development status of euvl blank and substrate6 2010 international workshop on euv lithography jun /...

1

2010 International Workshop on EUV Lithography Jun / 21-25 / 2010

Development Status of

EUVL Blank and Substrate

Asahi Glass Company

Toshiyuki Uno

2


Contents

1. Introduction

2. Blank defect reduction

1. Inspection capability

2. Substrate

3. ML blank

4. Absorber

3. Integrated & Best performance of ML Blank

4. New material development

1. New capping film

2. 193nm ARC

5. Road map & Summary

3


1. Introduction

4


1-1 EUVL mask blank structure

EUVL mask blank has a stack of reflective, capping, absorber and anti-reflective layers on its front side for a patterning and a conductive layer on its back side for mask chucking.

Glass substrate

Conductive layer for

electrostatic chuck

Reflective multilayer

Capping (and/or Buffer) layer

Absorber layer

Antireflective layer

Structure of EUV mask blank

Resist film

5


Material synthesis

Lapping&Polishing Lap and polish to smooth & flat

surfaces

1-2 EUVL mask blank manufacturing process

Flame hydrolysis of Si and Ti source materials to

form the Ti-doped SiO2 glass ingot

Slicing Slice the glass ingots into 6025 plates

Cleaning Remove particles and polishing slurries

Film deposition Deposit reflective, capping, absorber and ARC

films

AGC is taking care of all processes essential to EUV mask blank: LTEM material, polishing, cleaning, film deposition and resist coat.

Resist coat

6


Metrology tools Pilot line

CTE

measurement

Dilatometer (ASET model)

Defect inspection Lasertec M1350A

Defect analysis SEM/EDS

Surface flatness Interferometer

Surface

roughness

AFM

Zygo Newview

EUV

reflectometer

Stand Alone

DUV~NIR

reflectometerHitachi, Shimadzu

In addition to these, we also have various kinds of defect characterization tools

(u-FTIR, u-Raman,…), film characterization tools (XRR, XRD,…) and dry etcher.

1-3 EUV mask blank pilot Line ~ Metrology tool

AGC possesses critical metrology tool sets, which make it possible to realize the total quality improvement of blanks.

7


2. Blank defect reduction

8


The real size of defect increases with the relative defect size (pixel) provided by the defect inspection tool (M1350A)

Pixel 3 corresponds to 34nm SEVD (Sphere equivalent volume diameter), Pixel 8 to 56nm SEVD

2-1 Inspection and size of native defects

Depth or

Height

SEVD

Full width

Native defect Equiv. Sphere

d

0

50

100

150

200

0 5 10 15 20

Actu

al defe

ct siz

e

(nm

)

Relative defect size (M1350A pixel)

Volume by AFM

M1350A pixel vs real defect sizeTypical defect map and histogram

of M1350A inspection

9


We have been optimizing the polishing conditions by changing processes parameters, conditioning and materials to reduce the substrate pit.

The results shown here were ML pit count & relative yield (10Q1-2 = 100) of QZ test plates. We demonstrated 0 pit @56nm SEVD ( sphere eqiv. volume diameter)

2-1-1 Blank defect reduction ~ Substrate pit ~

0 0

48

100

0

47

64

100

0

20

40

60

80

100

Rel

ativ

e Y

ield

0 pit yield @ 56nm SEVD

<=15 pit yield @ 34nm SEVD

08Q4 09Q1 09Q3 10Q1-2

2 1 0 0

16

8 97

0

5

10

15

20

ML p

it c

ount

Best pit count @ 56nm SEVD

Best pit count @ 34nm SEVD

08Q4 09Q1 09Q3 10Q1-2

ML Pit defect count trend ML Pit defect yield trend

10


The average defects counts on LTEM substrates @ p1+ and relative yield of 0 unremoval particle were shown below.

To reduce un-removal particles, we installed the new intermediate cleaning tool in 09Q3. Consequently, the number of un-removal particles on LTEM substrate has been drastically decreased.

We expect that “added particle” will be decreased further by optimizing the final cleaning processes.

2-1-2 Blank defect reduction ~Substrate particle ~

Average defect counts of LTEM substrates

0

2

4

6

8

10

12

14

16

18

08H1 08H2 09H1 09H2 10Q2

Ave

rage

de

fect

co

un

t@p

1+

added particle

unremoval particle

Pit

11


2-2 Blnak defect reduction ~ LTEM-ML blank ~

Target

This is the best defect trend of the LTEM-ML blank where

the substrate flatness is <150mm.

