiconductor. In an n-channel enhancement MOSFET there is no channel at \( V_{GS} = 0 \); a positive gate voltage above the threshold \( V_T \) inverts the surface and creates an electron channel. With \( k = \mu_nC_{ox}W/L \):

$$I_D = k\left[(V_{GS} - V_T)V_{DS} - \tfrac{1}{2}V_{DS}^2\right] \quad (\text{triode},\ V_{DS} < V_{GS} - V_T)$$$$\boxed{I_D = \tfrac{1}{2}k(V_{GS} - V_T)^2} \quad (\text{saturation},\ V_{DS} \geq V_{GS} - V_T)$$$$g_m = k(V_{GS} - V_T) = \sqrt{2kI_D}$$

A depletion MOSFET has a built-in channel and conducts at \( V_{GS} = 0 \); it can be operated with either sign of gate voltage. Channel-length modulation adds a factor \( (1 + \lambda V_{DS}) \), the FET analogue of the Early effect.

CMOS: an n-channel and a p-channel enhancement MOSFET in series form an inverter. In either logic state one transistor is off, so no static current flows; power is drawn only while switching. This is why CMOS dominates digital logic and the microprocessors built from it.

PropertyBJTFET
Carriersboth (bipolar)one (unipolar)
Controlcurrent (base current)voltage (gate field)
Input impedancemoderate (\( r_\pi \))very high
Transconductance at equal currenthigher (\( I_C/V_T \))lower
Transfer lawexponentialsquare law (saturation)
Thermal runawaypossibleresists it (mobility falls with \( T \))

Heterojunctions

A heterojunction joins two different semiconductors. Because band gaps differ, the band edges must jump at the interface. In the simplest model (Anderson’s rule) the vacuum level is continuous, so the offsets come from the electron affinities \( \chi \):

$$\boxed{\Delta E_c = \chi_1 - \chi_2}, \qquad \Delta E_v = \Delta E_g - \Delta E_c$$

(with material 2 the wider-gap one, \( \Delta E_g = E_{g2} - E_{g1} \)). Anderson’s rule ignores interface dipoles and is only approximate; measured offsets are used in practice.

AlignmentPictureConsequence
Type I (straddling)narrow gap lies inside wide gapelectrons and holes both confined in the narrow-gap layer (e.g. GaAs in AlGaAs)
Type II (staggered)both edges shifted the same wayelectrons in one layer, holes in the other: spatially indirect
Type III (broken gap)conduction band of one below valence band of the othersemimetal-like interface (e.g. InAs/GaSb)

Applications in the syllabus:

  • Double-heterostructure laser and LED: a thin narrow-gap active layer between wide-gap cladding confines carriers (band offsets) and light (the narrow-gap layer has the higher refractive index). Threshold currents fall by orders of magnitude, which made room-temperature continuous-wave diode lasers practical. Alferov and Kroemer shared the 2000 Nobel Prize in Physics for this work.
  • HEMT (modulation doping): donors are placed in the wide-gap layer; their electrons fall into the narrow-gap side and collect in a triangular well at the interface, forming a two-dimensional electron gas spatially separated from the ionised donors. Less impurity scattering means much higher mobility, especially at low temperature.
  • HBT (heterojunction bipolar transistor): a wide-gap emitter blocks back-injection of holes from the base. Emitter efficiency no longer depends on the emitter being far more heavily doped than the base, so the base can be heavily doped (low resistance) and thin (fast).
  • Quantum wells: a thin narrow-gap layer between barriers gives particle-in-a-finite-well subbands, shifting emission to higher energy than the bulk gap.

