Accelerating the next technology revolution

Copyright ©2009

SEMATECH, Inc. SEMATECH, and the SEMATECH logo are registered servicemarks of SEMATECH, Inc. International SEMATECH Manufacturing Initiative, ISMI, Advanced Materials Research Center

and AMRC are servicemarks of SEMATECH, Inc. All other servicemarks and trademarks are the property of their respective owners.

Mechanisms of EUV reflectivity loss

from the multiple cleanings

Abbas Rastegar

Byunghoon Lee,

Arun John Kadaksham,

SEMATECH Albany

October 2011

Special thanks to Thomas Laursen (ASML)

SEMATECH Confidential

Outline

• EUV reflectivity loss from multiple cleanings

• Mask blank surface changes from cleaning

– Si diffusion and oxidation

– Ru oxidation

• Contribution of cleaning processes to EUV

reflectivity loss

– Radiation effects

– Chemistry effects

– Megasonic effects

– Combined effects

25 October 2011 2

SEMATECH Confidential

How many times can an EUV mask be

cleaned?

25 October 2011 3

• EUV reflectivity (Rmax) dropped below target after cleaning 20X

±0.5%

Ru layer

+ 4 ML

removed

• What is the mechanism of EUV reflectivity loss?

SEMATECH Confidential 25 October 2011 4

Which cleaning steps oxidize/remove Ru?

• Which steps oxidize Ru and/or Si?

• Which steps remove Ru or Si?

• Is there any combined effects?

– (VUV+ SPM), (VUV+ APM), (SPM+APM)

• Does megasonic contribute to the chemistry effects?

• Verification method:

– Change in EUV reflectivity at each step

Oxidizing steps Etching steps

Physical interactions

APM: (NH4OH/H2O2/H2O)

SPM: (H2SO4/H2O2)

VUV: ( Vacuum UV =172 nm)

SEMATECH Confidential 25 October 2011 5

EUV reflectivity loss from individual

cleaning steps (50X)

• EUV reflectivity loss is maximum when SPM follows VUV

exposure

After 50 x clean

0 0.5 1 1.5 2 2.5 3 3.5

VUV

SPM

VUV+APM

VUV+SPM

VUV+SPM+APM

EUV Reflectivity loss (%)

VUV+SPM+ APM

VUV+SPM+ APM

VUV+SPM+ APM

VUV+SPM+ APM

Standard process

+ Megasonic

+ Megasonic

SEMATECH Confidential 25 October 2011 6

EUV reflectivity loss by individual cleaning

steps (50X)

• EUV reflectivity drops linearly with exposure time

After 50 x clean

0 0.5 1 1.5 2 2.5 3 3.5

VUV

SPM

VUV+APM

VUV+SPM

VUV+SPM+APM

EUV Reflectivity loss (%)

VUV+SPM+ APM

VUV+SPM+ APM

VUV+SPM+ APM

VUV+SPM+ APM

Standard process

+ Megasonic

+ Megasonic

y = -0.0577x - 0.5037

R2 = 0.7154

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

5X 10X 15X 20X 25X 30X 35X 40X 45X 50X

Cleaning Cycle

Ch

an

ge

in

EU

V R

efl

ec

tiv

ity

(%

)VUV(Ar)

SEMATECH Confidential 25 October 2011 7

Si oxidation on top of Ru from

exposure to VUV( =172 nm)

TOFSIMS

Surface oxidation

• Diffused Si on top of Ru is further oxidized by VUV radiation

( ratio of SiO2 / SiO increases)

SEMATECH Confidential 25 October 2011 8

VUV(172 nm)+ NH4OH/H2O2/H2O

(APM)

• VUV + APM is more aggressive than VUV alone.

• Why EUV reflectivity loss is less for (VUV+APM)?

SEMATECH Confidential

Si oxidation from cleaning process

25 October 2011 9

• Greater Si oxidation from (VUV+ APM) than VUV

• APM oxidizes Si and etches SiO2

VUV+SPM

SPM

VUV+ APM

VUV

No clean

SiO2 increase

deeper by

VUV+SPM

With contribution of Thomas Laursen

SEMATECH Confidential

Study of Ru oxidation by TOFSIMS and Auger

( freshly deposited Ru cap multilayer)

25 October 2011 10

• Auger depth profile analysis shows Ru, Si, and O on the

surface, indicative of both Ru and Si oxides

• Ru oxide in all 3 oxidation valances is found on surface of the

fresh Ru blank after deposition (RuO2>Ru2O3>RuO): (9.5:8:1)

