Chemistry Reference
In-Depth Information
origin, where we choose to write the general solution for a state at energy
E in the form
A e i kx
B e i kx
ψ(
x
) =
+
0
<
x
<
a
(3.10)
with k 2
2 , as before. The general solution to Schrödinger's
equation in the first barrier to the left of the origin is given by
=
2 mE
/
C e κ x
D e κ x
ψ(
x
) =
+
b
<
x
<
0
(3.11)
2
2 , and we assume E
where
V 0 .
We now introduce four boundary conditions to determine the allowed
solutions of Schrödinger's equation. The first two are easily chosen, requir-
ing that the wavefunction
κ
=
2 m
(
V 0
E
)/
<
ψ
and its derivative d
ψ/
d x are continuous at the
=
well/barrier interface at x
0:
ψ(
0
)
: A
+
B
=
C
+
D
(3.12a)
ψ (
0
)
:
i kA
i kB
= κ
C
κ
D
(3.12b)
The two remaining boundary conditions are derived using Bloch's theo-
rem. We apply eq. (3.2b) at x
e i q ( a + b ) and
=
a ,sothat
ψ
(
a
) = ψ
(
b
)
w
b
ψ w (
) = ψ b (
e i q ( a + b ) :
likewise for the derivatives
a
b
)
: A e i ka
B e i ka
C e κ b
D e κ b
e i q ( a + b )
ψ
(
a
)
+
= (
+
)
(3.12c)
w
ψ w (
i kA e i ka
i kB e i ka
C e κ b
D e κ b
e i q ( a + b )
a
)
:
=
κ
)
(3.12d)
We can then derive the condition for allowed energy levels by solving the
4
4 determinant involving the four unknowns A , B , C , and D ,aswas
done with eq. (1.52) in Chapter 1. The energy levels associated with Bloch
wavenumber q are then found to be solutions of the equation
×
k
sin
1
2
k
κ
(
(
+
)) =
(
)
) +
(
)
)
cos
q
a
b
cos
ka
cosh
b
ka
sinh
b
(3.13)
We can deduce one particularly important result from this equation,
namely, that there will always be energy gaps between the allowed energy
bands in the K-P model. To see this, we first note that the left-hand side of
eq. (3.13), cos
(
q
(
a
+
b
))
, can only take values between
1 and 1. If we choose
=
π
±
)
(
π) =
ka
n
, then the right-hand side reduces to
cosh
b
, as sin
n
0
and cos
is always greater than 1 for
states which are bound within the well. Hence, there can be no solutions
to eq. (3.13) for k
(
n
π)
1. The magnitude of cosh
b
)
=
π/
n
a , implying no solutions are possible with energy
2 , and likewise for neighbouring values of the energy.
This is illustrated in fig. 3.4(a), where we plot an example of the variation
2
E
= (
/
2 m
)(
n
π/
a
)
 
Search WWH ::




Custom Search