Such Dependence Is One Most Important in Semiconductor
Such Dependence Is One Most Important in Semiconductor
2
Solved Examples
Example (31) Example (32)
Band gap of Si depends on the temperature Calculate the intrinsic carrier density in
as Eg = 1.17 eV − 4.73 × 10−4 T2 / T + 636 . germanium, silicon and gallium arsenide at
Find a concentration of electrons in the 300, 400, 500 and 600 K.
conduction band of intrinsic (undoped) Si at Solution
T = 77 K if at 300 K ni = 1.05 × 1010 cm−3.
The intrinsic carrier density in silicon at 300 K equals:
Solution: − 𝑬𝒈
𝑬𝒈 𝒏𝒊 𝟑𝟎𝟎𝑲 = 𝑵𝒄 𝑵𝒗 𝒆𝒙𝒑
𝒏𝟐𝒊= 𝑵𝒄 𝑵𝒗 𝒆𝒙𝒑 − 𝟐 𝑲𝑻
𝑲𝑻 𝟏𝟗 𝟏𝟗
− 𝟏. 𝟏𝟐
𝑬𝒈 = 𝟐. 𝟖𝟏 𝒙 𝟏𝟎 𝒙 𝟏. 𝟖𝟑 𝒙 𝟏𝟎 𝒆𝒙𝒑
𝟑 𝟐 𝒙 𝟎. 𝟎𝟐𝟓𝟖
≈ 𝑻 𝐞𝐱𝐩 − = 𝟖. 𝟕𝟐 𝒙 𝟏𝟎 𝒎𝟗 −𝟑
𝑲𝑻
Therefore, Similarly, one finds the intrinsic carrier
𝒏𝟏 𝑻𝟐 density for germanium and gallium arsenide
𝟑ൗ
𝟐 𝑬𝒈 𝑻𝟐 𝑬𝒈 𝑻𝟏
at different temperatures, yielding:
𝑻𝟐
= 𝒏𝟏 𝑻𝟏 𝐞𝐱𝐩 − + Germanium Silicon Gallium
𝑻𝟏 𝟐 𝑲 𝑻𝟐 𝟐 𝑲 𝑻𝟏
Arsenide
Putting the proper values in the formula we 300 K 2.02 x 1013 8.72 x 109 2.03 x 106
obtain that ni(77K) ≈ 10−20cm−3. 400 K 1.38 x 1015 4.52 x 1012 5.98 x 109
500 K 1.91 x 1016 2.16 x 1014 7.98 x 1011
600 K 1.18 x 1017 3.07 x 1015 2.22 x 1013 3
1.10. The Concept of Mobility
When an electron Figure1.24. Random Electron Path at no
Field
introduced without
subjected external
electric field at absolute
temperature in a perfect
semiconductor sample.
The electron will move
freely random through
the crystal as shown in
figure 1.24.
Figure1.25. Directed Electron Path under Field 4
This is a general property associated with perfect periodic
structures. At room temperature, valance electrons are
liberated by random motion through the crystal. The
random velocity of an electron at room temperature is
about 107 cm/sec. The mean free path of the electrons
between two collisions is in the order of 10-5 about 500
times the distance between neighboring atoms. The
average time between collisions is about 10-12 second.
Under thermal equilibrium condition, the random motion
of electrons leads to zero current in any direction. Thermal
vibration introduced treated as a particle (phonons). Its
collisions with electrons and holes called scattering. The
scattering phenomena increased as temperature increases
for a maximum scattering velocity.
5
If we apply electric field to the crystal, the electrons gain a force
q toward the positive pole as shown in fig.1.25, and moves with
acceleration given by:
𝒒𝜺
𝒂= − 𝑬𝒒. 𝟏. 𝟓𝟎
𝒎
After time () , the electrons suffer of collisions decreases
velocity (𝝑) between each two consecutive collisions as ∆ 𝝑 = a ,
produces what is called drift velocity (𝝑𝒅 ) .
𝝑𝒅 = ∆ 𝝑 = 𝝁 𝜺 𝑬𝒒. 𝟏. 𝟓𝟏
The term (q/m) gives the mobility of the charge carriers, then
(Eq.1.51) becomes:
𝝑𝒅 = 𝝁 𝜺 𝑬𝒒. 𝟏. 𝟓𝟐
Normally, n > p for any material. For example, in silicon n =
1800, p = 400 cm2/V.sec, and in germanium we find that n =
3800, p = 1800 cm2/V.sec.
