Upload
holland
View
26
Download
0
Embed Size (px)
DESCRIPTION
Ultrafast processes in molecules. VI – Transition probabilities. Mario Barbatti [email protected]. Fermi’s golden rule. Transition rate:. Quantum levels of the non-perturbed system. Perturbation is applied. Transition is induced. Fermi’s Golden Rule. which solves:. and. - PowerPoint PPT Presentation
Citation preview
2
Fermi’s golden rule
3
Fermi’s Golden Rule
kpk kHiW 22
Transition rate:
i
kk ,
Quantum levels of the non-perturbed
system
Perturbation is applied
Transition is induced
4
Derivation of Fermi’s Golden Rule
Time-dependent formulation
pHHH 0H0 – Non-perturbated Hamiltonian
Hp – Perturbation Hamiltonian
0
Ht
i
n which solves: 00 nEH n
ijji and ijji EE
5
Derivation of Fermi’s Golden Rule
n
tiEn neta n
pHHH 0
0
Ht
i
Prove it!
Multiply by at left and integrate
n
tinp
k knetanHkttai
k
Note that the non-perturbated Hamiltonian is supposed non-dependent on time.
6
An approximate way to solve the differential equation
n
tinp
k knetanHkttai
Guess the “0-order” solution: tan)0(
Use this guess to solve the equation and to get the 1st-order approximation: tak)1(
n
tinp
k knetanHkdt
tdai )0()1(
Use the 1st-order to get the 2nd-order approximation and so on.
n
tipnp
pk knetanHkdt
tdai )1()(
7
First order approximation
Guess the “0-order” solution: nin ta )0(
n
tinp
k knetanHkdt
tdai )0()1(
Suppose the simplified perturbation:
0
'kn
pH
nHk Constant between 0 and tOtherwise
t0 t
'knH
0
nHk p
8
First order approximation
0
')1(
n
tininmk
kneH
dttdai
Between 0 and tOtherwise
ki
kiiik
tiikk
knki eHdteHai
tt tt 2/sin2 2/'
0
')1(
It was used:
a ikxikx
kkxedxe
0
2/ 2/sin2
9
Transition probability
22
22'2 2/sin4ki
kiikkik HaP
tt
0
0.0
0.2
0.4
0.6
0.8
1.0
sin(
)2 /
2
In this derivation for constant perturbation, only transitions with ~ 0 take place. If the perturbation oscillates harmonically (like a photon), ≠ 0 can occur.
The final result for the Fermi’s Golden Rule is still the same.
10
Physically meaningful quantity
kneark
ikk PW'
'1t
i
kk ,
Near k: density of states dEdn
k
kikikkkikkneark
ikk dPdEPPW t
tt
11'
'
11
Physically meaningful quantity
kikikkkikkneark
ikk dPdEPPW t
tt
11'
'
Using
22
22' 2/sin4ki
kiikik HP
t
kikk HW 2'2
kiki
kikk dHWki
tt
2
22' 2/sin141
2/
12
Fermi’s Golden Rule: photons and molecules
kti
k dtekiW ik t
2
0
2 Ed
Transition rate:
i
kk ,
Ed 0HH H0 – Non-perturbated molecular Hamiltonian – Light-matter perturbation HamiltonianEd
13
Transition dipole moment
kti
k dtekiW ik t
2
0
2 Ed
i
kk ,
kikZeiki eikN
N
n
N
mmnn
at el
ddrRd
1 1
0
Electronic transition dipole moment
14
Einstein coefficients
i
k
ik
iN
Rate of absorption i → k iika
ik NBW
Einstein coefficient B for absorption
22
061 ki
ggB e
i
kik d
ik
ng - degeneracy of state n
15
Einstein coefficients
i
k
ki
kN
Rate of stimulated emission k → i kkist
ki NBW
Einstein coefficient B for stimulated emission
kik
iik B
ggB
ki
16
Einstein coefficients
i
k
ki
kN
Rate spontaneous decay k → i 2NAW kisp
ki
Einstein coefficient A for spontaneous emission
kiki
ki Bc
A 32
3
ki
17
Einstein coefficient and oscillator strength
kiki
ki Aemcf 22
302
In atomic units:
kiki
ki AE
cf 2
3
2
18
Einstein coefficient and lifetime
kikiki fE
cA 2
3
21
t
1
2
R
E
If E21 = 4.65 eV and f21 = -0.015, what is the lifetime of the excited state?
au
eVE
.1708840211396.27/65.4
65.421
au10
2
3
21
1029.0
)015.0()170884.0(2)137(
t
nss
70104189.21029.0 1710
21
t
Converting to nanoseconds:
19
non-adiabatic transition probabilities
20
Non-adiabatic transitions
11,
22 ,
21,
12 ,
0
x
E
0 H11
H22 E2
E1
Problem: if the molecule prepared in state 2 at x = ∞ moves through a region of crossing, what is the probability of ending in state 1 at x = ∞?
