USING AN OPTICAL CAVITY

surprisesameΗμιαγωγοί

1 Νοε 2013 (πριν από 3 χρόνια και 7 μήνες)

88 εμφανίσεις

6/1/2009

CLEO/IQEC’09 Baltimore, Maryland

PICOSECOND ULTRASONIC MEASUREMENTS

USING AN OPTICAL CAVITY

Yanqiu

Li,
Qian

Miao,
Arto

Nurmikko
, Humphrey Maris

Department of Physics and Division of Engineering

Brown University


OUTLINE


Introduction/Motivation


Optical Cavity


Experiment Result and Analysis


Summary



1








Film

(d~200nm)

Time delayed Probe

light pulses.

Detect returning of the
strain pulse by monitor
the change in the
reflectivity of the probe.

Introduction/Motivation


Picosecond

Ultrasonics

6/1/2009

CLEO/IQEC’09 Baltimore, Maryland

z=
0

z=d

Pump light pulse

sends out strain pulse,
that induces change in
the optical properties of
the film (
piezo
-
optic
effect).










Result of measurement of
D
R/R

of an Aluminum film

PROBE TIME DELAY (
ps
)

0

100

200

300

2x10
-
5

4x10
-
5

Launching
of the
longitudinal
strain pulse

1
st

round
trip of the
strain pulse

2
nd

round
trip of the
strain pulse







Substrate

SMALL



even zero, e.g. at
l
=800nm, of Cu (key
material for interconnect
wiring in the semiconductor
industry).

SMALL




e.g. for Cu, with
10mW absorbed
laser power, focal
spot size of 20
m
m


max

~10
-
5

Piezo
-
optical
coefficients

Returning
strain pulse


Picosecond ultrasonics has become a standard
technique for non
-
destructive ultrasonic
investigation of features of thin films and
nanostructures, such as film thickness, internal
cracks, etc.


Plays important role in wafer inspection in the
semiconductor industry.


Two significant challenges include small strain
and piezo
-
optic effect.

2

6/1/2009

CLEO/IQEC’09 Baltimore, Maryland








Film

(d~200nm)

z=
0

z=d







Substrate











DBR

z=
-
w

The Optical Cavity

q

Probe,
k
0

R


0


0.2


0.4


0.6


0.8


1

0.1

0.4

0.7

0.95


0


p


-
p


a

0.85

Enhancement factors

-
20

-
10


0


10


20


30

0.1

0.4

0.95

0.8

0.7


0


-
p/4


a

-
15

-
10

-
5


0


5


10


15

0.1

0.4

0.95

0.7

0.8

0.8

0


p/4


a


-
p/4

Sample
R
s

=|r
s
|
2
=0.85, varying
R
R
, plotting

Need to maximize cavity Q

factor

Only exists in
the cavity case

3

CAVITY SPACING

(nm
)

1900

1950

2000

2050

2100

0.0

0.2

0.4

0.6

0.8

1.0



R
pump

R
probe

Shifts pump resonance (maximum
energy absorption) 5nm away from
probe resonance to 2005nm, where
D
R
probe

is largest.

2005

Measured Reflectivity of

Al film (
R
Al
=0.793) with cavity

1.
Choose
R
R

(0.84 at 800nm)
close to
R
S

2.
Minimize spread in
l, q
, and
w


Maximizing the Enhancement Factors

Cavity R

Spectrum of Laser Light

CAVITY SPACING (nm
)

1500

1600

1700

1800

1900

2000

2100

0.4

0.6

0.8

1.0

Dl
12.5 nm

Laser line filter

Dl
3.2 nm

Dl
12.5 nm

Dl
3.2 nm



Convert spread in the angle to
the spread in the wavelength, by
c
onsidering
k
z

= 2
p
cos
q
/
l
0


Dll
0
tan
q

Dq


Measured Reflectivity of

Cu film (
R
Cu
=0.955) with cavity

Decrease the spread in wavelength
using a laser line filter

Decrease the spread in the
incidence angle of the probe beam

For example
l
0
800
nm


qp/4,
Dq
0.1, Dl  80
nm


q0,
Dq
0.1, Dl  1
nm



Probe needs to be at
normal incidence


Decrease the spread in the spacing
of the cavity


The angle between the two
reflecting surfaces of the cavity needs
to be less than 10
-
4

rad



Probe

Incidence angle 0
o

Pump

Incidence angle 4
o

6/1/2009

CLEO/IQEC’09 Baltimore, Maryland

4

Result of Al film (190 nm) with/
wo

cavity

6/1/2009

CLEO/IQEC’09 Baltimore, Maryland


Enhancement factor
with the use of the
cavity is bigger than
170
.

PROBE TIME DELAY (
ps
)

0

100

200

300

2x10
-
5

4x10
-
5

-
1.5x10
-
3

-
1x10
-
3

-
0.5x10
-
3

(a
) Al film

(b
) Same Al film
with cavity

Measured at
w
=2005nm

1900

1950

2000

2050

2100

0.0

0.2

0.4

0.6

0.8

1.0

R
pump

R
probe

2005



CAVITY SPACING

(nm
)

5


Measured values of
D
R(t)/R for Aluminum film

at 9 cavity spacing in the vicinity of the 5
th

resonance mode.

