1

VIII.3. Bipolar Transistors

Consider the npn structure shown below.

COLLECTOR

n -
BASE +
+ p - IC -

+
+ n -
IB
-
EMITTER

The base and emitter form a diode, which is forward biased so that a
base current IB flows.

The base current injects holes into the base-emitter junction.

As in a simple diode, this gives rise to a corresponding electron

current through the base-emitter junction.

If the potential applied to the collector is sufficiently positive so that

the electrons passing from the emitter to the base are driven towards
the collector, an external current IC will flow in the collector circuit.

The ratio of collector to base current is equal to the ratio of

electron to hole currents traversing the base-emitter junction.
In an ideal diode
I C I nBE Dn / N A Ln N D Dn L p
= = =
I B I pBE D p / N D L p N A D p Ln

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 2

If the ratio of doping concentrations in the emitter and base regions

ND /NA is sufficiently large, the collector current will be greater than
the base current.

⇒ DC current gain

Furthermore, we expect the collector current to saturate when the

collector voltage becomes large enough to capture all of the minority
carrier electrons injected into the base.
Since the current inside the transistor comprises both electrons and
holes, the device is called a bipolar transistor.

Dimensions and doping levels of a modern high-frequency transistor

(5 – 10 GHz bandwidth)

0 0.5 1.0 1.5

Distance [µm]
(adapted from Sze)

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 3

High-speed bipolar transistors are implemented as vertical structures.

(from Sze)

The base width, typically 0.2 µm or less in modern high-speed

transistors, is determined by the difference in diffusion depths of the
emitter and base regions.

The thin base geometry and high doping levels make the
base-emitter junction sensitive to large reverse voltages.

Typically, base-emitter breakdown voltages for high-frequency

transistors are but a few volts.

As shown in the preceding figure, the collector region is usually

implemented as two regions: one with low doping (denoted “epitaxial
layer in the figure) and the other closest to the collector contact with a
high doping level. This structure improves the collector voltage
breakdown characteristics.

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 4

Quantitative analysis of the bipolar junction transistor (BJT)

1. Holes are injected into the base through the external contact.

The potential distribution drives them towards the emitter.

Since they are majority carriers in the base, few will recombine.

Holes entering the base-emitter depletion region will either

a) pass through the depletion region into the emitter

b) be lost due to recombination

2. As shown in the discussion of the pn-junction, coupled with the

hole current is an electron current originating in the emitter. This
electron current will flow towards the collector, driven by the more
positive potential.

These electrons either

a) enter the collector to become the collector current

b) recombine in the base region. The holes required for the
recombination are furnished by the base current.

3. Thus, the base current is the sum of the

a) hole current entering the emitter

b) hole losses due to recombination in the base-emitter depletion
region
c) electron losses due to recombination in the base during
transport to the collector

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 5

The transport of minority carriers in the base is driven by diffusion, so

d 2n p n p − n p0
Dn − =0
dx 2 τn

At the boundary to the base-emitter depletion region

n p ( 0) = n p 0 e qeVbe / k BT

The equilibrium concentration of electrons in the base is determined

by the base acceptor doping level NAB

ni2
n p0 =
N AB

At the collector boundary all minority carriers will be immediately

swept away by the reverse bias field, so that the boundary condition
becomes
n p (WB ) = 0

Then the solution of the diffusion equation is

 x  WB − x
 sinh sinh
Ln  Ln
n p ( x ) = n p 0 1 −  + ( n p ( 0) − n p 0 )
 sinh WB  W
sinh B
 Ln  Ln

If VBE >> 4kBT/qe the non-equilibrium concentration will dominate

n p (0) >> n p 0
so this simplifies to
WB − x
sinh
Ln
n p ( x ) = n p ( 0)
W
sinh B
Ln

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 6

Since the base width WB in good transistors is much smaller than the
diffusion length Ln, the concentration profile can be approximated by
a linear distribution.

 x 
n p ( x ) = n p (0) 1 − 
 WB 

Now we can evaluate the individual current components.

