Vous êtes sur la page 1sur 49

WEEK 3-2017

Semiconductor Devices for


Electronics

ELEC2133: Analogue Electronics


Semiconductor Basics
Charge Carriers
Electrons and holes

Intrinsic silicon pure silicon only


At T=0K,
S
S
S
S
s
2
Semiconductor Basics
Intrinsic silicon pure silicon only
At T>0K,
hole
S
S
S
S
s
CB (Conduction Band) electrons and VB (Valence Band)
holes can move through semiconductor crystal
VB holes move by VB electrons hopping

3
Semiconductor Doping
Extrinsic (Doped) silicon
contains small amounts of impurities. Typical doping levels
are 1018 cm-3 - 1019cm-3.
N-type ( e.g doping with 15P)

Group V impurities have 5


valence electrons they
donate the extra electron to
the crystal: e.g. phosphorus,
arsenic, antimony

Donor Impurities the fifth electron is weakly bound to the


donor impurity, i.e. only a small energy is needed to move the
electron into the CB.
This is called ntype because the electrons are extra negative
carriers.
n = ND + p ND for ND >> p. (D = donors) np = ni2

4
Semiconductor Doping


B - a material dependent parameter (7.3x1015cm-3K-3/2 for silicon)


Eg - a parameter known as the bandgap energy (1.12 electron volt
(eV) for silicon). It refers to the minimum energy required to
break a covalent bond and thus generate an electron-hole pair.
K is Boltzmanns constant (8.62x10-5 eV/K).

Extrinsic (Doped) silicon


contains small amounts of impurities. Typical doping levels
are 1018 cm-3 - 1019cm-3.
P-type ( e.g doping with 3B)

5
Semiconductor Doping
Group III impurities have 3 valence
electrons they accept a VB electron
from the crystal: e.g. boron, gallium,
indium.

Acceptor Impurities -
a valence band electron is accepted by the impurity leaving
a weakly bound VB hole, i.e. only a small energy is needed
to move this hole into the VB.
This is called ptype because the extra hole is a positive carrier
p = NA + n NA for NA >> n. (A = acceptor) np = ni2

6
Current Flow in Semiconductors
Two distinctly different mechanisms contribute to
current flow in semiconductor
Drift current
Current flow due to the presence of electric field in the
semiconductor
= +

Where J is current density, E is electric field, q is electron


charge, p is concentration of holes, n is concentration of
electrons, p is mobility of holes and n is mobility of
electrons
= /V.s
= /V.s

= = +
is conductivity

=
is resistivity
7
Current Flow in Semiconductors
Diffusion current
Current flow due to concentration gradient of dopants in
semiconductor

= +

Where Dp is diffusion constant ( diffusivity) of holes and Dn


is diffusion constant (diffusivity) of electrons.
= /s = /s

Diffusion constants are related to mobility ( Einstein relationship)



= =

Where VT is known as Thermal Voltage and about 25.6mV at


room temperature

=

8
PN Junctions
PN Junctions

When ntype and ptype semiconductors are brought together, they form
a PN Junction.
To start with, we will consider the case at Thermal Equilibrium, where
there is no applied voltage bias or light or temperature.
Diffusion Current consists of majority charge carriers moving from one
side of the junction to the other due to a concentration gradient.
In PN junctions, electrons will diffuse from the ntype region to the p-
type region while holes move from the ptype to ntype material.
Donor atoms are fixed and cannot diffuse.

9
PN Junctions
PN Junctions

Movement across the junction causes recombination of electrons


and holes, leaving the charged donors and acceptors. The electric
field that results presents a barrier to the diffusion current.
The region where this occurs is called the depletion region and
has few electrons or holes.

