INTRODUCTION TO

Semiconductor Module
Introduction to the Semiconductor Module
© 1998–2019 COMSOL
Protected by patents listed on www.comsol.com/patents, and U.S. Patents 7,519,518; 7,596,474; 7,623,991; 8,457,932;
8,954,302; 9,098,106; 9,146,652; 9,323,503; 9,372,673; 9,454,625; and 10,019,544. Patents pending.
This Documentation and the Programs described herein are furnished under the COMSOL Software License
Agreement (www.comsol.com/comsol-license-agreement) and may be used or copied only under the terms of the
license agreement.
COMSOL, the COMSOL logo, COMSOL Multiphysics, COMSOL Desktop, COMSOL Compiler, COMSOL Server,
and LiveLink are either registered trademarks or trademarks of COMSOL AB. All other trademarks are the property
of their respective owners, and COMSOL AB and its subsidiaries and products are not affiliated with, endorsed by,
sponsored by, or supported by those trademark owners. For a list of such trademark owners, see www.comsol.com/
trademarks.
Version: COMSOL 5.5

Contact Information
Visit the Contact COMSOL page at www.comsol.com/contact to submit general inquiries, contact
Technical Support, or search for an address and phone number. You can also visit the Worldwide
Sales Offices page at www.comsol.com/contact/offices for address and contact information.

If you need to contact Support, an online request form is located at the COMSOL Access page at
www.comsol.com/support/case. Other useful links include:

• Support Center: www.comsol.com/support

• Product Download: www.comsol.com/product-download
• Product Updates: www.comsol.com/support/updates
• COMSOL Blog: www.comsol.com/blogs
• Discussion Forum: www.comsol.com/community
• Events: www.comsol.com/events
• COMSOL Video Gallery: www.comsol.com/video
• Support Knowledge Base: www.comsol.com/support/knowledgebase

Part number. CM024103

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Semiconductor Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
The Semiconductor Module Physics Interface Guide . . . . . . . . 12
Physics Interface Guide by Space Dimension and Study Type . . . 16
Tutorial Example: DC Characteristics of a MOSFET. . . . . . . . . 18

|3
4|
Introduction

Device engineers and physicists use the Semiconductor Module to design and
optimize semiconductor devices. For many years semiconductor device design has
been closely associated with the use of simulation tools, due to the high cost of
prototyping new devices and processes. The advance of nanotechnology and
organic semiconductors has helped create many novel devices. Researchers in
these fields also use simulation to assist the fundamental understanding and design
optimization. For all of the systems mentioned above, multiphysics effects often
play an important role, and COMSOL Multiphysics is the ideal platform for
investigating these effects.
The Semiconductor Module includes a predefined Semiconductor interface,
which is based on the conventional drift-diffusion formulation. An optional
density-gradient implementation is also available to provide a computationally
efficient method to add the effect of quantum confinement to the drift-diffusion
equation system.
In addition, a predefined Schrödinger Equation interface and a predefined
Schrödinger-Poisson Equation multiphysics interface allow more detailed
modeling of quantum-confined systems such as quantum wells, wires, and dots.
The Semiconductor Module enables the stationary and dynamic performance of
devices to be modeled in one, two and three dimensions, together with
circuit-based modeling of active and passive devices. In the frequency domain, it
is possible to model devices driven by a combination of AC and DC signals. A
broad range of semiconductor devices can be modeled, and phenomena such as
heat generation, electrochemical reaction, and optoelectronic effects can be
straightforwardly included using predefined or manual couplings.
The Semiconductor interface solves the drift-diffusion and Poisson’s equations by
either the finite volume or the finite element method. The physics interface solves
a set of coupled partial differential equations for the electric potential and for the
electron and hole concentrations (or their logarithm in the case of the finite
element method log formulation, or their quasi-Fermi levels in the case of the
quasi-Fermi level and density-gradient formulation). The corresponding initial
and boundary conditions are easily specified in the physics interface.
The COMSOL Multiphysics design emphasizes the physics by providing users
with the equations solved by each feature and by offering full access to the
underlying equation system. The Semiconductor Module User’s Guide provides
complete information on the theory underlying the Semiconductor interface.
There is also tremendous flexibility to add user-defined equations and expressions
to the system. For example, user-defined mobility models can be readily specified
simply by typing appropriate expressions into the user defined feature, no scripting

|5
or coding is required. These user-defined mobility models can be combined
arbitrarily with the predefined mobility models built into the software. When
COMSOL Multiphysics compiles the equations, the complex couplings generated
by these user-defined expressions are automatically included in the equation
system. The equations are then solved using a range of state-of-the-art solvers.
Once a solution is obtained, a large range of result analysis tools are available to
interrogate the data, and predefined plots are automatically generated to show the
device response. COMSOL Multiphysics offers the flexibility to evaluate a wide
range of physical quantities including predefined quantities such as the electron
and hole currents (including current components from drift, diffusion and thermal
diffusion), the electric field, the generation/recombination rate, and the
temperature, all available through easy-to-use menus, as well as arbitrary
user-defined expressions.
To model a Semiconductor device, the geometry is first defined in the software.
Then appropriate materials are selected and the Semiconductor interface is added.
The dopant distribution can be computed separately using a diffusion equation
calculation, imported from third party software, or specified empirically using the
built-in doping features. Initial conditions and boundary conditions are set up
within the physics interface. Next, the mesh is defined and a solver is selected.
Finally the results are visualized using a wide range of plotting and evaluation
tools. All of these steps are accessed from the intuitive COMSOL Desktop
graphical user interface.
The Schrödinger Equation interface solves the Schrödinger equation for the wave
function of a single particle in an external potential. This can be applied to general
quantum mechanical problems, as well as for the electron and hole wave functions
in quantum-confined systems under the envelope function approximation.
Appropriate boundary conditions and study types are implemented for the user to
easily set up models to compute relevant quantities in various situations, such as
the eigenenergies of bound states, the decay rates of quasi bound states, the
transmission and reflection coefficients, the resonant tunneling condition, and the
effective band gap of a superlattice structure. The Double Barrier 1D model and
the Superlattice Band Gap Tool app in the Application Libraries illustrate the
usage of the various built-in functionalities.

