on IntensityDiffracti

Nanometrology: enabling applications Nanometrology: enabling applications

of nanotechnology

C M Sotomayor Torres*, T Kehoe, C M Sotomayor Torres , T Kehoe,

N Kehagias, V Reboud, D Dudek

*ICREA

Phononic and Photonic Nanostructures Groupp Catalan Institute of Nanotechnology (CIN2-CSIC)

Barcelona SPAIN

Collaborators & Support

J Bryner, J Vollman, J Dual,

C Mechanics, ETH, Zurich

S Landis,

CEA, LETI , Grenoble

C Gourgon

LTM CNRS Grenoble

C Gourgon,

LTM-CNRS Grenoble

S Arpiannen, J Ahopelto,

VTT Espoo, Finland

W Khunsin, S G Romanov, A Amann, E P O’Reilly, G Kocher

R Zentel,

^{U Mainz}

S Pullteap and H C Seat, ENSEEIHT, Toulouse

Outline

1. Introduction

2. Nanometrology for Nanoimprint lithography: g p y

a. Sub-wavelength diffraction b Photoacoustic metrology b. Photoacoustic metrology

3. Nanometrology for Self-assembly

a. Opposite partners/ elements b. Rotational diffraction

4. Conclusions

Nanometrology and Nanotechnology

• Critical to enable the industrial uptake of nanotechnology

• Necessary to measure:

– Product characteristics device performance toxicology Product characteristics, device performance, toxicology (potential public health risks), product lifetime, security

• Requirements for manufacturable technology

• Requirements for manufacturable technology

– Standardisation, regulation – repeatable and universal – Easy to operate

– Developed in coordination with manufacturing techniquesp g q

• Integrated, in-line, real-time, advanced process control

• Relevant measurands

Nanometrology Challenges

• Miniaturisation – things are getting smaller

• Heterogeneous integration – things are getting more complex

– 3^rd Dimension increasingly used – Dimensions and material properties

• Insufficient standardisation of techniques or reference samples

sa p es

• Existing methods are slow, often destructive and not optimised for 3D

optimised for 3D

Metrology challenges for Nanofabrication

• Critical dimension measurement < 50 nm

S i d t lith h d 32 (2011)

Intel SRAM

Semiconductor lithography node 32 nm (2011)

Critical dimensions and physical properties ^45nm

• 3D Structure

Complex: Typical of heterogeneous integration, interconnects

• In-situ, inline, real-time

S. Landis et al, Nanotechnology (2006) Ye et al, Langmuiir 22 7378 B.Chao, Proc SPIE 6921 (2008) (2006).3D Photonic crystal in Si

, ,

Advanced Process Control of systematic drift and process behaviour

Metrology techniques for nanoscale – limitations

• SEM

 For height, slope, profile, requires destructive cross section

destructive cross-section.

• AFM

 Difficult to access sidewalls, corners, l ti l l

1m

A.E. Vladar et al, Proc SPIE 69220H

relatively slow

• TEM

 Resolution ~ 0.1 nm.

 Destructive, slow

• X-ray Imaging

 Requires synchrotron x-ray B.C. Park et al, Proc. SPIE 651819

X ray Imaging

 Requires synchrotron x ray source

• Optical Scatterometry

• Optical Scatterometry

Wi t h N T h l Ltd

 Requirement of wavelength, polarization or angle variability.

C. David et al, Proc. MNE 2008

Wintech Nano-Technology Ltd

Nanoimprint Lithography (NIL)

Advantages

• Resolution (sub 10 nm)

Stamp (Si, Quartz, etc)

Resist (polymer monomer) Resolution (sub 10 nm)

• Fast (sec/cycle)

• Low cost ($0.2M vs $25M)

• Simple

Substrate

Resist (polymer, monomer)

mp

• Flexible (UV, heat) Applications

Imprint

(Pressure +heat or UV light)

• Semiconductors

• Optics

• Bio

Release (cool down )

Bio

• Organic electronics

• Sensors

Hi h l ti C l tt F ti l d i

RIE of residual layer

High resolution Complex patterns Functional devices

N.Kehagias, Nanotechnology 18 (2007) V.Reboud, Jpn. J. Appl. Phys., 47 (2008)