Best defect trend of LTEM-ML blank

(Lasertec M1350A inspection)

12


We have improved the absorber adder defect performance.

• The adder defect count has been decreased by optimizing

coating equipment.

• We have obtained several plates with no adder defect

larger than 100nm SEVD.

Adder defect map

& Histogram

2-3 Blank defect reduction ~ Absorber ~

100nm SEVD

Average adder defect counts

09H2 10H1

~200nm SEVD

89nm SEVD

13


3. Integrated & Best performances

14


Structure

CrN/LTEM/ML/Ru

CTE of LTEM

Mean CTE -3.5 ppb/K

CTE variation (PV) 1.7 ppb/K

Substrate flatness

front 89 nm

back 76 nm

Blank Flatness(Bow) 0.38 um

Reflectivity

Peak R 66.2 %

R range(Abs.) 0.3 %

centroid l 13.535 nm

l range 0.026 nm

3-1 Integrated performance of LTEM-ML blank

0.12cm-2 @ >34nm SEVD

Defect map and histogram

0.05 cm-2 @ >56nm SEVD

15


Requirements

(SEMI P37,38) current best

Substrate

CTE mean +/-5 ppb/K 0 ppb/K

CTE variation +/-3 ppb/K +/- 1ppb/K

flatness (front) <30 nm 38 nm on LTEM

flatness (back) <30 nm 48 nm on LTEM

HSFR (RMS) <0.15 nm <0.10 nm

ML/Capping

Defect (ML/Cap)0.00 c/cm2

@30 nm0.04@56nm (SEVD)

Peak Reflectivity >67 % 67.1

Mean Centroid λ n/a 0 to target

FWHM >0.50 nm >0.50 nm

3-2 Best performance of LTEM -ML BlankWe can mostly achieve the requirements other than defect.

Green : met SEMI spec. yellow : TBD Red : need development

16


4-1. New capping film

AGC has developed the new Ru film, which showed better

chemical & thermal durabilities than the conventional (current) Ru film.

17


4-1-1 EUV reflectivity change upon O3-DIW cleaning

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

12.5 13.0 13.5 14.0 14.5

Wavelength (nm)

EU

V R

efle

ctivity

as-depo

post-cleaning

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

12.5 13.0 13.5 14.0 14.5

Wavelength (nm)

EU

V R

efle

ctivity

as-depo

post-cleaning

66.6 %

66.0 %

-0.6 %

66.8 %

63.7 %

-3.1 %

Fig. EUV reflectivity spectrum before and after O3-DIW cleaning (~30ppm x 10min)

Conventional Ru

(~2.5nm)

New Ru (~2.5nm)

The new Ru capped ML showed the better durability upon the wet cleaning using

ozonated water (~30ppm) than the conventional Ru (2.5nm).

In the new Ru capped ML blank, the reflectivity drop was <1% and the reflectivity after

cleaning was >65%. This effect of the new Ru has been confirmed repeatedly.

18


No clear difference of roughness was observed in both new Ru and

conventional Ru films.

4-1-2 Roughness change upon O3-DIW cleaning

Conventional Ru (~2.5nm)New Ru (~2.5nm)

As-depo

Post-Cleaning

Rms=0.94nm Rms=1.02nm

Rms=0.82nm Rms=0.99nm

Fig. Surface roughness of

QZ/ML/Ru

19


4-1-3 EUV reflectivity change upon thermal treatment

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

12.5 13.0 13.5 14.0 14.5

Wavelength (nm)

EU

V R

efle

ctivity

as-depo

post-bake

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

12.5 13.0 13.5 14.0 14.5

Wavelength (nm)

EU

V R

efle

ctivity

as-depo

post-bake

Conventional Ru (~2.5nm)New Ru (~2.5nm)

66.6 %

62.6 %

-4.0 %

66.4 %

58.1 %

-8.5 %

Fig. EUV reflectivity spectrum before and after the thermal annealing (210oCx10min, in air)

The new Ru film also showed the better thermal durability in air than the

conventional Ru film.