In this approximation the diffusion current of electrons in the base

region becomes
ni2
J nB = qe DnB e qeVBE / kBT
N ABW B

where DnB is the diffusion constant of electrons in the base.

Similarly, the diffusion current of holes injected into the emitter, under
the assumption that the emitter depth is much smaller than the
diffusion length, is
ni2
J pE = qe D pE e qeVBE / kBT
N DEWE

For the moment, we’ll neglect recombination of holes in the base-

emitter depletion region. Under this assumption, the base current is
I B = J pE AJE

where AJE is the area of the emitter junction.

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 7

The collector current is primarily the electron current injected into the
base, minus any losses due to recombination during diffusion. The
collector transport factor

dn p
electron current reaching collector dx x =WB
αT = =
electron current injected from emitter dn p
dx x =0

Using the above result this becomes

1
αT =
WB
cosh
Ln

Recalling that the base width is to be much smaller than the diffusion
length, this expression can be approximated as

2
1 W 
αT ≈ 1 −  B 
2  Ln 

The resulting collector current is the diffusion current of electrons in

the base times the transport factor

I C = J nB AJEαT

One of the most interesting parameters of a bipolar transistor is the

DC current gain which is the ratio of collector current to base current
IC
β DC =
IB

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 8

Using the above results

ni2
qe DnB e qeVBE / kBT
N ABWB
β DC =
ni2
qe D pE e qeVBE / k BT
N DEWE

N DE DnBWE
β DC = ⋅
N AB D pEWB

Primarily, the DC current gain is determined by the ratio of doping

concentrations in the emitter and the base.

This simple result reflects the distribution of current in the

forward biased diode between electrons and holes.

Transistors with high current gains always have much higher doping
levels in the emitter than in the base.

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 9

The result of this simple analysis implies that for a given device the
DC current gain should be independent of current. In reality this is not
the case.
DC CURRENT GAIN vs. COLLECTOR CURRENT - 2N918

250

200
DC CURRENT GAIN

150

100

50

0
0.1 1.0 10.0 100.0
COLLECTOR CURRENT [mA]

The decrease at low currents is due to recombination in the base-

emitter depletion region.

Since the most efficient recombination centers are near the middle of
the bandgap, we set Et=Ei to obtain the recombination rate

dN R pn − ni2
= σvth N t
dt n + p + 2ni

Due to quasi-equilibrium in the space charge region

pn = ni2 e qeVBE / kBT

This gives us the product of electron concentrations, but we also

need the individual concentrations.

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 10

For a given forward bias the maximum recombination rate will

coincide with the minimum concentration, i.e.

d ( p + n) = 0

and since pn= const.

pn
dp = −dn = dp
p2
or
p=n

at the point of maximum recombination.

Hence, the carrier concentrations are

pn = n 2 = p 2 = ni2 e qeVBE / k BT
or

n = p = ni e qeVBE / 2 kBT

Inserting these concentrations into the expression for the

recombination rate yields
dN R ni2 ( e qeVBE / kBT − 1)
= σvth N t
dt 2ni (e qeVBE / 2 kBT + 1)

This maximum recombination rate will not prevail throughout the

depletion region, but only over a region where the potential changes
no more than ~kBT/qe. If we assume that the average field is
VBE /WBE, a suitable averaging distance is
k B T WBE
w=
qe VBE

Nevertheless, for simplicity assume that recombination is uniform

throughout the depletion region.

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 11

For VBE >> kBT/qe this yields the recombination current to be made
up by holes from the base current
dN R 1
J pr = qe ≈ qeσvth N t niWBE e q eV BE / 2 k B T
dt 2

Now the base current becomes

I B = ( J pE + J R ) AJE ,

so the DC current gain with recombination is

ni2
qe DnB e q eV BE / k B T
N ABWB
β DC =
ni2 1
qe D pE e q eV BE / k B T + qeσvth N t niWe q eV BE / 2k B T
N DEWE 2

Because of the factor ½ in the exponent of the recombination term

ni2
DnB
N ABWB
β DC = ,
ni2 ni
D pE + WBE e − q eV BE / 2 k B T
N DEWE 2τ 0

which can be rewritten in a more informative form as

1 1 N ABWBE N t σvth
= + WB
β DC β0 DnB ni e q eV BE / 2 k B T

where β0 is the DC current gain without recombination. Increasing the

concentration of traps Nt decreases the current gain βDC, whereas
decreasing the base width WB reduces the effect of traps. Since a
smaller base width translates to increased speed (reduced transit
time through the base), fast transistors tend to be less sensitive to
trapping.