6|
Semiconductor Devices

The Semiconductor Module can be applied to solve a range of device simulation

problems. The Semiconductor interface can be straightforwardly coupled with
other physics interfaces, such as the Electromagnetic waves interfaces (using the
predefined Semiconductor Optoelectronics multiphysics coupling), the Heat
Transfer in Solids interface and the Electrical Circuits interface. Coupling to a
circuit is straightforward using the terminals included with appropriate boundary
conditions. Figure 1 shows results obtained from a 2D p-n junction model in
which a device model of a diode is coupled to an electrical circuit to produce a
rectifier. The electron and hole concentrations are shown in the plots when
different voltages are applied to the circuit.

Holes Electrons Log of carrier concentration (1/cm3)

Bias

+5 V

0V

-5 V

Figure 1: Electron and hole concentrations in a p-n junction diode connected to a series resistor under
different bias conditions. This plot shows clearly the changing geometry and extent of the depletion region
under reverse bias.

A range of common device types can be simulated with the module, including
MOSFETs, MESFETs, JFETs, diodes, and bipolar transistors. These devices can
be analyzed for the steady state, in the time domain or in the frequency

|7
domain (with mixed DC and AC signals, using the small signal analysis study
type).
A number of standard analyses are illustrated with the MOSFET model series,
which shows how to include a range of increasingly complicated semiconductor
physics effects using features included within the Semiconductor interface. The
first model in the MOSFET series is described in the section Tutorial Example: DC
Characteristics of a MOSFET, below. Figure 2 shows some of the results obtained
from this analysis.

Vg=4 V, Vd=1 V

Logarithm of Electron Concentration (cm-3)

Figure 2: Stationary analysis of a MOSFET. The plot shows the drain current plotted against the drain
voltage (Vd) for a range of different values of the gate voltage (Vg) The inset shows the logarithm of the
electron concentration (in units of cm-3) for a gate voltage of 4 V and for a drain voltage of 1 V. The
channel is clearly visible.

In addition to the MOSFET model series, a wide array of other example models
are available. The Bipolar Transistor model sequence gives an example of how to
set up a multiphysics device simulation. Initially, a model of a 2D cross section of
a bipolar transistor device is created using only the Semiconductor interface. This
model is then extended by two other models in the sequence, one adds coupling
to the Heat Transfer interface and the other demonstrates how to create a full 3D
simulation of the same device.
The MOSCAP model series shows different ways of analyzing this fundamental
building block of many devices. The ISFET model demonstrates the coupling to

8|
the electrochemistry physics. The Heterojunction Tunneling model shows how to
add tunneling current contributions using the WKB approximation. The Interface
Trapping Effects of A MOSCAP model does what its name suggests. The pair of
models Reverse Recovery of a PIN Diode and Forward Recovery of a PIN Diode
demonstrates the modeling of carrier dynamics with a time dependent study.
Several individual models are also provided, including a MESFET, an EEPROM
device, a solar cell, a photodiode, and some LED models. Figure 3 shows the
electron and hole currents flowing in a simple 2D Bipolar transistor.

Figure 3: Log of the current density in A/cm2 (color) and direction of the current flow for electrons (black
arrows) and holes (white arrows) in a simple 2D bipolar transistor.

|9
The density-gradient formulation is demonstrated in a 1D inversion layer, a 2D
FET, and a 3D nanowire model. The last one is shown in Figure 4.

Figure 4: Electron concentration (color), electric potential (isosurface), and total current density (arrow)
of a 3D nanowire model with the quantum confinement effect included using the density-gradient
formulation.

10 |
The Schrödinger Equation interface enables the modeling of various
quantum-confined systems. The following graph shows the wave functions shifted
by the respective energy levels for the resonant tunneling conditions of a double
barrier structure.

The Schrödinger-Poisson Equation interface adds the Electrostatics physics to

take into account the effects of the charge density of the carriers. The following
graph shows the self-consistent result of electrons confined in a quantum wire.

| 11
The Semiconductor Module Physics Interface Guide

Each COMSOL Multiphysics physics interface (for example, the Semiconductor

interface or the Schrödinger Equation interface) expresses the relevant physical
phenomena in the form of sets of partial or ordinary differential equations,
together with appropriate boundary and initial conditions. Each feature added to
the physics interface represents a term or condition in the underlying equation set.
These features are usually associated with a geometric entity within the model,
such as a domain, boundary, edge, or point.
Figure 5 uses the MOSFET application library example to show the Model Builder
tree structure, and the Settings window for the selected Semiconductor Material
Model 1 feature node. This node adds the semiconductor equations to the
simulation within the domains selected. In the Model Inputs section the
temperature of the material is specified. It is straightforward to link this
temperature to a separate Heat Transfer interface to solve nonisothermal problems
— the Semiconductor interface automatically defines an appropriate heat source
term that can be readily accessed in a Heat Transfer interface. In the Material
Properties section the Settings window indicate that the relative permittivity and
the band gap are inherited from the material properties assigned to the domain.
The material properties can be set up as functions of other dependent variables in
the model, for example, the temperature. The dopant density is specified by means
of multiple additive doping features, which can be used to combine Gaussian and
user-defined dopant density profiles to produce the desired profile. Several
boundary conditions are also indicated in the model tree. The Ohmic Contact
boundary condition is commonly used to model nonrectifying interconnects. The
Thin Insulating Gate feature models a gate with a thickness smaller than the
typical length scale of the mesh. It is also possible to model gates explicitly, solving
Poisson’s equation within the dielectric.