10
PN Junctions
In addition to diffusion current, drift current also exists due to
the minority carriers being swept through the depletion region.
Normally, the drift current is very small unless the carrier
concentration is increased due to temperature, optical generation
or injection of carriers.
Under equilibrium conditions, the diffusion current equals the
drift current for both electron and holes. Therefore, the net
current is zero.
The barrier voltage, Vo , across the pn junction can shown to be
given by

=

Typically, for silicon at room temperature, Vo is in the range of
0.6V to 0.9V.

11
PN Junctions
Width of and charge stored in the depletion region depend on
acceptor and donor concentrations in P and N silicon,
respectively.

12
PN Junctions
The magnitude of the charge on either sides of the junction is the
same, and therefore
=

The width, W, of the depletion layer can be shown to be given by



= + = +

13
PN Junctions
Open circuited PN junction ( ID= Is)
=


= +

Reverse biased PN junction


= + ( + )

<

14
PN Junctions
Forward biased PN junction

= + ( )



= +

LP is diffusion length of holes in the n material
Ln is diffusion length of electrons in the p material

15
PN Junctions
The depletion layer narrows and the depletion barrier
voltage reduces. Thus, the diffusion current increases.

Holes are injected across the junction into n region and


electrons are injected across the junction into the p region.

16
PN Junctions
The concentration of injected holes ( minority carriers) is
highest at the edge of the depletion region and can be
related to the forward voltage as ( law of the junction)
( ) = ( )

The distribution of excess hole concentration in the n


region decays exponentially as a function of distance
= +[ ](( ) )

Where LP is the diffusion length of holes in the n-type


silicon

The holes diffusing in the n-region will give rise to a hole


current density

=
)(( ) )

17
PN Junctions
Note that JP decays exponentially. The decay is due to the
recombination with the majority electrons.
In the steady state, the majority carriers will have to be
replenished, and thus electrons will be supplied from the
external circuit to the n- region at a rate will keep the
current constant at the value it has at x = xn. Thus, the
current density due to hole injection is

= ( )

Similarly, the current density due to electron injection is
given by

= ( )

Where Ln is the diffusion length of electrons in the p-type


silicon.
18
PN Junctions
The total current I is

= ( + )( )


= +

19
Reverse Bias Breakdown
Two possible breakdown mechanisms

Zener breakdown
It occurs when the electric field in the depletion layer
increases to the point of breaking covalent bonds and
generating electron-hole pairs
Once it occurs, large current can be obtained without any
increase in reverse biasing voltage

Avalanche breakdown
Reverse bias increases the electric field at the junction.
Increasing electric field in depletion region will increase
force with which carriers collide with lattice atoms.
At a certain critical energy, the collision between the carrier
and the atoms ionises the atom ( removes a carrier, such that
it is free). This is called impact ionisation.
20
Reverse Bias Breakdown

21
Diodes as Voltage Limiter
Diodes can be used to fix a voltage as shown in these
two diagrams of a diode and zener diode.

22
Bipolar Junction Transistors (BJTs)
A bipolar junction transistor is a three terminal device which
consists of two pnjunctions. A simplified cross-section is
simply two backtoback pn junctions.

The actual cross section of the device would look more like:

23
Bipolar Junction Transistors (BJTs)
The key to the operation of a BJT is that the central region is
thin enough such that minority carriers can cross the base,
allowing the emitter and collector to communicate.
Thin enough base layer is to allow transmission of minority
carriers without recombination.
BJTs are extensively used in many applications including
both analogue circuits and digital ICs.

24
Bipolar Junction Transistors (BJTs)
Depending on the biasing of the two junctions, the device can
be operating in one of three basic modes:

Active mode (one junction forward biased and one reverse biased): If
EB is forward biased, called forward active. If other way around
called reverse active. Transistor acts like a current controlled
constant current source. Most commonly used mode of operation.
Saturation (both junctions forward biased): Not many applications
actually use the saturation mode intentionally, but in many circuits
the transistors enter the saturation region while switching between
active and cutoff.
Cutoff (both junctions reverse biased): no current flows in the
transistor.

