Solutions For The Exercices (J J Sakurai)

Part II Solutions
23

1. FUNDAMENTAL CONCEPTS

25

1 Fundamental Concepts

1.1 Consider a ket space spanned by the eigenkets {Ia'n of a Hermitian operator A. There is no degeneracy.

(a) Prove that

II(A - a')

a'

is a null operator.

(b) What is the significance of

(A - all)?

II a' - a" . a"#a'

(c) Illustrate (a) and (b) using A set equal to Sz of a spin ~ system.

(a) Assume that la) is an arbitrary state keto Then

II(A - a')la)

a'

II (A - a') L la") (a"la) = L Call II (A - a') la")

a' a" "-v-' a" a'

L Call II(a" - a') I a") a"E{a'} o.

a" a'

(1.1 )

(b) Again for an arbitrary state la) we will have

[ II (A - ali)] '" la"')(a'" la)

a' - a" Z::

a"#a' a'"

1

(", ")

'" (a"'la) II a - ala"') =

Z:: a' - a"

a'" a"#a'

L(a"'la)6all'a,la"') = (a'la)la') =?

a"'

[(A - ali)] =

II a' - a" a"#a'

la')(a'i Aa,.

(1.2)

So it projects to the eigenket la').

26

(c) It is Sz h/2(1+)(+I-I-)(-I). This operator has eigenkets 1+) and 1-) with eigenvalues h/2 and -t.r: respectively. So

II(Sz - a') - II(Sz-a'l)
-
a' a'
- [~(I+)(+I-I-)(-I) - ~(I+)(+I + 1-)(-1)]
-
X [~(I+)(+I-I-)(-I) + ~(I+)(+I + 1-)(-1)]
0
~
- [-hl-)(-IHhl+)(+IJ = _h21_) (-1+)(+1 = 0, (1.3)
- where we have used that 1+)(+1 + 1-)(-1 = 1.

For a' = h/2 we have

(Sz - all)

II a' - a" a"f:a'

II (Sz-aIl1) = Sz+~l

a"f:n/2 h /2 - a" h/2 + h /2

- ;, [~(I+)(+I- HH) + ~(I+)(+I + 1-)(-1)]

1

- hhl+)(+1 = 1+)(+1·

(1.4)

Similarly for a' = -t.tz we have

(Sz - all)

II a' - a" a"f:a'

(Sz-aIl1) Sz-~l

- a"En/2 -h/2 - a" - -h/2 - h/2

- -;, [~(I+)(+I-I-)(-Il - ~(I+)(+I + I-)(-Il]

1

- - h (-hl-)(-I) = 1-)(-1· (1.5)

1.2 A spin ~ system is known to be in an eigenstate of S . it with eigenvalue h/2, where it is a unit vector lying in the xz-plane that makes an angle I with the positive z-axis.

(a) Suppose Sx is measured. What is the probability of getting +h/2?


27

(b) Evaluate the dispersion in Sx, that is,

(For your own peace of mind check your answers for the special cases, = 0, tt /2, and 7r.)

Since the unit vector n makes an angle, with the positive z-axis and is lying in the xz-plane, it can be written in the following way

(1.6)

So

5 . n S, cos, + S; sin, = [(S-1.3.36),(S-1.4.18)]

[~ (1+)(+1-1-)(-1)] cos! + [~ (1+)(-1 + 1-)(+1)] siw)('1.7)

Since the system is in an eigenstate of 5 . n with eigenvalue 17,/2 it has to satisfay the following equation

5 . nl5 . n; +) = 17,/215 . n; +).

From (1.7) we have that

(1.8)

5. n . 17, ( c~s,

2 SIn,

sin v )

- cos,

(1.9)

The eigenvalues and eigenfuncions of this operator can be found if one solves the secular equation

det(5. n - )..I) = 0 ::::}- det ( 17,/2 cos.' - ).. 17,/2 sin, ) = 0 ::::}-

17,/2 SIn, -17,/2 cos, - )..

17,2 2 2 17,2. 2 2 17,2 17, ()

- 4 cos ,+).. - 4 SIn ,= 0 ::::}-).. - 4 = 0 ::::}- ).. = ± 2 . 1.10

Since the system is in the eigenstate IS· n; +) (~) we will have that

~ ( :~~ ~ ~~~71) ( ~ ) = ~ ( ~ ) * {

1 - cos, 2 sin 2 ~ ,

b = a = a - a tan

sin, 2 sin ~ cos ~ - "2.

a cos, + b sin, = a a sin, - b cos, = b

}*

(1.11)

28

But we want also the eigenstate IS . n; +) to be normalized, that is

where the real positive convention has been used in the last step. This means that the state in which the system is in, is given in terms of the eigenstates of the Sz operator by

IS . n; +) = cos 'I +) + sin '1-)·

2 2

(1.13)

(a) From (8-1.4.17) we know that

(1.14)

80 the propability of getting +Ii /2 when Sx is measured is given by

(_1 (+1 + _1 (_I) (cos '1+) + sin '1_)) 2

V2 V2 2 2

2

1 , 1.,

- cos - + - SIn -

V2 2 V2 2

1 2' 1.2, ,.,

- cos - + - sm - + cos - sIn -

2 2 2 2 2 2

1 + 1 . 1 (1 +. )

"2 "2 sin v = "2 sm, .

(1.15)

For, = 0 which means that the system is in the ISz; +) eigenstate we have

(1.16)

For, = tt /2 which means that the system is in the ISx; +) eigenstate we have

(1.17)

For, = tt which means that the system is in the ISz; -) eigenstate we have

(1.18)


29

(b) We have that
( ( 5 x - (5 x) ) 2) = (5;) - (( 5 x) ) 2 . (1.19)
As we know
n
5x - 2 (1+)(-1 + 1-)(+1) =*
-
52 n2
- 4 (1+)(-1 + 1-)(+1) (1+)(-1 + 1-)(+1) =*
-
x
52 n2 n2 (1.20)
- - (1+)(+1 + 1-)(-1) = -.
-
x 4 4
1 So
(5x) -
-
-
-
((5x) ? -
-
(5;) -
- [cos ~ ( + 1 + sin ~ ( -I] ~ (I + ) (-I + 1-) ( + 1 ) [cos ~ 1 +) + sin ~ 1- ) ]

ti ,., n., , n.

- cos - SIn - + - S'l Tl - cos - = - sm, =*

2222222

n2

- sin2, and 4

[, .'] n2 [, .']

cos 2 ( + 1 + sin 2 (-I 4 cos 2 1 +) + SIn 21- )

_ n2 [cos2 , + sin2 '] = n2•

4 2 2 4

(1.21)

So substituting in (1.19) we will have

and finally
(( ~5x )2)I'=o;lsz;+) n2 (1.23)
-
- 4'
(( /s.S; )2)1'=11"/2;lsx;+) - 0, (1.24 )
-
(( ~5x)2 )1'=o;ISz;-) n2 (1.25)
- -
-
4 30

1.3 (a) The simplest way to derive the Schwarz inequality goes as follows. First observe

(\al + A*\pl)· (Ia) + Alp)) > 0

for any complex number A; then choose A in such a way that the preceding inequality reduces to the Schwarz inequility.

(b) Show that the equility sign in the generalized uncertainty relation holds if the state in question satisfies

~Ala) = A~Bla)

with A purely imaginary.

(c) Explicit calculations using the usual rules of wave mechanics show that the wave function for a Gaussian wave packet given by

satisfies the uncertainty relation

Prove that the requirement

(x/l~xla) = (imaginary number)(x/l~pla)

is indeed satisfied for such a Gaussian wave packet, in agreement with (b).

(a) We know that for an arbitrary state Ie) the following relation holds

(ele) > o.

(1.26)

This means that if we choose Ie) = la) + Alp) where A is a complex number, we will have

((al + A*(pl)· (Ia) + Alp)) > 0 =? (ala) + A\alp) + A*(pla) + IAI2(plp) > o.

(1.27) (1.28)


31

If we now choose A = -(pia) / (pip) the previous relation will be / ala) - (pla)(alp) _ (pla)(alp) + l(plaW > 0 ::::;,.

\ (pip) (pip) (plp)-

(ala)(plp) > l(pla)12. (1.29)

Notice that the equality sign in the last relation holds when

Ie) = la) + Alp) = 0 ::::;,. la) = -Alp)

(1.30)

that is if la) and 113) are colinear. (b) The uncertainty relation is

(1.31)

To prove this relation we use the Schwarz inequality (1.29) for the vectors la) = ~Ala) and 113) = ~Bla) which gives

(1.32)

The equality sign in this relation holds according to (1.30) when

~Ala) = A~Bla).

(1.33)

On the other hand the right-hand side of (1.32) is

1(~A~B)12 = ~ I([A, B])12 + ~ I({~A, ~B})12 (1.34)

4 4

which means that the equality sign in the uncertainty relation (1.31) holds if

~ I({~A, ~B})12 = 0 ::::;,. ({~A, ~B}) = 0 4

::::;,. (al~A~B + ~B~Ala) = 0 (1J.3) A*(al(~B?la) + A(al(~B)2Ia) = 0

::::;,. (A + A*)(al(~B?la) = O. (1.35)

Thus the equality sign in the uncertainty relation holds when

~Ala) = A~Bla)

(1.36)

with A purely imaginary.

32

(c) We have (x'l~xla)

(x'l(x - (x))la) = x'(x'la) - (x)(x'la) (x' - (x) )(x'la).

(1.37)

On the other hand

(x'l~pla)

(x'i (p - (p)) la)

-ihaa (x'la) - (p)(x'la)

x'

(1.38)

But

aa (x'la)

x'

/ x'la) _i_ [i(P)X' _ (x' - (x) )2]

\ ax' h 4d2

(x'la) [i~) - 2~2 (X' - (X))]

(1.39)

So substituting in (1.38) we have

(x'l~pla)

(p)(x'la) + ;:2 (x' - (x)) (x'la) - (p)(x'la) ;:2 (x' - (x)) (x'la) = ;:2 (x'l~xla) :::}

-i2d2

h (x'l~pla). (1.40)

(x'l~xla)

1.4 (a) Let x and Px be the coordinate and linear momentum in one dimension. Evaluate the classical Poisson bracket

[x, F(Px)]classical .

(b) Let x and Px be the corresponding quantum-mechanical operators this time. Evaluate the commutator

( c) Using the result obtained in (b), prove that

(Zpxa) ,

exp h Ix),

( x I x') = x'I x') )


33

is an eigenstate of the coordinate operator x. What is the corresponding eigenvalue?

(a) We have

[x, F(Px)]classical

8x 8px 8px 8x

8F(px)

8px .

(1.41)

(b) When x and Px are treated as quantum-mechanical operators we have

[ ~(ia)np~l=~]_(ia)n[ n]

x,~ tn , ~,tn x,px

n=O rc n. n=O n. Ii

~ 1 (ia)n ~ k [ ] n-k-l

~-,~ z:». x,px v;

n=O n. k=O

~ ]_ (ia )n (ih) ~ k n-k-l = ~ !!_ (ia )n-l n-l (-a)

c: ,tn ~PxPx ~, tn-l Px

n=l n. Ii k=O n=l n. rt.

00 1 (ia )n-l (ipxa)

-a];(n_1)! r;Px =-aexp ------,;:-. (1.42)

(c) We have now

x [exp (ip~a)] Ix') (b) exp (ip~a) xix') _ aexp (ip~a) Ix') x' exp (ip~a) Ix') _ aexp (ip~a) Ix')

(x' - a) exp (ip~a) Ix'). (1.43)

So exp (iP'ha) Ix') is an eigenstate of the operator x with eigenvalue x' - a. So we can write

lx' - a) = C exp (ip~a) Ix'),

(1.44 )

where C is a constant which due to normalization can be taken to be 1.

34

1.5 (a) Prove the following:

( i) (p' I x I a)

(ii) (plxla)

where ¢a(P') = (p'la) and ¢(3(p') = (p'lp) are momentum-space wave functions.

(b) What is the physical significance of

( ix'::)

exp 11, ....... ,

where x is the position operator and :=: is some number with the dimension of momentum? Justify your answer.

(a) We have (i)

1

(p' Ix la)

(p'lx J dx'lx')(x'l a) = J dx'(p'lxlx')(x'la)

J dx' x' (p'lx')(x'la) (5-1.7.32) J dx' x' Ae- i<xl (x'la)

;1 .f dx' a: (e- ;P~"') (iii.) (x'la) = iii. :p' [J dx' ;1 «: ;P~"' (x'la) 1

iii a: [.f dx' (p'lx')(x'la) 1 = iii a~' (p'la) =}

ihaa (p'la). (1.45)

p'

(ii)

(Pix la) = .f lip' (Pip') (p' Ix la) = .f dp'rI/;' (p')ili a~,rPa (p'), (1.46)

where we have used (1.45) and that (pip') = ¢~(p') and (p'la) = ¢a(P').


35

(b) The operator exp (i~:=:) gives translation in momentum space. This can be justified by calculating the following operator

[p, exp C~3) 1 = [p, ~ ~l;3rl = ~~! c~r [p, xn]

~ ]_ (i'3)n.;:-.. n-k[ ] k-l

L...J , 11, L...J x p, x X

n=l n. k=l

CXl 1 (.~)n n CXl 1 (.~)n

L -, z=" L( -ih)Xn-1 = L -, z=.. n( -ih)Xn-1

n=l n. 11, k=l n=l n. 11,

t 1 ,(i'3)n-l xn-1( -ih) (i'3) = '3 t ~ (ix'3)n

n=1(n-1).h 11, n=on. 11,

~ (ix'3)

z.exp T .

(1.47)

So when this commutator acts on an eigenstate Ip') of the momentum operator we will have

p [exp (i~'3) Ipl)] _ [exp (i~'3)] p/lp') :::} P [exp (i~'3) Ipl)] _ p' [exp (i~'3)] Ip') :::}

(p' + '3) [exp (i~'3) Ipl)] . (1.48)

Thus we have that

(ix'3) I ') I I ~)

exp T p Ap+=..,

(1.49)

where A is a constant which due to normalization can be taken to be 1.

36

2 Quantum Dynamics

2.1 Consider the spin-procession problem discussed in section 2.1 in Jackson. It can also be solved in the Heisenberg picture. Using the Hamiltonian

H = - (~) s, = »s.,

write the Heisenberg equations of motion for the time-dependent operators Sx(t), Sy(t), and Sz(t). Solve them to obtain Sx,y,z as functions of time.

Let us first prove the following

(2.1 )

Indeed we have

[UtAsU,UtBsU] = UtAsBsU - UtBsAsU u' [As, Bs]U = u'c,u = CR.