12 |
Highlighted
node shown
in settings
window

Additive
doping
features
{
Boundary
conditions
{
Equations
added by
the feature

Material
temperature

Material properties are obtained

from the built-in material library
in this model

Figure 5: The Model Builder (to the left), and the Settings window for Semiconductor Material Model1
for the selected feature node (to the right). The Equation section in the Settings window shows the model
equations.

| 13
The Semiconductor interface is the starting point for most simulations. The
Semiconductor Module also includes physics interfaces to enable modeling of
different physical situations encountered in device design. When a new model is
started, these physics interfaces are selected from the Model Wizard.
Figure 6 shows the physics interfaces included with the Semiconductor Module.
The two Semiconductor Optoelectronics interfaces are only available with an
additional Wave Optics module license.

Figure 6: The Semiconductor Module interfaces as displayed in the Model Wizard for a 3D model. The
Semiconductor Optoelectronics interfaces are only available with an additional Wave Optics Module
license.

Also see Physics Interface Guide by Space Dimension and Study Type. Below, a brief
overview of each of the Semiconductor Module physics interfaces is given.

E LECTROSTATICS
The Electrostatics interface ( ), found under the AC/DC branch in the Model
Wizard, solves for the electric potential given the charge distribution in the
domain and the voltages applied to boundaries. It is used to model electrostatic
devices under static or quasi-static conditions, that is at frequencies sufficiently low
that wave propagation effects can be neglected. Many of the features of the
Electrostatics interface are included in the Semiconductor interface, where they
affect the solution of the electric potential.

E LECTRICAL C IRCUIT
The Electrical Circuit interface ( ), found under the AC/DC branch in the
Model Wizard, has the equations to model electrical circuits with or without
connections to a distributed fields model. The interface solves for the voltages,
currents, and charges associated with the circuit elements. Circuit models can
contain passive elements like resistors, capacitors, and inductors as well as active
elements such as diodes and transistors. Circuits can be imported from an existing
SPICE net list. A typical application of this interface would be to consider the
effect of series or parallel lumped components on device behavior.

14 |
S EMICONDUCTOR
The Semiconductor interface ( ), found under the Semiconductor branch in the
Model Wizard, solves the drift-diffusion and Poisson’s equations. The physics
interface allows both insulating and semiconducting domains to be modeled. The
equations account fully for thermal effects, and the interface can be coupled to a
heat transfer interface using the temperature model input and the predefined heat
source term. This physics interface is appropriate for modeling semiconductor
devices.

S EMICONDUCTOR O PTOELECTRONICS , B EAM E NVELOPES

The Semiconductor Optoelectronics, Beam Envelopes ( ) multiphysics interface
combines the Semiconductor interface with the Electromagnetic Waves, Beam
Envelopes interface. The coupling occurs through the Optical Transitions feature,
which adds a stimulated emission generation term (appropriate for direct
band-gap materials) on domains in the Semiconductor interface. This term is
proportional to the optical intensity in the corresponding Wave Equation, Beam
Envelopes feature in the Electromagnetic Waves, Beam Envelopes interface.
Additionally spontaneous emission (for direct band-gap materials) can be
accounted for. The effect of the light adsorption or emission is accounted for by a
corresponding change in the complex permittivity or refractive index in the Wave
Equation, Beam Envelopes feature.
This multiphysics interface can be used for modeling devices such as photodiodes,
light emitting diodes and laser diodes without quantum wells in direct band gap
materials.

S EMICONDUCTOR O PTOELECTRONICS , F REQUENCY D OMAIN

The Semiconductor Optoelectronics, Frequency Domain ( ) multiphysics
interface combines the Semiconductor interface with the Electromagnetic Waves,
Frequency Domain interface. The coupling occurs through the Optical
Transitions feature, which adds a stimulated emission generation term
(appropriate for direct band-gap materials) on domains in the Semiconductor
interface. This term is proportional to the optical intensity in the corresponding
Wave Equation, Electric feature in the Electromagnetic Waves, Frequency
Domain interface. Additionally spontaneous emission (for direct band-gap
materials) is accounted for. The effect of the light adsorption or emission is
accounted for by a corresponding change in the complex permittivity or refractive
index in the Wave Equation, Electric feature.
This multiphysics interface can be used for modeling devices such as photodiodes,
light emitting diodes and laser diodes without quantum wells in direct band gap
materials.

| 15
S CHRÖDINGER E QUATION
The Schrödinger Equation interface ( ), found under the Semiconductor
branch in the Model Wizard, solves the Schrödinger equation for a single particle
in an external potential. This physics interface is useful for general quantum
mechanical problems as well as for quantum confined systems such as quantum
wells, wires, and dots (with the envelope function approximation).