6
1.10.1. FIELD DEPENDENCE ON MOBILITY
By increasing the subjected field, the drift velocity increases, due to the
increasing of kinetic energy of the electrons. When electric field is higher
increased, a critical value developed and reaches maximum scattering
velocity; this velocity is not more increased by increasing electric field, figure
1.26.
Maximum Scattering
(M/S) Velocity For Si
5x104
Missed
12
Solved Examples
Example (38) Example (39)
Calculate the electron and hole densities in an n-type silicon
wafer (Nd = 1017 cm-3) illuminated uniformly with 10 mW/cm2 of
What are the approximate thermal
red light (Eph = 1.8 eV). The absorption coefficient of red light in velocities of electrons and holes in silicon
silicon is 10-3 cm-1. The minority carrier lifetime is 10 ms. at room temperature?
Solution SOLUTION:
The generation rate of electrons and holes Assume T = 300 K and recall mn = 0.26 m0.
equals: 𝟏 𝟑
𝑷𝑶𝑷𝑻 𝟏𝟎 −𝟐 𝑲𝒊𝒏𝒆𝒕𝒊𝒄 𝒆𝒏𝒆𝒓𝒈𝒚 = 𝒎𝒏 𝑽𝟐𝒕𝒉 = 𝑲𝑻
𝑮𝒏 = 𝑮𝒑 = ∝ = 𝟏𝟎 −𝟑 𝟐 𝟐
𝑬𝒑𝒉 𝒒 𝟏. 𝟖 𝒙 𝟏. 𝟔 𝒙 𝟏𝟎−𝟏𝟗 𝑽𝒕𝒉
= 𝟑. 𝟓 𝒙 𝟏𝟎𝟐𝟑 𝒄𝒎−𝟑 𝒔−𝟏 𝟑 𝑲𝑻 𝟑𝟎𝟎 𝑲
𝑱
= 𝟑 𝒙 𝟏. 𝟑𝟖 𝒙 𝟏𝟎−𝟐𝟑 ൗ𝑲 𝒙 𝒙 𝟗. 𝟏 𝒙 𝟏𝟎
Where, α is absorption coefficient, Popt = 𝒎 𝟎. 𝟐𝟔
illumination power , Ept is the red light , = 𝟐. 𝟑 𝒙 𝒎𝟓 𝒎Τ𝒔 = 𝟐. 𝟑 𝒙 𝟏𝟎𝟕 𝒄𝒎Τ𝒔
where the photon energy was converted
into Joules. The excess carrier densities Note that 1 J = 1 kg·m2/s2. Using mp = 0.39
then obtained from: m0 instead of mn, one would find the hole
𝜹𝒏 = 𝜹𝒑 = 𝝉𝒑 𝑮𝒑 = 𝟏𝟎 𝒙 𝟏𝟎−𝟑 𝒙 𝟑. 𝟓 𝒙 𝟏𝟎𝟐𝟑 thermal velocity to be 2.2 × 107cm/s.
= 𝟑. 𝟓 𝒙 𝟏𝟎𝟐𝟏 𝒄𝒎−𝟑 Therefore, the typical thermal velocity of
electrons and holes is 2.5 × 107cm/s, which
The excess carrier densities then obtained is about 1000 times slower than the speed
from: So that the electron and hole of light and 100 times faster than the sonic
densities equal: speed.