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 12

The current dependence enters through the exponential, which

relates the injected carrier concentrations to the base-emitter voltage
VBE. With decreasing values of base-emitter voltage, i.e. decreasing
current, recombination becomes more important. This means that the
DC current gain is essentially independent of current above some
current level, but will decrease at lower currents.

Furthermore note that - although not explicit in the above expression -

the “ideal” DC current gain depends only on device and material
constants, whereas the recombination depends on the local density
of injected electrons and holes with respect to the concentration of
recombination centers. Consider two transistors with different emitter
areas, but operating at the same current, so the larger transistor
operates at a lower current density. The larger area transistor will
achieve a given current at a lower base-emitter voltage VBE than the
smaller device, which increases the effect of trapping. Thus, the
relative degradation of DC current gain due to recombination
depends on the current density.

This means that within a given fabrication process, a large transistor

will exhibit more recombination than a small transistor at the same
current. Stated differently, for a given current, the large transistor will
offer more recombination centers for the same number of carriers.

The figures below show the DC current gain of npn and pnp
transistors over a wide range of current densities.

100 50
NPN PNP
90
80 40
DC CURRENT GAIN

DC CURRENT GAIN

70
PRE-RAD PRE-RAD
60 30
50
40 20
POST RAD
30
20 10 POST-RAD

10
0 0
-5 -4 -3 -2 -1 0 1 2 3 -5 -4 -3 -2 -1 0 1 2 3
10 10 10 10 10
10 10 10 10 10 10 10 10 10
10 10 10 10
EMITTER CURRENT DENSITY [µA/(µm)2 ] EMITTER CURRENT DENSITY [µA/(µm)2 ]

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 13

As shown in the figures, the DC current gain in modern devices is

quite uniform over orders of magnitude of emitter current. The
decrease in DC current gain at low current densities due to increased
recombination is apparent. The figures also show the degradation of
current gain after irradiation, here after exposure to the equivalent of
1014 minimum ionizing protons/cm2. Radiation damage increases the
concentration of trapping sites Nt proportional to fluence Φ, so the
above equation can be rewritten to express the degradation of current
gain with fluence for a given current density (fixed VBE ).

1 1
= + KΦ
β DC β0

where the constant K encompasses the device constants and

operating point. In a radiation-damaged transistor the reduction in
current gain for a given DC current will be less for smaller devices
and faster transistors tend to be less sensitive to radiation damage.

Why does the DC current gain drop off at high currents?

a) with increasing current the high field region shifts towards the
collector, effectively increasing the base width (“Kirk effect”).

b) at high current levels the injected carrier concentration becomes

comparable with the bulk doping.
a) reduction in injection efficiency
b) at very high current densities band-gap narrowing

c) voltage drops in base and emitter resistance

d) Auger recombination
Auger effect:
In an atomic transition, instead of photon emission an electron
is emitted.
Assume a Kα transition (i.e. from the L to the K shell).

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 14

If the ionization of an L electron is less than the Kα transition

energy, instead of a photon an L electron can be emitted with
an energy
E el = E Kα − E L = ( E K − E L ) − E L = E K − 2 E L

In semiconductors the Auger effect manifests itself as

recombination of an electron-hole pair with emission of an
energetic majority carrier.

Summary of effects that degrade DC current gain:

(from Sze)

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 15

Bipolar Transistors in Amplifiers

Although the bipolar transistor is a current driven device, it is often

convenient to consider its response to input voltage.