Auger TOFSIMS

With c contribution of Thomas Laursen

SEMATECH Confidential

Ru oxidation from the cleaning process

25 October 2011 11

• (VUV+ SPM) leaves the most Ru2O3 on the surface

• All steps with wet chemistry oxidized the Ru surface

• VUV in Ar atmosphere has a small contribution to Ru oxidation

VUV+SPM

SPM

VUV+ APM

VUV No clean

With contribution of Thomas Laursen

SEMATECH Confidential 25 October 2011 12

• Megasonic cleaning of blanks creates pits on the blank

surface

Megasonic induced pits

Average EUV Reflectivity Change by Cleaning

60

61

62

63

64

65

66

3 9 15 24 30 36 42 48 54 60 66 72

Cleaning Cycle

Pe

ak

Re

fle

cit

vit

y (

Rm

ax

av

era

ge

)

Megasonic contribution to EUV reflectivity loss

F/B process

TEM: Jena Harris Jones

SEMATECH Confidential 25 October 2011 13

Dependence of Megasonic pits on Megasonic Power

y = 11.005e0.163x

y = 17.691e0.1076x

0

500

1000

1500

2000

2500

3000

3500

4000

0 10 20 30 40 50 60

Cleaning cycle

Ad

de

r (p

it)

co

un

t

High power MS

Low Power MS

Cleaning-induced pit dependency on

megasonic power

• Higher megasonic powers result in faster pit creation

F/B process

SEMATECH Confidential

Improvement in mask lifetime using softer

cleaning processes

25 October 2011 14

• Softer cleaning processes can clean the mask up to 80X

• Softer clean processes:(no VUV) + lower megasonic powers

±0.5%

SEMATECH Confidential

Chemistry vs. megasonic contributions

25 October 2011 15

• VUV( =172 nm) and megasonic contribute the most to EUV

reflectivity loss

SEMATECH Confidential

Summary • The top surface of the Ru cap is a mixture of Ru, oxygen, Si and all

possible combined oxides and silicides (RuxOy, SixOy,RuxSiyOz)

• VUV radiation

– Oxidizes Si; therefore, EUV reflectivity drops when the Si oxide layer

becomes thicker

– Oxidizes Ru slightly

• (VUV+APM)

– APM will etch the silicon oxide layer formed by VUV radiation; therefore,

EUV absorption will be minimal

• SPM

– Mainly oxidizes Ru

• (VUV+SPM)

– Strongly oxidizes Ru and severely impacts EUV reflectivity

– Not clear which Ru-containing compound has highest EUV absorption

• Megasonic exposure compromises Ru layer integrity, which results in

oxidizing chemistries attacking the multilayer

25 October 2011 16

SEMATECH Confidential

Conclusions • UV light radiation and cleaning chemistries modify the surface of the

Ru-capped multilayer, increasing both the scattering and absorption of

EUV light

• VUV light not only oxidizes the Si on the top surface but also modifies

Ru oxides and silicides such that it will be further oxidized by strong

oxidant agents such as sulfuric acid and hydrogen peroxide

• At this time, it is not clear which chemical compound (Ru oxides, Ru

silicides, or a combination) severely reduces EUV reflectivity

• Therefore, the combined effects of VUV+ chemistry and exposure to megasonic are responsible for degrading capping layer integrity

25 October 2011 17

SEMATECH Confidential 25 October 2011 18

Thank You

Acknowledgement:

Cleans Program: Matt House, Zak Robinson, Thomas Laursen (ASML)

Metrology: CC. Lin, Tonmoy Chakraborty, Jana Harris Jones,

Steve Novak (CNSE)

Operation Support: Anne Rudack and all MBDC technicians

SEMATECH Confidential 25 October 2011 19

EUV reflectivity loss by individual steps

• Only VUV in Ar atmosphere and VUV+ SPM reduce reflectivity

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

5X 10X 15X 18X 23X 28X 33X 38X 43X 48X

Cleaning Cycle

EU

V r

efl

ec

tiv

ity

Ch

an

ge

(%

)SPM VUV(Ar)+APM

y = -0.0577x - 0.5037

R2 = 0.7154

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

5X 10X 15X 20X 25X 30X 35X 40X 45X 50X

Cleaning Cycle

Ch

an

ge

in

EU

V R

efl

ec

tiv

ity

(%

) VUV(Ar) VUV(Ar)+SPM