(2.2)

The Heisenberg equation of motion gives

ss, dt as,

dt .is,

dt

1 1 (S-1.4.20) w .

in [Sx, H] = in [Sx, wSz] = in (-znSy) = -wSy, (2.3)

1 1 (S-1.4.20) w .

in [Sy, H] = in [Sy, wSz] = in (znSx) = »s; (2.4)

1 1 (S-1.4.20)

in [Sz, H] = in [Sz, wSz] = 0 ::::;:. S, = constant. (2.5)

Differentiating once more eqs. (2.3) and (2.4) we get

d2Sx dS

y (2.4) 2 ( ) . ( )

dt2 -WTt = -w Sx::::;:' Sx t = Acoswt + BSlnwt::::;:. Sx 0 = A

d2 Sy as, (2.3) 2 ( ) . ( )

dt2 wTt = -w Sy ::::;:. Sy t = C coswt + D sin cor ::::;:. Sy 0 = C.

But on the other hand

ss, dt -Awsinwt + Bwcoswt

A=D

-wSy::::;:'

-Cw cos wt - Dw sin wt ::::;:.

C= -B.

(2.6)

2. QUANTUM DYNAMICS

So, finally

37

Sx(O) coswt - Sy(O) sinwt Sy(O) coswt + Sx(O) sinwt Sz(O).

(2.7) (2.8) (2.9)

2.2 Let x(t) be the coordinate operator for a free particle in one dimension in the Heisenberg picture. Evaluate

[x(t), x(O)] .

The Hamiltonian for a free particle in one dimension is given by p2

H=-

2m

(2.10)

This means that the Heisenberg equations of motion for the operators x and p will be

ap(t) at p(t) ax(t) at

x( t)

Thus finally

i~' [P(t), H(t)] = i~' [p(t), p~~:)] = 0 =?

p(O) (2.11)

;_[x, H] = ;_ [x(t), p2(t)] = 1. 2p(t)iti = p(t) (2.11) p(O) =?

ili ili 2m 2mzti m m

t

-p(O) + x(O). (2.12)

m

[ t ] t itit

[x(t), x(O)] = -p(O) + x(O), x(O) = - [p(O), x(O)] = --.

m m m

(2.13)

2.3 Consider a particle in three dimensions whose Hamiltonian is given by

:c2

H = _E_ + V(i). 2m

38

By calculating [x· p, H] obtain

!:_\i. if> = (p2) - \x· VV).

dt m

To identify the preceding relation with the quantum-mechanical analogue of the virial theorem it is essential that the left-hand side vanish. Under what condition would this happen?

Let us first calculate the commutator [i· p, H]

Ix· p, H] = [x. p, ~, + V(X)] = [t. XiPi, t. :!" + V(X)]

2:= [Xi, :,~] Pi + L XdPi' V(i)]. (2.14)

~J ~

The first commutator in (2.14) will give

1 2 1 1 . .

2m [Xi,Pj] = 2m (pj[Xi,pj] + [Xi,pj]pj) = 2m (PjZhOij + ZhOijPj)

1 .t.\ ih .\ ( )

-2ZIiUijPJ = -UiJPj. 2.15

2m m

The second commutator can be calculated if we Taylor expand the function V(x) in terms of Xi which means that we take V(i) = Ln anxi with an independent of Xi. So

[ p2]

X. _J

z, 2m

[pi, V(i)] = [Pi, f= anxi] = L an [Pi, xi] = L an I: x7 [Pi, Xi] Xi-k-1

n=O n n k=O

n-l a

Lan L ( -ih )xi-1 = -ih L annxi-1 = -ih-a L anxi

n k=O n Xi n

-ih~ V(x). (2.16)

aXi

The right-hand side of (2.14) now becomes

[i· p, H] = L ih OijPjPi + L( -ih)Xi~V(i)

.. m . aXi

ZJ Z

ih -Q ·t -. nv(-')

-P - ZltX· v x.

m

(2.17)

2. QUANTUM DYNAMICS

39

The Heisenberg equation of motion gives

d -> ->

-x·p dt

!:_(x.j5) dt

1 [-> -> H] (2_27) r -> nv(-»

~ X· p, - - - x . v X::::}

ui m

/ p2) ->

\m - (x· VV),

(2.18)

where in the last step we used the fact that the state kets in the Heisenberg picture are independent of time.

The left-hand side of the last equation vanishes for a stationary state.

Indeed we have

! (nix· pin) = i~ (nl [x· p, H] In) = i~ (En(nlx· pin) - En(nlx· pin)) = o.

So to have the quantum-mechanical analogue of the virial theorem we can take the expectation values with respect to a stationaru state.

2.4 (a) Write down the wave function (in coordinate space) for the state

You may use

(-zpa)

exp =t: 10).

(/10) - -1/4 -1/2 [1 (XI)2]

X - 7r Xo exp - 2 - ,

Xo

(b) Obtain a simple expression that the probability that the state is found in the ground state at t = O. Does this probability change for t > 07

(a) We have

la, t = 0)

(x'la, t = 0)

(-zpa)

exp -h- 10)::::}

(x' exp (-t,a) 10) (Pr.1.4.c) (x' - alO)

-1/4 -1/2 [1 (Xl - a) 2]

7r Xo exp -2 .

Xo

(2.19)

40

(b) This probability is given by the expression

I (0 I a, t = 0) 12 = I (exp (- ~a) 10) 12.

(2.20)

It is

(-zpa)

(exp =r: 10)

J dx'(Olx')(x'l exp (-~a) 10)

J d ' -1/4 -1/2 [1 (x') 2] -1/4 -1/2

X 'IT Xo exp -"2 - 'IT Xo

Xo

X exp [ - ~ ( X' x~ a ) 2]

.f dx',,-1/2xr;' exp [ - 2:6 (X'2 + xl2 + a2 - 2ax') 1

1 J' [ 2 ('2 , a a 2 a 2 ) 1

dx exp - - x - 2x - + - + -

ftxo 2x6 2 4 4

( a2) 1 ( a2)

exp --2 ~xo = exp --2 .

4xo ftxo 4xo

(2.21 )

So

I(Ola, t = 0)12 = exp (-~) .

2xo

(2.22)

For t > 0

2 ( iHt) 2

I(OIU(t)la, t = 0)1 = 1(01 exp -T la, t = 0)1

le-iEot/h(Ola, t = 0)12 = I(Ola, t = 0)12. (2.23)

2.5 Consider a function, known as the correlation junction, defined by

C(t) = (x(t)x(O)),

(2.24)

where x(t) is the position operator in the Heisenberg picture. Evaluate the correlation function explicitly for the ground state of a one-dimensional simple harmonic oscillator.

2. QUANTUM DYNAMICS

41

The Hamiltonian for a one-dimensional harmonic oscillator is given by

p2(t)

H = -- + lmw2x2(t). 2m 2

So the Heisenberg equations of motion will give

(2.25)

dp(t)

--

dt

1 1 [ p2 ( t ) 1 2 2 1

in [x(t), H] = in x(t), 2m + 2mw x (t)

2~in [X(t),p2(t)] + ~mw2 i~ [x(t), x2(t)]

~in p( t) = p( t)

2'l,nm m

;_ [p(t), H] = ;_ [p(t), p2(t) + ~mw2x2(t)l

'l,n 'l,n 2m

2 2

mw [ 2] mw. 2

2in p(t), x (t) = 2in [-2'l,nx(t)] = -mw x(t).

(2.26)

dx(t) dt

(2.27)

Differentiating once more the equations (2.26) and (2.27) we get

d2x(t) dt2 d2p( t)

dt2

1 dp( t) (2.27) 2 .

--d- = -w x(t) =? x(t) = Acoswt + B sin coz =? x(O) = A m t

1 dx(t) (2.26) 2

- d = -w p(t) =? p(t) = C coswt + D sinwt =? p(O) = C.

m t

But on the other hand from (2.26) we have

dx(t) dt

p(t) -=? m

p(O) D .

- coswt + - sin cor =?

m m

-wx(O) sin wt + Bw cos wt

B = p(O)

mw

D = -mwx(O).

(2.28)

So

p(O) .

x(t) = x(O) coswt + - sin coz mw

(2.29)

and the correlation function will be

C(t) = (x(t)x(O)) (2.29) (X2(0)) coswt + (p(O)x(o))_l- sinwt. (2.30)

mw

42

Since we are interested in the ground state the expectation values appearing in the last relation will be

1i 1i 1i

(Ol-(a + at)(a + at)IO) = -(OlaatIO) = ~2.31)

2mw 2mw 2mw

.~1iW{I;

i - -(Ol(at - a)(a + at)IO)

2 2mw

.1i/ I tl) .1i

-z-\O aa 0 = -z-.

2 2

(2.32)

(p(O)x(O))

Thus

C ( ) 1i . 1i. 1i -iwt

t = -- coswt - z-- sin cor = --e .

2mw 2mw 2mw

(2.33)

2.6 Consider a one-dimensional simple harmonic oscillator. Do the following algebraically, that is, without using wave functions.

(a) Construct a linear combination of 10) and 11) such that (x) is as large as possible.

(b) Suppose the oscillator is in the state constructed in (a) at t = o. What is the state vector for t > 0 in the Schrbdirigcr picture? Evaluate the expectation value (x) as a function of time for t > 0 using (i) the Schrodinger picture and (ii) the Heisenberg picture.

(c) Evaluate (( ~x)2) as a function of time using either picture.

(a) We want to find a state 10:) = colO) + clll) such that (x) is as large as possible. The state 10:) should be normalized. This means

(2.34)

We can write the constands Co and Cl in the following form

Co

(2.35)

2. QUANTUM DYNAMICS

43

The average (x) in a one-dimensional simple harmonic oscillator is given

by

(2.36)

where we have used that x = ~(a + at).

What we need is to find the values of leo I and 01 - 00 that make the average (x) as large as possible.

But for (x) maximum we want also

< 0 ::::;:. ti = 2k,

k E Z.

(2.39)

So we can write that

10:) = ei8o_1 10) + ei(8o+2k1r)_1 11) = ei8o_1 (10) + 11)). (2.40)

V2 V2 V2

We can always take 00 = o. Thus

1

10:) = V2(IO) + 11)).

(2.41 )

44

(b) We have la, to) = la). So

U(t, to = O)la, to) = e-iIIt/hla) = ~e-iEot/hIO) + ~e-iE't/hll) ~ (e-iWI/210) + e-iW31/211)) = ~e-iWI12 (10) + e-iWIII)) .(2.42) (i) In the Schrodinger picture

la, to; t)

(X)s (a, to; tlxsla, to; t)s

[~ (eiwt/2(01 + eiw3t/2(11)] x [~ (e-iwt/210) + e-iw3t/211))] ~ei(wt/2-w3t/2)(0Ix 11) + ~ei(W3t/2-wt/2) (11x 10)

. {I; . {I; {I;

le-zwt -- + lezwt -- = -- cos wt.

2 2mw 2 2mw 2mw

(2.43)

(ii) In the Heisenberg picture we have from (2.29) that XH(t) = x(O) cos wt + p(O) sinwt.

mw

So

(X)H (alxHla)

[ ~ (0 I + ~ (II] (x (0) cos wt+ ~:~ sin Wi) [~ 10) + ~ 11 ) ]

~ coswt(Olxll) + ~ coswt(llxIO) + ~_1_ sinwt(Olpll)

mw

1 .

+~- slnwt(llpIO)

mw

l{I; l{I; 1 ()JmhW

- -- coswt + - -- coswt + -- sinwt -i --

2 2mw 2 2mw 2mw 2

I . .JmhW +--Slllwtz --

2mw 2

.;;;!f; cos cot,

(2.44)

( c) It is known that

(2.45)

2. QUANTUM DYNAMICS

45

In the Schodinger picture we have

which means that

(x)~ (a, to; tlx2la, to; t)s

[~ (eiWI/2(OI + eiW31/2(11) 1 x2 [~ (e-iWI/210) + e-iw3I/2 I 1)) 1

[lei(wt/2-wt/2)(0IaatI0) + lei(w3t/2-w3t/2)(1Iaatll) + l(llatall)] _fi_ 2 2 2 2mw

[1 1 1] fi fi

- + 2- + - -- = --

2 2 2 2mw 2mw·

(2.4 7)

So

/ ( ) 2 ) ( 2 .43) fi fi 2 fi. 2

\ ~x S = -- - --cos wt = --SIn wt.

2mw 2mw 2mw

(2.48)

In the Heisenberg picture

x'j,(t) = [x(O) coswt + ~~ SillWtJ'

p2 (0) .

X2(0) cos2 wt + 2 2 sin ' wt m w

x(O)p(O) . p(O)x(O) .

+ COS wt smwt + COS wt smwt

mw mw

fi 2 t2 t t 2

-- (a + a + aa + a a) COS wt

2mw

mluo 2 t2 t t 2

- 2 2 (a + a - aa - a a) sin wt

2m w

i Jhmhw( t)( t )sin2wt

+- a+a a -a ---

mw 4mw 2

i Jt.mt.w (t )( t) sin 2wt

+- a - a a+ a

mw 4mw 2

Ii 2 t2 t t 2

-- (a + a + aa + a a) COS wt

2mw

46

- _n_( a2 + at2 _ aat _ at a) sin2 wt + ___!!_:_( at2 - a2) sin 2wt

2mw 2mw

Ii t t Ii 2 Ii t2

--(aa + a a) + --a cos 2wt + --a cos 2wt

2mw 2mw 2mw

in t2 2

+--( a - a ) sin 2wt, (2.49)

2mw

which means that

(alx~la)H

2::W [~(Ol + ~(lll

X [aat + ata + a2cos2wt + at2 cos2wt + i(at2 - a2)sin2L

(2.

So

( ( ) 2 ) (2.44) n n 2 n. 2

~x H = -- - --cos wt = --sm wt.

2mw 2mw 2mw

(2.51 )

2.7 A coherent state of a one-dimensional simple harmonic oscillator is defined to be an eigenstate of the (non-Hermitian) annihilation operator a:

al).) = ).1).),

where). is, in general, a complex number.

(a) Prove that

is a normalized coherent state.

(b) Prove the minimum uncertainty relation for such a state.

2. QUANTUM DYNAMICS

47

(c) Write I).) as

00

I).) = L f(n)ln).

n=O

Show that the distribution of If(n)12 with respect to ti is of the Poisson form. Find the most probable value of n., hence of E.

(d) Show that a coherent state can also be obtained by applying the translation (finite-displacement) operator e-ipl/n (where p is the momentum operator, and l is the displacement distance) to the ground state.

( e) Show that the coherent state I).) remains coherent under timeevolution and calculate the time-evolved state I).(t)). (Hint: directly apply the time-evolution operator.)

(a) We have

(2.52)

since alO) = o. The commutator is

[a, eAa'] = [a, ~ ~! (Aat)" 1 = ~ ~! An [a, (at)"]

f= ~).n t(at)k-l [a, at] (att-k = f= ~).n t(att-1

n=l n. k=l n=l n. k=l

~ (n 21)!An(at)"-' = A ~ ~!(Aat)n = AeAa'. (2.53)

So from (2.52)

(2.54)

which means that I).) is a coherent state. If it is normalized, it should satisfy also ().I).) = 1. Indeed

().I).) = (0IeA*ae-1AI2 eAat 10) = e-1A12 (OleA*aeAat 10)

48

(2.55)

(b) According to problem (1.3) the state should satisfy the following relation

~xl).) = c~pl).),

(2.56)

where ~x x - ().Ixl).), ~p p - ().Ipl).) and c is a purely imaginary

number.

