UNIVERSITY OF CALIFORNIA College of Engineering Department of Electrical Engineering and Computer Sciences EE 130 Spring 2003 Homework Assignment

#6 Solutions Problem 1: a) We have an abrupt pn junction with Nd=1x1016 cm-3, and Na=5x1016 cm-3. Since the doping concentrations on both sides are comparable, we use the standard full expressions: i) Vbi = ii) W = [2 Prof. King

N N (5 1016 )(1016 ) kT ln[ a 2 d ] = 0.026V ln[ ] = 0.76V (1010 ) 2 q ni

Si
q

Vbi (

1 1 1/ 2 (10 12 F / cm ) 1 1 + )] = [2 (0.76V )( + 16 3 )]1 / 2 19 3 16 Na Nd 1.6 10 C 5 10 cm 10 cm

5

Therefore, W = 3.45 10 cm = 0.345m

iii) The electric field curve (see diagram below) is linear, and has its maximum at the point of the metallurgical junction. The total area under the curve is equal to the potential difference i.e., the built-in potential. Therefore: 2V 1 2(0.76V ) max W = Vbi max = bi = = 4.4 10 4 V / cm 4 2 W 0.345 10 cm iv)
where:

' max =
Si
q

2(Vbi Va ) 2(0.76V + 5V ) = = 1.3 105V / cm 4 W' 0.87 10 cm

1 1 1/ 2 + )] = 8.7 10 5 cm = 0.87 m Na Nd

W ' = [2

(Vbi Va )(

Note that both the depletion region width and |max| are larger than with Va=0 V. b) The requested plots are given on the next page.

i) No applied bias (Va=0)

ii) Applied bias (Va!=0)

----Depletion Approximation

c) In this case, the p-side is degenerate. As a result, Efp=Ev => Ei-Efp=0.56eV. Therefore, N kT Vbi = p + n = 0.56V + ln[ d ] = 0.89V q ni
Additionally, because the p-side of the pn junction is doped more heavily than n-side, the depletion region width of the highly-doped region can be neglected:

W = X n + X p = [2

Si
q

Vbi (

1 1 1/ 2 1 1/ 2 + )] X n = [2 Si Vbi ( )] = 3.4 10 5 cm = 0.34 m Na Nd q Nd

Note that this depletion region is a little larger than the one obtained in part (a).

Problem 2: (a) From the charge neutrality condition, we have an equal amount of charge at either side of the junction; therefore 1/ 2 2N A X p 1 2(1016 )(0.5 10 4 ) qN A X p = q (aX n ) X n X n = = = 10 4 cm = 1m 20 2 10 a

(b) W X d = X n + X p = 1m + 0.5m = 1.5m (c) From Poissons equation (for the region xp < x < 0):

2V = 2 Si X X
Therefore, since:

E ( x)

=
0

xp

qN A

Si

x ( x) =

qN A

Si

(x + x p )

V (x) X

V =

qN A

Si

( x + x p )x V ( x) =

qN A x 2 ( + x p x) + C1 Si 2

Boundary Condition: Let V(x=0)=0; therefore, C1=0. As a result, the potential drop from x=-xp to x=0 is just V(-xp). Thus:

qN A qN A x p (1.6 10 19 )(1016 ) 2 2 V ( x p ) V ( 0) = V ( x p ) = ( xp ) = (x p ) = (0.5 10 4 ) 2 = 1.93Volts 2 2 Si Si 2(1.036 10 12 )

Note that the voltage drop is negative since the x=-xp position is at a lower potential than the x=0 position.

(d) From Poissons equation (for the region 0< x < xn): 2V qax qa 2 = = x ( x) = ( x ) + C2 2 X 2 Si Si Si X

But, since (xn)=0 Now, since:

=> C 2 =

qax n qa 2 2 ( x ) = ( x xn ) 2 Si 2 Si

V (x) X

V =

qa 2 qa x 3 2 2 ( x x n )x V ( x) = ( x n x) + C3 2 Si 2 Si 3

Boundary Condition: As before, V(x=0)=0; therefore, C3=0. As a result, the potential drop from x=0 to x=xn is just V(xn). Thus:

 V ( x n ) V (0) = V ( x n ) =

qa x n qa (1.6 10 19 )(10 20 ) 3 3 ( xn ) = (xn ) = (110 4 ) 3 = 5.15Volts 2 Si 3 3 Si 3(1.036 10 12 )

Note that the voltage drop is positive since the x=xn position is at a higher potential than the x=0 position.

(e) Vbi Va = V(xn) V(xp) = 5.15 V (-1.93 V) = 7.08 Volts (f) From part c):
( x ) = qN A (x + x p )
.for xp < x < 0

Si

max occurs at x=0. Thus:

max = ( x = 0) =

qN A

Si

(1.6 10 19 )(1016 )(0.5 10 4 ) = 7.73 10 4 V / cm (x p ) = 12 1.036 10

1.036 10 12 F / cm [100(10 4 cm) 2 ] = 6.9 10 15 F = 6.9 fF ( A) = (g) C = 4 W 1.5 10 cm

Si

Problem 3: (a) Avalanche breakdown: Impact ionization occurs due to transfer of energy from an energetic electron to an atom in the lattice which is sufficient to free a valence electron from the atom represented by vertical transitions on the band diagram.