20


4-2 193nm ARC

21


Standard Candidate 1 Candidate 2

Structure

R%@193nm > 20% < 7% < 7%

Feature ARC for 257nnmThinner ARC

Lower Etching Rate

Thicker ARC

Higher Etching Rate

4-2-1 193nm ARC candidate material

ML

TaON ARC 7nm

Ru

TaN Absorber

ML

HfON ARC ~10nm

Ru

TaN Absorber

ML

SiN ARC ~10nm

Ru

TaN Absorber

TaON ARC ~1nm

ML

Ru

TaN Absorber

TaON ARC 7nm

ML

TaON ARC 7nm

Ru

TaN Absorber

ML

TaON ARC 7nm

Ru

TaN Absorber

ML

Ru

TaN Absorber

ARC-1 8nm

ML

TaON ARC 7nm

Ru

TaN Absorber

ML

Ru

TaN Absorber

ARC-2 12nm

AGC released the blanks with 2 types of New ARC layers for 193nm mask inspection.

22


4-2-2 Candidate 1 ~DUV performances~

This the measurement result of DUV spectra of TaN(51nm)/ARC-1(6~10nm). The samples more than 8nmthickness shown here had <7% reflectivity at 193nm.

Fig. DUV reflectivity spectra

0

5

10

15

20

25

30

190 210 230 250 270 290

wavelength (nm)

Re

flectiv

ity (

%)

TaN/HfON=51/6nm

TaN/HfON=51/8nm

TaN/HfON=51/10nm

193nm

257nm

TaN(51nm)/ARC-1（6nm）


TaN(51nm)/ARC-1(10nm)

23


0

5

10

15

20

25

30

190 210 230 250 270 290

Re

fle

ctivity (

%)

wavelength (nm)

4-2-3 Candidate 2 ~DUV performances~

This is the measurement result of DUV spectra of TaN(51nm)/ARC-2(11~13nm). All three samples shown here had <7% reflectivity at 193nm.Absorber with ARC-2 can be etched by one-step with Cl2 chemical.

Fig. DUV reflectivity spectra

193nm

257nm

TaN(51nm)/ARC-2 （11nm）



24


4-2-4 Durability to wet cleaning & thermal

Durability performances for 60min were investigated .Both materials candidates showed no degradation upon H2SO4, SC1 and 150oC thermal

annealing for 60min.

0

1

2

H2SO4(98%) SC1 Heat(150C)

Re

fle

ctivity C

ha

ng

e (

%)

TaN/ TaON (Standard) @ 257nm

TaN/ SiN/TaON @ 200nm

TaN/ HfON @ 200nm

Fig. Durability performances : DUV bottom reflectivity changes

Durability Test Condition

SC1 : 40C 60min Dip

H2SO4(98%) : 40C 60min Dip

Heat: 150C 60min in Air Better

WorseTaN/TaON (standard) @ 257nm

TaN/ ARC-2 @ 200nm

TaN/ ARC-1 @200nm

0

25


5. Roadmap of EUVL Blank development

26


5-1 Roadmap

2009 2010 2011 2012 2013 2014

H1 H2 H1 H2 H1 H2 H1 H2 H1 H2 H1 H2

1st generationSubstrate : LTEM w/zero CTE@22oC

(Flatness<100nm)

Capping : Ru or New Ru

Absorber : TaN (51nm or 77nm)

ARC : TaON or 193nm ARC

2nd generationSubstrate : LTEM w/ zero CTE

@ high temperature?

Capping : New Ru

Absorber : New material (30~40nm)

ARC : TaON

or 193nm ARC

or no ARC

HVMProcess development

at Pilot line

HVMProcess developmentMaterial

development

Any needs for LTEM w/ zero CTE at elevated temperature?

AGC is carrying out the process development of 1st generation blank. AGC is also

developing the material of 2nd generation blank which will be suitable for 16nm hp and off-

axis illumination.

27


5-2 Summary

The tool upgrade and improvement of polishing and

intermediate cleaning were carried out, which resulted in the

significant improvement of substrate defect performance

and yield.

Continuous improvement and integration are in progress

at AGC and single defect counts on ML blank @ 54nm

SEVD has been achieved.

AGC successfully introduced New-Ru capping film and

193nm ARC to the industry.

development status of euvl blank and substrate6 2010 international workshop on euv lithography jun /...

Documents