Since I).) is a coherent state we have

(2.57)

U sing this relation we can write

and

(x) - ().Ixl).) = J Ii ().I(a + at)I).) = J Ii (().Ial).) + ().Iatl).))

2mw 2mw

- {;!f(). + ).') (2.59)

and so

I'.xl).) = (x - (x))I).) = J Ii (at - ).')1).).

2mw

(2.60)

Similarly for the momentum p = iff(at - a) we have

r~ ~

pi).) = V zy ~-2-(at - a)I).) = iy ~-2-(at - ).)1).)

(2.61 )

2. QUANTUM DYNAMICS

49

and

and so

(p - (p)ll-') = iJm;w (at - -"ll-') '*

.:

-z -~pIA).

mluo

(2.63)

So using the last relation in (2.60)

I'lxl-') = J Ii (-ilV : I'lpl-') =

2mw mruo

i

(2.64)

mw "-v--"

purely imaginary

and thus the minimum uncertainty condition is satisfied.

(c) The coherent state can be expressed as a superposition of energy eigenstates

oo oc

IA) = L In)(nIA) = L f(n)ln).

(2.65)

n=O

n=O

for the expansion coefficients f(n) we have

f(n) (nIA) = (nle-IAI2/2eAatI0) = e-IAI2/2(nleAatI0)

e-IAI2/2(nl f= ~(Aat)mIO) = e-IAI2/2 f= ~Am(nl(at)mIO)

m=O m. m=O m.

If(n)12 =

(2.66)

(2.67)

which means that the distribution of If(n)12 with respect to n is of the Poisson type about some mean value n = IAI2.

50

The most probable value of n is given by the maximum of the distribution

If(n)12 which can be found in the following way

If(n + 1)12 (IAI2)n~1 exp( _1.-\12) 1.-\12

____ = (n+l). = > 1

If(n)12 (IAl~)n exp( _1.-\12) ti + 1 -

(2.68)

which means that the most probable value of n is 1.-\12.

(d) We should check if the state exp (-ipljn) 10) is an eigenstate of the annihilation operator a. We have

aexp (-ipljn) 10) = [a, e(-iPI/h)] 10)

(2.69)

since alO) = O. For the commutator in the last relation we have

t ~ (_il)n [a,pn] = t ~ (_il)n tpk-l[a,p]pn-k

n=O n. n n=l n. n k=l

t ~ (_il)n t pn-liJrnhW

n=l n. Ii k=l 2

i ~ (-il) t 1 (_ilP)n-l

V~ n n=1(n-1)! n

l rmw e(-ipl/h)

V2f '

(2.70)

where we have used that

[a,p] = iff[a, at - a] = iln~w.

So substituting (2.70) in (2.69) we get

a [exp (-ipljn) 10)] = l~ [exp (-ipljn) 10)]

(2.71 )

(2.72)

which means that the state exp ( -ipl jn) 10) is a coherent state with eigenvalue l j!fi.

(e) Using the hint we have

2. QUANTUM DYNAMICS

51

oc (>. -iwt)n

-iwt/2,"""" _1>.e-iwtI2/2 e 1 ) (2.66) -iwt/21 \ -iwt)

e L...J e q n = e /\e

n=O yn:

Thus

al>.(t))

(2.73)

e-iwt/2al>'e-iwt) = >'e-iwte-iwt/21>'e-iwt)

>'e-iwt 1>'( t)).

(2.74)

2.8 The quntum mechanical propagator, for a particle with mass m, moving in a potential is given by:

T( .) -1= d iEt/ti .T( . ) _ A'"""" sin( nix) sin( n,y)

K x,y,E - te 11 x,y,t,O - L...J ti22

o E - _r n2

n 2m

where A is a constant.

(a) What is the potential?

(b) Determine the constant A in terms of the parameters describing the system (such as m, I etc. ).

We have

«i., y; E)

10= dteiEt/ti 1« x, y; t, 0) 10= dteiEt/ti(x, t Iy, 0)

10= dteiEt/ti(xle-iHt/tily)

1= dteiEt/ti L(xle-iHt/tiln)(nly)

o n

1= dteiEt/ti L e-iEnt/ti(xln)(nly)

o n

L cPn( x )cP~(y) 1= ei(E-En)t/tidt

n 0

52

L<Pn(x)<p:(y)lim [ -ih. ei(E-En+ic)t/h]OO

n c-tO E - En + ZE 0

L<Pn(X)<P:(Y)E~hE. (2.75)

n n

So

~-.L. ()-.L.*() ih =A~sin(nlx)sin(nIY)

L...J Y"n X Y"n Y E _ E L...J E _ h2r2 2 ::::}

n n n 2m n

(2.76)

For a one dimensional infinite square well potential with size L the energy eigenvalue En and eigenfunctions <Pn(x) are given by

E = 11,2 ('IT)2 n2•

n 2m L

(2.77)

Comparing with (2.76) we get L = 1 ::::} L = ~ and

v={~

for 0 < x < 21:_

r

otherwise

(2.78)

while

A = 2, ::::} A = i2hl

ih 'IT 'IT .

(2.79)

2.9 Prove the relation

dB ( x) = a (x) dx

where B( x) is the (unit) step function, and 0'( x) the Dirac delta function. (Hint: study the effect on testfunctions.)

For an arbitrary test function f( x) we have

1+00 dB( x) f( x )dx = 1+00 ~ [B( x )f( x)] dx -1+00 B( x) df( x) dx

-00 dx -00 dx -00 dx

2. QUANTUM DYNAMICS

53

d8(x) dx

8(x)j(x) ~: - f= d~l~) dx

lim f(x) - f(x)I+CXl = f(O)

X--++CXl 0

I:CXl 5( x )f( x )dx :::}

5(x ).

(2.80)

2.10 Derive the following expression

mw [2 2 ]

Scl= . ( ) (XO+2XT)cos(wT)-xOXT

2 SIn wT

for the classical action for a harmonic oscillator moving from the point Xo at t = 0 to the point XT at t = T.

The Lagrangian for the one dimensional harmonic oscillator is given by

1'( .) 1 ·2 1 2 2

J..., x, X = 2mx - 2mw x .

(2.81 )

From the Lagrange equation we have

a£ d a£ (2.81) 2 d .

- - --- = 0 :::} -mw x - -(mx) = 0 :::}

ax dt ax dt

•• 2 0

x+wx= .

(2.82)

which is the equation of motion for the system. This can be solved to give

x ( t) = A cos wt + B sin wt

(2.83)

with boundary conditions

x(t = 0) x(t = T)

Xo = A

(2.84)

XT = Xo cos wT + B sin wT :::} B sin wT = XT - Xo cos wT :::}

XT - Xo cos wT B=-----sinwT

(2.85)

54

So

X( t)

XT - Xo coswT .

Xo cos wt +. smwt

sInwT

Xo cos wt sin wT + XT sin wt - Xo cos wT sin wt

sinwT

XT sinwt + Xo sinw(T - t)

. T =?

sin «:

XTW cos wt - XoW cos w(T - t)

sinwT

(2.86)

x(t) =

(2.87)

With these at hand we have

iT dt12(x, x) = iT dt (~mx2 - ~mw2x2)

iT d [ 1 d ( . ) 1 .. 1 2 2]

t - m - xx - -mxx - - mw x

o 2 dt 2 2

-~m rT dtx[x + w2x] + m xxlT

Jo 2 0

; [x(T)x(T) - x(O)x(O)]

S

(2.82)

m [ XTW XoW ]

-. (XT cos wT - xo) -. (XT - Xo cos wT)

2 SIn wT sm wT

~w [x} cos wT - XOXT - XOXT + X~ cos wT]

2 sInwT

~w [( x} + x~) cos wT - 2XOXT] .

2 sInwT

(2.88)

2.11 The Lagrangian of the single harmonic oscillator is

1 '2 1 2 2

£ = -mx - -mw x

2 2

(a) Show that

where Sel is the action along the classical path Xel from (xa, ta) to (Xb' tb) and G is

G(O, tb; 0, ta) =

2. QUANTUM DYNAMICS

55

h - tb-ta

were E - (N+l).

[Hint: Let y(t) = x(t) - Xcl(t) be the new integration variable, Xcl(t) being the solution of the Euler-Lagrange equation.]

(b) Show that G can be written as

(N+l)

G = lim ( ~n) 2 J dYl ... dYNexp( _nT an) N ---too 2?TZ E

where n = [ '" land nT IS its transpose. Write the symmetric

y~

matrix a.

( c) Show that

Jd d (T )-JdN -nTan ?TN/2

Yl . .. YNexp -n an = ne = --~-

vdeta

[Hint: Diagonalize a by an orthogonal matrix.]

(d) Let (2;;;0:) N deio deta~ PN. Define j x j matrices aj that consist of the first j rows and j columns of a~ and whose determinants are Pj. By expanding aj+l in minors show the following recursion formula for the Pj:

j = 1, ... ,N

(2.89)

(e) Let ¢(t) = EPj for t = t; + je and show that (2.89) implies that in the limit E ---+ 0, ¢( t) satisfies the equation

with initial conditions ¢(t = ta) = 0, d¢(~~ta) = 1.

56

(f) Show that

mw {zmw 2 2 }

"tt . ( ) exp h· ( ) [(Xb + Xa) cos(wT) - 2XaXbJ

2HZ SIn wT 2 sin wT

where T = tb - tao

(a) Because at any given point the position kets in the Heisenberg picture form a complete set, it is legitimate to insert the identity operator written as

J dxlxt)(xtl = 1

(2.90)

So

lim J dXldx2 ... dXN(XbtblxNtN)(XNtNlxN-1tN-l) ... X

N-+co

(Xi+lti+llxiti) ... (Xltllxata). (2.91)

It is

( . I -iR(ti+1-td/til .) - ( . I -iRe/til·)

Xz+l e x; - Xz+l e x;

(since e is very small)

(2.92)

For the second term in this last equation we have

2. QUANTUM DYNAMICS

57

(2.93)

Substituting this in (2.92) we get

1

( m ) 2 i [Tn (_ _)2 1 2_]

/ _ i: I -to) - h ~ X,+l-X, -"2cmw x,

\X~+l ~+l X~ i - 2.r_ e

tt u LE

(2.94)

and this into (2.91):

(Xbtblxata) J Dx exp {~ S[x]} =

Let y(t) = x(t) - Xel(t) =? x(t) = y(t) + Xel(t) =? x(t) = y(t) + Xel(t) with boundary conditions y(ta) = y(tb) = 0. For this new integration variable we have Dx = Dy and

S[X] =

So

J { i i ltb [1 ·2 1 2 2] }

Dyexp n Sel + n ta 2my - 2mw y dt

[iSel]

exp T G(O, tb; 0, ta)

(2.95)

with

(N+l) { . N }

. m -2- i m 2 1 2 2

Inn J dYl ... dYN ( ·n) exp t L [-(Yj+l - yj) - -Emw yj] .

N -+= 27rZ E rt. j=O 2E 2

58

where the last step is written in such a form so that the matrix a will be
symmetric. Thus we have
(N+l)
G = lim ( "!n) -2 - J dYl ... dYNexp( _nT an) (2.97)
N ---too 2nz E
with
2 -1 0 0 0 1 0 0 0 0
-1 2 -1 0 0 0 1 0 0 0
m 0 -1 2 0 0 iEmw2 0 0 1 0 0
a=-- + 2.98)
2Ein 2n
0 0 0 2 -1 0 0 0 1 0
0 0 0 -1 2 0 0 0 0 1 (c) We can diagonalize a by a unitary matrix U. Since a is symmetric the following will hold

a = ut aDU ::::;:. aT = UT aD(Ut)T = UT aDU* = a ::::;:. U = U*. (2.99)

So we can diagonalize a by an orthogonal matrix R. So

a = RT aDR and det R = 1 (2.100)

which means that J dNne-nTan

J dNne-nTRTaRn Rn=( J dN(e-ea(

[J «, e-(fa1] [J «: e-(~a2] ... [J d(Ne-(haN]

Ifff;···ff=

nN/2

VTI~l a;

(2.101 )

2. QUANTUM DYNAMICS

59

where a; are the diagonal elements of the matrix oti.
(d) From (2.98) we have
CihE)N d
-- eta =
m
2 -1 0 0 0 1 0 0 0 0
-1 2 -1 0 0 0 1 0 0 0
0 -1 2 0 0 _ E2W2 0 0 1 0 0
det
0 0 0 2 -1 0 0 0 1 0
0 0 0 -1 2 0 0 0 0 1
det a~ PN. (2.102) We define j X j matrices aj that consist of the first j rows and j columns of a~. So

2 - E2W2 -1 0 0 0
-1 2 - c:2W2 0 0 0
0 -1 0 0 0
det aj+1 = det
0 0 2 - c:2W2 -1 0
0 0 -1 2 - c:2W2 -1
0 0 0 -1 2 - c:2W2 From the above it is obvious that

(2 - c:2W2) det aj - det aj_1 ::::;,.

(2 - c:2W2)pj - pj-1 for j = 2,3, ... , N

(2.103)

(e) We have

cjJ(t) cjJ(ta+jc:)

::::;,. cjJ(ta + (j + 1)c:)

::::;,. cjJ( t + e )

C:Pj

C:Pj+1 = (2 - c:2W2)c:pj - C:pj-1

2cjJ(ta + jc:) - c:2w2cjJ(ta + jc:) - cjJ(ta + (j - 1)c:)

2cjJ(t) - c:2W2cjJ(t) - cjJ(t - e ). (2.104)

60

So

¢(t+c)-¢(t) ¢(t)-¢(t-c)

_-----'c"------- __ -------'c'-------_ = _ W 2 cjJ ( t) ::::;,.

E

lim cjJ'( t) - cjJ'( t - E) = _w2cjJ( t) ::::;,. d2cjJ = _w2cjJ( t).

c~o E &2

From (c) we have also that

and

cjJ(ta + E) - cjJ(ta) E(PI - Po)

= = PI - Po

E E

2 - E2W2 - 1 --+ 1.

The general solution to (2.105) is

cjJ(t) = A sin(wt + 5)

and from the boundary conditions (2.106) and (2.107) we have

(2.105)

(2.106)

(2.107)

(2.108)

cjJ(ta) = 0 ::::;,. A sin(wta + 5) = 0 ::::;,. 5 = -wta + nst ti E Z (2.109)

which gives that cjJ(t) = A sinw(t - ta), while

dcjJ , (2.107)

dt = Aw cos(t - ta) ::::;,. cjJ (ta) = Aw ::::;,.

1

Aw=l::::;,.A=-

w

Thus

cjJ(t) = sinw(t - ta).