S CHRÖDINGER -P OISSON E QUATION

The Schrödinger-Poisson Equation multiphysics interface ( ) combines the
Schrödinger Equation interface with the Electrostatics interface to model charge
carriers in quantum-confined systems. The electric potential from the
Electrostatics contributes to the potential energy term in the Schrödinger
Equation. A statistically weighted sum of the probability densities from the
eigenstates of the Schrödinger Equation contributes to the space charge density in
the Electrostatics. A dedicated Schrödinger-Poisson Study Type is available to
automatically generate the solver sequence iterations for the self-consistent
solution of the two-way coupled system.
This multiphysics interface can be used for modeling quantum confined devices
such as quantum wells, wires, and dots.

Physics Interface Guide by Space Dimension and Study Type

The table below lists the physics interfaces available specifically with this module
in addition to the COMSOL Multiphysics basic license.

PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE

DIMENSION

AC/DC

Electrical Circuit cir Not space stationary; frequency

dependent domain; time dependent;
small signal analysis,
frequency domain
Electrostatics1 es all dimensions stationary; time dependent;
stationary source sweep;
eigenfrequency; frequency
domain; small signal analysis,
frequency domain

16 |
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE
DIMENSION

Semiconductor

Semiconductor semi all dimensions small-signal analysis,

frequency domain;
stationary; time dependent;
semiconductor equilibrium
Semiconductor — 3D, 2D, and frequency-stationary;
Optoelectronics, Beam 2D frequency-transient;
Envelopes2 axisymmetric small-signal analysis,
frequency domain
Semiconductor — 3D, 2D, and frequency-stationary;
Optoelectronics, Frequency 2D frequency-transient;
Domain2 axisymmetric small-signal analysis,
frequency domain
Schrödinger Equation schr all dimensions eigenvalue; stationary; time
dependent
Schrödinger-Poisson — all dimensions Schrödinger-Poisson
Equation
1 This physics interface is included with the core COMSOL package but has added functionality
for this module.
2 Requires both the Wave Optics Module and the Semiconductor Module.

| 17
Tutorial Example: DC Characteristics of a MOSFET

This tutorial calculates the DC characteristics of a MOS (metal-oxide

semiconductor) transistor. The MOSFET (metal oxide semiconductor field-effect
transistor) is by far the most common semiconductor device, and the primary
building block in all commercial processors, memories, and digital integrated
circuits. Since the first microprocessors were introduced approximately 40 years
ago this device has experienced tremendous development, and today it is being
manufactured with feature sizes of 22 nm and smaller.
The MOSFET is essentially a miniaturized switch. In this example the source and
drain contacts (the input and output of the switch) are both ohmic (low
resistance) contacts to heavily doped n-type regions of the device. Between these
two contacts is a region of p-type semiconductor. The gate contact lies above the
p-type semiconductor, slightly overlapping the two n-type regions. It is separated
from the semiconductor by a thin layer of Silicon oxide, so that it forms a capacitor
with the underlying semiconductor. Applying a voltage to the gate changes the
local band structure beneath it through the Field Effect. A sufficiently high voltage
can cause the semiconductor to change from p-type to n-type in a thin layer (the
channel) underneath the gate. This is known as inversion and the channel is
sometimes referred to as the inversion layer. The channel connects the two n-type
regions of semiconductor with a thin n-type region under the gate. This region
has a significantly lower resistance than the series resistance of the n-p/p-n
junctions that separated the source and the drain before the gate voltage produced
the inversion layer. Consequently, applying a gate voltage can be used to change
the resistance of the device. The gate voltage where a significant current begins to
flow is called the threshold or turn-on voltage. Figure 7 shows a schematic
MOSFET with the main electrical connections highlighted. Figure 8 shows an
electron microscope image of a modern MOSFET device.

insulator VGS VDS

gate
source drain
n+ n +
p

base
VBS
Figure 7: Schematic diagram of a typical MOSFET. The current flows from the source to the drain through
a channel underneath the gate. The size of the channel is controlled by the gate voltage.

18 |
Figure 8: Cross-section TEM (transmission electron microscopy) image of a 50 nm gate length MOSFET
fabricated at KTH Electrum laboratory by P.E Hellström and coworkers within the ERC advanced grant
OSIRIS research project headed by Prof. M. Östling.

As the voltage between the drain and the source is increased the current carried by
the channel eventually saturates through a process known as pinch-off, in which
the channel narrows at one end due to the effect of the field parallel to the surface.
The channel width is controlled by the gate voltage. Typically a larger gate voltage
results in wider channel and consequently a lower resistance for a given drain
voltage. Additionally, the saturation current is larger for a higher gate voltage.

Source Gate Drain

Base

Figure 9: Model geometry showing the external connections.

Figure 9 shows the model geometry, indicating how the geometry elements
correspond to features in Figure 7. In this model both the source and the base are
connected to ground and the voltages applied to the drain and the gate are varied.
In the first study a small voltage (10 mV) is applied to the Drain and the Gate
voltage is swept from 0 to 5 V. A plot of the current flowing between the source
and the drain is used to determine the turn-on voltage of the device. The second
study sweeps the drain voltage from 0 to 5 V at three different values of the gate

| 19
voltage (2, 3 and 4 V). The drain current versus drain voltage is then plotted at
several values of the gate voltage.

Model Wizard

Note: These instructions are for the user interface on Windows but apply, with
minor differences, also to Linux and Mac.
1 To start the software, double-click the COMSOL icon on the desktop. When
the software opens, you can choose to use either the Model Wizard to create a
new model, or Blank Model to create one manually. For this tutorial, click the
Model Wizard button.
If COMSOL is already open, you can start the Model Wizard by selecting
New from the File menu and then click Model Wizard .
The Model Wizard guides you through the first steps of setting up a model. The
next window lets you select the dimension of the modeling space.
2 In the Select Space Dimension window, click the 2D button .
3 In the Select Physics tree, under the Semiconductor branch, click
Semiconductor (semi) .
4 Click Add button. Observe that the Semiconductor interface is added in the
“Added physics interfaces” pane. Click the Study button .
5 In the Select Study tree, under General Studies branch, click Stationary .
6 Click Done .

