Silicon VLSI Technology Fundamentals, Practice and Modeling © 1999

by Plummer, Deal and Griffin

P+ N P+ N+ P N+

N Well P Well

Fig. 1.11: Schematic cross-section of a modern silicon integrated circuit. The particular
technology shown is CMOS with an NMOS device on the right and a PMOS device on
the left. There are two levels of wiring shown. We will discuss this technology in detail
in Chapter 2.

Fig. 1.12: Actual cross-section of a modern IC (IBM’s PowerPC chip). Note the multiple layers of
metal for wiring above the silicon surface. The active parts of the transistors are barely
visible at the bottom of the photograph. Reprinted with permission of Integrated Circuit
Engineering Corporation.

1.3 Semiconductors

This section provides a brief review of the basic properties of semiconductor materials. Many
of the ideas discussed here will be used throughout this book. For readers already familiar with
this material, this section can be skimmed or skipped.

12
Silicon VLSI Technology Fundamentals, Practice and Modeling © 1999
by Plummer, Deal and Griffin

Semiconductors are a class of materials which have the unique property that their electrical
conductivity can be controlled over a very wide range by the introduction of dopants. While this
property can easily be observed in crystalline, polycrystalline or amorphous semiconductor
materials, crystalline materials provide the most reproducible properties and the highest
performance devices and are almost always used in integrated circuits. Dopants are atoms that
generally contain either one more or one fewer electrons in their outermost shell than the host
semiconductor. They provide one extra electron or one missing electron (a “hole”) compared to
the host atoms. These excess electrons and holes are the carriers, which carry current in
semiconductor devices. The key to building semiconductor devices and integrated circuits lies in
the ability to control the local doping and hence the local electronic properties of a semiconductor
crystal.
Consider silicon as a representative semiconductor. As illustrated in Fig. 1.13, silicon has
four electrons in its outermost shell. These are known as the valence electrons since they are the
participants in chemical reactions and chemical bonding. When silicon atoms combine to form a
solid crystal, a particularly stable electronic arrangement can be formed if the silicon atoms form
covalent or shared electron bonds with their four nearest neighbors as illustrated in Fig. 1.13. This
arrangement is favored because each silicon atom can then “fill up” its outermost shell to a total
of eight shared electrons. (In the periodic table, elements with full outer electron shells are the inert
gases like Ar and Ne which are chemically rather inert and very stable.) The covalent bonding or
sharing arrangement in essence populates the entire outer shell for each Si atom, resulting in a
stable structure in which all electrons are bound to atoms, at least at very low temperatures.

Si •• Si •• Si •• Si

• • • •
• • • •
Si •• Si •• Si •• Si + 14

• • • •
• • • •
Si •• Si •• Si •• Si

Fig. 1.13: Simple representation of silicon atoms bonded in a crystal (left). The dotted areas are
covalent or shared electron bonds. The electronic structure of a single Si atom is shown
conceptually on the right. The four outermost electrons are the valence electrons that
participate in the covalent bonds.

Elemental semiconductors, all of which have this same bonding arrangement, lie in Column
IV of the periodic table as shown in Fig. 1.14. This same type of bonding arrangement can be
produced using mixtures of elements from other columns of the periodic table. For example GaAs
consists of alternating Ga (column III) and As (column V) atoms which have an average of four
electrons per atom and so the same covalent bonding arrangement works. More complex examples
like HgxCd1-xTe are also possible. Thus nature provides many possible materials which can act as
semiconductors.

13
Silicon VLSI Technology Fundamentals, Practice and Modeling © 1999
by Plummer, Deal and Griffin

Fig. 1.14: Portion of the periodic table relevant to semiconductor materials and doping.
Elemental semiconductors are in column IV. Compound semiconductors are
combinations of elements from columns III and V (or II and VI).

At temperatures above absolute zero, thermal energy can break some of the Si-Si bonds as
illustrated in Fig. 1.15. This creates both a free or mobile electron and a mobile hole (or missing
electron). The concentrations of electrons and holes are exactly equal in pure semiconductors and
are referred to as the intrinsic carrier concentration ni. Fig. 1.16 plots this concentration as a
function of temperature. At room temperature, ni has a value of about 1.4 x 1010 cm-3 in silicon.
Since there are about 5 x 1022 cm-3 atoms in silicon, fewer than 1 in 1012 cm-3 bonds are broken at
room temperature. As a result, pure silicon is a very poor conductor. The mobile carriers carry
electrical currents in devices. With so few of these present in pure silicon, the currents would be
far too small to be useful in devices.

Si •• Si •• Si •• Si Fr ee
Electron
•
• • M obile
• •
• • Hole • •
Si •• Si • Si •• Si

• • • •
• • • •
Si •• Si •• Si •• Si

Fig. 1.15: Electron (-) and hole (+) pair generation represented by a broken bond in the crystal.
Both carriers are mobile and can carry currents in devices.

14
Silicon VLSI Technology Fundamentals, Practice and Modeling © 1999
by Plummer, Deal and Griffin

1020
1018
Ge Si
1016
1014
GaAs
1012

ni (cm-3)
1010
108
106
104
102
1
0 400 800 1200 1600
Temperature in K
Fig. 1.16: Intrinsic carrier concentration versus temperature in common semiconductors. After
[1.9].

