MOSFET: Mathematical Modeling
Keywords: mosfet equations,mosfet modeling,threshold voltage,drain current,NMOS PMOS,short channel effects,subthreshold,device physics equations
MOSFET: Mathematical Modeling
Metal-Oxide-Semiconductor Field-Effect Transistor (MOSFET) Comprehensive equations, mathematical modeling, and process-parameter relationships
1. Fundamental Device Structure
1.1 MOSFET Components
A MOSFET is a four-terminal semiconductor device consisting of:
- Source (S) : Heavily doped region where carriers originate
- Drain (D) : Heavily doped region where carriers are collected
- Gate (G) : Control electrode separated from channel by dielectric
- Body/Substrate (B) : Semiconductor bulk (p-type for NMOS, n-type for PMOS)
1.2 Operating Principle
The gate voltage modulates channel conductivity through field effect:
1.3 Device Types
| Type | Substrate | Channel Carriers | Threshold |
|---|---|---|---|
| NMOS | p-type | Electrons | $V_{th} > 0$ (enhancement) |
| PMOS | n-type | Holes | $V_{th} < 0$ (enhancement) |
2. Core MOSFET Equations
2.1 Threshold Voltage
The threshold voltage $V_{th}$ determines device turn-on and is highly process-dependent:
Component Equations
- Flat-band voltage :
- Fermi potential :
- Oxide capacitance per unit area :
- Work function difference :
Parameter Definitions
| Symbol | Description | Typical Value/Unit |
|---|---|---|
| $V_{FB}$ | Flat-band voltage | $-0.5$ to $-1.0$ V |
| $\phi_F$ | Fermi potential | $0.3$ to $0.4$ V |
| $\phi_{ms}$ | Work function difference | $-0.5$ to $-1.0$ V |
| $C_{ox}$ | Oxide capacitance | $\sim 10^{-2}$ F/m² |
| $Q_{ox}$ | Fixed oxide charge | $\sim 10^{10}$ q/cm² |
| $N_A$ | Acceptor concentration | $10^{15}$ to $10^{18}$ cm⁻³ |
| $n_i$ | Intrinsic carrier concentration | $1.5 \times 10^{10}$ cm⁻³ (Si, 300K) |
| $\varepsilon_{Si}$ | Silicon permittivity | $11.7 \varepsilon_0$ |
| $\varepsilon_{ox}$ | SiO₂ permittivity | $3.9 \varepsilon_0$ |
2.2 Drain Current Equations
2.2.1 Linear (Triode) Region
Condition : $V_{DS} < V_{GS} - V_{th}$ (channel not pinched off)
Simplified form (for small $V_{DS}$):
Channel resistance :
2.2.2 Saturation Region
Condition : $V_{DS} \geq V_{GS} - V_{th}$ (channel pinched off)
Without channel-length modulation ($\lambda = 0$):
Saturation voltage :
2.2.3 Channel-Length Modulation
The parameter $\lambda$ captures output resistance degradation:
Output resistance :
Where $V_A$ is the Early voltage (typically $5$ to $50$ V/μm × L).
2.3 Subthreshold Conduction
2.3.1 Weak Inversion Current
Condition : $V_{GS} < V_{th}$ (exponential behavior)
Characteristic current :
Thermal voltage :
2.3.2 Subthreshold Swing
The subthreshold swing $S$ quantifies turn-off sharpness:
Numerical values :
- Ideal minimum: $S_{min} = 60$ mV/decade (at 300K, $n = 1$)
- Typical range: $S = 70$ to $100$ mV/decade
- $n = 1 + \frac{C_{dep}}{C_{ox}}$ (subthreshold ideality factor)
2.3.3 Depletion Capacitance
2.4 Body Effect
When source-to-body voltage $V_{SB} eq 0$:
Body effect coefficient :
Typical values : $\gamma = 0.3$ to $1.0$ V$^{1/2}$
2.5 Transconductance and Output Conductance
2.5.1 Transconductance
Saturation region :
Alternative form :
2.5.2 Output Conductance
2.5.3 Intrinsic Gain
3. Short-Channel Effects
3.1 Velocity Saturation
At high lateral electric fields ($E > E_{crit} \approx 10^4$ V/cm):
Saturation velocity :
3.1.1 Modified Saturation Current
Note: Linear (not quadratic) dependence on gate overdrive.