Global Definitions

Parameters
1 On the Home ribbon/toolbar click Parameters and select Parameters 1 .
Note: On Linux and Mac, the Home ribbon/toolbar refers to the specific set of
controls near the top of the Desktop.

20 |
2 In the Settings window for Parameters, locate the Parameters section. In the
table, enter the following settings:

NAME EXPRESSION DESCRIPTION

Vd 10[mV] Drain Voltage

Vg 2[V] Gate Voltage

These are examples of global parameters that we can use to parameterize the
model. The parameters can be any global quantity, such as the geometric
dimensions, material properties, or doping concentrations, just to name a few.

G e o me t r y 1

The geometry can be specified using the built-in tools in COMSOL Multiphysics.
First choose to define geometry objects using micrometer units.
1 In the Model Builder tree, under Component 1, click Geometry 1 .
2 In the Settings window for Geometry, locate the Units section. From the
“Length unit” list (drop-down menu), choose μm.

Rectangle 1
Next create a rectangle to define the geometry extents.
1 In the Model Builder, right click Geometry 1 and choose Rectangle .
2 In the Settings window for Rectangle, locate the “Size and Shape” section. In
the Width text field, type 3. In the Height text field, type 0.7.

Polygon 1
Add a polygon that includes points to define the source, drain and gate contacts.
It will also include a line to help create the mesh.
1 Right click Geometry 1 and choose Polygon .
2 In the Settings window for Polygon, locate the Object Type section. From the
Type list (drop-down menu), choose “Closed curve”.

| 21
3 Locate the Coordinates section. From the “Data source” list (drop-down
menu), select Vectors.
- In the x text field, type 0 0 0.5 0.7 2.3 2.5 3 3.
- In the y text field, type 0.67 0.7 0.7 0.7 0.7 0.7 0.7 0.67.

Polygon 1
1 On the Geometry ribbon/toolbar, click Virtual Operations and choose
Mesh Control Edges .
2 In the Settings window for Mesh Control Edges, locate the Input section. For
the “Edges to include:” input panel, select Boundary 4 (the interior line that
was just created by the Polygon step). This tells the software that Boundary 4 is
used for mesh generation only, and is ignored by the physics settings.

3 Click the Build All button to build the entire geometry. Use the Zoom
Extents button at the top of the Graphics window to zoom to the full
geometry if desired.

22 |
Mate rials

Add Material
Next the material properties are added to the model.
1 On the Home ribbon/toolbar click Add Material .
2 Go to the Add Material window. In the tree, under the Semiconductors branch,
click Si .
3 Click Add to Component.
4 On the Home ribbon/toolbar click Add Material to close the Add Material
window.

Semiconductor

Next define the physics settings. Start with the carrier statistics and doping.
1 In the Model Builder window, click Semiconductor .
2 In the Settings window for Semiconductor, locate the Model Properties section.
From the “Carrier statistics” list (drop-down menu), choose Fermi-Dirac.

Analytic Doping Model 1

First a constant background acceptor concentration is defined.

| 23
1 On the Physics ribbon/toolbar, click the Domains menu and choose Analytic
Doping Model .

2 In the Settings window for Analytic Doping Model locate the Domain Selection
section. From the Selection list (drop-down menu), choose All domains.
3 Locate the Impurity section.
- From the “Impurity type” list (drop-down menu), choose “Acceptor doping
(p-type)”.
- In the NA0 text field, type 1e17[1/cm^3].

Analytic Doping Model 2

Add a second doping feature to define the implanted doping profile for the source.
1 On the Physics ribbon/toolbar click the Domains menu and choose Analytic
Doping Model .
2 In the Settings window for Analytic Doping Model, locate the Domain
Selection section. From the Selection list (drop-down menu), choose All
domains.
3 Locate the Distribution section. From the list (drop-down menu), choose Box.
When defining a Box doping distribution, a rectangular region of constant
doping is defined, with a Gaussian drop off away from the edges of this
rectangular region.
Define the location of the lower left corner of the uniformly doped rectangular
region.

24 |
1 Locate the Uniform Region section.
- Specify the r0 vector as

0[um] X

0.6[um] Y

2 Then define the width and height of the uniformly doped region.
- In the W text field, type 0.6[um].
- In the D text field, type 0.1[um].

Choose the dopant type and the doping level in the uniformly doped region.
3 Locate the Impurity section.
- From the “Impurity type” list (drop-down menu), choose “Donor doping
(n-type)”.
- In the ND0 text field, type 1e20[1/cm^3].

Next, specify the length scale over which the Gaussian drop off occurs. If
doping into a background dopant distribution of opposite type (as in this case),
this setting specifies the junction depth. In this model, different length scales
are used in the x and y directions.
4 Locate the Profile section. Select the “Specify different length scales for each
direction” check box.
5 Specify the dj vector as

0.2[um] X

0.25[um] Y

Finally specify the constant background doping level.

6 From the Nb list (drop-down menu), choose “Acceptor concentration (semi/
adm1)”.

Analytic Doping Model 3

Add a similar Gaussian doping profile for the drain.
1 Right-click Analytic Doping Model 2 and choose Duplicate .

| 25
2 In the Settings window for Analytic Doping Model, locate the Uniform Region
section.
- Specify the r0 vector as

2.4[um] X

0.6[um] Y

Next set up boundary conditions for the contacts and gate.

