This study material has been compiled from a lecture audio transcript and a copy-pasted text document.

---

📚 The p-n Junction: A Comprehensive Study Guide

📝 Introduction

The p-n junction is a fundamental building block in solid-state electronics, crucial for understanding individual components and more complex devices like Bipolar Junction Transistors (BJTs) and Metal-Oxide-Semiconductor Field-Effect Transistors (MOSFETs). This guide explores its basic physics, operating principles, and electrical characteristics, starting from semiconductor fundamentals.

1️⃣ Device Structure and Simplifying Assumptions

To simplify the analysis and focus on essential physics, several assumptions are made:

• Monodimensional (1-D) Device ✅: The p-n junction is treated as a 1-D device with two semiconductor regions (p-doped and n-doped) in direct contact. This implies an infinite junction area and parameter changes only along the direction orthogonal to the junction (x-direction).
• Abrupt Junction ✅: The change in doping polarity from p to n is assumed to be instantaneous, occurring at a well-defined coordinate (e.g., x=0).
• Uniform Doping ✅: Doping concentrations (Nd for n-region, Na for p-region) are constant within each region and do not vary with position.
• Neglect of Compensation Effects ✅: While real devices involve "counterdoping" (introducing dopants of opposite polarity) leading to "compensated" regions with both donors and acceptors, for this analysis, the n-region is treated as having only donors (Nd) and the p-region only acceptors (Na). The "net" concentration (e.g., Nd - Na) is what matters for electrostatics, but for simplicity, we assume only one type of dopant per region.

💡 Examples of Integrated p-n Junctions

Real-world p-n junctions can be more complex, often created by counterdoping a silicon wafer.

• (a) Superficial n-doped region and p-doped substrate: An n-doped region at the surface contacts the p-doped substrate.
• (b) Superficial p+-doped region and n-well: A p+-doped region at the surface contacts an n-doped well within a p-doped substrate.
• (c) Superficial p-doped region and n-doped epitaxial layer with n+-buried layer: A p-doped region at the surface is within an epitaxial n-doped region, which has an n+-doped buried layer underneath. This buried layer provides a low-resistance path, improving vertical device operation and reducing non-uniformity issues like current crowding.

2️⃣ Device Electrostatics Under Thermodynamic Equilibrium

Under thermodynamic equilibrium, the Fermi level (EF) must be constant throughout the device. Solving the Poisson equation is key to understanding the electrostatic potential (ϕ) and band diagram.

2.1 📊 Band Diagram for Isolated p and n Regions

When p-doped and n-doped regions are considered in isolation and under thermodynamic equilibrium, their band diagrams are flat.

• Key Features (Figure 3 - Conceptual):
  • Conduction Band (EC) and Valence Band (EV): Have the same vertical position on both sides (assuming same material and electron affinity).
  • Intrinsic Fermi Level (Ei): Flat across each region.
  • Fermi Level (EF):
    • In the p-region: EF is closer to EV (E(p)F).
    • In the n-region: EF is closer to EC (E(n)F).
  • Equations for EF position:
    • Ei - E(p)F ≈ kT ln(Na/ni)
    • E(n)F - Ei ≈ kT ln(Nd/ni)
    • Where ni is intrinsic carrier concentration, kT is thermal energy.

2.2 📈 Band Diagram Under Thermodynamic Equilibrium (Contacted)

When the p and n regions are brought into contact, EF must become constant across the entire system. This causes the energy bands to bend.

• Built-in Potential (qϕbi) 📚: The total band bending, or vertical shift, between the p-side and n-side. It represents the voltage drop across the junction at equilibrium.
  • qϕbi = (E(n)F - E(p)F) = kT ln(Na Nd / ni^2)
  • ϕbi is typically close to the material's energy gap (EG). For silicon, it's around 0.9 V.
• Neutral Regions ✅: Far from the junction, the p-doped and n-doped regions remain electrically neutral, with flat bands.
• Transition Region ✅: Near the junction, bands bend to accommodate the constant EF.