𝒏 = 𝒏𝒐 + 𝜹𝒏 = 𝟏𝟎𝟏𝟕 + 𝟑. 𝟓 𝒙 𝟏𝟎𝟐𝟏
= 𝟑. 𝟓 𝒙 𝟏𝟎𝟐𝟏 𝒄𝒎−𝟑 13
1.11. THE DRIFT CURRENT
Transport of charge carriers under electric field produces a
drift current. The current flow in a sample of figure
(1.27), having electron concentration (n) given by:
𝑰𝒏 = −𝒒 𝒏 𝝑𝒅 𝑨 𝑬𝒒. 𝟏. 𝟓𝟕
Substitute in the drift velocity, gives:
L Cross Sectional
𝑰𝒏 = −𝒒 𝒏 𝝁𝒏 𝜺 𝑨 Area
Drift
𝑬𝒒. 𝟏. 𝟓𝟖 Electron
Current In
V
.Figure 1-27. Current Conduction in a
Semiconductor Bar
14
𝑽 𝑽
we know, 𝜺= , 𝑹 = , 𝝈 = 𝒒𝒏 𝝁𝒏
𝑳 𝑰
We can find the resistance of the sample due to the
electron component as:
𝜺𝑳 𝑳 𝑳
𝑹𝒏 = = = 𝝆𝒏 𝑬𝒒. 𝟏. 𝟓𝟗
𝒒 𝑨 𝒏 𝝁𝒏 𝜺 𝒒 𝑨 𝒏 𝝁𝒏 𝑨
The same steps can be done to determine the drift current
due to the hole component in p-type material,
𝑰𝒑 = 𝒒 𝒑 𝝁𝒏𝒑 𝜺 𝑨
𝝈𝒑 = 𝒒𝒑 𝝁𝒑
𝟏
𝝆𝒑 = 𝑬𝒒. 𝟏. 𝟔𝟎
𝒒𝒑𝝁 15
In semiconductor, both carriers are included, so:
𝝈𝑻 = 𝝈𝒏 + 𝝈𝒑 = 𝒒 𝒏 𝝁𝒏 + 𝒑 𝝁𝒑
𝑰𝑻 = 𝒒 𝑨 𝜺 𝒏 𝝁𝒏 + 𝒑 𝝁 𝒑 𝑬𝒒. 𝟏. 𝟔𝟏
In intrinsic, 𝒏 = 𝒑 = 𝒏𝒊 , then,
𝝈𝒋 = 𝒒 𝒏𝒊 𝒏 𝝁𝒏 + 𝒑 𝝁𝒑 𝑬𝒒. 𝟏. 𝟔𝟐
Note that n >p, so in an intrinsic material the electrons
contribute more to conductivity than holes. It seems that
the conduction is due to electrons, so the conduction in
intrinsic material considered as in n-type material. For
extrinsic material, the conductivity given by:
𝝈𝒏 = 𝒒 𝒏𝒊 𝝁𝒏 𝒘𝒉𝒆𝒓𝒆 𝒏𝒏 > 𝒑𝒏 &
𝝈𝒑 = 𝒒 𝒑 𝝁𝒑 𝒉𝒆𝒓𝒆 𝒑𝒑 > 𝒏𝒑
16
1.12. THE DIFFUSION CURRENT
Transport of charges in semiconductors called diffusion. In a
semiconductor bar, the concentration of charge carrier is not
uniform. Diffusion of electrons or holes results from their
movement from higher concentration to lower concentration with
gradient d/dx where concentration of carriers is not distributed
uniform and varies with distance x. sectional area A is,
N/m3 Diffusion
Of Electron
Electron
Concentration Current In
Figure 1.28
Electron Diffusion Gradient
(m2)
17
Charges (electrons or holes) are in random motion because of
their thermal energy. Their motion gives rise to a current flow
known as diffusion current as shown in figure 1.28, diffusion flux
obeys Fick’s first law:
𝝏𝑵
𝑭 = −𝑫 𝑬𝒒. 𝟏. 𝟔𝟑
𝝏𝒙
F is flux of carriers and defined as the number of carriers passing
through m2/sec. Net transport of charges across the surface
constitutes current , it is proportional to the concentration
gradient. Current passing through cross
𝑰=𝒒𝑨𝑭 𝑬𝒒. 𝟏. 𝟔𝟒
So that, the diffusion current due to electrons given by:
𝜕𝑁
𝑰𝒏 = 𝒒 𝑨 𝑫𝒏 𝑬𝒒. 𝟏. 𝟔𝟓 18
And, the diffusion current due to holes is given by:
𝜕𝑃
𝑰𝒑 = − 𝒒 𝑨 𝑫𝒑 𝑬𝒒. 𝟏. 𝟔𝟔
𝜕𝑥
Where Dn , Dp are the diffusion constants for electrons and holes
respectively
The negative sign in (Eq.1.66) indicates that the hole current flows in the
direction opposite to the gradient of the holes. The diffusion constants
related with the mobility by Einstein relationship as:
𝑫𝒑 𝑫𝒏 𝑲𝑻