25
Forward Active Mode
The basecollector junction is reverse biased and has a strong
electric field across it which will sweep any minority carriers
reaching the BC junction into the collector.
In the collector, the carriers are majority carriers and the
collector current will be the number of minority carriers
injected into the base at the EB junction which make it
across the base.
Because there are not many carriers recombining in the base
region, the base current will be small.

26
Forward Active Mode

The emitter current is a combination of


1. Electrons injected into base from emitter ( 3 and 1 )
2. Electrons replacing electrons recombining with holes
injected from base into emitter ( 2 on diagram).

27
Forward Active Mode

The base current is comprised of


1. Holes injected into emitter from forward bias of EB junction
( 2 )
2. Holes replacing holes which have recombined with electrons
in base. ( 3 )
3. Holes generated in collector (usually due to thermal
generation) and swept across BC junction into base ( 4 ).
This term is very small term and is often neglected.
28
Forward Active Mode

The collector current is comprised of


Electrons in the base being swept from base to collector due
to E-field at BC junction. ( 1 )

A BJT is similar to a currentcontrolled current source.


The base current is the controlling current, with the emitter and
collector currents related to the base current by the current gain or
beta value ( or hFE), which is typically quite large (50100), providing
current amplification.
29
Forward Active Mode
For an npn, we have
= +

= = = =

= ( + )

The factor is a measure of how much of the emitter current


makes it across the base region to the collector. The value for
is typically > 0.99 for a highquality BJTs.
The higher the value of , the higher the value of .
The emitter current IE can be approximated, as for a diode,
consisting of the emitter and base regions:

= =

Typically, VBE = 0.60.7 V for a silicon BJT when the emitter
current is flowing. Typically, the saturation current is IS ~
1014 A.
30
Forward Active Mode
For an npn, we have
= +

= = = =

= ( + )

The factor is a measure of how much of the emitter current


makes it across the base region to the collector. The value for
is typically > 0.99 for a highquality BJTs.
The higher the value of , the higher the value of .
The emitter current IE can be approximated, as for a diode,
consisting of the emitter and base regions:

= =

Typically, VBE = 0.60.7 V for a silicon BJT when the emitter
current is flowing. Typically, the saturation current is IS ~
1014 A.
31
Forward Active Mode
The value of depends on collector current,
collectortoemitter voltage and temperature, amongst other
things. As a result, it often varies enormously.

A circuit design using a BJT must ensure that the variation of


in specified range is taken into account and operations are
immune to changes in , which can be achieved using feedback.

32
Graphical Representation: I-V
Characteristics

in reverse mode is smaller than F due to doping gradients


Note that IC = IB only applies in the active region.
Since the emitter and collector junction are not identically the
same, IC=0 when VCE 0;
Enlarging the low voltage
current region:

33
Ebers-Moll Model of BJT
Ideal D.C. properties of a transistor result in terminal
properties which can be described by the EbersMoll
equations:

IE IC

IB
= + =
+ (
)

= + =
+ (
)

= ( + )
34
Base-Width Modulation
From the VCE-IC characteristics in the active region, Ic is
increases with VCE. When extrapolated, the characteristic lines
meet at a point on the negative VCE axis at VCE= -VA.
This voltage is called Early Voltage, and typically ranges from
10V-100V.

=
+ (
)
=
+ (
)
35
Base-Width Modulation
At a given value of VBE, increasing VCE increases the reverse-
bias voltage on the collector base junction and thus increases
the width of the depletion region of this junction.

This in turn results in a decrease in the effective base width W.