2.3 ⚠️ Depletion Approximation

The band bending significantly reduces free carrier concentrations (electrons and holes) in the transition region.

• Concept 📚: In the region where bands bend, the concentration of free carriers (n and p) becomes much lower than the doping concentration (Na and Nd).
• Approximation ✅: For electrostatic calculations (Poisson equation), the free carrier concentration is neglected compared to the doping concentration in this region.
• Depletion Layer 📚: The transition region, where free carriers are "depleted," is called the depletion layer.
• Importance 💡: This approximation simplifies solving the Poisson equation, allowing for analytical solutions. It holds for electrostatics but not for carrier transport or generation/recombination processes.

2.4 📊 Approximate Analytical Solution of the Poisson Equation

Using the depletion approximation, the Poisson equation can be solved to find the charge density, electric field, and electrostatic potential profiles.

• Charge Density Profile (Figure 5 - Conceptual):
  • In the depletion layer, the charge density (ρ) is piecewise constant:
    • +qNd on the n-side (ionized donors).
    • -qNa on the p-side (ionized acceptors).
  • It's assumed to drop abruptly at the edges of the depletion layer.
• Electric Field Profile (Figure 6 - Conceptual):
  • The electric field (F) has a triangular profile within the depletion layer.
  • It is zero in the neutral regions and reaches its maximum magnitude at the metallurgical junction (x=0).
  • F(x) = - (qNd/ϵSi) * (xn - x) for 0 < x < xn (n-side)
  • F(x) = (qNa/ϵSi) * (x + xp) for -xp < x < 0 (p-side)
  • Condition for continuity at x=0: Ndxn = Naxp (The depletion layer is wider on the less-doped side).
• Electrostatic Potential and Band Diagram (Figure 7 - Conceptual):
  • The electrostatic potential (ϕ) has a parabolic profile within the depletion layer.
  • ϕ(x) = ϕ(xn) - (qNd / 2ϵSi) * (xn - x)^2 for 0 < x < xn
  • ϕ(x) = ϕ(-xp) + (qNa / 2ϵSi) * (x + xp)^2 for -xp < x < 0
  • The band bending is also parabolic.
• Depletion Layer Width (Wd) 📚:
  • xn = sqrt( (2ϵSi / q) * ϕbi * (Na / (Nd * (Na + Nd))) )
  • xp = sqrt( (2ϵSi / q) * ϕbi * (Nd / (Na * (Na + Nd))) )
  • Wd = xn + xp = sqrt( (2ϵSi / q) * ϕbi * (1/Nd + 1/Na) )
  • The depletion layer is wider on the less-doped side. This side mainly determines Wd.

3️⃣ Ideal Analysis of the p-n Junction Under Bias

Applying an external voltage (bias) perturbs the thermodynamic equilibrium, leading to current flow.

3.1 Device Electrostatics Under Bias

An applied voltage (V) changes the band diagram and depletion layer characteristics.