Metal Contact 1
First add a contact for the source.
1 On the Physics ribbon/toolbar click the Boundaries menu and choose Metal
Contact .
The Metal Contact feature is used to define Metal-Semiconductor interfaces of
various types. In this instance use the default Ideal Ohmic contact type to define
the source.
2 Select Boundary 3 only.

Note: The Metal Contact feature in COMSOL Multiphysics is a so called

Terminal boundary condition. By default a fixed potential of 0 V is applied,
which is appropriate in this instance, since the source is grounded. The terminal
can also be set up to specify an input current, input power, or to connect to a
voltage or current source from an external circuit.

Metal Contact 2
Add a second Metal Contact feature to define the drain.
1 On the Physics ribbon/toolbar click the Boundaries menu and choose Metal
Contact .

26 |
2 Select Boundary 7 only.

Set the drain voltage to be specified by the previously defined parameter.

3 In the Settings window for Metal Contact, locate the Terminal section. In the
V0 text field, type Vd.

Metal Contact 3
Add a third Metal Contact to set the body voltage to 0 V.
1 On the Physics ribbon/toolbar click the Boundaries menu and choose Metal
Contact .
2 Select Boundary 2 only.

Thin Insulator Gate 1

Set up the Gate. The Gate dielectric is not explicitly represented in the model,
instead the Thin Insulator Gate boundary condition represents both the gate
contact and the thin layer of oxide.

| 27
1 On the Physics ribbon/toolbar click the Boundaries menu and choose Thin
Insulator Gate .
The Thin Insulator Gate feature is also a terminal, but in this instance it is
possible to fix the voltage or charge on the terminal, as well as to connect it to
a circuit.
Note: The charge setting determines the charge on the conductor and does not
relate to trapped charge at the oxide-semiconductor interface.
The voltage applied to the gate is specified by the parameter added previously.
2 In the Settings window for Thin Insulator Gate, locate the Terminal section. In
the V0 text field, type Vg.
3 Locate the Gate Contact section.
- In the εins text field, type 4.5.
- In the dins text field, type 30[nm].
4 Select Boundary 5 only.

Trap-Assisted Recombination 1
A range of recombination/generation mechanisms are available to be added to the
model. In this case, we simply add trap-assisted recombination, using the default
Shockley-Read-Hall model.
1 On the Physics ribbon/toolbar click the Domains menu and choose
Generation-Recombination section > Trap-Assisted Recombination .
2 In the Settings window for Trap-Assisted Recombination, locate the Domain
Selection section. From the Selection list (drop-down menu), choose All
domains.

28 |
Me sh 1

We will use a user-defined mesh for this model.

1 In the Model Builder window, under Component 1 (comp1) click Mesh 1 .
2 In the Settings window for Mesh, locate the Mesh Settings section. From the
“Sequence type” list (drop-down menu), choose User-controlled mesh.

Size
1 In the Model Builder window, under Component 1 (comp1)>Mesh 1 click
Size .
2 In the Settings window for Size, locate the Element Size section. Click the
Custom button.
3 Locate the Element Size Parameters section. In the “Maximum element growth
rate” text field, type 1.05.

Size 1, Size 2 and Free Triangular 1

1 In the Model Builder window, under Component 1 (comp1)>Mesh 1
right-click Size 1 and choose Delete. In the Confirm Delete pop-up
window, click Yes.
2 Right-click Size 2 and choose Delete. In the Confirm Delete pop-up
window, click Yes.
3 Right-click Free Triangular 1 and choose Delete. In the Confirm Delete
pop-up window, click Yes.

Edge 1
1 Right-click Mesh 1 and choose More Operations>Edge .
2 In the Settings window for Edge, locate the “Boundary Selection” section.
Select Boundaries 3–7 only.

| 29
3 Click to expand the “Control Entities” section. Clear the “Smooth across
removed control entities” check box.

Size 1
1 Right-click Edge 1 and choose Size .
2 In the Settings window for Size, locate the Element Size section.
- From the “Calibrate for” drop-down menu, choose Semiconductor.
- Click the Custom button.
3 Locate the Element Size Parameters section.
- Select the “Maximum element size” check box.
- In the associated text field, type 0.03.

Mapped 1
1 In the Model Builder window, right-click Mesh 1 and choose Mapped .
2 In the Settings window for Mapped, locate the Domain Selection section.
- From the “Geometric entity level” drop-down menu, choose Domain.
- Select Domain 2 only.

3 Click to expand the “Control Entities” section. Clear the “Smooth across
removed control entities” check box.
4 Click to expand the Reduce Element Skewness section. Select the “Adjust edge
mesh” check box.

Distribution 1
1 Right-click Mapped 1 and choose Distribution .

30 |
2 Select Boundary 9 only.

3 In the Settings window for Distribution, locate the Distribution section.

- From the “Distribution type” drop-down menu, choose Predefined.
- In the “Number of elements” text field, type 8.
- In the “Element ratio” text field, type 9.
- From the “Growth formula” drop-down menu, choose “Geometric
sequence”.
- Select the “Reverse direction” check box.

Free Triangular 1
1 In the Model Builder window, right-click Mesh 1 and choose Free
Triangular .
2 In the Settings window for Free Triangular, click to expand the “Control
entities” section. Clear the “Smooth across removed control entities” check
box.
3 Click Build all .
The user-defined mesh is shown in the images below. The mapped mesh with the
specific distribution helps create layers of thin elements underneath the gate,

| 31
where the large gradient of the carrier concentration needs to be resolved by the
mesh.