H12
21
Models for non-adiabatic transitions
xFHH 122211 constant, 1212 FH
1. Landau-Zener
12
2122expF
Hv
P
2. Demkov / Rosen-Zener
constant21 EE /sech1212 xhh x
xvhEEP
12
212
4sech
3. Nikitin
xAH
xAHH
expsin2
expcos
012
02211
02
02
02
cottan21exp
12/cosexp
v
P
4. Bradauk; 5. Delos-Thorson; 6. …
22
Derivation of Landau-Zener formula
nn
t
nnn dttHita
exp
Multiply by at left and integrate
kn
nknk netaHdt
tdai
k
nkkn HH
t
nnn dtHi
0
Ht
i
In the deduction it was used:
nn
t
nn HdtHdtd
23
kn
nknk netaHdt
tdai
Since there are only two states:
21212
1 etaHidt
tda
12121
2 etaHidt
tda
jiij
(i)
(ii)
Solving (i) for a2 and taking the derivative:
dt
tdadt
tadHei
dtda
HeHH 2
21
2
12
1
12
22111212
(iii)
Substituting (iii) in (ii):
011
2122
122112
12
aHdtdaHHi
dtad
24
Zener approximation: tHH 2211
21,
x
0
E
0 11,
22 ,
12 ,
222 F
dxdH
111 F
dxdH
vtFxFH 1111
vtFxFH 2222
vtFFHH 212211
vF12
011
2122
122112
12
aHdtdaHHi
dtad
25
01
01
22
1222
1222
2
12
1221
1221
2
aHdt
davtFidt
ad
aHdtdavtFi
dtad
Problem: Find a2(+∞) subject to the initial condition a2(∞) = 1.
The solution is:
0x
E
0
12 a
01 a ?2 a
?1 a
12
212
2 expF
Hv
a
26
The probability of finding the system in state 1 is:
12
2122
22exp
FH
vaPnad
The probability of finding the system in state 2 is:
12
2122exp11F
Hv
PP nadad
0
0
1
Pro
babi
lity
|H12
|2
Pnad
Pad
0
0
1
|F12
| or v
Pad
Pnad
27
212
** 2expH
P
Example: In trajectory in the graph, what are the probability of the molecule to remain in the * state or to change to the closed shell state?
0 2 4 6 8 10 12 14
-224.85
-224.80
-224.75
-224.70
-224.65
*
cs
H11
= t
Ene
rgy
(au)
Time (fs)
H22
= t
H12
= au
cs
*
au0.001au/f0435.0
01126.003224.0
s
fs104.2au1 2
time
tHH 2211
vF12 1
28
Example: In trajectory in the graph, what are the probability of the molecule to remain in the * state or to change to the closed shell state?
43.0001.0
011577.02exp2
**
P
57.01 *** PP cs
0 2 4 6 8 10 12 14
-224.85
-224.80
-224.75
-224.70
-224.65
*
cs
Ene
rgy
(au)
Time (fs)
cs
*
0.43
0.57
29
0
x
E
0
0
x
E
0
vv
For the same H12, Landau-Zener predicts:
Non-adiabatic (diabatic) Adiabatic
12
2122expF
Hv
Pnad
H12
30
0
x
E
0
0
x
E
0
For the same , Rosen-Zener predicts:
Non-adiabatic (but not diabatic!) Adiabatic
vv
xnad vhEEP
12
212
4sech
xh12
31
0
x
E
0
0
x
E
0
For the same 0 (H12), Nikitin predicts:
Non-adiabatic (diabatic) Adiabatic
v v
02
02
02
cottan21exp
12/cosexp
v
Pnad
32
Marcus Theory
AB → A+B 2022 1 exp
44ABBB
l Gk H
k Tk T
Ener
gy
Reaction coordinateA-B A+-B
2HAB
G0
33
The problem with the previous formulations is that they only predict the total probability at the end of the process.
If we want to perform dynamics, it is necessary to have the instantaneous probability.