Sub-wavelength diffraction metrology

• Test structures of blazed gratings  asymmetric diffraction pattern

• Individual lines < diffraction limit Groups of lines > diffraction limit

• Individual lines < diffraction limit. Groups of lines > diffraction limit

• Sub-wavelength features in grating eg Line-width, height, defects, sidewall angle, curvature. Linewidths: 50, 100, 150, … 350nm

• Defects affect relative intensity of diffraction orders in far-field

• Suitable for transparent or opaque structures

• Non-destructive, Fast collection of data



Gap 58 nm on Intensity -2 -1 0 +1 +2 Gap 58 nm

Diffracti

Line 53 nm 1 m

T. Kehoe et al, Proc. SPIE 69210F, 2008

Modelling of Sub-wavelength diffraction

• Rigorous Coupled Wave Analysis

(RCWA) ^4.0

4.5

n

Height

( )

• Finite difference time domain

(FDTD) ^2.02.5

3.0 3.5

ve diffraction fficiency

+first/-first

0.0 0.5 1.0 1.5

0.07 0.08 0.09 0.10 0.11 0.12 0.13

Relativ ef first/ first

+first/+second

500nm

line height/ um

Width

20 25

units)

FDTD vs. RCWA normalised to first order

FDTD RCWA

4.0 5.0 6.0

raction cy

+first/-first +first/+second

10 15 20

(arbitrary ^RCWA

1.0 2.0 3.0

elative diffr efficienc

0 5 10

Efficiency

0.0

0.00 0.01 0.02 0.03 0.04 0.05 0.06

Re

Line width increase / um

-3 -2 -1 0 1 2 3

E

Diffracted Orders

Sub-wavelength diffraction measurement system

Laser

• In-line design

• Sub wavelength diffraction

Sample

• Sub-wavelength diffraction

metrology with surface imaging by microscope optics

Sample

CCD A – diffraction  pattern z x

y

Microscope  objective

• Enables centring of the laser spot on the gratings

j

CCD B–

image of surface 250

150 200 250

bitrary units +1

50 100

ntensity / arb

-2 -1

+2

25 m

0

0 200 400 600 800 1000 1200 1400

In

Pixels

Sub-wavelength – Detection of defects

1 1.2

au

Measured (+1/-1) = 2.85 Simulated (+1/-1) = 2.68

• Defect detection

– missing 50 nm & 100 nm lines

0.6 0.8

n Efficiency /

• Agreement between simulated and measured diffraction efficiencies

0 0.2 0.4

Diffraction

Simulated ^0.6

0.8 1 1.2

n efficiency /au

No Defect (RHS/LHS) = 2.59 Defect (RHS/LHS) = 1.44

0

-3 -2 -1 0 1 2 3

Diffraction Order 0

0.2 0.4

-3 -2 -1 0 1 2 3

Diffraction

1 m

0 8 1 1.2

cy / au

No defect (RHS/LHS)

= 3.53

Defect (RHS/LHS) =

Measured 2 11

Diffraction Orders

0.2 0.4 0.6 0.8

Diffraction eficienc 2.11

100 nm 50 nm

0

-3 -2 -1 0 1 2 3

Diffraction Order T. Kehoe, Microelec. Eng. 86 (2009)

Sub-wavelength – Detection of defects

Line 53

Gap 58 nm • Grating stamps made with average line-width from 193 –

214 nm

1 m

Line 53

nm • Measured and modelled diffraction efficiencies (1^st & 2^nd order) decrease with increasing line-width, by approximately 5% per 5 nm

1/ 2 l ti diff ti ffi i d t

-2 Simulated on

• +1/+2 relative diffraction efficiency decreases at approximately 4% per 5 nm

3

on Efficiency

-2 Simulated -1 Simulated +1 Simulated +2 Simulated -2 Measured -1 Measured

1 M d

1.1

Diffractio

+1/+2 Simulated +1/+2

M d

1 2

zd Diffractio +1 Measured

+2 Measured 1

d Relative Efficiency Measured

0

195 200 205 210 215

Normaliz

0.9

195 200 205 210 215

Average Linewidth / nm

ormalized E

95 00 05 0 5

Average Linewidth / nm

g

No

T. Kehoe, NNT 08 conference, Kyoto, Japan

Diffraction from 3D Structures

• Polymer relaxes and partially reflows, creating rounded line profiles

• Measured diffracted order intensities

• Measured diffracted order intensities before and after reflowing of the lines

Comparison of measured and simulated diffraction intensities for imprinted lines and reflowed lines