W

(f) Gathering all the previous results together we get

(2.110)

(2.111)

2. QUANTUM DYNAMICS

61

[ N ]-1/2

1/2 2'17,

C;;r,) J~ c ( ~ £ ) <leta

mw

(2.112)

27rih sin( wT) .

So from (a)

[iScl] ( )

exp T G 0, tb; 0, t;

(2088)

mw [Z mw [2 2 ] ]

017, • ( T) exp h' T (Xb + Xa) cos wT - 2XbXa .

27rZ SIn w 2 SIn w

2.12 Show the composition property

where J<j(X1' t.: Xo, to) is the free propagator (Sakurai 2.5.16), by explicitly performing the integral (i.e. do not use completeness).

We have

62

{ im 1 [X2(tl - to) + XO(t2 - tl)]2}

exp --

2ft (t2 - td(tl - to) (t2 - to)

1

m

[X~(tl - to)(t2 - to) + X6(t2 - td(t2 - to) _ (t2-tO)

X~(tl - to)2 - X6(t2 - td2 - 2X2Xo2(tl - to)(t2 - td]} (t2-tO)

m

. x

2n'lft( t2 - td

{ im [X~(tl - to)(t2 - to - tl + to) + X6(t2 - h)(t2 - to - t2 + td

exp - -

2ft (t2 - to)(t2 - td(tl - to)

2X2Xo2( tl - to) (t2 - td ]}

(t2 - to)(t2 - td(tl - to)

m exp [im(X2 - XO)2]

2nift( t2 - to) 2ft( t2 - to)

J<j(X2' t2; Xo, to).

(2.

2. QUANTUM DYNAMICS

63

2.13 (a) Verify the relation

where IT m! = p - ef and the relation

d2i dIT [_, 1 (dX _, _, dX)]

m dt2 = dt = e E + 2c dt x B - B x dt .

(b) Verify the continuity equation

ap V' -t 0

at + . J =

with; given by

(a) We have

(2.114)

We have also that

64

(2.115)

d2Xi m dt2

d2x m dt2

(b) The time-dependent Schrodinger equation is

ina a (x'la, to; t) = (x'IHla, to; t) = (x'I_1_ (p _ eA) 2 + e<pla, to; t)

t 2m c

1 [ . _, I eA( x') 1 [ . _, I eA( x') 1 / 'I ) ( -;/) / 'I )

2m -zh\l - c . -zn\l - c \x a, to; t + e¢ x \x a, to; t

_1_ [-n2VI . V' + ':_inV' . A( x') + in':_ A( x') . V' + e2 A 2( Xl)] ?jJ( x', t)

2m c c c2

+e¢(xl)?jJ(x', t)

_1 [_n2\l12?jJ(xl, t)' + ':_in (V' . A) ?jJ(X', t) + ':_inA(xl) . VI?jJ(X', t)

2m c c

+ in':_A(x') . VI?jJ(X', t) + e~ A2(XI)?jJ(X', t)] + e¢(xl)?jJ(x', t)

c c

_1 [_n2\l12?jJ + ':_in (V'. A)?jJ + 2in':_A. V'?jJ + e2 A2?jJ] + e¢?jJ. (2.116)

2m c c c2

Multiplying the last equation by ?jJ* we get

in?jJ*!!_?jJ = at

2. QUANTUM DYNAMICS

65

The complex conjugate of this eqution is

-ifL1fJ !l_1fJ* = at

1 [ 2,2 e ( -; , -;) 2 e -; -; , e2 2 2] 2

- -fL 1fJ\l 1fJ* - -ifL \1 . A 11fJ1 - 2ifL-A . 1fJ\l 1fJ* + -A 11fJ1 + e4>11fJ1 .

2m c c c2

Thus subtracting the last two equations we get

- ;~ [4/\7",/) - 4;\7" 4)* 1

+ (~c) ifL (V' . A) 11fJ 12 + (~c) ifLA . (1fJ*V'1fJ + 1fJ V'1fJ*)

ifL (1fJ*!l_1fJ + 1fJ !l_1fJ*) ::::;:.

at at

- ;::, V' . [,j;'V'4) - 4) V'4J * 1 + ('~J in. (V' . 11)1,/) I' + ('~J inA· (V' 14) I')

ifL!l_ 11fJ 12 ::::;:.

at

:t11fJ12 = - ~ V'· [2f(1fJ*V'1fJ)] + (~c) V'· [AI1fJ12] ::::;:.

%tI4)I' + v'· [~ 'S(,j;'v',/)) - c,:J ;\14JI'] = 0 =}

ap -;,-t

at+\l·J=O (2.117)

with] = (!) 2f(1fJ*V'1fJ) - C~J A11fJ12. and p = 11fJ12

2.14 An electron moves in the presence of a uniform magnetic field in the z-direction (B = Bz).

(a) Evaluate

where

eAx

IIx Px --,

c

_ eAy

IIy = py --. c

66

(b) By comparing the Hamiltonian and the commutation relation obtained in (a) with those of the one-dimensional oscillator problem show how we can immediately write the energy eigenvalues as

E = fi2 k2 (leBlfi) (n ~)

k,n 2 + + 2 '

m me

where hk is the continuous eigenvalue of the P» operator and ti is a nonnegative integer including zero.

The magentic field B = Bi: can be derived from a vector petential A(x) of the form

A __ By

x - 2'

A _ Bx y - 2 '

(2.118)

Thus we have

[ _ eAx _ eAy] (2.~8) [ eBy _ eBx]

Px , py - Px + 2 ' py 2

e e e e

eB eB ifieB ifieB

-~[Px,x] + ~[y,py] = ~ + 2e

ifi eB.

e

(2.119)

(b) The Hamiltonian for this system is given by

( -» 2

1 -> eA 1 2 1 2 1 2

H = - P - - = -IIx + -II + -pz = HI + H2

2m e 2m 2m y 2m

(2.120)

h H - 1 II2 + 1 II2 d H - 1 2 S'

were 1 = 2m x 2m Y an 2 = 2mPz, mce

(2.121 )

there exists a set of simultaneous eigenstates Ik, n) of the operators HI and H2. So if lik is the continious eigegenvalue of the operator P» and Ik, n) its eigenstate we will have

(2.122)

2. QUANTUM DYNAMICS

67

On the other hand HI is similar to the Hamiltonian of the one-dimensional oscillator problem which is given by

H = _1_p2 + 1mw2x2

2m 2

(2.123)

with [x, p] = in. In order to use the eigenvalues of the harmonic oscillator En = lu» (n + ~) we should have the same commutator between the squared operators in the Hamiltonian. From (a) we have

. eB (IIxc) .

[IIx, IIy] = zn- =? [ - ,IIy] = ih,

C eB

So HI can be written in the following form

(2.124 )

_1_II2 + _1_II2 = _1_II2 + _1_ (IIxc) 2 leBI2

2m x 2m y 2m y 2m eB c2

_1 II2 + 1m (~) 2 (IIxc) 2

2m y 2 mc eB

(2.125)

In this form it is obvious that we can replace w with leEI to have

me

Hlk, n) =

2.15 Consider a particle of mass m and charge q in an impenetrable cylinder with radius R and height a. Along the axis of the cylinder runs a thin, impenetrable solenoid carrying a magnetic flux CP. Calculate the ground state energy and wavefunction.

--+

In the case where B = 0 the Schrodinger equation of motion in the

cylindrical coordinates is

(2.127)

68

If we write W(p, cjJ, z) = (cjJ)R(p)Z(z) and k2 = 2;:;E we will have ( cjJ) Z ( z ) d2 R + ( cjJ) Z ( z ) ~ d R + R (p ) Z ( z) d2 

dp2 P dp p2 dcjJ2

d2Z

+R(p)(cjJ) dz2 + k2 R(p)(cjJ)Z(z) = 0::::;:.

1 d2 RId R 1 d2 1 d2 Z 2

R(p) dp2 + R(p)p dp + p2<1>(cjJ) dcjJ2 + Z(z) dz2 + k = 0(2.128)

with initial conditions W(Pa,cjJ,z) = W(R,cjJ,z) = W(p,cjJ,0) = W(p,cjJ,a) = O. So

1 d2 Z l2 d2 Z l2 ( ) () A ilz -ilz ( )

Z(z) dz2 = - ::::;:. dz2 + Z Z = 0::::;:. Z z = Ie + Ble 2.129

with

Z(O) - 0::::;:. Al + BI = 0::::;:. Z(z) = Al (eilz - e-ilz) = Csinlz

Z(a) - O::::;:.Csinla=O::::;:.la=n7r::::;:.l=ln=n7r n=±1,±2, ...

a

So

Z (z) = C sin lnz

(2.130)

N ow we will have

(2.131 )

with

 ( cjJ + 27r) = ( cjJ) ::::;:. m E Z.

(2.132)

So the Schrodinger equation is reduced to

2. QUANTUM DYNAMICS

69

In the case at hand in which pa ----7 0 we should take B3 = 0 since Nm ----7 00 when P ----7 O. From the other boundary condition we get

(2.134 )

where /\'mv is the v-th zero of the m-th order Bessel function Jm. This means that the energy eigenstates are given by the equation

while the corresponding eigenfunctions are given by

(2.136)

with n = ±1,±2, ... and m E Z.

Now suppose that jj = Bz. We can then write

A = (BP~) ~ = (___!_) ~.

2p 27rp

(2.137)

The Schrodinger equation in the presence of the magnetic field jj can be written as follows

(2.138)

70

Making now the transformation D¢ ;¢ - ~~ 2~ we get

(2.139)

h D2 ( 8 ie <I» 2 L t' A e t

were ¢ = 8¢ - "lie 211" . e lng = "lie 211" we ge

2 ( 82 2ie el> 8 2) ( 82 . 8 2)

D¢ = 8(jJ2 - ;;;; 27'1 8¢ - A = 8¢2 - 2zA 8¢ - A .

Following the same procedure we used before (i.e. 'ljJ(p, ¢, z) = R(p )el>( ¢)Z (z)) we will get the same equations with the exception of

(2.140)

[ 82 . 8 2] 2 d2el> . del> (2 2)

8¢2 - 2zA 8¢ - A el> = -m el> ::::;,. d¢2 - 2zA d¢ + m - A el> = O.

The solution to this equation is of the form el¢. So

l2el¢ - 2iAlel¢ + (m2 - A2)el¢ = 0 ::::;,. l2 - 2iAl + (m2 - A2)

2iA ± J -4A2 - 4(m2 - A2) 2iA ± 2im

::::;,. l= = =i(A±m)

2 2

which means that

(2.141)

But

el>(¢ + 27'1) = el>(¢) ::::;,. A ± m = m' m' E Z

::::;,. m = ±(m' - A) m' E Z. (2.142)

This means that the energy eigenfunctions will be

0/' (-»-Al (K,mv )im'¢' (7rZ)

'f/nmv X - e m RP e SIn =:

(2.143)

but now m is not an integer. As a result the energy of the ground state will be

(2.144)

2. QUANTUM DYNAMICS

71

where now m = m' - A is not zero in general but it corresponds to m' E Z such that 0 < m' - A < 1. Notice also that if we require the ground state to be unchanged in the presence of B, we obtain flux quantization

'A e , 27rm'fic,

m - = 0 ::::} -- = m ::::} = m E Z.

fic27r e

(2.145)

2.16 A particle in one dimension (-00 < x < (0) is subjected to a constant force derivable from

v = AX, (A > 0).

(a) Is the energy spectrum continuous or discrete? Write down an approximate expression for the energy eigenfunction specified by E.

(b) Discuss briefly what changes are needed if V is replaced be V = Alxl.

(a) In the case under construction there is only a continuous spectrum and the eigenfunctions are non degenerate.

From the discussion on WKB approximation we had that for E > V (x)

A i

[E - V(X)]1/4 exp C J j2m[E - V(x)]dx)

B i

+ [E _ V(X)1'/4 exp ( - It J j2m[E V(x )]dx)

[E - ~(X)]1/4 sin G t j2m[E - V(x)]dx - :)

[E_~(X)]1/4sin(~ (~P,/'(~ _X)'/2dX_ :)

[E _ ~(X)]1/4 S111 [-~ (~ - X t2 f¥ - :]

(2.146)

72

where q = a [~ - x] and a = e;;y\ f/3.

On the other hand when E < V( x)

'ijJ II( x) = [A C2 )1/4 exp (- ~ lX V2m( Ax - E)dX)

x - E rc x'=E/>.

[A C2 )1/4 exp [- h 1 A lx V2m( Ax - E)d(2mAx)]

x - E 2m x'=E/>.

C3 [2( )2/3]

[_q)1/4 exp -3 -q . (2.147)

We can find an exact solution for this problem so we can compare with the approximate solutions we got with the WKB method. We have

Hla) = Ela) ::::;,. (pIHla) = (piEla)

p2

::::;,. (pi 2m + Axla) = E(pla)

p2 d

::::;,. -a(p) + ihA-d a(p) = Ea(p)

2m p

d -i ( p2 )

::::;,. dpa(p) = hA E - 2m a(p)

::::;,. da(p) = -i (E _ L) dp

a(p) hA 2m

-i ( p3 )

::::;,. In a(p) = hA Ep - 6m + Cl

* ag(p) = cexp [hiA (:,:, - EP) ].

(2.148)

We also have

(EIE') = J dp(Elp)(pIE') = J a'E;(p)aE,(p)dp

IcI2 J dpexp [hiA (E - E')p] IcI227rhA6(E - E') ::::;,. 1

6(E - E')

~.

(2.149)

C

So

(2.150)

2. QUANTUM DYNAMICS

73

These are the Hamiltonian eigenstates in momentum space. For the eigenfunctions in coordinate space we have

y6(x) =

jdp(xlp)(pIE) (2.150) 1 jdpei~x eh\.(::-EP)

27rhV>-

1 j [ . p3 i (E )]

27rhV>- dpexp i h)..6m - h --:\ - x p .

(2.151 )

U sing now the substitution

P p3 u3

U = (h2m)..)1/3 ::::;:. h)..6m 3

(2.152)

we have

y6( x) (h2m).. )1/31+CXl du exp [iu3 _ j_ (E _ x) u(h2m).. )1/3]

27r h V>- -CXl 3 h)..

:x 1+CXl du exp [iu3 - iuq] , (2.153)

27r -CXl 3

a 1+CXl (u3 ) a l+CXl (u3 )

y6(x)= V>- ducos --uq = 1\ cos --uq du

27r ).. -CXl 3 7rY).. ° 3

since J!";: sin (~3 - uq) du = O. In terms of the Airy functions

1 r= (u3 )

Ai( q) = yI7r Jo cos 3 - uq du

(2.154 )

we will have

a .

y6(x) = ,c-\Az( -q). v 7r)..