Since IS is inversely proportional to W, we see that IS will
increase proportionally causing base-width modulation effect.
= + ( )
=
+ (
)
36
Base-Width Modulation
The linear dependence of IC on VCE can be explicitly accounted
for by assuming that IS remains constant and including the
factor (1+VCE/VA) in the equation from IC as follows:


= +

The output resistance looking into the collector is




=

+
= =

=

VCE and IC the DC bias voltage and current at the CE terminal


of the BJT
37
Base-Width Modulation
The linear dependence of IC on VCE can be explicitly accounted
for by assuming that IS remains constant and including the
factor (1+VCE/VA) in the equation from IC as follows:


= +

The output resistance looking into the collector is




=

+
= =

=

VCE and IC the DC bias voltage and current at the CE terminal


of the BJT
38
Punch-Through
Punchthrough is an extreme of basewidth modulation which
occurs when the entire base region becomes depleted.

When the base region is depleted, there is no barrier to current


flow such that the emitter and collector are effectively shorted.

When punchthrough occurs, current increases suddenly and


sharply.

Occurs primarily in transistors with narrow and/or lightly


doped base

39
MOS Field Effect Transistors
A MOSFET consists of two back to back diodes.

If a voltage is applied across the source and drain, the current


flow between the drain and source is very small (typically on
the order of the reverse current in a diode). This region of
operation is called cut off.
The current from the source to the drain can be increased by
connecting the two with a conductive channel.
How?
40
MOS Field Effect Transistors
A channel between the source and drain can be formed by
creating an inversion layer under the gate.

For ptype substrates, this voltage should be positive such that


negative charges are attracted under the gate. ( N-channel
between source and drain can be created)
Note that an nchannel device refers to a channel with electrons
(i.e. a ptype substrate) and similarly, pchannel refers to a
channel formed from holes (i.e., an n type substrate).

41
MOS Field Effect Transistors

For ptype substrates, this voltage should be positive such that


negative charges are attracted under the gate. ( N-channel
between source and drain can be created)
Note that an nchannel device refers to a channel with electrons
(i.e. a ptype substrate) and similarly, pchannel refers to a
channel formed from holes (i.e., an n type substrate).

42
MOS Field Effect Transistors
The threshold voltage (VT) is the minimum voltage to create an
inversion layer under the gate.

When there is an inversion layer across the entire channel and


there is an applied voltage across the source and drain, the
device is in the ohmic or triode region.
The resistance of the conducting channel depends on the length
of the channel (i.e., the sourcedrain distance) and on the
depth or number of carriers in the channel (which in turn
depends on the applied gate voltage).
In this region, the device acts as a voltage controlled resistance.
43
MOS Field Effect Transistors
In order to maintain an inversion layer, the gate voltage must
always be LARGER than the drain voltage by at least VT.

When VG VD < VT, then there is no longer an inversion layer


near the drain and the MOSFET is in the saturation region.

44
MOS Field Effect Transistors
As VD is further increased, the voltage in the semiconductor
surrounding the drain also increases. Therefore, the voltage
where VG - VD= VT moves further away from the drain, into
the channel.

The voltage at the edge of the inversion layer always equals VT


and the voltage at the source remains constant as well.
Therefore, once the device is in saturation, the voltage across
the channel is constant.
45
MOS Field Effect Transistors
The change in distance at which V=VT is relatively small,
since in the depletion region of a semiconductor the voltage
drop can be large. Therefore, the channel length, to first order
does not change.

Since neither the channel resistance nor voltage change, the


current through the inversion region does not change.
Consequently, as VD is further increased past the onset of
saturation, the current flowing through the device remains
constant.

The electric field at the boundary does not provide a barrier to


current flow. The depleted region is a reverse biased diode,
and incident minority carriers are swept across this region
very quickly.
46
MOS Field Effect Transistors
The I-V curve for n-channel MOSFETs

47
Enhancement and Depletion Mode: MOSFET
Inversion is defined as the condition at which the device has
the same number of carriers as the substrate, but of opposite
polarity.

Threshold voltage is voltage required to achieve this


condition.

The chart on the next page shows how the threshold voltage
determines whether a MOSFET is depletion or enhancement
mode based on the type of charge carrier in the channel.

48
Enhancement and Depletion Mode: MOSFET

49