PHY 251, Introduction

The semiconductor industry is one of the most rapidly growing industries.

Many other industries have sprung from this industry, for example, electronics,
communications, computer, internet, biomedical etc. Also, the rapid growth is
also due to the rapid growth in the integrated circuit (IC) capability. It has
given rise to the ability to pack a humongous number of transistors on a single
chip. As of 2016, over 7 billion transistors can be put on a single chip, called a
monolithic integrated circuit. The integrated circuit design (ICD) industry
is a highly important industry; it is responsible for the design and production
of computer chips needed for the modern world. The knowledge of this course
is your first step in gathering the knowledge required to become an IC design
engineer.

1 Periodic Table and Bandgaps

1.1 Periodic Table, Mole, and Atomic Weight
The periodic table was discovered and put together by Mendeleev (1834-1907).
Before him, scientist knew that different elements have different atomic weights.
But Mendeleev saw the rhyme and reason behind them, and organized them in
columns and rows: that was the beginning of the periodic table.

1
 Figure 1: The complete periodic table showing all the elements.

The above shows a complete periodic table, but for this course, only the
column with four valence electrons is the most important, plus the elements
from the adjacent columns.

Figure 2: The sub-periodic table showing elements around carbon (C) column
which has four valence electrons. The important elements are in the shaded
area of the table (courtesy of Jaeger and Blalock).

Figure 2 shows a sub periodic table around the carbon column which has
four valence electrons. But elements from columns with less than four valence
electrons can combine with elements from column with larger than four valence
electrons to make a semiconductor. The silicon-germanium (SiGe) compound
semiconductor has recently found useful applications in many fronts such as
mixed signal circuits (analog plus digital circuits). It was introduced by IBM in
1989.

2
Figure 3: Bandgaps of various semiconductors. Some are called III-V or II-VI
compounds because they are formed with elements from columns III and V, or
columns II and VI (courtesy of Jaeger and Blalock).

Figure 3 shows the bandgaps of various semiconductors. Most are from

column IV.
One other concept you should know is Avagadro’s number (1776-1856) which
is ≈ 6 × 1023 . It is a humongous number and is the number of atoms in one
mole of elements. For instance, one gram of hydrogen has about 6 × 1023 atoms.
This concept can be used to define mole and atomic weight (also called atomic
mass).

3
1.2 Bandgaps, Insulators, Semiconductors, and Conduc-
tors

Figure 4: Bandgaps of various materials. Metals have no bandgap, insulators

have large bandgaps, and semiconductors have small bandgaps.

Most materials have a bandgap. It is the energy gap between the conduction and
valence band of a material. When the gap is large, it is hard for electron to jump
from the valence band to the conduction band. Hence, there is no conduction
electrons around: electric current cannot flow. This is the mark of an insulator.
But when the gap is small, the electron may jump from the valence band to
the conduction band even at room temperature. This material can conduct
electricity but not well. This is the mark of a semiconductor. However, when
the bandgap is small, or nonexistent, then the conduction band is replete with
electrons, which form the conduction electrons. These electrons can conduct
electricity well. This is the mark of a conductor.

• Electronic materials fall into three categories:

– Insulators, resistivity ρ > 105 Ω-cm

−3 5
– Semiconductors, 10 < ρ < 10 Ω-cm
−3
– Conductors, ρ < 10 Ω-cm

• Elemental semiconductors are formed from a single type of atom, typically

silicon.
• Compound semiconductors are typically formed from combinations of col-
umn III and V elements or columns II and VI.
• Germaniums was used in many early devices. The first transistor was
build with germanium.

4
 • Silicon quickly replaced germanium due to its higher bandgap energy,
lower cost, and is easily oxidized to form silicon-dioxide insulating layers.

Figure 5: Figures showing bandgap in a semiconductor at 0 ◦ K and room tem-

perature (courtesy of Jaeger and Blalock).

