Lecture XIII 37

Lecture XIII: Double Well Potential: Tunneling and Instantons

How can phenomena of QM tunneling be described by Feynman path integral?
No semi-classical expansion!
E.g QM transition probability of particle in double well: G(a, a; t) a|e
i

Ht/
| a
Potential
x
Inverted
potential
V(x)
Feynman Path Integral:
G(a, a; t) =
_
q(t)=a
q(0)=a
Dq exp
_
i

_
t
0
dt

_
m
2
q
2
V (q)
_
_
Stationary phase analysis: classical e.o.m. m q =
q
V
only singular (high energy) solutions Switch to alternative formulation...
Imaginary (Euclidean) time Path Integral: Wick rotation t = i
N.B. (relative) sign change! V V
G(a, a; ) =
_
q()=a
q(0)=a
Dq exp
_

_

0
d

_
m
2
q
2
+ V (q)
_
_
Saddle-point analysis: classical e.o.m. m q = +V

(q) in inverted potential!

solutions depend on b.c.
(1) G(a, a; ) q
cl
() = a
(2) G(a, a; ) q
cl
() = a
(3) G(a, a; ) q
cl
: rolls from a to a
Combined with small uctuations, (1) and (2) recover propagator for single well
Inverted
potential
-a a
-V(x)
x
a
-a
x

1/
(3) accounts for QM tunneling and is known as an instanton (or kink)
Lecture Notes October 2005
Lecture XIII 38
Instanton: classically forbidden trajectory connecting two degenerate minima
i.e. topological, and therefore particle-like
For large, q
cl
0 (evident), i.e. rst integral m q
cl
2
/2 V (q
cl
) =

0
precise value of xed by b.c. (i.e. )
Saddle-point action (cf. WKB
_
dqp(q))
S
inst.
=
_

0
d

_
m
2
q
2
cl
+ V (q
cl
)
_

_

0
d

m q
2
cl
=
_
a
a
dq
cl
m q
cl
=
_
a
a
dq
cl
(2mV (q
cl
))
1/2
Structure of instanton: For q a, V (q) =
1
2
m
2
(q a)
2
+ , i.e. q
cl

(q
cl
a)
q
cl
()

= a e

, i.e. temporal extension set by

1

Imples existence of approximate saddle-point solutions

involving many instantons (and anti-instantons): instanton gas

1
a
-a
x

5

4

3

2
Accounting for uctuations around n-instanton conguration
G(a, a; )

n even / odd
K
n
_

0
d
1
_

1
0
d
2

_

n1
0
d
n
A
n,cl.
A
n,qu.
..
A
n
(
1
, . . . ,
n
),
constant K set by normalisation
A
n,cl.
= e
nS
inst.
/
classical contribution
A
n,qu.
quantum uctuations (imported from single well): G
s.w.
(0, 0; t)
1

sin t
A
n,qu.

n

i
1
_
sin(i(
i+1

i
))

i
e
(
i+1

i
)/2
e
/2
G(a, a; )

n even / odd
K
n
e
nS
inst.
/
e
/2

n
/n!
..
_

0
d
1
_

1
0
d
2

_

n1
0
d
n
=

n even / odd
e
/2
1
n!
_
Ke
S
inst.
/
_
n
Using e
x
=

n=0
x
n
/n!, N.B. non-perturbative in !
G(a, a; ) Ce
/2
cosh
_
Ke
S
inst.
/
_
G(a, a; ) Ce
/2
sinh
_
Ke
S
inst.
/
_
Lecture Notes October 2005
Lecture XIII 39
Consistency check: main contribution from
n = n

n
nX
n
/n!

n
X
n
/n!
= X = Ke
S
inst.
/
no. per unit time, n/ exponentially small, and indep. of , i.e. dilute gas
h
S
A
S
A
x
Oscillator States
Exact States

Physical interpretation: For innite barrier two independent oscillators,

coupling splits degeneracy symmetric/antisymmetric
G(a, a; ) a|Se

S
/
S| a +a|Ae

A
/
A| a
|a|S|
2
= a|SS| a =
C
2
, |a|A|
2
= a|AA| a =
C
2
Setting:
A/S
= /2 /2
G(a, a; )
C
2
_
e
()/2
e
(+)/2
_
= Ce
/2
_
cosh(/)
sinh(/)
.
Remarks:
(i) Legitimacy? How do (neglected) terms O(
2
) compare to ?
In fact, such corrections are bigger but act equally on |S and |A
i.e. = Ke
S
inst.
/
is dominant contribution to splitting
V(q)

V(q)
q

q q
q
m
m
(ii) Unstable States and Bounces: survival probability: G(0, 0; t)? No even/odd eect:
G(0, 0; ) = Ce
/2
exp
_
Ke
S
inst
/

=it
= Ce
it/2
exp
_

2
t
_
Decay rate: |K|e
S
inst
/
(i.e. K imaginary) N.B. factor of 2
Lecture Notes October 2005