= = 𝑽𝑻 = = 𝑻ൗ𝟏𝟏𝟔𝟎𝟎 𝑬𝒒. 𝟏. 𝟔𝟕
𝝁𝒑 𝝁𝒏 𝒒
For lightly doped Si at room temperature = 39D. The total current of
electrons and holes components given by the summing of its diffusion
and drift currents as:
𝝏
𝑰𝒏 = 𝒒𝑨 𝒏 𝝁𝒏 𝑬 + 𝑫𝒏 𝒏ൗ𝝏𝒙 𝑬𝒒. 𝟏. 𝟔𝟖
𝝏𝒏ൗ
𝑰𝒑 = 𝒒𝑨 𝒑 𝝁𝒑 𝑬 + 𝑫𝒑 𝝏𝒙 𝑬𝒒. 𝟏. 𝟔𝟗
19
Solved Examples
Example (43) Note that the hole diffusion current occurs
An abrupt silicon p-n junction (Na = 10 cm and Nd = 4 x in the "short" n-type region and therefore
16 -3
1016 cm-3) is biased with Va = 0.6 V. Calculate the ideal diode depends on the quasi-neutral width in that
current assuming that the n-type region is much smaller
than the diffusion length with Wn = 1 µm and assuming a region. The electron diffusion current
"long" p-type region. Use µn = 1000 cm2/V-s and µp = 300 occurs in the "long" p-type region and
cm2/V-s. The minority carrier lifetime is 10 µs and the therefore depends on the electron
diode area is 100 µm by 100 µm.
Solution:
diffusion length in that region.
Example (44)
The current calculated from:
𝑫𝒏 𝒏𝒑𝒐 𝑫𝒑 𝒑𝒏𝒐 𝑽𝒐 For a p+-n Si junction the reverse current at
ൗ𝑽 room temperature is 0.9 nA/cm2. Calculate
𝑰=𝒒𝑨 + 𝒆 𝒕 −𝟏
𝑳𝒏 𝑾𝒏 the minority-carrier lifetime if Nd = 1015
cm−3, ni = 1.05 × 1010 cm−3, and μp = 450
With cm2 V−1 s−1.
Dn = µnVt = 1000 x 0.0258 = 25.8 cm2/V-s Solution:
𝝏 𝒏𝒏 𝒏𝒏 − 𝒏𝒏𝟎 𝝏 𝒏𝒏 − 𝒏𝒏𝟎
= − = 𝑬𝒒. 𝟏. 𝟕𝟒
𝝏𝒕 𝝉 𝝏𝒕
The quantity pn - pno or nn - nno represents the excess carrier
density, and the rate of change of excess density is proportional to
the density itself. The (-) sign indicates that the change is decreases
in case of recombination.
22
1-14 CONTINUITY EQUATIONS
carrier concentration is a function of both time and distance, upon
fact that charge created or destroyed. Consider a volume of area A
and length dx, fig.1.29, in which the hole concentration is P, and
P/p is the decreasing of hole concentration by recombination, so,
the change in hole
P Holes/m3
+
Cross Section
Area A
x dx
25
The general continuity equation 1.81 be considered for three
special cases.
1.When concentration is independent of x at zero bias:
If = 0 and concentration is independent of x, Eq.1.81 will
be rewrite as:
𝝏𝒑𝒏 𝒑𝒏 − 𝒑𝒏𝟎
= − 𝑬𝒒. 𝟏. 𝟖𝟐
𝝏𝒕 𝝉𝒑
The solution of (Eq.1.82) is:
𝒑𝒏 − 𝒑𝒏𝟎 = 𝒑− 𝒏𝟎 − 𝒑𝒏𝒐 𝒆− 𝒕Τ𝝉
𝑬𝒒. 𝟏. 𝟖𝟑
26
2.When concentration is independent of t at zero bias:
If = 0, and a steady state has been reached at no time,
so that pn/t = 0, Eq.1.81 will be rewritten again as:
𝟐 𝟐
𝝏 𝒑𝒏 𝒑𝒏 − 𝒑𝒏𝟎 𝝏 𝒑𝒏
𝑫𝒑 𝟐
= − =
𝝏𝒙 𝝉𝒑 𝝏 𝒙𝟐
𝒑𝒏 − 𝒑𝒏𝟎
= 𝑬𝒒. 𝟏. 𝟖𝟒
𝑫𝒑 𝝉𝒑
The solution of Eq.1.84 is:
− 𝒙Τ𝑳𝒑 𝒙Τ𝑳𝒑
𝒑𝒏 − 𝒑𝒏𝟎 = 𝑲𝟏 𝒆 + 𝑲𝟐 𝒆 𝑬𝒒. 𝟏. 𝟖𝟓
Where K1, K2 are integration constants, 𝑳𝒑 = 𝑫𝒑 𝝉𝒑 and
it represents the distance into the semiconductor at which
the injected concentration falls to 1/e of its value at
distance x = 0.