Figure 5 shows the bandgap of a semiconductor at 0 ◦ K and at room tem-

perature. At 0 ◦ K, the electrons are “frozen” into the valence band, and the
semiconductor cannot conduct since there is no conduction electron in the con-
duction band. At room temperature, some electrons are excited into the con-
duction band making a semiconductor being able to conduct electricity.

Figure 6: The lattice structure of diamond. (Left) A unit cell. (Center) A

carbon atom connected to the next nearest four diamond atoms. (Right) The
diamond lattice structure viewed from another angle, along the crystallographic
axis (courtesy of Jaeger and Blalock).

5
2 Intrinsic Semiconductors

◦
Figure 7: A cartoon picture of the silicon crystal lattice at 0 K (courtesy of
Sedra and Smith).

Figure 8: At room temperatures, some of the covalent bonds are broken, freeing
some electrons, leaving behind holes (courtesy of Sedra and Smith).

6
2.1 Electron-hole Concentrations
When a semiconductor is undoped, there will be holes and electrons charge
carriers generated in the material due to thermal agitation of the environment.
The movement of these charge carriers gives rise to electric current flow.

• A vacancy is left when a covalent bond is broken due to thermal agitation

of the environment.

• The electron is free to roam, and the vacancy is called a hole. This is
called electron-hole pair generation.
• A hole moves when the vacancy is filled by an electron from a nearby
broken bond (hole current).
• The hole density or concentration is represented by p. The electron carrier
concentration is represented by n.
• For intrinsic silicon with no doping, n = p = ni .
• The product of electron and hole concentrations is pn = n2i .

• The pn product above holds when a semiconductor is in thermal equilib-

rium (not with an external voltage applied).

The electron-hole pairs are generated by thermal agitation or breaking of

the covalent bonds. They are removed when the electron-hole pairs recombine.
The electron-hole pairs recombine with a rate proportional to pn while they are
generated with a rate proportional to n2i . At thermal equilibrium, these two
rates are the same; hence np = n2i . It can be shown that this formula remains
valid for doped semiconductors.
According to Boltzmann’s law (1844-1906), the following formula can be
derived: The number of valence bond per cm3 that is broken forms electron-hole
pair. The electron concentration (with similar number for hole concentration)
is

ni = BT 3/2 e−Eg /(2kB T ) (2.1)

where

B = 7.3 × 1015 K−3/2 cm−3 (2.2)

T is the temperature in degree Kelvin K, Eg is the bandgap energy, and kB

is 8.62 × 10−5 K−1 eV is the Boltzmann constant. The hallmark of the above
formula is that the higher the temperature T , the larger the number of bro-
ken bonds, and hence, more the electron-hole pair generated. The larger the
bandgap energy Eg , the harder the bond is to be broken, and hence, the less
electron-hole pair generated. A sketch for the derivation of the above formula
is given in the Appendix.

7
Figure 9: Intrinsic carrier density as a function of temperature. The higher the
temperature, the higher the intrinsic carrier density as more covalent bonds are
brokenm (courtesy of Jaeger and Blalock).

The room temperature Troom temp is usually taken at 300◦ K. The bandgap
of silicon Eg-si is 1.12 eV where eV stands for electron volt which is a unit of
energy.
For intrinsic silicon, ni = 1.5 × 1010 carriers/cm3 , whereas the number of
silicon bonds is about ≈ 5 × 1022 /cm3 . Therefore, ni is very small, or one
electron per 1013 Si bond.

8
3 Doped Semiconductors

Figure 10: The doping of a column IV semiconductor by a donor impurity from

column V with five valence electrons. As a result, an extra electron is free to
move around (courtesy of Sedra and Smith).

Figure 11: The doping of a column IV semiconductor by an acceptor impurity

from column III with three valence electrons. As a result, an extra unfulfilled
bond is generated, resulting in a ‘hole’ in the lattice (courtesy of Sedra and
Smith).