Fortunately, semiconductors have the property that they can be doped. This process is
illustrated in Fig. 1.17. Doping could be accomplished by the gas phase diffusion process we
discussed in connection with Figs. 1.7 and 1.8. However, today it is generally accomplished by a
process called ion implantation that we will discuss in detail in Chapter 8. Doping results in a
column V (As for example) or a column III (B for example) atom replacing a silicon atom in the
crystal structure. Such dopants either contribute an extra electron (column V) to the crystal and
become N type dopants, or they contribute a hole (column III) and become P type dopants. The
electrons or holes are introduced on a one for one basis by the dopants. As a result, to the extent
that we can control the doping concentration accurately, we can precisely control the free electron
and hole concentrations and therefore the conductivity of the silicon. ND and NA are used to refer
to the N type (donor) and P type (acceptor) concentrations respectively. In semiconductor devices,
this doping is done locally as illustrated in Fig. 1.11 using photolithography and masking.

Fr ee
E lectr on •
Si •• Si •• Si Si •• Si •• Si

• • • • M obil e
H ole • •
• • • • •
Si •• As+ •• Si Si •• B- •• Si

• • • • • •
• • • • • •
Si •• Si •• Si Si •• Si •• Si

Fig. 1.17: Doping of group IV semiconductors using elements from column V (arsenic) or III
(boron) of the periodic table.

15
Silicon VLSI Technology Fundamentals, Practice and Modeling © 1999
by Plummer, Deal and Griffin

Modern integrated circuit technology generally uses ion implantation to dope the
semiconductor. This permits the controlled introduction of parts per million to parts per hundred
of dopant atoms. As a result, the conductivity of silicon can be controlled over a very wide range,
permitting many types of semiconductor devices to be fabricated. Fig. 1.18 illustrates the range
of conductivity it is possible to achieve in silicon by doping. The terms N+, N-, P++ etc. are often
used in semiconductor devices to describe relative levels of doping in a particular region. While
there are no formal definitions of the ranges of doping each of these terms represents, the following
approximate definitions are often used:

N- - or P- - : ND or NA < 1014 cm-3

N- or P- : 10 cm < ND or NA
14 -3
< 1016 cm-3
N or P : 1016 cm-3 < ND or NA < 1018 cm-3
N+ or P+ : 1018 cm-3 < ND or NA < 1020 cm-3
N++ or P++ : ND or NA > 1020 cm-3

Since the lattice density in silicon is 5 x 1022 cm-3, even the heaviest doped regions, N++ and P++,
normally use doping concentrations less than 1%.
The resistivity plotted in Fig. 1.18 is inversely proportional to the carrier concentrations and
is defined as

1
ρ= (1.1)
qµnn + qµpp

where n and p are the electron and hole concentrations and µn and µp are the electron and hole
mobilities respectively. µn and µp vary with temperature, doping and electric field in
semiconductors. In silicon at room temperature, with low doping concentration and small fields,
maximum values of µn ≈ 1500 cm2 volt-1 sec-1 and µp ≈ 500 cm2 volt-1 sec-1 are observed. n and p
are determined by the dopant concentrations ND and NA through relationships we will describe
shortly. In general only one of the two terms in Eqn. 1.1 is significant since n >> p in N doped
material and p >> n in P doped material.
The electrical properties of semiconductors are often described through the use of two types
of models. The first of these is the “bond” model that we used beginning with Fig. 1.13. The second
model is the “band” model. Both of these approaches model the same physical phenomena. They
are simply different ways of explaining the same underlying mechanisms. We will make use of
both approaches in this book because some concepts are easier to understand with the bond model,
others with the band model.
Fig. 1.19 illustrates both models for the case of pure silicon. In the band model representation,
the vertical direction is the potential energy of electrons in the crystal. From the bond model, we
can see that in a perfect crystal, electrons are either bound to host silicon atoms or they are free.
A finite amount of energy is required to free the electron (that is to break a Si-Si bond). In the
band diagram, we represent this with the bound electron energy levels below the valence band (EV)
and the free electron energy levels above the conduction band (EC). The energy gap between EV
and EC represents the energy needed to free the electron. In a perfect crystal, there are no allowed
energy levels between EV and EC (an electron is either free or it is bound). Thus the breaking of a

16
Silicon VLSI Technology Fundamentals, Practice and Modeling © 1999
by Plummer, Deal and Griffin

Si-Si bond, which creates one free electron and one free hole, is illustrated in the band model by
elevating an electron up to the conduction band.

Fig. 1.18: Resistivity versus doping for n and p-type silicon. After [1.10].

Free E lectr ons

EC
Si• • Si • • Si E lectron
•
I ntrinsic • • •
EG Ca rrier • H ole • •
e
Generation Si • Si • • Si
H ole • • •
EV • • •
Si • • Si • • Si
Bound E lectrons

Band M odel Bond M odel

Fig. 1.19: Simple band and bond diagram representations of pure silicon. Bonded electrons lie at
energy levels below EV; free electrons are above EC. The process of intrinsic carrier
generation is illustrated in each model.

In Fig. 1.20, the introduction of dopants is represented in the bond model and in the band
diagram by the ED and EA energy levels. Consider the ED level first, which represents a donor atom

17