Avalanche

Zener

Zener breakdown: the potential barrier between a filled state on one side of the barrier and an empty state on the other side is thin enough for an electron to tunnel (quantum mechanically) from the valence band on the p-side to the conduction band on the n-side (at the same energya horizontal transition on the band diagram). (b) Avalanche breakdown is the dominant mechanism for pn diodes with moderate doping because the depletion width is wider, so quantum mechanical tunneling (which is necessary for Zener breakdown) is not possible. ( c)
As temperature increases, VBR for avalanche, but VBR for Zener. 2. Breakdown is softer for Zener. (Current approaches infinity more slowly.) 1.

Problem 4: (a) The built-in potential, Vbi, can be determined as follows: 2 n N N kT kT kT [ln( n ( xn )) ln( n ( x p ))] = [ln( N d ) ln( i )] = ln[ a 2 d ] = 0.66V Vbi = n + int + p = Na q q q ni
In other words, the built-in potential is the same as in the normal pn junction. Now, since the area under the curve of the (x) plot provides the potential dropped in each section, then: qN d 2 qN d qN a 2 Vbi = n + int + p = wn + wn wint + w p = 0.66V (equation 1) Si 2 Si 2 Si (The middle section expression can easily be found from int=Eintwint). With respect to the intrinsic region, we can make the assumption that the charge density is zero in that region. This assumption is accurate because the density of free carriers in this region is negligibly small. N Therefore, by charge neutrality: qN d wn = qN a w p w p = d wn (equation 2) Na Substitute equation (2) into equation (1), and solve for wn, wp. By doing this, we get:

wn= 2.5x10-6 cm=0.025 m

wp= 2.5x10-5 cm=0.25 m

The requested -field plot is given below (additional plots are included for completeness):

Nd Na (cm-3)
-2 -1

1016 0.5 1015 1 2

X (m)

(x) xp

xn

(x)
wint wn wp V(x) n int p Vbi

b) In a pin junction, max, pin =

qN d

Si

wn = 4000V / cm 2Vbi = 14500V / cm W

For a pn junction with the same doping concentration: max, pn =

2 V 1 1 + ] where: W = Si bi [ Na Nd q

1/ 2

= 9.1 10 5 cm = 0.91m

Thus, the pn junction provides a larger max.

c) Much of the built-in potential is dropped across the intrinsic region, so that the depletion regions in the n- and p-type materials need not extend as far as in a pn junction. The maximum field is therefore reduced. Alternatively, we may say that the total depletion region width is the intrinsic region plus some depleted regions in the n- and p-type materials. However, the intrinsic region remains neutral. The depletion approximation means that the density of free carriers is negligible, and that when solving Poissons equation for this region, we can treat it as containing no carriers (i.e., the charge density in this region, (x), is approximately 0; therefore, the electric field is constant there).

Problem 5:
a) Using figure 3.8 in the book, the dopant densities corresponding to 2-cm p-type and 0.5cm n-type materials are Na=7x1015 cm-3 and Nd=1x1016 cm-3, respectively. Therefore, assuming low-level injection (and from equations 6.15, 6.18 in the book), the minority carrier concentrations at the edge of the depletion region are determined as follows:

n p ( x p ) n ( x p ) no = n po ( x p )[exp(
Similarly,

qVa n qV (1010 ) 2 ) 1] i [exp( a )] = [exp( 23)] = 1.39 1014 cm 3 15 kT Na kT 7 10

2

p n ( x n ) p( x n ) po = p no ( x n )[exp(

qVa n qV (1010 ) 2 ) 1] i [exp( a )] = [exp( 23)] = 9.7 1013 cm 3 kT Nd kT 1 1016

b) For a long-base diode, the hole current distribution on the n-side and the electron current distribution in the p-side are predicted by equations 6.24, 6.26 respectively (assuming that the field is negligible in the quasi-neutral regions). Note that in the expressions below, the x=0 point is at the junction; x>0 is the n-side, and x<0 is the p-side. Since Dn kT n = (.0258V )(1290cm 2 / V sec) = 33.3cm 2 / s Ln Dn n = .00577cm = 57.7 m q kT p = (.0258V )(437cm 2 / V sec) = 11.3cm 2 / s L p D p p = 3.36 10 4 cm = 3.36 m Dp q

the values for each mobility where obtained from figure 3.5 in the book.

Thus:
J p ( x ) = qD p ni qV ( x xn ) ( x xn ) [exp( a ) 1] exp[ ] = 0.52 exp[ ] A / cm 2 N d Lp kT Lp Lp (x + xp ) (x + xp ) ni qV [exp( a ) 1] exp[ ] = 0.13 exp[ ] A / cm 2 N a Ln kT Ln Ln
2 2

(in n-type region, x>0)

J n ( x ) = qDn

(in p-type region, x<0)

Note that both of these expressions are exponentially-decaying functions. The total current, Jt, is then: J t = J n ( x p ) + J p ( xn ) = 0.13 + 0.52 = 0.63 A / cm 2
As expected, the total current is constant. The plots of the current distributions is given below:
total = 0.63 A/cm JJ=1.37 A/cm2 t 2

p-Si
Jn (x)

Jp (x)

n-Si

x
-xp 0 xn

c) According to the sketch in part (b), we can only find equal minority-carrier and majoritycarrier currents (in magnitude) in the n-type region. Therefore: J p = J n = 0.5J t = 0.315 = 0.52 exp[ ( x xn ) ] Lp

( x x n ) = L p (0.501) = 1.7 10 4 cm = 1.7 m