WHY ???? Ultrashort laser pulses. (Very) High field physics Highest peak power, requires highest...

Post on 21-Dec-2015

213 views 0 download

Tags:

Transcript of WHY ???? Ultrashort laser pulses. (Very) High field physics Highest peak power, requires highest...

WHY????

Ultrashort laser pulses

(Very) High field physics

Highest peak power, requires highest concentration of energy

E L I

Create … shorter pulses (attosecond)Create x-rays (point source)Imaging

High fields high nonlinearities high accuracy

F=ma0~ 31 Å

1015 W/cm2, 800 nm

20

Electrons ejected by tunnel ionization can be re-captured by the next half optical cycle of opposite sign. The interaction of the returning electron with the atom/molecule leads to high harmonic generation and generation of single attosecond pulses.

-1

0

1

To do this you need to control a single cycle

Resolve very fast events

- “Testing” Quantum mechanics

Probing chemical reactions

Pump probe experiments

All applicatons require propagation/manipulation of pulses

-1

0

1

MANIPULATION OF THIS PULSE

Chirped pulse

LEADS TO THIS ONE:

Propagation through a medium with time dependent index of refraction

Pulse compression: propagation through wavelength dependent index

Why do we need the Fourier transforms?

Construct the Fourier transform of

“Linear” propagation in frequency domain

“Non-Linear” propagation in time domain

Actually, we may need the Fourier transforms (review)

0

Properties of Fourier transforms

Shift

Derivative

Linear superposition

Specific functions: Square pulse Gaussian Single sided exponential

Real E(E*(-

Linear phase

Product Convolution

Derivative

Description of an optical pulse

Real electric field:

Fourier transform:

Positive and negative frequencies: redundant information Eliminate

Relation with the real physical measurable field:

Instantaneous frequency

0 z

t t

What is important about the Carrier to Envelope Phase?

Slowly Varying Envelope Approximation

Meaning in Fourier space??????

Forward – Backward Propagation

Maxwell Equation

s = t – n/c Zr = t + n/c z

20 0

02

n nE P

z c t z c t t

2 2 2 20

02 2 2 2

nE P

z c t t

No scattering

i s i rE e e FΕ Ε

No coupling between EF & EB

22 2

024

cE P

r s n s r

22 2i s i s

0 F2

ce 2 i P e

r s r 4 n s r

F FΕ Ε

No linear assumption

Slowly varying envelope

22

02

22

0

2

1

2

F

F

ci P

r n

Pn

22

2F FP P

t

Study of linear propagation

(Maxwell second order)

Solution of 2nd order equation

22

02

( ) ( , ) 0E zz

0( )P E 0( ) (1 ( ))

( )( , ) ( , ) ik zE z E 0 e

( ) ( )2 20k

Propagation through medium

No change in frequency spectrum

To make F.T easier shift in frequencyExpand k value around central freq l

l

( )( , ) ( , ) lik zz 0 e ε ε

( )1( , ) ( ,0) ( )

2lik z i tt z e e d

z

Z=0

1( , ) ( , ) ( )

2i tE t z E z e d

Dispersion includedk real

10

gz v t

ε ε

Study of linear propagation

Expansion orders in k(Material property

l

l

2| 22

1( , ) ( ,0) (1 | ( ) ) ( )

2l

dkiik z i td d k

t z e e e i z dd

ε ε

II)( )( , ) ( , ) lik zz 0 e ε ε

ll| ( )| ( )( , )

22

2 l

1 d kdk i zi z ik z2d d0 e e e

ε

l

l

| ( )( , ) ( | ( ) ) l

dk 2i z 2 ik zd2

1 d k0 e 1 i z e

2 d

ε

22

2

( ) 1( ) ( )

2ixtt

x x e d xt

ε ε

2 2

2 2

10

2g

i d k

z v t d t

ε ε ε

Second

Study of linear propagation

Propagation in the time domain

PHASE MODULATION

n(t)or

k(t)

E(t) = (t)eit-kz

(t,0) eik(t)d (t,0)

DISPERSION

n()or

k()() ()e-ikz

Propagation in the frequency domain

Retarded frame and taking the inverse FT:

PHASE MODULATION

DISPERSION

Application to a Gaussian pulse

Evolution of a single pulse in an ``ideal'' cavity

Dispersion

Kerr effect

Kerr-induced chirp

20 0

02

n nE P

z c t z c t t

2 2 2 20

02 2 2 2

nE P

z c t t

22

2F FP P

t

Study of propagation from second to first order

From Second order to first order (the tedious way)

( ) ( )kz kz

2 2 2 20 i t i t

02 2 2 2

ne P e

z c t t

2 2 22

2 2 2 2 2

22

0 0 02

1 2ik 2ik

c z c t c t z

P i P Pt t

01 i cP

z c t 2

(Polarization envelope)

Pulse duration, Spectral width

Two-D representation of the field: Wigner function

-2 -1 0 1 2

-2

-1

0

1

2

-2 -1 0 1 2

-2

-1

0

1

2

Time TimeF

requ

ency

Fre

quen

cy

-2 -1 0 1 2

-2

-1

0

1

2

-2 -1 0 1 2

-2

-1

0

1

2

Time TimeF

requ

ency

Fre

quen

cyGaussian Chirped Gaussian

Wigner Distribution

Wigner function: What is the point?

Uncertainty relation:

Equality only holds for a Gaussian pulse (beam) shape free of anyphase modulation, which implies that the Wigner distribution for aGaussian shape occupies the smallest area in the time/frequencyplane.

Only holds for the pulse widths defined as the mean square deviation