• Forward Bias (V > 0) ✅:
  • The effective potential barrier across the junction is reduced to (ϕbi - V).
  • Energy bands flatten.
  • Significant injection of majority carriers across the junction (electrons from n-side to p-side, holes from p-side to n-side).
  • Leads to a large current.
• Reverse Bias (V < 0) ✅:
  • The effective potential barrier across the junction increases to (ϕbi - V).
  • Energy bands bend more steeply.
  • Carrier flow is severely limited.
  • Results in a very small current (leakage current).
• Rectifying Behavior 📚: The p-n junction allows large current flow in one direction (forward bias) and very small current in the other (reverse bias).
• Depletion Layer Width Under Bias 📚:
  • Wd = sqrt( (2ϵSi / q) * (ϕbi - V) * (1/Nd + 1/Na) )
  • Forward bias (V > 0) decreases Wd.
  • Reverse bias (V < 0) increases Wd.
• Quasi-Neutral Regions ✅: Regions near the contacts where charge neutrality is maintained, and majority carrier quasi-Fermi levels are almost flat.
• Low-Level Injection ✅: Assumes minority carrier concentration increase is much smaller than doping concentration.
• Band Diagram Under Bias (Figure 9 - Conceptual):
  • Shows the effective potential barrier q(ϕbi - V).
  • Quasi-Fermi levels (EFn, EFp) split by qV across the depletion layer.
  • EFn is almost flat in the quasi-neutral n-region and depletion layer.
  • EFp is almost flat in the quasi-neutral p-region and depletion layer.
  • Spatial variation of EFn mainly occurs in the quasi-neutral p-region (bottleneck for electrons).
  • Spatial variation of EFp mainly occurs in the quasi-neutral n-region (bottleneck for holes).

3.2 Ideal Investigation of Current Transport

Current transport in the quasi-neutral regions is primarily due to the diffusion of minority carriers.

• Diffusion Equation 📚: Describes the change in minority carrier concentration due to diffusion and recombination.
  • For electrons in the quasi-neutral p-region (stationary conditions): d^2(Δn)/dx^2 - Δn/Ln^2 = 0
  • Ln = sqrt(Dnτn) is the electron diffusion length.
• Boundary Conditions ✅:
  • At the edge of the depletion layer (x=0), Δn(0) ≈ np0 (e^(qV/kT) - 1).
  • At the contact (x=Wp), Δn(Wp) = 0.
• Electron Current Density (Jn(0)) 📚:
  • Jn(0) = - qDn np0 / (Ln tanh(Wp/Ln)) * (e^(qV/kT) - 1)
• Hole Current Density (Jp(0)) 📚:
  • Jp(0) = qDp pn0 / (Lp tanh(Wn/Lp)) * (e^(qV/kT) - 1)
• Shockley Ideal Diode Equation 📚: The total current density (Jid) is the sum of electron and hole current densities.
  • Jid = J0,id * (e^(qV/kT) - 1)
  • J0,id is the reverse saturation current, which depends on material properties, doping, and diffusion lengths.
  • Dominant Contribution 💡: The current is mainly determined by minority carriers on the less-doped side. The heavily doped side is often called the "emitter," and the less-doped side the "base."

📊 Current-Voltage Characteristics (Jid vs. V)

The Shockley equation describes the fundamental rectifying behavior.

• Logarithmic Scale (Figure 11a - Conceptual):
  • Forward Bias (V > 0):
    • Current increases exponentially: Jid ≈ J0,id * e^(qV/kT).
    • Slope: Approximately 60 mV/decade at room temperature (RT) for m=1 (ideality factor). This means for every 60mV increase in voltage, the current increases by a factor of 10.
    • J0,id is typically in the 10^-12 to 10^-9 A/cm^2 range for silicon.
  • Reverse Bias (V < 0):
    • Current rapidly reaches a constant, small value: Jid ≈ -J0,id. This is the reverse saturation current.
• Linear Scale (Figure 11b - Conceptual):
  • Turn-on Voltage 📚: A well-defined voltage (typically 0.6 V to 0.8 V for silicon) where the current density suddenly and steeply increases. Below this, current is very low.
  • Working Current Density: Typically 10^1 to 10^3 A/cm^2 in circuit applications.

📊 Excess Carrier Concentration (Δn) and J0,id vs. Wp/Ln

The behavior of minority carriers in the quasi-neutral regions depends on their width relative to the diffusion length.