27
3 . When concentration varies sinusoidal with t at zero bias :
If = 0, and the injected concentration varies sinusoidal with
angular frequency (),
𝒑𝒏 𝒙 , 𝒕 = 𝒑𝒏 𝒙 𝒆 𝒋 𝝎 𝒕 𝑬𝒒. 𝟏. 𝟖𝟔
If Eq.1.86 substituted in to the continuity Eq.1.81, results:
𝒑𝒏 𝒙 𝝏𝟐 𝒑𝒏
𝒋 𝝎 𝒑𝒏 𝒙 = − + 𝑫𝒑 𝟐
𝑬𝒒. 𝟏. 𝟖𝟕
𝝉𝒑 𝝏𝒙
𝝏𝟐 𝒑𝒏 𝟏 + 𝒋 𝝎 𝝉𝒑
𝟐
= 𝟐
𝒑𝒏 𝑬𝒒. 𝟏. 𝟖𝟖
𝝏𝒙 𝑳𝒑
And at F = 0, Eq.1.88 rewritten again as:
𝝏𝟐 𝒑𝒏 𝒑𝒏
𝟐
= 𝟐 𝑬𝒒. 𝟏. 𝟖𝟗
𝝏𝒙 𝑳𝒑
28
Solved Example
Example (48)
𝒕
Consider n-type silicon with Nd = 1015 cm−3 at T = 300◦K. 𝜹𝒑 𝒕 = 𝑮𝒑 𝝉𝒑 𝟏 − 𝒆𝒙𝒑 −
light source is turned on at t = 0. The source illuminates 𝝉𝒑
the semiconductor uniformly, generating carriers at the
rate of Gn = Gp = 1019cm−3s−1. There is no applied field. (b) As t >∞, the system will approach steady state.
When the steady state excess carrier concentration
(a) Write down the continuity equation and solve it to get is 5×1013cm−3, get the minority carrier lifetime, 𝝉𝒑 .
the expression for the excess minority carrier
concentration, 𝜹𝒑 𝒕 , as a function of time for t ≥ 0. Solution:
Solution: The system will approach steady state as t >∞,
When there is no applied electric field, the carrier Evidently, the steady state carrier density is given
distribution is diffusion driven. The continuity 𝜹𝒑 𝒕 |𝐭 → ∞ = 𝑮𝒑 𝝉𝒑 .
equation for the minority carrier
𝝏𝒑 𝒙 , 𝒕 𝟏 𝝏 𝑱𝒑 𝒙 , 𝒕 The 𝝉𝒑 must then take on the value
= + 𝑮𝒑 𝒙 , 𝒕 − 𝑹𝒑 𝒙 , 𝒕 𝟏𝟑
𝝏𝒕 𝒒 𝝏𝒙 𝝉𝒑 = 𝟓 𝒙 𝟏𝟎 ൗ𝑮𝒑 .
Then reduces to , With Gp = 1019 cm−3s−1, then the minority (hole)
𝝏𝜹𝒑 𝝏𝜹𝒑 carried lifetime must be 𝝉𝒑 = 5 × 10−6 s.
= 𝑮𝒑 −
𝝏𝒕 𝝉𝒑 (c) Determine the time at which the excess carrier
With the general solution concentration becomes half of the steady state
𝒕 value, 𝜹𝒑 𝒕 |𝐭 → ∞ that you calculated in (b).
