Example: Fermi levels in semiconductors

An n-type Si wafer has been doped uniformly with 1016 phosphorus (P) atoms cm–3. Calculate the
position of the Fermi energy with respect to the Fermi energy EFi in intrinsic Si. The above n-type Si
sample is further doped with 21017 boron atoms cm–3. Calculate position of the Fermi energy with
respect to the Fermi energy EFi in intrinsic Si at room temperature (300 K), and hence with respect to
the Fermi energy in the n-type case above.
Solution

P (Group V) gives n-type doping with Nd = 1016 cm–3, and since Nd >> ni ( = 1010
cm–3 from Table 3.1), we have n = Nd = 1016 cm–3. For intrinsic Si,
ni = Ncexp[−(Ec − EFi)/kBT]
whereas for doped Si,
n = Ncexp[−(Ec − EFn)/kBT] = Nd
where EFi and EFn are the Fermi energies in the intrinsic and n-type Si. Dividing
the two expressions
Nd /ni = exp[(EFn − EFi)/kBT]
so that
EFn − EFi = kBTln(Nd/ni) = (0.0259 eV) ln(1016/1010) = 0.358 eV
Example: Fermi levels in semiconductors

Solution (Continued)

When the wafer is further doped with boron, the acceptor concentration, Na =
21017 cm–3 > Nd = 1016 cm–3. The semiconductor is compensation doped and
compensation converts the semiconductor to a p-type Si. Thus,
p = Na − Nd = 21017−1016 = 1.91017 cm–3.
For intrinsic Si,
p = ni = Nvexp[−(EFi − Ev)/kBT],
whereas for doped Si,
p = Nvexp[−(EFp − Ev)/kBT] = Na − Nd
where EFi and EFp are the Fermi energies in the intrinsic and p–type Si
respectively Dividing the two expressions,
p/ni = exp[−(EFp − EFi)/kBT]
so that
EFp − EFi = −kBT ln(p/ni)
= −(0.0259 eV)ln(1.91017/1.01010) = −0.434 eV
Example: Conductivity of n-Si

Consider a pure intrinsic Si crystal. What would be its intrinsic conductivity at 300K? What is the
electron and hole concentrations in an n-type Si crystal that has been doped with 1016 cm–3
phosphorus (P) donors. What is the conductivity if the drift mobility of electrons is about 1200 cm2
V-1 s-1 at this concentration of dopants.

Solution

The intrinsic concentration ni = 1×1010 cm-3, so that the intrinsic conductivity is

 = eni(e + h) = (1.6×10-19 C)( 1×1010 cm-3)(1450 + 490 cm2 V-1 s-1)
= 3.1×10-6 -1 cm-1 or 3.1×10-4 -1 m-1
Consider n-type Si. Nd = 1016 cm-3 > ni (= 1010 cm-3), the electron concentration
n = Nd = 1016 cm-3 and
p = ni2/Nd = (1010 cm-3)2/(1016 cm-3) = 104 cm-3
and negligible compared to n.
The conductivity is
 = eN d  e = (1.6 10 −19 C)(11016 cm −3 )(1200 cm 2 V −1s −1 ) = 1.92 Ω −1cm −1
EXAMPLE: LED spectral linewidth
We know that a spread in the output wavelengths is related to a spread in the emitted photon energies.
The emitted photon energy hυ = hc / λ. Assume that the spread in the photon energies (h υ)  3kBT
between the half intensity points. Show that the corresponding linewidth Δλ between the half intensity
points in the output spectrum is
3k BT
 = o2 LED spectral linewidth (3.11.3)
hc
where λo is the peak wavelength. What is the spectral linewidth of an optical communications LED
operating at 1310 nm and at 300 K?
Solution

First consider the relationship between the photon frequency υ and λ,

c hc
= =
 h
in which h υ is the photon energy. We can differentiate this
d hc 2
=− =− (3.11.4)
d (h ) (h ) 2
hc
The negative indicates that increasing the photon energy decreases the
wavelength. We are only interested in changes, thus Δ λ / Δ(hυ)  |dλ /d(hυ)|, and
this spread should be around λ = λo , so that Eq. (3.11.4) gives,
EXAMPLE : LED spectral linewidth
Solution (continued)

o2 3k BT
 = (h ) = o2
hc hc
where we used (hυ) = 3kBT. We can substitute λ = 1310 nm, and T = 300 K to
calculate the linewidth of the 1310 nm LED

