PH12b 2010 Solutions HW#7

1.

a) Our state is
p 1
j (dt)i = dt^
aj i j1i + I^ ay a
dt^ ^ j i j0i
2
and
p 1
h (dt)j = ay h j
dt^ h1j + I^ ay a
dt^ ^ h j h0j
2
then
2
1
h (dt)j (dt)i = dt h j a ^ j i h1j 1i + h j I^
^y a ay a
dt^ ^ j i h0j 0i
2
= ^y a
dt h j a ^j i+h j i ^y a
dt h j a ^ j i + O(dt2 )
= h j i = 1:

b) The coherent state is

1
X n
j j2 =2
j i=e p jni :
n=0 n!
Then, working in linear order in dt we get
1
X n
1 j j2 =2 1
I^ ay a
dt^ ^ j i = e p I^ ay a
dt^ ^ jni ;
2 n=0 n! 2
X1 n
j j2 =2 1
= e p 1 dtn jni ;
n=0 n! 2
X1 n n
j j2 =2 dt=2
e p e jni ;
n=0 n!
1 h in
X1
j j2 =2 dt=2
e p e jni ;
n=0 n!
1 2 E
j j2 =2 dt=2 dt=2
e exp e e ;
2
" #
1 1
2 E
j j2 =2 dt=2
e exp 1 dt e ;
2 2
1 E
2 dt=2
exp dt j j e :
2

^y a
where we used a ^ jni = n jni and
h i 1
dt=2
e 1 dt + O(dt2 ):
2

c) We want to verify that

ED
2
dt^ ^y
aj ih ja dt j j e dt=2
e dt=2

to linear order in dt. First, because a

^j i = j i the left-hand side gives
2
lhs = dt j j j i h j :

1
 dt=2
Now, for the right hand side we should expand e to linear order in dt, this is
E 1 2 X1
1 h in
dt=2 dt=2 dt=2
e = exp j j e p e jni
2 n=0 n!
1
X n
1 2 1 1
exp j j 1 dtn p 1 dtn jni
2 2 n=0 n! 2
1
X n
j j2 =2 1 1
e 1 dt p 1 dtn jni j i (1 + O( dt)) :
2 n=0 n! 2
Then ED
2 dt=2 dt=2 2
rhs = dt j j e e dt j j j i h j = lhs:
Using this, the result in a) and Eq.(1) the evolution of j i h j is given by
1 1
j ih j ! dt^ ^y + I^
aj ih ja dt^ ^ j i h j I^
ay a ay a
dt^ ^ ;
2 2
ED 1
2 ED
2 dt=2 dt=2 2 dt=2 dt=2
dt j j e e + exp dt j j e e ;
2
ED ED
2 dt=2 dt=2 2 dt=2 dt=2
dt j j e e + (1 dt j j ) e e ;
ED
dt=2 dt=2
= e e :

then ED
dt=2 dt=2
j ih j ! e e
to linear order in dt:
To evolve the state a …nite time t = ndt (n >> 1) we just have to evolve the state n times by applying
the last formula n times. It is easy to see that this gives
ED ED ED
dt=2 dt=2 dt dt (ndt)=2 (ndt)=2 t=2 t=2
j ih j ! e e ! e e ! ::: ! e e = e e ;

then the initial coherent state evolve as E

t=2
j i! e :
d)
" 1
#" 1
#
X m X n
j j2 =2 j j2 =2
h j i = e p hmj e p jni ;
m=0 m! n=0 n!
1 X
X 1 m n
j j2 =2 j j2 =2
= e e p p hmj jni ;
m=0 n=0 m! n!
X1 X 1 m n
j j2 =2 j j2 =2
= e e p p mn ;
m=0 n=0 m! n!
1
X n
j j2 =2 j j2 =2 ( )
= e e ;
n=0
n!
j j2 =2 2
= e e j j =2 e ;
1 2 2 1
= exp j j +j j exp ( ) ;
2 2
1 2 1
= exp j j exp ( ) ;
2 2
1 2
= exp j j exp [i Im ( )] :
2