Because np is approximately a constant, the ratio of electron-hole concentration

can be changed by doping. A donor atom contributes to an electron while an

9
acceptor atom contributes to a hole (or a vacancy in a valence bond). The
donor concentration is usually denoted by ND while the acceptor concentration
is denoted by NA .

3.1 Donor Impurities in Silicon

When silicon is doped with donor impurities from column V elements, the fol-
lowing happens:

• Phosphorous (or other column V elements) atom replaces the silicon atom
in the crystallin lattice.
• Since phosphorous has five outer shell electrons, an ‘extra’ electron is
released in the lattice structure.

• The material is still charge neutral, but little energy is needed to move
electron for conduction since it is not bonded to the atoms.

Such a silicon is called an n-type doped silicon. The electron carrier concen-
tration (density) n is increased. Here, nn ≈ ND , or that the electron concen-
tration is approximately equal to the donor concentration (where the subscript
n in nn implies that it is the electron carrier concentration for n-type silicon).
With this in mind, doping can be used to change carrier density, to increase
the conductivity of the semiconductor. Given that pn nn = n2i , and that nn ≈
ND , then
n2
pn ≈ i (3.1)
ND
Since ND  ni by doping or material processing, it implies that pn  nn .
Hence, the holes become minority carriers, while the electrons become majority
carriers.

3.2 Acceptor Impurities in Silicon

For doping with column III elements, called acceptor impurities, similar things
happen.

• Boron (or other column III elements) has been added to the silicon lattice.
• With only three valence electrons, there is now an incomplete bond pair,
creating a vacancy for an electron similar to the hole in a broken covalence
bond.
• Little energy is needed to move a nearby electron into this vacancy, cre-
ating a hole in its original bond.
• As the hole propagates, charge is moved across the silicon.

10
 Needless to say, such a silicon is a p-type doped silicon. The hole car-
rier concentration (density) p is increased. Here, pp ≈ NA , or that the hole
concentration is approximately equal to the acceptor concentration, where the
subscript p in pp implies that it is the hole concentration for p-type silicon.
If we assume an acceptor concentration of NA ≈ 1016 /cm3 , then a silicon
wafer is p-doped (or hole-doped or doped with boron), and the hole concentra-
tion is
3
p ≈ NA ≈ 1016 /cm (3.2)
The electron concentration is
n2i 1020
n= ≈ 16 ≈ 104 /cm3 (3.3)
p 10
Hence, the electron concentration is much much less than the hole concentration.

Appendix A Boltzmann’s Law and Electron-Hole

Pair Generation
Boltzmann’s law is the underpinning law for many observations in semiconduc-
tor physics. It will not be derived but just stated here. If a system has many
possible energy states Ei , i = 1, . . . , N , when the system is in thermodynamic
equilibrium with another system such as a heat bath, then the probability of
the system in a state with energy Ei is proportional to
Pi ∼ Ce−Ei /(kB T ) (A.1)
−5 −1
where kB is Boltzmann constant with value 8.62 × 10 K eV, and T is the
temperature in degree Kelvin. The above formula makes physical sense that the
higher the energy of the state is, the less likely is the system going to be in that
state.
The energy it takes to generate electron-hole pair (EHP) is Eg . After EHP,
the system is in a higher energy state. Hence, the probability of electron-hole
pair generation is proportional to
PEHP ∼ Ce−Eg /(kB T ) (A.2)
Also the probability of electron-hole pair generation is proportional to the EHP
recombination at equilibrium. The EHP recombination rate is proportional to
the product of the densities of the electrons and holes. Hence,
PEHP ∼ ni pi (A.3)
where ni and pi are the intrinsic electron and hole concentration in the semi-
conductor. Since ni = pi in an undoped semiconductor, we deduce that
ni ∼ C 0 e−Eg /(2kB T ) (A.4)
Deriving the constant C 0 is a lot more advanced and is given in some advanced
books. The above is the partial rationale for (2.1). To show that pn = n2i will
need the understanding of Fermi level which is outside the scope of this course.

11