−23
3k T 3(1.38  10 )(300)
 = 2 B −9 2
= (1310  10 )
hc (6.626  10−34 )(3  108 )
= 1.07 × 10-7 m or 107 nm

The spectral linewidth of an LED output is due to the spread in the photon
energies, which is fundamentally about 3kBT. The only option for decreasing Δ λ
at a given wavelength is to reduce the temperature. The output spectrum of a
laser, on the other hand, has a much narrower linewidth.
EXAMPLE: LED spectral width
Consider the three experimental points in Figure 3.32 (b) as a function of temperature. By a suitable
plot find m and verify
3k BT
 = o2
hc LED spectral linewidth (3.11.3)
Solution
From Example,3.11.1, we can use the Eq. (3.11.3). with m instead of 3 as follows

  mk B  LED linewidth and temperature (3.11.5)

=  T
o  hc 
2

and plot 2 vs. T. The slope of the best line forced through zero should give
mk/hc and hence m. Using the three o and  values in the inset of Figure 3.32(b),
we obtain the graph in Figure 3.34. The best line is forced through zero to follow
Eq. (3.11.5), and gives a slope of 1.95×10-7 nm-1 K-1 or 195 m-1 K-1. Thus,

m(1.38 10−23 J K -1 )
−1
slope = 195 m K =
so that m = 2.81 (6.626 10−34 J s)(3 108 m s −1 )
EXAMPLE: LED spectral width
Solution (continued)

8
AlGaAs LED
7
104  6
o 5
(m-1) 4
Slope = 195 m -1 K-1
3
m = 2.8
2
1
0
0 50 100 150 200 250 300 350 400
Temperature (K)

The plot of plot 2 vs. T for an AlGaAs infrared LED, using the peak wavelength
o and spectral width  at three different temperatures, using the data shown in the
table.
EXAMPLE: Dependence of the emission peak and linewidth on temperature

Using the Varshni equation, Eq. (3.11.2), Eg = Ego − AT2/(B + T), find the shift in the peak
wavelength (λo) emitted from a GaAs LED when it is cooled from 25 C to -25 C. The Varshni
constants for GaAs are, Ego = 1.519 eV, A = 5.41 × 10−4 eV K−1, B = 204 K.
Solution

At T = 298 K, using the Varshni equation

Eg = Ego − AT2/(B + T)
= 1.519 eV − (5.41 × 10−4 eV K-1)(298 K)2/(204 K + 298 K) = 1.423 eV.

At 298 K, (1/2) kBT = 0.0128 eV. The peak emission is at ho  Eg + (1/2) kBT. Using
o = c/o, we get
ch (3 108 ms −1 )(6.626 10−34 J s) /1.602 10−19 eV J -1 )
o = = = 864.2 nm
( Eg + 2 k BT )
1
(1.4223eV + 0.0128eV)

At −25 C or , 248 K, (1/2) kBT = 0.0107 eV, repeating the above calculation,
Eg = 1.519 eV − (5.41 × 10−4 eV K-1)(248 K)2/(204 K + 248 K) = 1.445 eV
EXAMPLE: Dependence on the emission peak and linewidth on temperature

Solution (continued)
0
and the new peak emission wavelength is
(3  108 m s−1 )(6.626  10−34 J s) /(1.602  10−19 eV J -1 )
o = = 852.4 nm
(1.445 eV + 0.01069 eV)
The change  = o − 0 = 864.2 − 852.4 = 11.8 nm over 50 C, or 0.24 nm / C.

The examination of the Figure

3.32(b) shows that the change in
the peak wavelength per unit
temperature in the range −40 C to
85 C is roughly the same.
Because of the small change, we
kept four significant figures in Eg
and o calculations.
EXAMPLE: LED brightness LED brightness
Consider two LEDs, one red, with an optical output power (radiant flux) of 10
mW, emitting at 650 nm, and the other, a weaker 5 mW green LED, emitting at
532 nm. Find the luminous flux emitted by each LED.
Solution
For the red LED, at  = 650 nm, Figure
3.41 gives V  0.10 so that from Eq.
(3.14.8)
 = Po× (633 lm W-1) × V
= (1010-3 W)(683 lm W-1)(0.10) = 0.68 lm

For the green LED,  = 532 nm, Figure

3.41 gives V  0.87 so that from Eq.
(3.14.8)
 = Po× (633 lm W-1) × V
= (510-3 W)(683 lm W-1)(0.87) = 3.0 lm

Clearly the green LED at half the power Figure 3.41

is 4 times brighter than the red LED.
Laser
REPORT