3.1.2 Critical Length
Velocity saturation dominates when:
3.2 Drain-Induced Barrier Lowering (DIBL)
The drain field reduces the source-side barrier:
DIBL coefficient :
Typical values : $\eta = 20$ to $100$ mV/V for short channels
3.2.1 Modified Threshold Equation
3.3 Mobility Degradation
3.3.1 Vertical Field Effect
Alternative form (surface roughness scattering):
3.3.2 Universal Mobility Model
Where $E_{eff}$ is the effective vertical field:
3.4 Hot Carrier Effects
3.4.1 Impact Ionization Current
Multiplication factor :
3.4.2 Ionization Rate
3.5 Gate Leakage
3.5.1 Direct Tunneling Current
Where:
3.5.2 Gate Oxide Field
4. Parameters
4.1 Gate Oxide Engineering
4.1.1 Oxide Capacitance
| Dielectric | $\kappa$ | EOT for $t_{phys} = 3$ nm |
|---|---|---|
| SiO₂ | 3.9 | 3.0 nm |
| Si₃N₄ | 7.5 | 1.56 nm |
| Al₂O₃ | 9 | 1.30 nm |
| HfO₂ | 20-25 | 0.47-0.59 nm |
| ZrO₂ | 25 | 0.47 nm |
4.1.2 Equivalent Oxide Thickness (EOT)
4.1.3 Capacitance Equivalent Thickness (CET)
Including quantum effects and poly depletion:
Where:
- $\Delta t_{QM} \approx 0.3$ to $0.5$ nm (quantum mechanical)
- $\Delta t_{poly} \approx 0.3$ to $0.5$ nm (polysilicon depletion)
4.2 Channel Doping
4.2.1 Doping Profile Impact
4.2.2 Depletion Width
4.2.3 Junction Capacitance
Where:
- $C_{j0}$ = zero-bias capacitance
- $\phi_{bi}$ = built-in potential
- $m = 0.5$ (abrupt junction), $m = 0.33$ (graded junction)
4.3 Gate Material Engineering
4.3.1 Work Function Values
| Gate Material | Work Function $\phi_m$ (eV) | Application |
|---|---|---|
| n+ Polysilicon | 4.05 | Legacy NMOS |
| p+ Polysilicon | 5.15 | Legacy PMOS |
| TiN | 4.5-4.7 | NMOS (midgap) |
| TaN | 4.0-4.4 | NMOS |
| TiAl | 4.2-4.3 | NMOS |
| TiAlN | 4.7-4.8 | PMOS |
4.3.2 Flat-Band Voltage Engineering
For symmetric CMOS threshold voltages:
4.4 Channel Length Scaling
4.4.1 Characteristic Length
For good short-channel control: $L > 5\lambda$ to $10\lambda$
4.4.2 Scale Length (FinFET/GAA)
4.5 Strain Engineering
4.5.1 Mobility Enhancement
Where:
- $\Pi$ = piezoresistive coefficient
- $\sigma$ = applied stress
Enhancement factors :
- NMOS (tensile): $+30\%$ to $+70\%$ mobility gain
- PMOS (compressive): $+50\%$ to $+100\%$ mobility gain
4.5.2 Stress Impact on Threshold
Where $\alpha_{th} \approx 1$ to $5$ mV/GPa
5. Advanced Compact Models
5.1 BSIM4 Model
5.1.1 Unified Current Equation
5.1.2 Effective Overdrive
5.1.3 Effective Saturation Voltage
5.2 Surface Potential Model (PSP)
5.2.1 Implicit Surface Potential Equation
5.2.2 Charge-Based Current
Where $Q_i$ is the inversion charge density:
5.3 FinFET Equations
5.3.1 Effective Width
For multiple fins:
5.3.2 Multi-Gate Scale Length
Double-gate :
Gate-all-around (GAA) :
Where $r$ = nanowire radius
5.3.3 FinFET Threshold Voltage
6. Process-Equation Coupling
6.1 Parameter Sensitivity Analysis
| Process Parameter | Primary Equations Affected | Sensitivity |
|---|---|---|
| $t_{ox}$ (oxide thickness) | $C_{ox}$, $V_{th}$, $I_D$, $g_m$ | High |
| $N_A$ (channel doping) | $V_{th}$, $\gamma$, $\mu$, $W_{dep}$ | High |
| $L$ (channel length) | $I_D$, SCE, $\lambda$ | Very High |
| $W$ (channel width) | $I_D$, $g_m$ (linear) | Moderate |
| Gate work function | $V_{FB}$, $V_{th}$ | High |
| Junction depth $x_j$ | SCE, $R_{SD}$ | Moderate |
| Strain level | $\mu$, $I_D$ | Moderate |
6.2 Variability Equations
6.2.1 Random Dopant Fluctuation (RDF)
Where $A_{VT}$ is the Pelgrom coefficient (typically $1$ to $5$ mV·μm).
6.2.2 Line Edge Roughness (LER)
6.2.3 Oxide Thickness Variation
6.3 Equations:
6.3.1 Drive Current
Where $\alpha = 2$ (long channel) or $\alpha \rightarrow 1$ (velocity saturated).
6.3.2 Leakage Current
6.3.3 CV/I Delay Metric
Constants:
| Constant | Symbol | Value |
|---|---|---|
| Elementary charge | $q$ | $1.602 \times 10^{-19}$ C |
| Boltzmann constant | $k$ | $1.381 \times 10^{-23}$ J/K |
| Permittivity of free space | $\varepsilon_0$ | $8.854 \times 10^{-12}$ F/m |
| Planck constant | $\hbar$ | $1.055 \times 10^{-34}$ J·s |
| Electron mass | $m_0$ | $9.109 \times 10^{-31}$ kg |
| Thermal voltage (300K) | $V_T$ | $25.9$ mV |
| Silicon bandgap (300K) | $E_g$ | $1.12$ eV |
| Intrinsic carrier conc. (Si) | $n_i$ | $1.5 \times 10^{10}$ cm⁻³ |
Equations:
Threshold Voltage
Linear Region Current
Saturation Current
Subthreshold Current
Transconductance
Body Effect
Source: ChipFoundryServices — Search this topic — Ask CFSGPT
Explore 500+ Semiconductor & AI Topics
From EUV lithography to CUDA optimization — search the full knowledge base or chat with our AI assistant.