2
Then
j j2 =2 i Im( )
h j i=e e ;
notice that the last factor is just a phase factor.
Now, we want to normalize
j i=N ; (j i + j i) ;
then

1 = h j i = N 2; (h j i + h j i + h j i + h j i) ;
h i
j2 =2
= N 2; 2 + e j ei Im( ) + ei Im( ) ;
j2 =2
= N 2; 2+e j
2 cos Im ( ) ;
1 1
) N ; =p q :
2 1+e j j2 =2 cos Im ( )

e) We want to compute

1 1
j i h j ! dt^ ^y + I^
aj ih ja dt^ ^ j i h j I^
ay a ay a
dt^ ^ ;
2 2

using the results in c) and b) we can write it like

ED
j2 =2 dtj j2 =2
j ih j ! dt ( ) j i h j + e dtj e e dt=2
e dt=2
;
ED ED
dtj j2 =2 dtj j2 =2
dt ( ) e dt=2 e dt=2
+e e e dt=2
e dt=2
;
ED h i ED
2 2
dt ( ) e dt=2 e dt=2
+ 1 dt j j =2 dt j j =2 + O( 2 dt2 ) e dt=2
e dt=2
;
dt ED
2 2 dt=2
1 j j +j j 2 e e dt=2 ;
2
dt ED
2 dt=2 dt=2
1 j j + e e ;
2
ED
j2 =2
ei dt Im( )
e dtj
e dt=2
e dt=2
;

then h i ED
j2 =2
j i h j ! ei dt Im( )
e dtj
e dt=2
e dt=2
;

Using the same procedure as in the last part of c) it is easy to see that for …nite time t the state will
evolve to
h i ED
j i h j ! ei (t) e (t) e t=2 e t=2 ;

where (t); (t), are some real functions of t that we will have to determined: So far we know that (0) = 0;
and (0) = 0: Now, evolve the state at t one more time a small dt we get
h i ED h i ED
2
ei (t) e (t) e t=2 e t=2 ! exp i (t) + dte t Im( ) exp (t) + dte t j j =2 e (t+dt)=2

then we can say that

t
(t + dt) = (t) + dte Im( );
t 2
(t + dt) = (t) + dte j j =2;

or
d (t) t d (t) t 2
= e Im( ); = e j j =2;
dt dt

3
solving this equations we get
t
(t) = Im( ) 1 e ;
2
j j t
(t) = 1 e
2
f) We know that r r
m! m!
= x; = x;
2h 2h
then r 2
2m! j j m! 2
( )= x) = x :
h 2 h
From e) we know that
2
j j m! 2 2
decohere = damp = x :
2 hQ
1 34
Now using, ! = 1 s ; m = 0:001 [kg] ; x = 0:01 [m], h = 10 m2 kg=s , we get

decohere 1018 s 1
:

This is lager than damp by a factor of 1027 :

2.
a)

' (x) = Aeikx + Be ikx

; x < 0;
' (x) = Ceikx + De ikx
; x < 0;

The …rst boundary condition imply

A + B = C + D;
and the second one

ik (A B) = ik(C D) + 2 (A + B);

A B = C D 2i (A + B);
k
) C D = (A B) + 2i (A + B);
k
where = =k; solving these equations we get

C = (1 + i ) A + i B;
D = (1 i ) B i A;

then
C 1+i i A
= ;
D i 1 i B
1+i i
M( ) = :
i 1 i

b) In general
1
a b 1 d b
=
c d a b c a
det
c d

4
since det M = 1 then
1 1 i i
M ( )= :
i 1+i
c) We have that

A 1 i i C 1 i
= = C;
B i 1+i 0 i
C 1 B BC i
) = ; = = :
A 1 i A CA 1 i
Then,
2
C 1
T( ) = = 2
;
A 1+
2 2
B
R( ) = = 2
;
A 1+
and
2
1
R+T = 2
+ 2
= 1;
1+ 1+
as expected.
d) We want
C 1
= ;
A 1 i
to diverge, then 1 i =0) = i= =k = i = ; where k = i then
= :
In this case the normalized bound state solution is given by
p
' (x) = e x; x < 0;
p x
' (x) = e ; x < 0;