(2.155)

For large Iql, leading terms in the asymptotic series are as follows

r--J 1 ( 2 3/2)

Ai( q) r-;» 2y17rq1/4 exp - 3"q ,q > 0

r--J 1 . [2 ( )3/2 7r]

Ai( q) r-;» yI7r( _q)1/4 sm 3" -q + 4 '

(2.156)

q<O

(2.157)

74

Using these approximations in (2.155) we get

~ ~_1_ sin [~q3/2 + 7r] for E > V(x)

7rv0: ql/4 3 4 '

"" 27rJx (-~l'/4 exp [-~( _q)3/2], for E < V( x) (2.158)

1jJ(q)

1jJ(q)

as expected from the WKB approximation.

(b) When V = Alxl we have bound states and therefore the energy spectrum is discrete. So in this case the energy eigenstates heve to satisfy the consistency relation

lX1 dxV2m[E - Alxl] = (n + ~) tth ; n = 0,1,2, ...

X2

(2.159)

The turning points are Xl = - ~ and X2 = ~. So

.: lElA

dxV2m[E - Alxl] = 2 V2m[E - Ax]dx

-EIA 0

rEIA (E )1/2

-2~Jo ~-x d(-x)

2 (E )3/2 EIA 2 (E)3/2

-2~- --x =2~- - =?

3 A 0 3 A

3 (n +~) 7rfi =? (E) = [3 (n +~) 7rfi]2/3 =?

4~ A 42/3(2mAF/3

[3 (n + ~) 7rfiA] 2/3

4~ (2.160)

3. THEORY OF ANGULAR MOMENTUM

75

3 Theory of Angular Momentum

3.1 Consider a sequence of Euler rotations represented by

V(i/2)(ex,{i,1) = exp (-i;3<» exp (-i;2/1) exp (-i;31)

( e-.i(a+'Y)/2 :os ~ _e_-i(a-"f)/2 sin ~ ) .

ez(a-"f)/2 SIn f}_ ez(a+"f)/2 COS f}_

2 2

Because of the group properties of rotations, we expect that this sequence of operations is equivalent to a single rotation about some axis by an angle cP. Find cP.

In the case of Euler angles we have

( e-i(a+"f)/2 cos «

V(1/2) ( j3 ) - 2

a, " - i(a-"f)/2· f3

e sm2

(3.1 )

while the same rotation will be represented by

Since these two operators must have the same effect, each matrix element should be the same. That is

e-i(a+~)/2 cos ~ = cos (~) - in; sin (n

::::} cos (cP) = cos (a + ,) cos j3 222

2j3 2(0'+,)

::::} cos cP = 2 cos "2 cos 2 - 1

'* rp = arccos [2COS2 ~ cos' (ex ~ 1) -1].

(3.3)

3.2 An angular-momentum eigenstate Ij, m = mmax = j) is rotated by an infinitesimal angle e about the y-axis. Without using the

76

explicit form of the d~~m function, obtain an expression for the probability for the new rotated state to be found in the original state up to terms of order E2.

The rotated state is given by

(3.4)

up to terms of order E2. We can write Jy in terms of the ladder operators

J+ Jx+~Jy }=?Jy=J+~J_.

J_ - Jx - 'lJy 2'l

Subtitution of this in (3.4), gives

Ij,j)n = [I - ~Y+ - L) + 8~~2 (.1+ - L )2]lj,j) (3.6)

(3.5)

We know that for the ladder operators the following relations hold

J+Ij, m) J_lj, m)

So

(J+ - J_)lj,j) (J+ - J_?lj,j)

n)(j - m)(j + m + l)lj, m + 1) n)(j + m)(j - m + l)lj, m - 1)

(3.7) (3.8)

-J_lj, j) = -nfijlj, j - 1) (3.9)

-nfij(J+ - J_)lj,j - 1)

-nfij (J+Ij,j - 1) - J_lj,j - 1))

-nfij [ fijlj, j) - )2(2j - 1) Ij, j - 2)]

and from (3.6)

2 2

lj,j)R Ij,j) + ~fijlj,j - 1) - E8 2jlj,j) + E8 2)j(2j - l)lj,j - 2)

(1- ~j) Ij,j)+~fijlj,j-I)+ ~Jj(2j-I)lj,j-2)


77

Thus the probability for the rotated state to be found in the original state will be

I (j, j Ij, j) R 12 = (1 _ : j) 2 = 1 _ C; j + 0 ( e 4)

(3.10)

3.3 The wave function of a particle subjected to a spherically symmetrical potential V(1") is given by

'ljJ (x) = (x + y + 3z) f (1" ).

(a) Is 'ljJ an eigenfunction of L? If so, what is the i-value? If not, what are the possible values of i we may obtain when L2 IS measured?

(b) What are the probabilities for the particle to be found in various m; states?

(c) Suppose it is known somehow that 'ljJ( x) is an energy eigenfunction with eigenvalue E. Indicate how we may find V(1").

(a) We have

'ljJ ( x) ( x 1 'ljJ) = (x + y + 3 z ) f ( 1" ) •

(3.11)

So

/ ->IL->210/') (5-3.6.15) _tc2 [ 1 !!_ _1_!!_ (. e!!_)] 0/'(-»

\x lj/ It sin2 e 8(jJ2 + sin e 8e SIn 8e lj/ x .

If we write 'ljJ (if) in terms of spherical coordinates (x = 1" sin e cos ¢, y = 1" sin e sin ¢, z = 1" cos e) we will have

(3.12)

'ljJ (x) = 1" f ( 1") (sin e cos ¢ + sin e sin ¢ + 3 cos e) .

(3.13)

Then

1 82 -> 1" f ( 1" ) sin e 8 . 1" f ( 1" ) •

. 28r1.2'ljJ(X)= ·2 8r1.(Cos¢-sIn¢)=-. e(cos¢+sIn¢03.14)

sin e If' sm e If' sm

78

and

1 B(. B) (--+) Tf(T)B[ .2 ( .). ]

sin 8 B8 SIn 8 B8 'l/J x = sin 8 B8 -3 SIn 8 + cos c/J + SIn c/J SIn 8 cos 8 =

Ts{~1 [-6 sin 8 cos 8 + (cos c/J + sin c/J)( cos2 8 - sin ' 8)13.15) Substitution of (3.14) and (3.15) in (3.11) gives

- 1i 2 T f ( T) [- _._1_ ( cos c/J + sin c/J) (1 - cos 2 8 + sin 2 8) - 6 cos 8]

sm8

1i 2 T f ( T) [si~ 8 2 sin 2 8 ( cos c/J + sin c/J) + 6 cos 8]

21i 2 T f (T) [sin 8 cos c/J + sin 8 sin c/J + 3 cos 8] = 21i 2 'l/J (if) ::::;:.

2 1i 2 'l/J (if) = 1 (1 + 1) 1i 2 'l/J (if) = l (l + l) 1i 2 'l/J ( if) ( 3 .16 )

which means that 'l/J(if) is en eigenfunction of I} with eigenvalue l = 1.

(b) Since we already know that l = 1 we can try to write 'l/J(if) in terms of the spherical harmonics y;_m (8, c/J). We know that

o [l; [l;z {fn 0

y;_ = - cos 8 = --::::;:. z = T -y;_

4n 4n T 3

_{l;(X~iY)} {x = T/¥ (Y;_-I _ Y;_+l)

(3(x-iy) ::::;:. y = ir (2; (Y,-l + Y,+l)

VB; r V3 1 1

So we can write

4) (x) = rf2f:.r(r) [3.;2y,° + y,-' - y,+1 + iy,+1 + iY,-']

f2f:".r(,,) [3.;2y,° + (I + i)y,-1 + (i - 1 )Y,+1]. (3.17)

But this means that the part of the state that depends on the values of m can be written in the following way

I'l/J)m = N [3v'2ll = I,m = 0) + (1 + i)ll = I,m = -1) + (1- i)ll = I,m = 1)]

and if we want it normalized we will have

1 IN12(18 + 2 + 2) = 1 ::::;:. N = j()(). v22

(3.18)


79

So

P(m = 0) P(m = +1) P(m = -1)

9·2 9 I(l = 1, m = 01y6)12 = 22 = U'

I(l = 1, m = +11y6)12 = 222 = 1\' 221

l(l=lm==IIy6)1 =-=-.

, 22 11

(3.19) (3.20) (3.21 )

V(r)

(c) If y6E( x) is an energy eigenfunction then it solves the Schrodinger equation

- n 2 [8822 y6 E (x) + ~ 88 y6 E (x) - ~ 2 y6 E (x)] + V ( r ) y6 E (x) =

2m r r r n r2

Ey6E(X)

_n2 [ d2 2 d 2]

=? 2m Yzm dr2[rf(r)] + ;:-dr[rf(r)] - r2[rf(r)] + V(r)rf(r)Yzm =

Er f(r )Yzm =?

E+ I~ )~[dd [f(r)+rj'(r)]+~[f(r)+rj'(r)]-~f(r)] =?

r r 2m r r r

E + I~)!!_ [[,(,.) + {,(,.) + "I"(,') + 2{'(,')]] =?

r r 2m

n2 rf"(r) + 4f'(r)

E + 2m rf(r) . (3.22)

V(r)

V(r)

3.4 Consider a particle with an intrinsic angular momentum (or spin) of one unit of li. (One example of such a particle is the gmeson). Quantum-mechanically, such a particle is described by a ketvector I g) or in x representation a wave function

(/ (x) = (x; i I g)

where lx, i) correspond to a particle at x with spin in the i:th direction.

(a) Show explicitly that infinitesimal rotations of gi(x) are obtained by acting with the operator

(3.23)

80

_, nA _, _,

where L = ---;-1 x V. Determine S !

z

(b) Show that Land S commute.

(c) Show that S is a vector operator.

(d) Show that V x e'( x) = ;d S . f) e' where p is the momentum operator.

(a) We have

3 3

Ig) = ?= J Ix; i)(x; ilg) = L J Ix; i)gi(x)d3x.

z=l z=l

(3.24)

Under a rotation R we will have

3

Il) = U(R)lg) = L J U(R) [Ix) Q9li)] gi(x)d3x

z=l

3 3

- ?= J IRx) Q9li)vg)(R)gl(x)d3x detR=l ?= J Ix; i)V~11)(R)gl(R-IX)d3X

z=l z=l

3

- ?= J Ix; i)gil x)d3x =?

z=l

gi/(X) - Vg\R)gl(R-1x) =? g'(X) = Re'(R-1x).

(3.25)

Under an infinitesimal rotation we will have

R( 6 c/J, it) r = r + 6 r = r + 6 c/J( it x r) = r + (x r. (3.26)

So

g'(x) R(6c/J)e'(R-1x) = R(6c/J)e'(x - (x x) e'( x - (x x) + {x e'( x - (x x).

(3.27)

On the other hand

e'( x - (x x) - e'( x) - ({ X x) . Ve'( x) = e'( x) - {. (x X V) e'( x)

- e'( x) - !_(. Le'( x) (3.28)

n


81

where V iJ( x) [V gi (i)] Ii). Using this in (3.27) we get

--+( --+) i --+ L--+ --+( --+) --+ [ --+( --+) i --+ L--+ --+( --+)]

gx -nE:- gx +EX gx -nE· gx

--+(--+) i --+ L--+--+(--+) --+ --+(--+)

gx -nE· gx +EXgX.

(3.29)

But

EX iJ= (Eyg3 - Ezl) ex + (Ezgl - Exgl) ey + (Exl- Eygl) ez

or in matrix form

(3.30)

which means that

--+ --+ 'l --+ S--+--+(--+)

E X g = --E· g x

n

with (Sp,hl = -inEp,kl.

Thus we will have that

e'(i) = Ui'iJ(x) = [1- ~f. (l+S)] iJ(x) =? Ui'= 1- ~E. (l+S). (3.31)

(b) From their definition it is obvious that land S commute since l acts only on the Ix) basis and S only on Ii).

( c) S is a vector operator since

[SiSj - SjSi]km = L [( -in)Eikl( -in)Ejlm - (-in)Ejkl( -in)Eilm] L [n2EiklEjml - n2EjklEiml]

n2 L (OijOkm - OimOjk - OijOkm + OjmOki) n2 L (OjmOki - OimOjk)

n2 L EijlEkml = L inEijl( -inEkml) = L inEijl(Slhm. (3.32)

82

(d) It is

v x e(x)

(3.33)

3.5 We are to add angular momenta i. = 1 and j2 = 1 to form j = 2,1, and 0 states. Using the ladder operator method express all (nine) i m eigenkets in terms of Ijd2; mlm2). Write your answer as

1 1

Ij = I,m = 1) = 01+,0) - 010,+), ... ,

(3.34)

where + and 0 stand for ml,2 = 1,0, respectively.

We want to add the angular momenta i. = 1 and j2 = 1 to form j = Ijl - j21, ... , jl + j2 = 0, 1,2 states. Let us take first the state j = 2, m = 2. This state is related to Ijl ml; j2m2) through the following equation

Ij, m) = L (jd2; mlm2Ijd2;jm)ljd2; mlm2) (3.35)

m=ml+m2

So setting j = 2, m = 2 in (3.35) we get

Ij = 2, m = 2) = (jd2; + + Ijd2;jm)1 + +) norm, I + +) (3.36)

If we apply the J _ operator on this statet we will get

J_lj = 2, m = 2) = (J1- + J2-)1 + +) ::::;,. nJ(j+m)(j-m+l)lj=2,m=I)=

nJ(jl + md(jl - ml + 1)10+) + nJr-()-'2-+-m-2-)(-j2--m-2-+-1-)1 + 0)

::::;,. AI] = 2,m = 1) = V210+) + V21 + 0)

1 1

::::;,. Ij = 2, m = 1) = 010+) + 01 + 0). (3.37)


83

In the same way we have

J_lj=2,m=l) V6lj = 2,m = 0)

1 1

- y'2(J1- + J2-)10+) + y'2(J1- + J2-)1 + 0) ::::;:.

- ~ [hl- +) + hIOO)] + ~ [hIOO) + hi + -)] '*

V6lj = 2, m = 0) - 2100) + I - +) + I + -) ::::;:.

Ij=2,m=0) - ~IOO)+ ~I+-)+ ~I-+)

J_lj = 2,m = 0)

V6lj=2,m=-I)

Ij=2,m=-I) Ij=2,m=-I)

J -Ij = 2, m = -1) V4lj = 2, m = -2) Ij = 2, m = -2)

(3.38)

Now let us return to equation (3.35). If j = 1, m = 1 we will have

Ij = I,m = 1) = al +0) +bIO+)

(3.41 )

This state should be orthogonal to all I), m) states and in particular to Ij = 2, m = 1). So

(j = 2, m = Ilj = 1, m = 1) = 0 ::::;:. ha + hb = 0 ::::;:.

a + b = 0 ::::;:. a = -b (3.42)

84

In addition the state Ij = 1, m = 1) should be normalized so

(j = I, m = Ilj = I, m = 1) = 1 =? lal2 + IW = 1 (34') 21al2 = 1 =? lal = ~.