• Wide-Base Diode (Wp >> Ln) ✅:
  • Δn(x) ≈ np0 (e^(qV/kT) - 1) e^(-x/Ln)
  • Excess electron concentration decreases exponentially into the quasi-neutral p-region.
  • Jn(0) ≈ - qDn np0 / Ln * (e^(qV/kT) - 1)
  • J0,id is inversely proportional to Ln and independent of Wp.
  • Generation/recombination processes are significant over the wide region.
• Narrow-Base Diode (Wp << Ln) ✅:
  • Δn(x) ≈ np0 (e^(qV/kT) - 1) * (Wp - x) / Wp
  • Excess electron concentration decreases linearly into the quasi-neutral p-region.
  • Jn(0) ≈ - qDn np0 / Wp * (e^(qV/kT) - 1)
  • J0,id is inversely proportional to Wp and independent of Ln.
  • Generation/recombination processes are negligible due to the short width.
• Figure 12a (Conceptual): Illustrates the exponential decay for wide-base and linear decay for narrow-base diodes, with intermediate cases.
• Figure 12b (Conceptual): Shows how J0,id changes with Wp/Ln, highlighting the transition between narrow-base and wide-base behavior.
• Modern Technologies 💡: Due to smaller device dimensions, narrow-base behavior is more typical in modern silicon technologies.

📊 Complete Band Diagram Under Bias (Figure 13 - Conceptual)

This diagram shows the full energy band profile, including the behavior of quasi-Fermi levels.

• Key Features:
  • Depletion Layer: Bands bend, EFn and EFp split by qV.
  • Quasi-Neutral Regions: Bands are almost flat.
  • Quasi-Fermi Level Variation:
    • Narrow-Base (Figure 13a): Minority carrier concentration decreases linearly, so their quasi-Fermi level variation is logarithmic.
    • Wide-Base (Figure 13b): Minority carrier concentration decreases exponentially, so their quasi-Fermi level variation is linear.
  • The diagram visually connects the electrostatic potential (ϕ) and the quasi-Fermi levels (EFn, EFp) across the device.

📊 Spatial Profiles of n, p, Jn, and Jp (Figure 14 - Conceptual)

These diagrams illustrate how carrier concentrations and current densities vary spatially under bias.

• Forward Bias (Figure 14a):
  • Minority Carriers (n in p-side, p in n-side): Exponential decrease from the depletion layer edge towards the contacts in wide-base diodes.
  • Majority Carriers: Nearly mirror the minority carrier profile to maintain quasi-neutrality (though their change is small relative to their equilibrium value).
  • Current Densities (Jn, Jp):
    • Depletion Layer: Jn and Jp are constant (no generation/recombination assumed in ideal model).
    • Quasi-Neutral Regions: Minority carrier current (e.g., Jn in p-region) decreases exponentially due to recombination. Majority carrier current (e.g., Jp in p-region) increases towards the contact to compensate, mainly by drift.
• Reverse Bias (Figure 14b):
  • Minority Carriers: Concentration decreases from contacts towards the depletion layer.
  • Current Densities:
    • Total current is much smaller.
    • Carriers move in the opposite direction.
    • Current is generated within the device (e.g., electrons generated in p-region flow to n-region).
    • Carriers move by drift through the depletion layer.
• Narrow-Base Case: In contrast to wide-base, carrier concentrations show a linear trend, and Jn and Jp are constant everywhere (negligible generation/recombination).

4️⃣ Corrections to the Ideal Diode Analysis

The ideal diode equation provides a good starting point, but real devices exhibit non-ideal behaviors.

📊 Gummel Plot (Figure 15 - Conceptual)

This semilogarithmic plot shows the current-voltage characteristics, highlighting ideal and non-ideal regimes.

• Ideal Region (m=1): At moderate forward bias, the current follows the 60 mV/decade slope predicted by the Shockley equation.
• Low Current Regime (m=2): At lower forward bias, the slope can be 120 mV/decade. This is due to generation/recombination in the depletion layer.
• High Current Regime (m > 2): At very high forward bias, the slope can exceed 120 mV/decade. This is due to high-injection effects and series resistance.