Memory Tricks

Don’t Confuse

ThisWith thisThe difference
Junction capacitance \( C_j \)Diffusion capacitance \( C_d \)\( C_j \) dominates in reverse bias, from depletion charge; \( C_d \) dominates in forward bias, from stored minority charge
Zener breakdownAvalanche breakdowntunnelling in heavily doped junctions, low voltage, negative TC; impact ionisation, higher voltage, positive TC
\( \alpha \)\( \beta \)\( I_C/I_E < 1 \); \( I_C/I_B \gg 1 \)
BJT saturationMOSFET saturationBJT: both junctions forward, switch on, not amplifying; MOSFET: channel pinched off, current set by \( V_{GS} \), the amplifying region
JFET \( V_P \)MOSFET \( V_T \)gate voltage that closes an existing channel; gate voltage that creates a channel
Schottky diodep-n diodemajority carriers, no minority storage, fast; minority injection, storage delay
Type I alignmentType II alignmentboth carriers in the same layer; carriers separated across the interface

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper. Take \( V_T = kT/q = 25.9 \) mV at 300 K unless stated.

Level 1: Foundation

  1. A BJT has \( \alpha = 0.99 \). What is \( \beta \)?
  2. Which junction is reverse biased in the active region of a BJT?
  3. Is the depletion region of a \( p^+n \) junction mostly on the p side or the n side?
Answers
  1. 99. \( \beta = \alpha/(1 - \alpha) = 0.99/0.01 \).
  2. The collector–base junction. The emitter–base junction is forward biased.
  3. The n side, the lightly doped side, because \( N_Ax_p = N_Dx_n \).

Level 2: Part B Style

Practice Q1. An ideal (\( \eta = 1 \)) silicon diode at 300 K carries 1 mA. By how much must the forward voltage increase to raise the current to 10 mA?

(a) 25.9 mV (b) 59.5 mV (c) 119 mV (d) 0.7 V

Answer: (b) 59.5 mV

\( \Delta V = V_T\ln(I_2/I_1) = 25.9 \times \ln10 = 59.6 \) mV, which is 59.5 mV with \( V_T = 25.85 \) mV. (a) forgets the logarithm; (c) uses \( \eta = 2 \); (d) confuses the turn-on voltage with a change.

Practice Q2. An n-channel JFET has \( I_{DSS} = 12 \) mA and \( V_P = -4 \) V. At \( V_{GS} = -2 \) V in saturation, the transconductance is:

(a) 1.5 mS (b) 3 mS (c) 6 mS (d) 12 mS

Answer: (b) 3 mS

\( g_{m0} = 2I_{DSS}/\lvert V_P\rvert = 24/4 = 6 \) mS; \( g_m = g_{m0}(1 - V_{GS}/V_P) = 6 \times (1 - 0.5) = 3 \) mS. (Coincidentally \( I_D = 12 \times 0.25 = 3 \) mA too; check units.) (c) is \( g_{m0} \), the value at \( V_{GS} = 0 \).

Practice Q3. A silicon junction has \( N_A = 10^{17} \) cm\( ^{-3} \), \( N_D = 10^{15} \) cm\( ^{-3} \) and \( n_i = 10^{10} \) cm\( ^{-3} \). At 300 K its built-in potential is closest to:

(a) 0.36 V (b) 0.60 V (c) 0.72 V (d) 1.12 V

Answer: (c) 0.72 V

\( V_{bi} = V_T\ln(N_AN_D/n_i^2) = 0.0259\ln(10^{12}) = 0.0259 \times 27.63 = 0.716 \) V. (a) uses \( \ln(N_AN_D/n_i^2) \) with half the exponent (a \( \sqrt{} \) slip); (d) is the band gap, an upper bound that \( qV_{bi} \) approaches only at degenerate doping.