Study 1

Before setting up the study, check that the doping was set up correctly. To do this,
first get the initial value for the study.

32 |
1 In the Model Builder click Study 1 .
Disable the default plots since these are not required.
2 In the Settings window for Study locate the “Study Settings” section. Click to
clear the “Generate default plots” check box.
3 On the Study ribbon/toolbar click Get Initial Value .

Results

2D Plot Group 1
Add a 2D plot group to check the dopant distribution in the model.
1 On the Home ribbon/toolbar click Add Plot Group and choose 2D Plot
Group .
2 Rename the Plot Group by typing “Signed Dopant Concentration” in the Label
field.
3 On the Signed Dopant Concentration ribbon/toolbar click Surface .
Plot the signed dopant concentration (Nd-Na). This quantity is positive for net
donor doping and negative for net acceptor doping.
4 In the Model Builder under Results>Signed Dopant Concentration, click
Surface 1 .
5 In the Settings window for Surface, locate the Expression section. In the
Expression text field, type semi.Nd-semi.Na.
6 Change the Unit field under Expression from 1/m^3 to 1/cm^3
7 Click the Plot button .

| 33
The doping distribution is shown below.

Study 1

Step 1: Stationary
Now set up a Stationary study to determine the turn-on voltage for the transistor.
In this study we set Vd to 10 mV and sweep over Vg.
1 In the Model Builder under Study 1, click Step 1: Stationary .
2 In the Settings window for Stationary click to expand the “Study Extensions”
section.
The study extensions panel can be used to set up a parametric sweep.
3 Under the “Study Extensions” section, select the “Auxiliary sweep” check box.
4 From the “Sweep type” list (drop-down menu), choose “All combinations”.
This option sweeps over every combination of the specified parameters.

34 |
5 Click Add . In the table, enter the following settings:

PARAMETER NAME PARAMETER VALUE LIST PARAMETER UNIT

Vd 0.01 V

The drain voltage is fixed at a constant value of 10 mV.

6 Click Add again to add Vg to the table.
7 In the table cell for the parameter value list for Vg, enter
range(0,0.2,1.4) 2 3 4. This sweeps the gate voltage between 0 and 4V
with uneven step sizes, to reduce computation time and file size.
By default the “Run continuation for” setting is set to “Last parameter” (Vg in
this case). This setting configures the solver to use the solution for the previous
value of Vg as the initial guess for the solution when solving for the next value
of Vg. It also allows the solver to take intermediate steps at values of Vg not
specified in the parameter value list if necessary. Using this continuation option
is a standard technique to achieve convergence for highly nonlinear equation
systems such as the semiconductor equation system.
8 On the Home ribbon/toolbar click Compute .

Results

1D Plot Group 2
Add a 1D plot group to plot the source current versus the gate voltage.
1 On the Home ribbon/toolbar click Add Plot Group and choose 1D Plot
Group .
2 Rename the Plot Group by typing “Id vsVg (Vd=10mV)” in the Label field.
3 On the “Id vsVg (Vd=10mV)” ribbon/toolbar click Global .
The postprocessing menus contain a wide range of quantities available for
plotting. Choose the current flowing into terminal 2.
4 In the Settings window for Global click Replace Expression (it is in
the upper-right corner of the “y-Axis Data” section). In the menu, expand the
tree to Model>Component1>Semiconductor>Terminals, and then double click
“semi.I0_2 - Terminal Current - A”.

| 35
5 Change the Unit to uA and click the Plot button .

From the plot it is clear that the threshold voltage, VT, of the transistor is
approximately 1.2 V. It is possible to compare this value with the theoretical value
given by (S. M. Sze and K. K. Ng, Physics of Semiconductor Devices, Wiley,
Hoboken, New Jersey, pp. 305–306, 2007):
1
---
2
d ox ( 4ε r ,s ε 0 qN a ψ B )
V T ≅ V FB + 2ψ B + -------------------------------------------------------
ε r ,ox ε 0

where dox is the thickness of the oxide and εr,ox is its relative permittivity, ε0 is the
permittivity of free space, εr,s is the relative permittivity of the semiconductor, q is
the electron charge and Na is the acceptor concentration under the gate. The flat
band voltage VFB and the potential difference between the intrinsic level and the
Fermi-level, ΨB, are given by the following equations:
kB T n eq
V FB = Φ m – χ + ----------- ln  --------
q  Nc 
kB T p eq
ψ B = ----------- ln  --------
q  ni 

where Φm is the work function of the metal contact, χ is the semiconductor

electron affinity, kB is Boltzmann’s contact, T is the absolute temperature, Nc is

36 |
the semiconductor density of states in the conduction band and ni is the intrinsic
carrier density. The equilibrium electron (neq) and hole (peq) densities are given
by:
1 1 2
n eq = --- ( N d – N a ) + --- ( N d – N a ) + 4n i
2 2
1 1 2
p eq = – --- ( N d – N a ) ± --- ( N d – N a ) + 4n i
2 2

Where Nd is the donor concentration under the gate. Note that these equations
assume both complete ionization and Maxwell Boltzmann statistics (reasonable
assumptions near the threshold condition). These equations give a threshold
voltage of 1.18 V, in good agreement with the simulation result. Follow the steps
below to compute the threshold voltage using the approximate formulas above.