By convention we take a to be real and positive so a = ~ and b = - ~. That is

1 1

Ij=l,m=I)= V"21+O)- V"2IO+).

(3.43)

U sing the same procedure we used before

J_I)=I,m=l) V2lj = 1, m = 0)

1)=I,m=O)

J_lj=l,m=O)

V2lj = I,m = -1)

Ij = 1, m = -1)

1 1

V"2(J1- + J2-)1 + 0) - V"2(J1- + J2-)10+) =?

~ [.;2100) +.;21 + -)]- ~ [.;21- +) + .;2100)] =?

1 1

V"21 + -) - V"21- +) (3.44)

Returning back to (3.35) we see that the state Ij = 0, m = 0) can be written as

This state should be orthogonal to all states I), m) and in particulat to I) = 2, m = 0) and to j = 1, m = 0). So

(. I. ) fi 1 1

) = 2, m = 0 ) = 0, m = 0 = 0 =? V 3C1 + yl6C2 + yl6C3

=? 2Cl + C2 + C3 = 0 (3.47)

.. 1 1

() = 1, m = 01) = 0, m = 0) = 0 =? V"2C2 - V"2C3

(3.48)


85

U sing the last relation in (3.47), we get

(3.49)

The state Ij = 0, m = 0) should be normalized so

By convention we take C2 to be real and positive so C2 C1 = - ~ Thus

y3·

_1 and y3

Ij = 0 m = 0) = _1 1+ -) + _1 1- +) - _1 100). (3.51)

, v'3 v'3 v'3

So gathering all the previous results together

Ij = 2, m = 2) Ij=2,m=1)

IJ = 2, m = 0) Ij=2,m=-1) Ij = 2, m = -2) IJ = 1, m = 1) lJ=l,m=O) Ij = 1, m = -1) IJ = 0, m = 0)

1++)

hIO+) + hi + 0)

v1100) + ~I + -) + ~I- +) hIO-) + hl- 0)

1- -)

hi + 0) - hIO+) hl+-)-hl-+) hIO-) - hl- 0)

~I + -) + ~I- +) - ~IOO).

(3.52)

3.6 (a) Construct a spherical tensor of rank lout of two different vectors tJ = (Ux, ti; Uz) and 11 = (Vx, Vy, Yz). Explicitly write Ti~,o in

terms of Ux,y,z and Vx,y,z.

(b) Construct a spherical tensor of rank 2 out of two different vectors tJ and 11. Write down explicitly Ti~,±l,O in terms of Ux,y,z and Vxyz.

, ,

86

( a) Since [J and 11 are vector operators they will satisfy the following commutation relations

(3.53)

From the components of a vector operator we can construct a spherical tensor of rank 1 in the following way. The defining properties of a spherical tensor of rank 1 are the following

[J U(l)] = hqU(l) [J±, UJl)] = hj(1 =t= q)(2 ± q)U~~l. (3.54)
z, q q ,
It is
[Jz,Uz] (3.53) OhUz (~4)
U, = U« (3.55)
[J+, Uo] (3~4) V2hU+1 = [J+, Uz] = [Jx + ii; Uz]
(3.53) -ihUy + i(ih)Ux = -h(Ux + iUy) ::::;,.
U+1 = - ~(Ux + iUy) (3.56)
[J_, Uo] (3~4) V2hU_1 = [J_, Uz] = [Jx - ii; Uz]
(3~3) -ihUy - i(ih)Ux = h(Ux - iUy) ::::;,.
U-1 = ~(Ux - iUy) (3.57) _, _,

So from the vector operators U and V we can construct spherical tensors

with components

u,

- ~(Ux + iUy) ~(Ux - iUy)

v,

- ~(Vx + IVy) ~(Vx - iVy)

(3.58)

It is known (S-3.10.27) that if XJ~d and ZJ;2) are irreducible spherical tensors of rank kl and k2 respectively then we can construct a spherical tensor of rank k

T?) = L(k1k2; qlq2lk1k2; kq)XJ;d ZJ:2)

ql q2

(3.59)


87

In this case we have

T(I) -
+1 -
(3~2)
-
-
-
TJl) -
-
(3~2)
- (3.60)

(11; 00111; 10)Uo Va + (11; -1 + 1111; 10)U_1 V+1 +(11; +1 - 1111; 10)U+1 V-I

1 1

- V2U-1 V+1 + V2U+1 V-I

1 1

V2~(Ux - iUy)(Vx + iVy) - V2HUx + iUy)(Vx - iVy)

1M [UxVx + iUxVy - iUyVx + UyVy - UxVx + iUxVy - iUyVx - UyVy] 2y2

i

V2(UXVy - UyVx) (3.61)

(11; -10111; II)U_1 Va + (11; 0 - 1111; II)Uo V-I

1 1

- V2U -1 Vo + V2UO V-I

-HUx - iUy)Vz + ~UZ(VX - iVy).

(3.62)

(b) In the same manner we will have

T(2) -
+2 -
-
-
T(2) -
+1 -
(3~2)
- (11; +1 + 1111; 2 + 2)U+1 V+1 (3.52) U+1 V+1 = ~(UX + iUy)(Vx + iVy)

-~(Ux Vx - u, Vy + iUx Vy + iu; Vx) (3.63)

(11; 0 + 1111; 2 + I)UOV+l + (11; +10111; 2 + I)U+1 Vo

1 1

V2UOV+1 + V2U+1 Va

-HUz Vx + o. Vz + u: Vy + iu; Vz)

(3.64)

(11; 00111; 20)Uo Va + (11; -1 + 1111; 20)U_1 V+1 +(11; +1 - 1111; 20)U+1 V-I

88

(3.52)

(3.52)

(3.65)

(11;0 -1111;2 -1)UOV-l + (II; -10111;2 + I)U-dlo

1 1

yl2UO V-I + yl2U -1 Va

~(Uz Vx + U: Vz - tu, Vy - ic, Vz)

(3.66)

T~~ (II; -1 - 1111; 2 - 2)U_1 V-I (3.52) U-1 V-I = HUx - iUy)(Vx - iVy)

~(Ux Vx - u, Vy - .u. Vy - iu; Vx). (3.67)

3.7 (a) Evaluate

m=-J

for any j (integer or half-integer); then check your answer for j = ~. (b) Prove, for any j,

J

L m2Id~Jm(6)12 = ~j(j + 1) sinp + m,2 + H3cos2 P - 1).

m=-J

[Hint: This can be proved in many ways. You may, for instance, examine the rotational properties of f; using the spherical (irreducible) tensor language.]


89

(a) We have

m=-J

m=-J

m=-J

m=-J

J

L \j m'l eiJy(3/nm Ij m) \j m I e -iJy(3/n Ij m')

m=-J

- ~. (jrn'leiJ,N"Jz [mti Ijrn)(jrnl] e-iJ,N"ljrn')

_ ~/. 'I iJy(3/nJ -iJy(3/nl· ')

Ii \Jm e ze jm

- ~ \jm'ID*(p; ey)JzD(p; ey)ljm').

(3.68)

But the momentum j is a vector operator so from (S-3.10.3) we will have that

trte, ey)JzD(p; ey) = L Rzj(p; ey)Jj.

J

(3.69)

On the other hand we know (S-3.1.5b) that

( cosp

R(p; ey) = 0

- sinp

o Sinp) 10.

o cosp

(3.70)

So

1 [ . u (. 'I J + + J - I· ') t' u]

- h - SIn {J J m 2 J m + lim cos {J

- m' cosp. (3.71)

90

For j = 1/2 we know from (S-3.2.44) that

( (3 . (3)

i1/2) (/3) = c~s"2 - SIn "2 .

mm SIn f}_ COS f}_

2 2

(3.72)

So for m' = 1/2

1/2

L Id~~/2(/3)12m

m=-I/2

l' 2/3 1 2/3

- - SIn - + - COS -

2 2 2 2

~ cos /3 = m' cos /3

(3.73)

while for m' = -1/2 1/2

L Id~~/2(/3)12m

m=-I/2

1 2/3 l' 2/3

- - COS - + - SIn -

2 2 2 2

-~ cos /3 = m' cos /3. (3.74)

(b) We have

m=-J

m=-J

m=-J

J

L m2 \jm'le-iJy(3/hljm) \jm leiJy(3/hljm')

m=-J

J

L \j m'l e -iJy(3 /hm 21j m) \j m I eiJy(3 /h IJ m')

m=-J

I:' (jm'le-iJ,Nh f; [mt.j Ijm)(jml] eiJ,Nhljm') :'2 (jm'le-iJy(3/h J;eiJy(3/hljm')

~ (jm'IV(/3; ey)J;Vt(/3; ey)ljm'). (3.75)


91

From (3.65) we know that

TJ2) = fe,(3.r; - .12)

(3.76)

where TJ2) is the O-component of a second rank tensor. So

(3.77)

and since V( R)J2Vt (R) = J2V( R)Vt (R) = J2 we will have L~=-j m2Id~~,(p)12 =

~~(jm'IJ2Ijm') + G~(jm'IV(p; ey)J;Vt(p; ey)ljm'Q3.78)

n 3 V 3n

We know that for a spherical tensor (S-3.10.22b)

k

V(R)Tq(k)Vt(R) = L V~~J(R)Tq(,k)

q'=-k

(3.79)

which means in our case that

2

(jm'IV(p; ey)J;Vt(p; ey)ljm') = (jm'l L Tq(?)V~;6(p; ey)ljm')

q'=-2

2

'" v(2)(r:l A )/. 'IT(2)1· ')

Z:: q'O p ; ey \Jm q' jm .

(3.80)

q'=-2

But we know from the Wigner-Eckart theorem that (jm'ITq('~oljm') = o. So

m=-J

_1_n2 .( . + 1) + ~ GV(2)(r:l. e )( ·m'IT.(2)1 ·m')

3 n 2 J J n 2 V 3 00 f-/ , y J 0 J

~j(j + 1) + ~d6~(p)(jm'IJ; - ~J2Ijm') ~j(j + 1) + ~d6~(P) [m,2 - ~j(j + 1)]

92

1 . ( . ) 1 ( 2 (3 ) [,2 1. ( . )]

"3 J J + 1 +"2 3 cos - 1 m -"3 J J + 1

-~j(j + 1) cos2 (3 + ~j(j + 1) + ~j(j + 1) + m,2 (3cos2 (3 - 1)

2 6 3 2

~j(j + 1) sin2 (3 + m,2 H3 cos2 (3 - 1) (3.81)

where we have used d6~ ((3) = P2 (cos (3) = H3 cos2 (3 - 1).

3.8 (a) Write xy, xz, and (x2 - y2) as components of a spherical (irreducible) tensor of rank 2.

(b) The expectation value

Q e(a,j, m = jl(3z2 - r2)la,j, m = j)

is known as the quadrupole moment. Evaluate

e(a,j, m'l(x2 - y2)la,j, m = j),

(where m' = j, j -1, j - 2, ... )in terms of Q and appropriate ClebschGordan coefficients.

(a) Using the relations (3.63-3.67) we can find that in the case where 0 = 11 = x the components of a spherical tensor of rank 2 will be

~(X2_y2)+ixy -(xz + izy)

ff,(2z2 - x2 - y2) = ff,(3z2 - r2)

~ (x2 - y2) - ixy

xz - 'lzy (3.82)

So from the above we have

T(2) _ T(2)

+2 -2

xy -

- 2i '

T(2) _ T(2)

x z = _-_1 _ ___;_+_1

2

(3.83)

(b) We have

Q e(a,j,m = jl(3z2 - r2)la,j,m = j)


93

r: / . _ ·1T.(2)1 . =i) (W:.=Eo) / 02. 0°102. 0 0) \ajIIT(2) Ilaj) 16

v 0 e \ a, J, m - J 0 a, J, m - J - \J, J J ,J J y!2 j + 1 V 0 e

o (2) 0 _ _9_ y!2j + 1

\aJIIT IlaJ)- ~ ("2001020"). (3.84)

V 6e J ; J J ; JJ

So

e (a,j, m'l(x2 - y2)la,j, m = j)

(3083) / 0 'IT(2)1 0 _ 0) / 0 'IT(2)1 0 _ 0)

- e\a,J, m +2 a,J, m - J + e\a,J, m -2 a,J, m - J

o

__ _.....- ( °IIT(2)11 0) ( °IIT(2)11 0)

( 0 0 I 0 0 ') aJ aJ (0 0 I 0 0 0 ) aJ aJ

= e J2· J2 J2· jm. + e6 I 0- J2· J - 2 J2· JJ - 2 -'-----~~=--'-

, , y!2j + 1 m .s 2, , y!2j + 1

(3084) .9_ (j2; j, - 21j2; j, j - 2) 6 I ° (3.85)

v'6 \j2;j,0Ij2;j,j) m,]-2·

94

4 Symmetry in Quantum Mechanics

4.1 (a) Assuming that the Hamiltonian is invariant under time reversal, prove that the wave function for a spinless nondegenerate system at any given instant of time can always be chosen to be real.

(b) The wave function for a plane-wave state at t = 0 is given by a complex function eipox/h. Why does this not violate time-reversal invariance?

(a) Suppose that In) in a nondegenerate energy eigenstate. Then

H01n) = 0Hln) = Enln) =? 01n) = ei8ln)

=? 01n, to = 0; t) = 0e-itH/hln) = 0e-itEn/hln) =

°tE /h o(Ent+s) .(2Ent+8)

et n 01n) = et --y;- In) = et -h- In, to = 0; t)

=? 0 [J d3xIX)(xl] In, to = 0; t) = eie~nt+8) [J d3xlX')(X'I] In, to = 0; t)

=? J d3x(X'ln, to = 0; t)*lx) = J d3xeie~nt+s)(xln, to = 0; t)IX')

=? cP~(X',t) = eie~nt+8)cPn(X,t). (4.1)

So if we choose at any instant of time 6 = - 2~nt the wave function will be real.

(b) In the case of a free particle the Schrodinger equation is

2 h2

;m In) = Eln) =? - 2m VcPn(x) = EcPn(x)

=? cPn(x) = Aeipox/h + Be-ipox/h (4.2)

The wave functions cPn( x) = e-ipox/h and cP~ (x) = eipox/h correspond to the same eigenvalue E = ;~ and so there is degeneracy since these correspond to different state kets IP) and 1- p). So we cannot apply the previous result.

4.2 Let cP(fJ) be the momentum-space wave function for state la), that is, cP(fJ) = (fJla).Is the momentum-space wave function for the

4. SYMMETRY IN QUANTUM MECHANICS

95

time-reversed state 81a) given by ¢(iJ) , ¢( -iJ), ¢*(iJ), or ¢*( -iJ)? Justify your answer.