Since the dependence of base current on base-emitter voltage is

given by the diode equation

I B = I R (e qeVBE / k BT − 1) ≈ I R e qeVBE / k BT

The resulting collector current is

I C = β DC I B = β DC I R e qeVBE / kBT

The transconductance, i.e. the change in collector current vs. base-

emitter voltage
dI C q q
gm ≡ = β DC I R e e qeVBE / kBT = e I C
dVBE k BT k BT

The transconductance depends only on collector current, so for any

bipolar transistor – regardless of its internal design – setting the
collector current determines the transconductance.

Since at room temperature kBT/qe= 26 mV

IC
gm = ≈ 40 I C
0.026

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 16

The obtainable voltage gain of an amplifier depends on the output

characteristics of the transistor.

At low collector voltages the field in the collector-base region is not

sufficient to transport all injected carriers to the collector without
recombination. At higher voltages the output current increases
gradually with voltage (saturation region), due to the change in
effective base width.

An interesting feature is that the extrapolated slopes in the saturation

region intersect at the same voltage VA for Ic= 0 (“Early voltage”).

The finite slope of the output curves is equivalent to a current

generator with a shunt resistance
C

IC Ro

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 17

where
VA
Ro = K
IC

K is a device-specific constant of order 1.

The maximum obtainable voltage gain is

dI C IC V VA
Av ,max = Ro = g m Ro = K A =K
dVBE k B T / qe I C k B T / qe

which at room temperature is about (40VA K).

Note that to first order the maximum obtainable voltage gain is

independent of current.

Transistors with large Early voltages will allow higher voltage gain.

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 18

Bipolar transistors can be implemented as pnp or npn structures.

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 19

In amplifier circuits transistors can be connected in three different

circuit configurations:

(from Sze)

a) Common base configuration

The input signal is applied to the emitter, the output taken from
the collector.

This configuration is used where a low input impedance is

required.

dVEB dVEB 1 k T 1
Ri = ≈ = = B
dI E dI C gm qe I C

Since at room temperature kBT/qe= 26 mV

0.026
Ri =
IC

i.e. Ri= 26 Ω at Ic= 1 mA.

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 20

b) Common emitter configuration

The input signal is applied to the base, the output taken from
the collector.

The input resistance is higher than that of the common base

stage.
dVBE dV β k T β DC
Ri = ≈ β DC BE = DC = B
dI B dI C gm qe I C

Since the base current is about βDC times smaller than the
emitter current, the input resistance of the common emitter
stage is βDC times larger than that of the common base stage.

For βDC = 100 and Ic= 1 mA, Ri= 2600 Ω.

c) Common collector configuration

The signal is applied to the base and the output taken from the
emitter (“emitter folower”).

The voltage gain of this configuration cannot exceed 1, but the

input resistance can be very high.

Since the load resistance RL introduces local negative

feedback, the input resistance depends on the load
Ri ≈ βRL

The output resistance of the emitter follower is

dVout d (Vin − VBE ) dVBE 1
Ro = = ≈ ≈
dI out dI E dI E gm

so at 1 mA collector current Ro= 26 Ω, i.e. although the stage

only has unity voltage gain, it does have current gain.

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors
 21

Although the bipolar transistor is often treated as a voltage-driven

device, the exponential dependence of current on input voltage
means that as an amplifier the device is very non-linear.

Collector Current vs. Base-Emitter Voltage - 2N918

2.0

1.5
Collector Current [mA]

1.0

0.5

0.0
0.65 0.66 0.67 0.68 0.69 0.70 0.71 0.72 0.73
Base-Emitter Voltage [V]

In audio amplifiers, for example, this causes distortion.

Distortion may be limited by restricting the voltage swing, which to
some degree is feasible because of the high transconductance.

With current drive the linearity is much better.

Collector Current vs. Base Current - 2N918

2.0

1.5
Collector Current [mA]

1.0

0.5

0.0
0 2 4 6 8 10
Base Current [µA]

Introduction to Radiation Detectors and Electronics Copyright  1998 by Helmuth Spieler

VIII.3. Bipolar Transistors