Variables 1
1 In Model Builder, right-click Component 1 > Definitions, select Variables .
2 In the Settings window for Variables, under the Geometric Entity Selection
section, for the “Geometric entity level” drop-down menu, select Point. Then
select Point 4.

3 In the Settings window for Variables, under the Variables section, enter the
Name and Expression of the three variables:

NAME EXPRESSION
VFB semi.Phi-semi.chi_semi+semi.Vth*log(semi.ni^2/semi.Na/semi.Nc)
PsiB semi.Vth*log(semi.Na/semi.ni)
VT VFB+2*PsiB+semi.d_ins*sqrt(4*semi.epsilonrs*e_const*
semi.Na*PsiB/epsilon0_const)/semi.epsilon_ins

Study 1
4 In the Model Builder, right-click Study 1, select Update Solution .

| 37
This allows the solution to recognize the three variables that we just defined,
without having to re-compute the study.

Results
5 In the Model Builder, right-click Results, select Evaluation Group.

Evaluation Group 1
6 In the Model Builder, click Results > Evaluation Group 1. In the Settings
window for Evaluation Group, under the Data section, for the two drop-down
menus “Parameter selection (Vd)” and “Parameter selection (Vg)”, select First,
since the evaluation of the threshold voltage is independent of the drain or gate
voltage.
7 In the Model Builder, right-click Results > Evaluation Group 1, select Point
Evaluation.

Point Evaluation 1
8 In the Settings window for Point Evaluation, under the Selection section, select
Point 4 only.

9 Under the Expressions section, enter the three variable names VFB, PsiB, and
VT into the Expression column of the table.
10Click Evaluate.
The results are -0.95 V, 0.43 V, and 1.18 V, respectively.

Study 2

Now add an additional study to plot the source current as a function of drain
voltage at a range of different gate voltages.
1 On the Home ribbon/toolbar click Add Study .

38 |
2 In the Add Study window, find the Studies subsection. In the tree under
General Studies, click Stationary .
3 Click Add Study in the window toolbar.
4 On the Home ribbon/toolbar click Add Study to close the Add Study
window.

Step 1: Stationary
1 In the Model Builder under Study 2, click Step 1: Stationary .
2 In the Settings window for Stationary locate the Study Extensions section.
3 Select the Auxiliary sweep
check box.
4 From the “Sweep type” list
(drop-down menu), choose
“All combinations”. Click
Add . Select Vg from the
Parameter name drop-down
menu.
Now set up the study to
sweep over Vg.
5 Click Range . Go to the
Range dialog box.
- In the Start text field, type 2.
- In the Step text field, type 1.
- In the Stop text field, type 4.
6 Click Replace.
7 Click Add . In this case the default Auxiliary Parameter Vd is the desired
parameter for the sweep.
8 In the table cell for the parameter value list for Vd, enter
range(0,0.25,1.5) 2 3 4 5.
The inner sweep over Vd, the last parameter, should be used for the
continuation solver (this is the default), since the solution will only change
slightly between close values of Vd.
Configure the solver to reuse the solution from the previous step of the
continuation parameter sweep as the initial value for the next step for the outer
sweep parameter Vg, using the first value of the inner parameter Vd
9 From the “Reuse solution for previous step” drop-down menu, choose Auto.
10On the Home ribbon/toolbar click Compute .

| 39
Results

By looking at the electron concentration at Vd values of 0, 1 and 5 V the pinch-off

of the channel can be clearly seen.

Electron Concentration (semi)

1 In the Model Builder under Results, click Electron Concentration (semi) .
By default the plot shows the results for the case Vg=4 V and Vd=5 V.
2 In the Settings window for 2D Plot Group locate the Data section. Change the
Vd Parameter value to 1 V and then to 0 V, each time selecting the Plot
button to see how the results change.

Electric Potential (semi)

1 In the Model Builder under Results, click Electric Potential (semi) .
2 Look at the plot for Vd values of 5 V, 1 V and 0 V.

40 |
The plots in Figure 10 show the results. The pinch off effect is apparent at 5 V.
Electron Concentration Electric Potential

Vd=5 V

Vd=1 V

Vd=0 V

Figure 10: Electron concentration and electric potential distribution at Vd values of 0, 1 and 5 V. The
pinch-off of the channel is apparent at 5 V.

Add another 1D plot group to plot the drain current versus the drain voltage.
3 On the Home ribbon/toolbar select Add Plot Group , and choose 1D Plot
Group .

| 41
4 Rename the plot group by typing “Id vs. Vd” in the Label field.
5 In the Settings window for 1D Plot Group locate the Data section. From the
“Data set” list (drop-down menu), choose Study 2/Solution 2.

1D Plot Group 6
1 From the “Id vs. Vd” ribbon/toolbar click Global .
2 In the Model Builder under Results>Id vs. Vd, click Global 1 .
3 In the Settings window for Global click Replace Expression in the
upper-right corner of the “y-Axis Data” section. In the menu, expand the tree
to Model>Component1>Semiconductor>Terminals, and then double click
“semi.I0_2 - Terminal Current - A”.
4 Change the Unit to uA for the current.
5 Click the Plot button .

The drain current versus drain voltage diagram shows a linear region for low bias,
followed by a nonlinear region. The drain current increases slightly with voltage
in the saturation region at higher drain voltages, as a result of short channel effects.
Short channel effects mean that the standard analytic expressions for the saturation
voltage and current do not apply, but the saturation voltages and currents are of a
similar magnitude to those predicted by the simple theory (S. M. Sze and K. K.
Ng, Physics of Semiconductor Devices, Wiley, Hoboken, New Jersey, pp. 305–
306, 2007).

42 |