Practice Q4. An n-channel enhancement MOSFET has \( V_T = 1 \) V and \( k = \mu_nC_{ox}W/L = 2 \) mA/V\( ^2 \). With \( V_{GS} = 3 \) V and \( V_{DS} = 1 \) V, the drain current is:

(a) 1 mA (b) 3 mA (c) 4 mA (d) 8 mA

Answer: (b) 3 mA

\( V_{GS} - V_T = 2 \) V \( > V_{DS} = 1 \) V, so the device is in the triode region: \( I_D = k[(2)(1) - \tfrac{1}{2}(1)^2] = 2 \times 1.5 = 3 \) mA. (c) is the saturation formula \( \tfrac{1}{2}k(2)^2 \), valid only for \( V_{DS} \geq 2 \) V; (d) drops the factor \( \tfrac{1}{2} \).

Level 3: Part C Style

Practice Q5. The capacitance of an abrupt \( p^+n \) silicon diode of area \( 10^{-3} \) cm\( ^2 \) is measured at 300 K (\( \varepsilon = 11.7\varepsilon_0 \)):

Reverse bias \( V_R \) (V)0135
\( C \) (pF)32.221.514.812.0

Consider:

  1. \( 1/C^2 \) is linear in \( V_R \), consistent with an abrupt junction.
  2. The built-in potential is about 0.8 V.
  3. The donor density of the n side is about \( 10^{16} \) cm\( ^{-3} \).
  4. The acceptor density of the p side can be found from the same data.

Which are correct?

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 2, 3 and 4 only (d) 1, 2, 3 and 4

Answer: (b) 1, 2 and 3 only

Compute \( 1/C^2 \) in pF\( ^{-2} \): 0.000964, 0.002163, 0.004565, 0.006944. Successive slopes are 0.001199, 0.001201, 0.001189 per volt: linear (statement 1 true).

Slope \( s \approx 1.195\times10^{-3} \) pF\( ^{-2} \)V\( ^{-1} = 1.195\times10^{21} \) F\( ^{-2} \)V\( ^{-1} \). Intercept: \( V_{bi} = (1/C^2)_{V_R=1}/s - 1 = 0.002163/0.001195 - 1 = 0.81 \) V (statement 2 true).

\( N_D = \dfrac{2}{q\varepsilon A^2 s} = \dfrac{2}{(1.602\times10^{-19})(1.036\times10^{-10})(10^{-7})^2(1.195\times10^{21})} = 1.0\times10^{22} \) m\( ^{-3} = 1.0\times10^{16} \) cm\( ^{-3} \) (statement 3 true; note \( A = 10^{-7} \) m\( ^2 \)).

In a one-sided junction the depletion region lies almost entirely in the n side, so the data carry no information on \( N_A \) beyond “much larger than \( N_D \)” (statement 4 false).

Practice Q6. An npn silicon transistor (\( \beta = 100 \), \( V_{BE} = 0.7 \) V) is biased with \( V_{CC} = 12 \) V, \( R_1 = 40 \) kΩ (to \( V_{CC} \)), \( R_2 = 10 \) kΩ (to ground), \( R_C = 4 \) kΩ and \( R_E = 1 \) kΩ, with \( R_E \) fully bypassed at signal frequencies. Consider:

  1. \( I_C \approx 1.56 \) mA and the transistor is in the active region.
  2. The small-signal voltage gain is about \( -240 \).
  3. If the bypass capacitor is removed, the gain magnitude drops to about 4.
  4. Neglecting base current altogether (taking \( V_B = 2.4 \) V exactly) overestimates \( I_C \) by less than 5%.

Which are correct?

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 1, 3 and 4 only (d) 1, 2, 3 and 4

Answer: (b) 1, 2 and 3 only

Thevenin: \( V_{Th} = 12 \times 10/50 = 2.4 \) V, \( R_{Th} = 8 \) kΩ. \( I_B = (2.4 - 0.7)/(8 + 101 \times 1)\ \text{k}\Omega = 1.7/109 = 15.6\ \mu\text{A} \), \( I_C = 1.56 \) mA. \( V_{CE} = 12 - 1.56(4) - 1.575(1) = 4.19 \) V \( > 0.2 \) V: active. Statement 1 true.