3.
a) This is the same procedure that we already did in 2a), you can solve the problem by brute force doing
the same calculations or you can to used the previews results as follow:
First, the general solution is
' (x) = Aeikx + Be ikx
; x < a;
' (x) = Ceikx + De ikx
; a < x < a;
' (x) = Eeikx + F e ikx
; x > a:
notice that we used an overbar in the coe¢ cients. To derive M notice that the only di¤erence with the
problem in 2a) is that the function potential is at x = a, then we only need to shift x ! x + a in 2a)
and relate the coe¢ cients to solve our problem. When we shift x ! x + a we get the following relations:
A = Aeika ; B = Be ika ; C = Ceika ; D = De ika : Then, using M from 2a) and making the substitutions
we get
Ce ika 1+i i Ae ika
= ;
Deika i 1 i Beika
ika ika
e 0 C 1+i i e 0 A
) ika = ;
0 e D i 1 i 0 eika B
C eika 0 1+i i e ika
0 A
) = ika ;
D 0 e i 1 i 0 eika B
C 1+i i ei2ka A
) = ;
D i e i2ka 1 i B

5
then
1+i i ei2ka
M ( ; a) = :
i e i2ka 1 i
To get N ( ; a) just notice that we have to shift now x ! x a; instead of x ! x + a; and (A ! C; B ! D;
D ! F , C ! E) then N ( ; a) = M ( ; a) ; this is

E C
= N ( ; a) ;
F D
1+i i e i2ka
N ( ; a) = :
i ei2ka 1 i

The inverse matrices are given by (since det M = 1 = det N )

1 1 i i ei2ka
M ( ; a) = i2ka ;
i e 1+i
1 1 i i e i2ka
N ( ; a) = :
i ei2ka 1+i

b) We have that

A 1 C
= M ;
B D
1 1 E
= M N ;
0
1 i i ei2ka 1 i i e i2ka E
= i2ka ;
i e 1+i i ei2ka 1+i F
1+ 2i + ( 1 + e4ika ) 2i ( cos (2ka) + sin (2ka)) E
= ;
2i (cos (2ka) sin (2ka)) 1 + 2i + ( 1 + e 4ika ) 0
1+ 2i + ( 1 + e4ika )
= E:
2i (cos (2ka) sin (2ka))

Then
B
= 2i (cos (2ka) sin (2ka)) ;
E
2
B R 2 2
) = =4 (cos (2ka) sin (2ka)) :
E T
Now, because R + T = 1;
R 1 T 1
= = 1;
T T T
1 R
) = + 1;
T T
2
= 4 2 (cos (2ka) sin (2ka)) + 1:

c) First, let’s simplify the result for 1=T:

p 1
cos (2ak) sin (2ak) = 1+ 2 p cos 2ka p sin 2ka ;
1+ 2 1+ 2
p
= 1+ 2 [cos cos 2ka sin sin 2ka] ;
p
= 1+ 2 cos (2ka + ) ;

6
where tan = : Then,
1 2 2
=1+4 1+ cos2 (2ka + ) :
T
From this simplify expression we get
1
Tmax ) min ) cos2 (2ka + ) = 0;
T
) 2ka + = (2n + 1) ; n = 0; 1; 2; 3 : : : ;
2
1 h i
) amax = (2n + 1) arctan ;
2k 2
and
1
Tmin ) max ) cos2 (2ka + ) = 1;
T
) 2ka + = n ; n = 0; 1; 2; 3 : : : ;
1
) amin = [n arctan ] :
2k