T. Kehoe, NNT 2010, Copenhagen, Denmark

Key steps for NIL – Polymer physical properties

1. Stamp fabrication

^Imprint (Pressure +heat or UV light)

2. Imprinting process: Temp, Pressure, UV

 Glass transition temperature, Viscosity

 Glass transition temperature, Viscosity 3. Demoulding: Mechanical strength

 Y ´ d l P i ’ ti

 Young s modulus, Poisson’s ratio Adhesion / anti-sticking coating

 f

Release (cool down )

 surface energy

4. Etching: Polymer etch resistance tc g o y e etc es sta ce

At nanoscale values may change

RIE of residual layer

Photoacoustic Metrology

• Thickness measurements, resolution ~10 nm

• Acoustic scattering from interfaces  changes surface reflectivity

• Acoustic scattering from interfaces  changes surface reflectivity

• Acoustic speed  Physical parameters – Modulus, Poisson’s

P P b L 70 f  810 Ti l ti 0 1

• Pump-Probe Laser, 70 fs,  = 810 nm, Time resolution 0.1 ps

Partial Reflection from interfaces

Propagating acoustic waves

Pump: 70fs laser pulse Probe: Optical

reflectivity change at y g surface

Laser absorption / generation

Al Polymer Si

p g

of thermal stress – acoustic wave

One-dimensional photoacoustic model

• Finite Element Simulation, including viscoelastic damping

• Measurement R + Thickness (Ellipsometry) + Optical / Physical properties (absorption, density)  Thermomechanical model

S laser power σ_excit stress, T temperature, ε strain z depth

t time  attenuation F sensitivity

ΔR reflectivity change

Photoacoustic Metrology of Nanoimprint Polymers

336 nm PMMA

• Nanoimprint polymers

13 – 586 nm thick layers

ity change

13 586 nm thick layers

• Damping in polymer not excessive

• Good acoustic impedence

Reflectivi R

difference  Strong signal

• Top interface: Al/polymer

• Bottom interface: polymer/Si

Time / ps

• Bottom interface: polymer/Si

• FilmThickness compared to ellipsometry and profilometry

ness / nm

• Physical parameters calculated using Finite Element model

c E 

Thickn

c_p (m/s)

E

(GPa)  

kg/m³)

mr-I PMMA 2603 3.2 0.4 1012

time of flight Time / ps

mr-NIL 6000 2504 2.95 0.4 1008

J. Bryner, 2007 IEEE Ultrasonics Symposium, 2007 p 1409

Nanoscale Effects

• Acoustic speed and Young’s modulus increase below 80 nm

• Acoustic speed (c ) increases by 12%

• Acoustic speed (c_p) increases by 12%

• Young’s modulus (E) increases by 26% at 13 nm.

P i l f HMDS (H th ldi il ) dd d  ll i

• Primer layer of HMDS (Hexamethyldisilazane) added  smaller increases

• Increase probably due to interface effects rather than confinement

i d

Acoustic speed

2500 3000 3500

/ m/s

5 6

/ GPa

Young’s modulus

1000 1500 2000 2500

stic speed

PMMA

2 3 4

's Modulus

PMMA 0

500 1000

0 20 40 60 80 100 120 140

Acous

PMMA with HMDS

0 1 2

0 20 40 60 80 100 120 140

Young'

PMMA

PMMA with HMDS

Thickness / nm ^{0 20 40 60 80} 100 120 140

Thickness / nm T. Kehoe, Proceedings of SPIE Vol. 6921 (2008)

Raised Temperature Measurements

90 deg

• Investigation of physical properties

approaching glass transition temperature, T_g app oac g g ass t a s t o te pe atu e, _g

• Acoustic speed inversely proportional to thickness

• Close to T_g increase of noise due to buckling of aluminium

Examples of self-assembled structures

Block 10 nm

VTT Protein crystals

Block

copolymer nanophase separation separation

3D periodic structures: eg. photonic crystals

Layer-by-layer Microassembly Autocloning

S Y Lin Sandial Lab 1998 K Aoki Semicon Lab 2003 S Y Lin, Sandial Lab, 1998 K. Aoki, Semicon Lab, 2003