4.1 Generation and Recombination Processes in the Depletion Layer

The ideal model neglects these processes, but they are significant, especially at low currents.

• Mechanism ✅: In the depletion layer, EFn and EFp are split by qV.
  • Forward Bias: EFn > EFp, leading to net recombination.
  • Reverse Bias: EFn < EFp, leading to net generation.
• Current Contribution (JDL) 📚: This adds a current component with a different voltage dependence.
  • JDL = J0,DL * (e^(qV/2kT) - 1)
  • Slope: 120 mV/decade (m=2) at RT, twice the ideal slope.
  • Impact: Dominates in the low current regime under forward bias, causing deviation from ideal behavior. Under reverse bias, J0,DL often determines the leakage current.

4.2 High-Injection and Series Resistance Effects

These effects become prominent at high forward currents.

• High-Injection 📚:
  • Occurs when minority carrier concentration at the depletion layer edge becomes comparable to the doping concentration.
  • The low-level injection approximation breaks down.
  • The current-voltage characteristic approaches a 120 mV/decade slope (m=2) again.
• Series Resistance 📚:
  • Resistance from contacts or the quasi-neutral regions.
  • Causes a voltage drop (IR drop) that reduces the effective voltage across the junction.
  • At high currents, this limits the current increase with applied voltage, causing the ideality factor m to become even larger than 2.

4.3 📊 Temperature Dependence of Electrical Characteristics (Figure 16 - Conceptual)

Temperature significantly impacts the p-n junction's electrical characteristics.

• Impact on Jid (Ideal Current):
  • Jid ≈ J0,id * e^(qV/kT)
  • kT term: Increases with temperature, reducing the exponential term's sensitivity to V (flatter slope in semilog plot).
  • J0,id term: Increases significantly with temperature (due to n_i^2 dependence).
  • Overall Effect (Figure 16): As temperature increases (e.g., from 300K to 400K):
    • The J-V curve shifts upwards (higher current at a given voltage).
    • The slope in the semilog plot becomes less steep (e.g., 79 mV/decade at 400K vs. 60 mV/decade at 300K).
• Voltage Sensitivity to Temperature (dV/dT) 📚:
  • For a constant current, the voltage required decreases with increasing temperature.
  • dV/dT ≈ (V - EG/q) / T - kγ/q + (1/q) * dEG/dT
  • Typically, dV/dT ≈ -1.9 mV/K for silicon at RT.
• Impact on JDL: J0,DL depends on ni (not ni^2), so it has a weaker temperature dependence than J0,id. At higher temperatures, Jid grows much faster than JDL, potentially masking JDL in the low current regime.

5️⃣ Small-Signal Model of the p-n Junction

The p-n junction is a non-linear device. A small-signal model provides a linearized description around a specific operating point.

📊 Equivalent Circuit (Figure 17 - Conceptual)

The most elementary small-signal model consists of a conductance and two capacitances in parallel.

• Small-Signal Conductance (g) 📚:
  • Represents the change in current with a small change in voltage.
  • g = ∂J / ∂V = J / (kT/q)
  • Proportional to the working current density J.
• Small-Signal Depletion Layer Capacitance (Cdep) 📚:
  • Arises from the change in depletion layer width (Wd) with voltage.
  • Cdep = ∂Qdep / ∂V = ϵSi / Wd (similar to a parallel plate capacitor).
  • Present under both forward and reverse bias.
• Small-Signal Diffusion Capacitance (Cdiff) 📚:
  • Results from the modulation of minority carrier charge stored in the quasi-neutral regions due to voltage changes.
  • Cdiff = ∂Qdiff / ∂V
  • Wide-Base Diode: Qdiff = Jn(0)τn, so Cdiff = g * τn.
  • Narrow-Base Diode: Qdiff = Jn(0)tp, so Cdiff = g * tp (where tp is minority carrier transit time).
  • Relevance: Dominant under forward bias. Becomes negligible under reverse bias.

---