𝜹𝒑 𝒕 = 𝑨 𝒆𝒙𝒑 − + 𝑮𝒑 𝝉𝒑 Solution:
𝝅𝒑
The value at which
Using the initial condition (before the light was 𝒕 𝟏
turned on) that 𝜹𝒑 𝒕 = 0, then we found that 𝑨 𝒆𝒙𝒑 − = 𝒔𝒐, 𝒕 = 𝒍𝒏 𝟐 𝝉𝒑
= − 𝑮𝒑 𝝉. The full solution then is, 𝝉 𝟐 −𝟔
= 𝟎. 𝟔𝟗 𝒙 𝟓 𝒙 𝟏𝟎 𝒔 29
1-15. Flux of Charged Spices
𝝏𝑵
To determine the formula of flux (F), 𝑭 = −𝑫 consider the motion of
𝝏𝒙
positively charged impurities in a crystal. The atom of the crystal forms a
series of potential hills, which hinder the motion of the charged impurities as
shown. The height of potential (a)
P
barrier (W) is in the order of o (W) Distance
T
some of electron volt. In most E Figure (a).
material, the distance between n F1 F2 F3 F4
T
successive potential barrier (a) I
is in the order of the lattice A
Fig. (1 - 30) Model of ionic motion within crystal, Potential distribution with and without bias
30
If a constant field is applied, the potential distribution as a
function of distance is shown in figure b, this make flow of
positively charged particles to the right easier than left. To
calculate the flux (F) at a position (x), it is average of fluxes at
positions (x – a/2) and at (x + a/2), these two fluxes given by
(f1 – f2), and (f3 – f4).Consider component (F1), is product of:
1. Intensity per unit area for impurities (charges) at the
potential valley at (x – a).
2. The probability of a jump of any of these impurities
(charges) to the next valley at position (x)
3. The frequency of attempted jump (ⱱ)
31
Thus, we can write:
𝒒 𝟏
𝑭𝟏 = 𝒂 𝑪 𝒙 − 𝒂 𝒆𝒙𝒑 − 𝑾− 𝒂 𝝐 (ⱱ) 𝐄𝐪 (𝟏 − 𝟗𝟎)
𝑲𝑻 𝟐
Where, 𝒂 𝑪 𝒙 − 𝒂 is the density per unit area of the particles
situated in the valley at (x – a). The exponential factor is the
probability of a jump from the valley at (x – a) to the valley at
position (x), and (ⱱ) is frequency of attempted jump note that,
the lowering of the barrier due to the electric field (ε).
Similar formulas can be written for (F2), (F3), and (F4). By
combined them to give a formula for the flux (F) at position (x),
with the concentration (C(x ± a)) approximated by
( 𝐂(𝐱) ± 𝐚 𝛛 𝐂Τ𝛛 𝐱 , We obtain,
32
𝟐 −𝒒𝑾Τ𝑲𝑻 𝝏𝑪 𝒒𝒂𝜺 −𝒒𝑾Τ𝑲𝒕
𝑭 𝒙 = - ⱱ𝒂 𝒆 𝒄𝒐𝒔𝒉 + 𝟐𝒂ⱱ𝒆 C
𝒒𝒂𝜺 𝝏𝒙 𝟐 𝑲𝑻
sinh - 𝟐 𝑲𝑻
𝑬𝒒 (𝟏 − 𝟗𝟏)
As extremely important, limiting from of this equation is obtained for the
case when the electric field is relatively small, i.e., ε «KT/q a. In this
case, we can expand the cosh and the sinh terms in the equation. Noting
that cosh (x) = 1 and sinh (x) = x for x → 0, this results in the limiting form
of the flux equation for a positively charged spices.
𝝏𝑪
𝑭 𝒙 = −𝑫 + 𝝁 𝜺 𝑪 𝑬𝒒. (𝟏 − 𝟗𝟐) , Where:
𝝏𝒙
𝑫 = 𝒗𝐚𝟐 𝒆−𝒒𝑾Τ𝑲𝑻
𝒗𝐚𝟐 𝒆−𝒒𝑾Τ𝑲𝑻
𝝁=
𝑲𝑻ൗ
𝒒
33
Note that, mobility (µ) and the diffusivity (D) related by:
𝑲𝑻
𝑫= 𝝁
𝒒
This is the well Known Einstein relationship. A similar derivation
made for the motion of negatively charged spices. In addition, we
can derive that,
−𝑬𝒂 Τ𝑲𝑻
𝑫 = 𝑫𝒐 𝒆
Where, (Do) is the diffusivity of impurities in oxide, (Ea) is the
activation energy. Thus, the activation energy corresponds to the
energy required to form a silicon space rather than to the energy
required to move the impurity. Since silicon – to – silicon, bunds
must be broken to form a space. In fact, activation energy of the
diffusivity of acceptor and donor type impurities, and it is in the
range of two and three electron volt in germanium.
34