Holography Direct laser writing Artificial Opals

S Kawakami, Tohoku, 1997

K. Aoki, Semicon Lab, 2003 M. Deubel, Karlsruhe, 2004 Tyndall/ICN-CIN2

FCC colloidal crystals: Improving structural order

Equilibrium

crystallisation Fast crystallisation Reduce interaction with substrate -> reduce crack

Shear force to improve crystal ordering

Improved quality using acoustic noise

Acoustic vibration

A .Amman et al Proc. SPIE Vol. 6603, 660321 (2007)

Quantifying order in self-assembly

• Define scale

M k h tibl ith i ti th d t l t

• Make approach compatible with existing methods or at least acceptable

• In-line or a posteriori?p

• Reliable?

• Suitable for a standard?

2 μm

L = 0 dB L = 20 dB L = 30 dB L = 40 dB

 L = 20 dB is calibrated to water displacement of 2.5 μm

Concept of “opposite beads”

p(r) – probability of finding an opposite beads within a radius r, for a given tolerance parameter ε for the exact location of g p

the spheres

At sphere ‘A’

ε _{( )}

^,

^{( ) ( )}

( )

A B C r

AB AB AC

p r AB

 









 



  

(  )

A B_



r

AB



1 if R

 

 

( ) 1

y 0

if R y R

else

 _{ }^{ }





 

 ( )

) ( )

) (

( N r

r p r r N

p

A A

A A

Global sum: A

weighted average



Conditions met by p(r)

• Scalar quantity (dependence on certain predefined orientation undesirable)

orientation undesirable)

• Integral measure of a locally observable quantity

• Based on actual position of sphere (not on pixel representation of SEM image: contrast & focus dependent)

dependent)

• Robust against missing spheres.

Perfectly ordered system

0 04  ( ) 1 00 

Th

( ) ( ) ( )

^A

N r p r

^{A A}

1

p r   

0.04  p r ( ) 1.00  0.06  p r ( ) 0.46 

Theory

Experiment

( ) 1

A A

( )

p r  N r 

 ^{Experiment p ( )}

SEM Characterisation

SEM images of size 65 μm x 40 μm (resolution: 3072 x 2304 pixels)

substrate

Drawing direction

Opal

Position (P)

Position (P)

Stochastic-resonance in photonic crystal growth

D = 368 nm

r = 5.5 μm ≈ 15D, and ε = 43 nm ≈ 0.12D

P7 P7

P10

Position (P)

Position (P)

3 D ordering - Experimental approach

3D analysis 2D analysis

θ – incident angle

SEM images g

φ – azimuth angle

SEM images

Transmission spectra

2* d (1 i (

2

)



Bragg’s Law

X

U ⁽¹¹¹⁾

(002)

(1 11)

2* n d

_{eff hkl}

(1 sin (

2 _hkl

)

   

L W U

X г

(111)

(200) K ⁽⁰²⁰⁾

L'

(200)

(11 1)

K ^{( )}

W.Khunsin, Adv. Funct.

Mater. 18 (2008)

Rotational symmetry of T(φ)

Without Noise Noise

X ⁽⁰⁰²⁾

L

U ⁽¹¹¹⁾

(111)

( ) (1 11)

L W U X

г

(200) K ⁽⁰²⁰⁾

L'_{(11 1)}

Noise

W Khunsin et al, J. Nonlinear Optical.

Physics & Materials 17 97 (2008)

Noise Susceptibility: lattice planes dependency

Without Noise

10x 5x

With NoiseWith

(311) (220) (111)

Conclusions

• New methods presented to characterise nanostructures fabricated by

N i i t Lith h (NIL) d lf bl

Nanoimprint Lithography (NIL) and self-assembly.

• Sub-wavelength diffraction found sensitive to defects, line-width and f

profile

– This is potentially in-line metrology method.

• Photoacoustic metrology demonstrated suitable for dimensional and physical measurement of printed structures

• We propose a robust and generic approach to analyse quantitatively two-dimensional lattice ordering.

– Opposite partners

– Rotational diffraction symmetry