In the momentum space we have

la) = J d3p' (iJ la) liJ) =* la) = J d3p' ¢(iJ) liJ)

=* 81a) = J d3p'8 [(iJla)liJ)] = J d3p'(iJla)*8IiJ). (4.3)

For the momentum it is natural to require

(alpla) = -(alpla) =* (aI8p8-lla) =* 8p8-1 = -p

( 4.4)

So

8plp) (4.4) -p8Ip) =* 81p) = I - p)

( 4.5)

up to a phase factor. So finally

81a) = J d3p'(pla)*I- p) = J d3p'(-pla)*lp)

=* (pI8Ia) = ¢(p) = (-pla)* = ¢*( -p). (4.6)

4.3 Read section 4.3 in Sakurai to refresh your knowledge of the quantum mechanics of periodic potentials. You know that the energybands in solids are described by the so called Bloch functions 'ljJn,k fullfilling,

where a is the lattice constant, n labels the band, and the lattice momentum k is restricted to the Brillouin zone [-IT/a, IT/a].

Prove that any Bloch function can be written as,

'ljJn,k( x) = L ¢n( x - Ri)eikRi

Ri

96

where the sum is over all lattice vectors Ri• (In this simble one dimensional problem H; = ia, but the construction generalizes easily to three dimensions.).

The functions ¢n are called Wannier functions, and are important in the tight-binding description of solids. Show that the Wannier functions are corresponding to different sites and/or different bands are orthogonal, i.e. prove

Hint: Expand the ¢ns in Bloch functions and use their orthonormality properties.

The defining property of a Bloch function 'ljJn,k( x) is

(4.7)

We can show that the functions LRi ¢n(x - Ri)eikRi satisfy the same relation

Ri

Ri

eika L ¢n(x - Rj)eikRJ

RJ

(4.8)

which means that it is a Bloch function

'ljJn,k(X) = L ¢n(x - Ri)eikRi.

Ri

(4.9)

The last relation gives the Bloch functions in terms of Wannier functions. To find the expansion of a Wannier function in terms of Bloch functions we multiply this relation by e-ikRJ and integrate over k.

Ri


97

But

ik(R-R.) 'If/a

e ' J

2sin [7r/a(Ri - Rj)] H; - Rj

(4.11 )

where in the last step we used that H; - Rj = na, with n E Z. So

(4.12)

So using the orthonormality properties of the Bloch functions

(4.13)

4.4 Suppose a spinless particle is bound to a fixed center by a potential V(X') so assymetrical that no energy level is degenerate. U sing the time-reversal invariance prove

(l) = 0

for any energy eigenstate. (This is known as quenching of orbital angular momemtum.) If the wave function of such a nondegenerate eigenstate is expanded as

L L F1m(r)Yzm(8, c/J),

I m

98

what kind of phase restrictions do we obtain on Flm(,)?

Since the Hamiltonian is invariant under time reversal

H0 = 0H.

(4.14)

So if In) is an energy eigenstate with eigenvalue En we will have

(4.15)

If there is no degeneracy In) and 01n) can differ at most by a phase factor. Hence

( 4.16)

For the angular-momentum operator we have from (S-4.4.53)

_, _, (416) _,

(nILln) = -(fiILlfi) . -(nILln) =*

(nilln) = 0

( 4.17)

We have

01n) - 0 J d3xli)(iln) = J d3x(iln)*0Ii) - J d3x(iln)*li) (4.16) eiSln) =*

(;'10In) - (:2ln)* = eiS(:2ln). (4.18)

So if we use (iln) = Ll Lm Flm(,)Yzm(8, 4Y)

L Fl"m(,)Yzm*(8, 4Y) = eiS L Flm(,)Yzm(8, 4Y)

ml

ml

ml

ml

J Yz;n'* L Fl"m(')( -1)mYz-m(8, 4Y)df! = eiS J Yz;n'* L F1m(,)Yzm(8, 4Y)df,

ml ml

L Fl"m(')( -l)mom',_mOl'l = eiS L Flm(,)om',mOl'l

ml

ml


99

4.5 The Hamiltonian for a spin 1 system is given by

Solve this problem exactly to find the normalized energy eigenstates and eigenvalues. (A spin-dependent Hamiltonian of this kind actually appears in crystal physics.) Is this Hamiltonian invariant under time reversal? How do the normalized eigenstates you 0 btained transform under time reversal?

For a spin 1 system l = 1 and m = -1,0, + 1. For the operator Sz we have

So

For the operator S; we have

(5-3.5.39)

S+ + S-I ) 1 I ) 1 I )

2 1, m = 2S+ 1, m + 2S- 1, m --+

~\1, nlS+ 11, m) + ~\1, nIS_ll, m)

~nj(1- m)(2 + m)On,m+l + ~nJ'-----(I-+-m-)-(2--m-)On,m-l'

Sxll, m) \1, nlSxll, m)

So

"CV2 0) ,,(0 0 n
- 0 0 V2 +- V2 0
2 0 0 0 2 0 V2
h(O V2 0)
- V2 0 V2 =?
2 0 V2 0
,,2 ( 2 ° 2) C 0 n
- 0 4 0 = n2 0 1 (4.21 )
4 2 021 0
2 100

In the same manner for the operator Sy = s, .: we find

Ii ( 0 V2 0) =}
Sx -. -V2 0
2z 0 -V2
r,? ( -2 0 2) (" 0 -~ )
~2 = li2 ~~
S2 - -- 0 -4 1 o . ( 4.22)
x 4 2
0 0 1
2 Thus the Hamiltonian can be represented by the matrix

(A 0 B)

H . fi2 0 0 0 .

BOA

( 4.23)

To find the energy eigenvalues we have to solve the secular equation

( Afi 2 - A 0 B fi 2 )

det(H - AI) = 0 =? det 0 -A 0 = 0

Bfi2 0 Afi2 - A

=? (Afi2 - A?(-A) + (Bfi2?A = 0 =? A [(Afi2 - A)2 - (Bfi2)2] = 0 =? A(Afi2 - A - Bfi2) (Afi2 - A + Bfi2) = 0

=? Al = 0, A2 = fi2(A + B), A3 = fi2(A - B). (4.24)

To find the eigenstate InAJ that corresponds to the eigenvalue Ac we have to solve the following equation

(~ 0 1 ) C ) = A, ( ~ ) .
fi2 0 (4.25)
0
For Al = 0
(~ 0 ~ ) ( ~ ) = 0 =} { aA + cB 0
fi2 0
0 A c aB + cA 0
=}{ B =}{ a=O
a = -cA ( 4.26)
B2 c=O
-cA + cA = 0 4. SYMMETRY IN QUANTUM MECHANICS

101

So
Ino) . (nno,m(n=}
Ino) = 110). ( 4.27)
In the same way for)' = Ji2(A + B)
(A 0 B) C ) C ) r: +cB - a(A + B)
-
o 0 0 b = (A + B) b =? 0 - b(A + B)
-
BOA c c aB + cA - c(A + B)
-
=}U ~ ( 4.28)
So
InA+B) - ( n rior m. ~ ( n =}
-
InA+B) 1 1 ( 4.29)
- V211, +1) + V211, -1).
-
For)' = Ji2(A - B) we have
(A 0 B) C ) C ) rA +cB - a(A - B)
-
o 0 0 b = (A - B) b =? 0 - b(A - B)
-
BOA c c aB + cA - c(A - B)
-
=} {\ ~ ( 4.30)
So
InA+B) - ( ~J nor-m ~ ( ~1 ) =}
-
InA-B) 1 1 (4.31 )
- V211, +1) - V211, -1).
- 102

N ow we are going to check if the Hamiltonian is invariant under time reversal

8H8-1 - A8S;8-1 + B(8S;8-1 - 8S~8-1)

- A8Sz8-18Sz8-1 + B(8Sx8-18Sx8-1 - 8Sy8-18Sy8-1)

- AS; + B(S; - S~) = H. (4.32)

To find the transformation of the eigenstates under time reversal we use the relation (S-4.4.58)

8Il,m) = (-I)mll, -m).

( 4.33)

So

8lno)

8110) (4.33) 110) Ino)

( 4.34) (4.35)

(4.33)

1 1

V2811, +1) + V2811, -1)

1 1

--11 -1) - -11 +1)

V2 ' V2 '

-lnA+B)

( 4.36) ( 4.37)

1 1

-811 +1) - -811 -1)

V2' V2'

1 1

--11 -1) + -11 +1)

V2' V2'

InA-B).

(4.38)

5. APPROXIMATION METHODS

103

5 Approximation Methods

5.1 Consider an isotropic harmonic oscillator in two dimensions.

The Hamiltonian is given by

(a) What are the energies of the three lowest-lying states? Is there any degeneracy?

(b) We now apply a perturbation

where 0 is a dimensionless real number much smaller than unity. Find the zeroth-order energy eigenket and the corresponding energy to first order [that is the unperturbed energy obtained in (a) plus the first-order energy shift] for each of the three lowest-lying states.

(c) Solve the Hi, + V problem exactly. Compare with the perturbation results obtained in (b).

[You may use (n'lxln) = V~1i/-2-m-w( vn + 10n',n+l + foon"n-d.]

Define step operators:

( iPx)

x+-

mw '

FW(x _ ipx), Vn mw

FW( ipy)

VnY+mw'

FW(y _ ipy). Vn mw

(5.1 )

From the fundamental commutation relations we can see that

104

Defining the number operators

we find

N

Ho

Ho

Nx + Ny = hw - 1 ::::;:.

hw(N + 1).

(5.2)

I.e. energy eigenkets are also eigenkets of N:

Nx I m,n) N; 1m, n) Nlm,n)

so that

mlm,n),

ti Im,n)::::;:. (m+n)lm,n)

(5.3)

Ho 1m, n) = Em,n 1m, n) = hw(m + n + 1) 1m, n).

(a) The lowest lying states are

state

Eo 0 = lu»

,

El 0 = Eo 1 = 2hw

, ,

E20 = Eo 2 = Ell = 3hw

, , ,

degeneracy 1

2

3

(b) Apply the perturbation V = 6mw2xy.

Full problem:

Unperturbed problem:

(tt; + V) Il) = Ell) HollO) = EO Ilo )

Expand the energy levels and the eigenkets as

E Il )

so that the full problem becomes

EO + ~l + ~2 + ...

Ilo ) + Ill) + ...

(5.4)


105

To 1 'st order:

(5.5)

Multiply with ( [0 I to find

\f0 I EO - n; 1[1 ) ~ 1 ( [0 IlO )

o = ( lO I v - ~ 1 1[°) =? ~ 1 = (to I V IlO )

(5.6)

In the degenerate case this does not work since we're not using the right basis kets. Wind back to (5.5) and multiply it with another degenerate basis ket

( rno I EO - tt; Ill) ~ 1( rno IlO)

o = (rnO I V - ~ 1 IlO) =? (rnO I V 1[°)·

(5.7)

Now, (rnO 1[°) is not necessarily 6kl since only states corresponding to different eigenvalues have to be orthogonal!

Insert a 1:

L (rnO I V I kO)( kO IlO) = ~ 1( rno IlO).

kED

This is the eigenvalue equation which gives the correct zeroth order eigenvectors!

Let us use all this:

1. The ground state is non-degenerate =?

2. First excited state is degenerate 11,0), 10,1). We need the matrix elements ( 1,0 I VII, 0), (1,0 I V 10,1), (0,11 VII, 0), (0,11 V 10,1 ).

and

ax I m., n) = Vm I ni - 1, ti )

a~ I m., n) = Vrn + 11 m + 1, n) etc.

106

Together this gives

VIO,lO

VIO,OI

VCn,lO

The V-matrix becomes

VOl,OI = 0,

Shu: t _ Shu:

-2-(1,0IaxayIO,I) - -2-'

Shu: t _ Shu:

-2- ( 1, ° I a x ay I 0, 1 ) - -2-·

(5.8)

o~w U ~)

and so the eigenvalues (= ~1) are

~I = ± 5hW. 2

To get the eigenvectors we solve

and get

I ... )-

1

V"2( 10,1) + 11, 0)),

1

V"2(IO,I) - 11,0)),

5 E + = hw (2 + 2)'

5

E - = hw (2 - 2). ( 5.9 )

3. The second excited state is also degenerate 12, 0), 11,1), 10,2), so we need the corresponding 9 matrix elements. However the only nonvanishing ones are:

5hw VII 20 = 1120 11 = VII 02 = V02 11 = M

' , , , y2

(5.10)

(where the V"2 came from going from level 1 to 2 in either of the oscillators) and thus to get the eigenvalues we evaluate

(-A

° = det ~

1 -A 1


107

which means that the eigenvalues are {O, ±5hw}. By the same method as above we get the eigenvectors

1 )+

1 )0

1 )_

HI2,0)+hI1,1)+ 10,2)), 1

V2(-12,0)+ 10,2)),

HI2,0)-hI1,1)+ 10,2)),

E+ = hw(3 + 5),

Eo = 3hW,

E_ = hw(3 - 5).

(c) To solve the problem exactly we will make a variable change. The potential is

mw2 [Hx2 + y2) + 5xy] =

mw2 [~((x + y)' + (x - y)') + ~(x + y)2 - (x - y)2)]. (5.11)

Now it is natural to introduce

,_ 1 ( )

x=V2x+y,

y' ~(x - y),

,_ 1 (' ')

Px = V2 v; + Py ,

1_ 1 (' ')

Py = V2 Px - Py .

( 5.12)

Note: [X',p~] = [yl,p~] = ih, so that (x', p~) and (y', p~) are canonically conjugate.

In these new variables the problem takes the form

So we get one oscillator with w~ = wVf+1' and another with w~ = w~. The energy levels are:

Eo,o hW,

El,o hw + hW~ = hw(l + Vf+1') =

hw(l + 1 + ~5 + ... ) = hw(2 + ~5) + 0(52),

108

EO,l hw + hW~ = = hw(2 - ~5) + 0(52),

E2,o lu» + 2hW~ = = hw(3 + 5) + 0(52),

El,l lu» + hW~ + hW~ = ... = 3hw + 0(52),

EO,2 lu» + 2hW~ = ... = hw(3 - 5) + 0(52).

( 5.13)

So first order perturbation theory worked!

5.2 A system that has three unperturbed states can be represented by the perturbed Hamiltonian matrix

where E2 > El. The quantities a and b are to be regarded as perturbations that are of the same order and are small compared with E2 - El. Use the second-order nondegenerate perturbation theory to calculate the perturbed eigenvalues. (Is this procedure correct?) Then diagonalize the matrix to find the exact eigenvalues. Finally, use the second-order degenerate perturbation theory. Compare the three results obtained.

( a) First, find the exact result by diagonalizing the Hamiltonian:

E, - E 0 a

o = 0 E, - E b

a" b* E2 - E

(El - E) [(El - E)(E2 - E) - Ib12] + a [0 - a*(El - E)] =

(El - E?(E2 - E) - (El - E)(lbI2 + laI2). (5.14)

So, E = El or (El - E)(E2 - E) - (lbl2 + lal2) = 0 i.e.

E2 (El + E2)E + ElE2 - (lal2 + Ib12) = 0 ::::;:.