\( g_m = 1.56/25.9 = 60.3 \) mS; \( A_v = -g_mR_C = -241 \). Statement 2 true.

Unbypassed: \( A_v = -\beta R_C/[r_\pi + (\beta + 1)R_E] \) with \( r_\pi = 100/0.0603 = 1.66 \) kΩ: \( -400/(1.66 + 101) = -3.9 \approx -R_C/R_E = -4 \). Statement 3 true.

Neglecting base current: \( I_C \approx 1.7/1 = 1.70 \) mA, which is 9% above 1.56 mA. Statement 4 false. (Here \( R_{Th} = 8 \) kΩ is not negligible against \( (\beta + 1)R_E = 101 \) kΩ at the 5% level.)

Practice Q7. Two semiconductors have (illustrative values) \( \chi_1 = 4.07 \) eV, \( E_{g1} = 1.42 \) eV and \( \chi_2 = 3.74 \) eV, \( E_{g2} = 1.80 \) eV. A thin layer of material 1 is sandwiched between thick layers of material 2. Using Anderson’s rule, consider:

  1. \( \Delta E_c = 0.33 \) eV and \( \Delta E_v = 0.05 \) eV.
  2. The alignment is type I, so both electrons and holes are confined in material 1.
  3. Hole confinement is weaker than electron confinement.
  4. The lowest optical transition of a thin layer lies below 1.42 eV because of quantum confinement.

Which are correct?

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 2 and 4 only (d) 1, 3 and 4 only

Answer: (b) 1, 2 and 3 only
  1. \( \Delta E_c = \chi_1 - \chi_2 = 0.33 \) eV; \( \Delta E_v = \Delta E_g - \Delta E_c = 0.38 - 0.33 = 0.05 \) eV. True.
  2. Material 1’s conduction band is 0.33 eV below material 2’s, and its valence band 0.05 eV above material 2’s: the narrow gap sits inside the wide gap. Type I. True.
  3. The hole barrier (0.05 eV, about \( 2kT \) at 300 K) is far smaller than the electron barrier. True.
  4. Confinement raises the subband energies of both carriers above the band edges, so the transition lies above 1.42 eV. False.

Quick Revision Table

IdeaFormula or fact
Mass action\( np = n_i^2 = N_cN_ve^{-E_g/kT} \)
Einstein relation\( D/\mu = kT/q \)
Built-in potential\( V_{bi} = V_T\ln(N_AN_D/n_i^2) \)
Depletion width\( W = \sqrt{\frac{2\varepsilon(V_{bi} - V)}{q}(\frac{1}{N_A} + \frac{1}{N_D})} \)
C–V profiling\( 1/C^2 = 2(V_{bi} + V_R)/(q\varepsilon N_DA^2) \)
Diode\( I = I_s(e^{V/\eta V_T} - 1) \); 59.5 mV/decade; \( r_d = \eta V_T/I \)
BreakdownZener < 5 V, negative TC; avalanche > 6 V, positive TC
BJT\( \beta = \alpha/(1 - \alpha) \); \( g_m = I_C/V_T \); \( r_\pi = \beta/g_m \)
CE gain\( -g_mR_C \) bypassed; \( \approx -R_C/R_E \) unbypassed
High frequency\( f_T = g_m/2\pi(C_\pi + C_\mu) \); Miller \( C_{in} = C_\pi + C_\mu(1 + g_mR_C) \)
JFET\( I_D = I_{DSS}(1 - V_{GS}/V_P)^2 \)
MOSFET saturation\( I_D = \tfrac{1}{2}k(V_{GS} - V_T)^2 \), \( V_{DS} \geq V_{GS} - V_T \)
Anderson’s rule\( \Delta E_c = \chi_1 - \chi_2 \), \( \Delta E_v = \Delta E_g - \Delta E_c \)
HeterostructuresDH laser (carrier + optical confinement), HEMT (2DEG), HBT (wide-gap emitter)

Key Takeaways

What’s Next