PROBE TIME DELAY (
ps
)

50

100

-
4x10
-
4

-
2x10
-
4

0

2x10
-
4

1

2

3

4

5

6

7

8

9

1980

1990

2000

2010

2020

0

2x10
-
4

4x10
-
4

6x10
-
4

CAVITY SPACING (nm)

|D
R/R|

In the vicinity of the 5
th

Resonance Mode

6/1/2009

CLEO/IQEC’09 Baltimore, Maryland

9

Cavity Spacing

Curve
#

W
eff

(nm)

R (probe)

D
R/R

1

1978.3

0.775

7.2E
-
6

2

1983.3

0.616

3E
-
5

3

1990.5

0.333

1.42E
-
4

4

1994.8

0.164

3E
-
4

5

1999.9

0.080

1E
-
8

6

2004.9

0.164

-
4E
-
4

7

2009.2

0.348

-
2.57E
-
4

8

2014.3

0.609

-
5.1E
-
5

9

2020.8

0.772

-
1.01E
-
5

Measured
D
R/R values

Calculated

6

-
30

-
20

-
10

0

10

20

30



0.0

3x10
-
6

1.5x10
-
6

-
1.5x10
-
6

PROBE TIME DELAY (
ps
)

D
R/R

-
30

-
20

-
10

0

10

20

30

-
4x10
-
4

-
2x10
-
4

0



D
R/R



PROBE TIME DELAY (
ps
)

1
st

echo of Al
film

Measured
result

Measured
result

1
st

echo of Al
film with cavity
at
w

= 2005nm

Data Fitting

6/1/2009

CLEO/IQEC’09 Baltimore, Maryland

Result of
best fit

Result of
best fit



-
200

-
100

0

100

200

z
(nm)



-
3x10
-
6

0

-
6x10
-
6


3x10
-
6

6x10
-
6

Deduced
strain pulse

We can also determine the
piezo
-
optical coefficients of the film

Compare with the theoretical model

When
z>0

When
z<0


(Thomsen
et al ,
1986)

7

-
30

-
20

-
10

0

10

20

30

-
4x10
-
4

-
2x10
-
4

0



D
R/R



PROBE TIME DELAY (
ps
)

Two contributions to
Da

6/1/2009

CLEO/IQEC’09 Baltimore, Maryland

D
R/R

Surface
displacement

Piezo
-
optic

effect


Surface displacement of
the film

Piezo
-
optical effect

Surface displacement of the
film
D
w

on the order of
10
-
3

nm
can be detected



high sensi瑩vi瑹 wi瑨 瑨e
use of the cavity

Measured
result at
w
=2005nm

8

Cu Thin Film (180
nm
)

6/1/2009

CLEO/IQEC’09 Baltimore, Maryland


Cu is the KEY material for interconnect
wiring in the semiconductor industry.


It is important and interesting to improve
picosecond ultrasonic measurements of
Cu at 800nm.


At
l
=800nm (wavelength of the laser
used in the present experiment, also the
most common wavelength of commercial
compact ultrafast lasers), the piezo
-
optic
coefficients of Cu are zero (Gerhardt,
1968).


When the cavity is not used, the only
contribution to
D
R/R comes from the
piezo
-
optic effect of the film.




NO acoustic signals can be see

0

20

40

60

0

2x
10
-
7

4x
10
-
7

6x10
-
7



PROBE TIME DELAY
(10
-
12

Sec)

D
R/R

9

Cu Thin Film (180
nm
) With Cavity

6/1/2009

CLEO/IQEC’09 Baltimore, Maryland


For Cu at
l
=800nm,
D|
r
S
|
=0,
D

S
=0, because
piezo
-
optical coefficients are zero.


Only contribution to
D
R
(
t
)
/R

from surface displacement
D
w


Theoretical model




where

1

is the returning strain


pulse using theoretical
(Thomsen) model,



’ the diffusion length.


Measured result smaller and
broader


due to attenuation or
approximation for the initial
stress


0


1x10
-
4


2x10
-
4


0


50


100


150


200


250

PROBE TIME DELAY (
ps
)

D
R/R

Calculated result
using

’=150nm

Calculated result
using

’=100nm

Measured result

at
w

= 1606nm
background removed

10

Summary

6/1/2009

CLEO/IQEC’09 Baltimore, Maryland


Significant signal enhancement,
bigger than 170
for Al film


Able to perform measurements on materials with small
piezo
-
optical
coefficients,
Cu, key material in semiconductor industry


Able to determine the actual shape of the returning acoustic pulse


enriching the total amount of information in the picosecond
ultrasonic measurements


Sensitive to very small displacement ~10
-
3
nm


The technique can be used for a wide variety of ultrafast optical
pump
-
and
-
probe experiments.

Acknowledgement


Zygo

Corporation for Grant Support


Fan Yang, Tom
Grimsley

for helpful discussions

11

Measurement of the Effective Cavity Spacing

6/1/2009

CLEO/IQEC’09 Baltimore, Maryland

2.8
m
m

3.0
m
m

3.2
m
m

12