E El + E2 ± .1 El + E22 _ ElE2 + lal2 + Ibl2 =

2 V 2


109

( 5.15)

Since lal2 + Ibl2 is small we can expand the square root and write the three energy levels as:

E E

E

( 5.16)

(b) Non degenerate perturbation theory to 2'nd order. The basis we use is

( 0 0 a)

The matrix elements of the perturbation V = 0 0 bare

a* b* 0

\ 11 V 13) = a,

\ 21 V 13) = b.

\ 1 I V 12) = \ k I V I k) = o.

Since ~11) = \ k I V I k) = 0 1 'st order gives nothing. But the 2'nd order shifts are

~ (2) 1

~ (2) 3

( 5.17)

110

The unperturbed problem has two ( degenerate) states 11) and 12) with energy El, and one (non-degenerate) state 13) with energy E2. Using nondegenerate perturbation theory we expect only the correction to E2 (i.e. ~~2)) to give the correct result, and indeed this turns out to be the case.

(c) To find the correct energy shifts for the two degenerate states we have to use degenerate perturbation theory. The V-matrix for the degenerate

subspace is (~ ~), so 1 'st order pert.thy. will again give nothing. We have to go to 2'nd order. The problem we want to solve is (Ho + V) Il) = E Il) using the expansion

Il) = IlO) + Ill) + . . . E = EO + ~ (1) + ~ (2) + ...

(5.18)

where HollO ) = EO IlO). Note that the superscript index in a bra or ket denotes which order it has in the perturbation expansion. Different solutions to the full problem are denoted by different l's. Since the (sub-) problem we are now solving is 2-dimensional we expect to find two solutions corresponding to l = 1,2. Inserting the expansions in (5.18) leaves us with

(EO - Ho) [llO) + Ill) + ... J =

(V - ~ (1) _ ~ (2) ... ) [llO) + Ill) + ... J .

(5.19)

At first order in the perturbation this says:

where of course ~ (1) = 0 as noted above. Multiply this from the left with a bra (kO I from outside the deg. subspace

\ kO I EO - n; Ill) = ( kO I V Ilo )

=? Ill) = L I kO ) (okO I V IlO) .

k::j:D E - Ek

This expression for 1[1) we will use in the 2'nd order equation from (5.19)

(5.20)

(EO _ Ho) Il2 ) = V Ill) - ~ (2) IlO ).

To get rid of the left hand side, multiply with a degenerate bra (rnO I (Hoi m ° ) = EO I m ° ) )

\ rno I EO - n; Il2) = 0 = \ rno I V Ill) - ~(2)( rno IlO).


111

Inserting the expression (5.20) for 1[1) we get

To make this look like an eigenvalue equation we have to insert a 1:

Maybe it looks more familiar in matrix form

L u.;»; = ~ (2)Xm nED

where

are expressed in the basis defined by 1[0). Evaluate M in the degenerate subspace basis D = { 11), 12)}

M - V13 V31 lal2 M - V13 V32 ab*
11 - E1 - Eg E1 - E2' 12 - E1 - Eg E1 - E2'
M _ V23V31 a*b M _ lv:d2 Ibl2
21 - E1 - Eg E1 - E2' 22 - E1 - Eg E1 - E2· With this explicit expression for M, solve the eigenvalue equation (define ). = ~(2)(E1 - E2), and take out a common factor El~E2)

o

( la 12 - x ab" )

det a*b Ibl2 _ ). =

(lal2 - ).)(lbI2 - ).) - lal21bl2 = ).2 _ (lal2 + IbI2).

::::;,. ). = 0, lal2 + Ibl2

::::;,. ~i2) = 0 ~~2) = lal2 + Ibl2

E1 - E2

(5.21 )

112

From before we knew the non-degenerate energy shift, and now we see that degenerate perturbation theory leads to the correct shifts for the other two levels. Everything is as we would have expected.

5.3 A one-dimensional harmonic oscillator is in its ground state for t < O. For t > 0 it is subjected to a time-dependent but spatially uniform force (not potential!) in the x-direction,

F(t) = Foe-t/T

(a) Using time-dependent perturbation theory to first order, obtain the probability of finding the oscillator in its first excited state for t > O. Show that the t ---+ 00 (T finite) limit of your expression IS independent of time. Is this reasonable or surprising?

(b) Can we find higher excited states?

[You may use (n'lxln) = Jh/2mw( yin + 16n',n+l + fo6n"n-d.]

(a) The problem is defined by

p2 mw2x2 av

Ho = - + V(t) = -Foxe-t/T (F = --)

2m 2 ax

At t = 0 the system is in its ground state 10:,0) = 10). We want to calculate

10:, t) L cn(t)e-Ent/1i In)

n

E~ hw(n + ~)

where we get c., (t) from its diff. eqn. (S. 5.5.15):

ih :t cn(t) L Vnmeiwnrntcm(t)

m

En - Em ( )

h =wn-m

( 5.22)

We need the matrix elements Vnm

Vnm ( n I - Foxe-t/T 1m) = - Foe-t/T ( n I x 1m) =

= -Foe-liT) Ii (vmbnm-l+v'm+lbnm+.).

2mw' ,


113

Put it back into (5.22)

ih:{n(t) = -Foe-'IT) 2r~:w (v'n + le-iw'cn+1(t) + Vneiw'cn_l(t)) .

Perturbation theory means expanding cn(t) = c~o) + C~l) + ... , and to zeroth order this is

To first order we get

We get one non-vanishing term for ti = 1, i.e. at first order in perturbation theory with the H.O. in the ground state at t = 0 there is just one non-zero expansion coefficient

- Fo) h (' di' eiWI'-I' IT v'!OI-I,a =

zh 2mw Jo

_ FO) h [. 1 1 e(iW-~)t/lt zh 2mw zw - -

T 0

Fa) h . 1 1 (1 _ e(iw-~)t)

zh 2mw zw - -

T

and

n

The probability of finding the H.O. in 11) is

I ( 1 I a, t ) 12 = ic~ 1) ( t ) 12.

As t --+ 00

C(I) -+ FO) h 1 const.

1 ih 2mw iw _ 1

T

This is of course reasonable since applying a static force means that the system asymptotically finds a new equilibrium.

114

(b) As remarked earlier there are no other non-vanishing en's at first order, so no higher excited states can be found. However, going to higher order in perturbation theory such states will be excited.

5.4 Consider a composite system made up of two spin ~ objects. for t < 0, the Hamiltonian does not depend on spin and can be taken to be zero by suitably adjusting the energy scale. For t > 0, the Hamiltonian is given by

(4~) _, _,

H= h,2 51.52.

Suppose the system is in 1 + -) for t < O. Find, as a function of time, the probability for being found in each of the following states 1++),1+-),1-+),1--):

(a) By solving the problem exactly.

(b) By solving the problem assuming the validity of first-order time-dependent perturbation theory with H as a perturbation switched on at t = O. Under what condition does (b) give the correct results?

(a) The basis we are using is of course 1 51z, 52z). Expand the interaction potential in this basis:

_, _,

51 . 52 51x52x + 51y52y + 51z52z = {in this basis}

h, 2 [

4 (1+)(-1 + 1-)(+lh(I+)(-1 + I-)(+Ih+

+ i2( -I +)( -I + 1-)( + 1 h (-I +)( -I + 1-)( + 1 h + + (I + )( + 1 - 1-)( -I ), ( 1 + )( + 1 - 1-)( -I), 1 =

h, 2 [

4 1++)(--1+1+-)(-+1+

+ 1 - +)( + - 1 + 1 - -)( + + 1 + + i2 (I + + ) ( - - 1 - 1 + - ) ( - + 1 +

-I - +)( + - 1 + 1 - -)( + + I) +


115

+ 1++)(++1-1+-)(+-1+

-I - + )( - + 1 + 1 - - )( - - 1 1 =

In matrix form this is (using 11) = I + +) I 2) = I + - ) 13) = 1-+) 14) = 1--))

1 0 o -1

o 2

o 0

o 0 2 0 -1 0 o 1

( 5.23)

This basis is nice to use, since even though the problem is 4-dimensional we get a 2-dimensional matrix to diagonalize. Lucky us! (Of course this luck is due to the rotational invariance of the problem.)

Now diagonalize the 2 X 2 matrix to find the eigenvalues and eigenkets

( -1 - ).

o = det 2

2 ) =(-1-).?-4=).2+2).-3

-1 -).

=?). = 1,-3

). = 1 :

1 =? -x + 2y = x =? x = Y = -

y'2

). = -3 :

1 =? -x + 2y = -3x =? x = -y = -

y'2

So, the complete spectrum is:

{I + +), I - - ), h(i + - ) + I - +) h(1 +-)-1-+)

with energy ~

with energy - 3~

116

This was a cumbersome but straightforward way to calculate the spectrum. A smarter way would have been to use S = 51 + 52 to find

_, _, 2 ( ) n2

We know that 5i = 5~ = 17, ~ ~ + 1 = 34 so

Also, we know that two spin~ systems add up to one triplet (spin 1) and one singlet (spin 0), i.e.

_, _, 1 2 3n2 1 2

5=1(3states)::::;,. 51.52=2(17, 1(1+1)--2 )=417,

(5.24)

5 = 0(1 state)::::;,.

. 4~ _, _,

Since H = 1j251 . 52 we get

E(spin=l)

E(spin=O)

(5.25)

From Clebsch-Gordan decomposition we know that {I + +), I - - ),

~ (I + - ) + I - + )) } are spin 1, and ~ (I + - ) - I - + )) is spin O!

Let's get back on track and find the dynamics. In the new basis H is diagonal and time-independent, so we can use the simple form of tthe time-evolution

operator:

U(t, to) = exp { - ~ H(t - to)} .

The initial state was I + -). In the new basis

{ 11) = I + +), 12) = I - -), 13) = ~( I + -) + I - +)), 14) = ~( I + - ) - I - + )) }


117

the initial state is

1

I + - ) = y'2( 13) + 14)).

Acting with U(t, 0) on that we get

I Q, t)

1 .

vl2exp { - :, Ht} (13) + 14)) =

~ [exp{ - J,L'>t} 13) + expC: L'>t} 14)] = [exp{-~~t} ~(I +-)+ 1_+))+

+exp{3if~t} ~( 1+ -) - I - + ))] =

~ [(e-iwt + e3iwt) I + -) + (e-iwt + e3iwt) I - +)]

where

w

( 5.26)

The probability to find the system in the state I (3) is as usual I (f31 Q, t ) 12

( + + I Q, t) = ( - - I Q, t ) = 0

I( + - I Q, t )12 = ~ (2 + e4iwt + e-4iwt) = ~ (1 + cos4wt) t"V 1 - 4(wt)2 ... I( - + I Q, t )12 = ~ (2 - e4iwt - e-4iwt) = ~ (1 - cos4wt) t"V 4(wt)2 ...

(b) First order perturbation theory (use S. 5.6.1 7):

C(o) - "'

n Uni,

c~1)(t) = hi 1: dt'eiWn' "v, (t').

( 5.27)

Here we have (using the original basis) Ho = 0, V given by (5.23)

Ii) = I + -),

118

If)

1- +),

En - Ei { }

--- = En = 0 = 0,

h

2~,

0, ri i- f.

Inserting this into (5.27) yields

(0) (0)

Ci C I +_) = 1,

(1) (1) _ i r _.

C f C I -+) - - h Jo dt2~ - - 2zwt.

(5.28)

as the only non-vanishing coefficients up to first order. The probability of finding the system in I - - ) or I + + ) is thus obviously zero, whereas for the other two states

P( I + - )) 1

to first order, in correspondence with the exact result.

The approximation breaks down when wt « 1 is no longer valid, so for a given t:

h wt « 1 =? ~ « -. t

5.5 The ground state of a hydrogen atom (n = l,l = 0) is subjected to a time-dependent potential as follows:

V(X', t) = Va cos(kz - wt).

Using time-dependent perturbation theory, obtain an expression for the transition rate at which the electron is emitted with momentum p. Show, in particular, how you may compute the angular distribution of the ejected electron (in terms of e and cP defined with respect to the z-axis). Discuss briefly the similarities and the differences between this problem and the (more realistic) photoelectric effect. (note: For the initial wave function use

( ..... ) _ 1 (Z)~ -Zr/ao

\).I n=l,l=O X - ft ao e .


119

If you have a normalization problem, the final wave function may be taken to be

WJ(X) = G~ )eiPoi!/h

with L very large, but you should be able to show that the observable effects are independent of L.)

To begin with the atom is in the n = 1, l = 0 state. At t = 0 the perturbation V = Va cos(kz - wt)

is turned on. We want to find the transition rate at which the electron is emitted with momentum Pf. The initial wave-function is

wi(i) = _1 (]_)3/2e_r/ao

yI7r ao

and the final wave-function is

,T, (-» _ ( 1 ) ipoi!/h

'Jf f x - L3/2 e .

The perturbation is

V Va [ei(kZ-wt) + e-i(kZ-wt)] Veiwt + vt «:".

( 5.29)

Time-dependent perturbation theory (8.5.6.44) gives us the transition rate Wi-+n = 2; IV~iI2 ste; - (Ei + hW))

because the atom absorbs a photon Iua. The matrix element is

and

(eikz)ni

( 5.30)

120

So (eikz) ni is the 3D Fourier transform of the initial wave-function (and some constant) with if = kf - kez. That can be extracted from (Sakurai problem 5.39)

( ikZ) 647r2 1

e ni = L3ag [a16 + (kf - kezp r

The transition rate is understood to be integrated over the density of states. We need to get that as a function of Pf = hkf. As in (S.5.7.31), the volume element is

2 2 dn

ti dndf2 = ti df2-d dp],

Pf

Using

we get

dn 1 2L2pf

dpf 2n (27rhP

27rh L2pf Lpf (27rhP

L 27rh

which leaves

L3k2 L3p2

n2dndf2 = (27r)lh df2dpf = (27rh)3df2dPf

and this is the sought density.

Finally,

Note that the L's cancel. The angular dependence is in the denominator:

[(lkficosB - k) ez + IkflsinB (coscpex + sincpey)]2 = Ikfl2cos2B + k2 - 2klkficosB + Ikfl2sin2B = k~ + k2 - 2klkficosB.

(5.31 )

In a comparison between this problem and the photoelectric effect as discussed in (S. 5.7) we note that since there is no polarization vector involved, w has no dependence on the azimuthal angle cP. On the other hand we did not make any dipole approximation but performed the x-integral exactly.

Solutions For The Exercices (J J Sakurai)

Încărcat de

Informații document

Descriere originală:

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Solutions For The Exercices (J J Sakurai)

Încărcat de

Drepturi de autor:

Formate disponibile

Part II Solutions

S-ar putea să vă placă și

Solutions For The Exercices (J J Sakurai)

Încărcat de

Informații document

Descriere originală:

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Solutions For The Exercices (J J Sakurai)

Încărcat de

Drepturi de autor:

Formate disponibile

﻿Part II Solutions

S-ar putea să vă placă și

Part II Solutions