EUV Endurance Testing of Ruthenium- Capped Multilayer …euvlsymposium.lbl.gov/pdf/2005/pres/43 3-CC-13 Hill.pdf2005 EUVL Symposium 12-2.0-1.5-1.0-0.5 0.0 0.5 0 1 01-8 2 10-8 3 10-8

2005 EUVL Symposium1

EUV Endurance Testing of Ruthenium-Capped Multilayer Mirrors

S. B. Hill, C. Tarrio and T. B. LucatortoNIST

T. E. MadeyRutgers, the State University of New Jersey

S. Bajt†

Lawrence Livermore National Laboratory

O. Wood II, S. Wurm and N. V. EdwardsSEMATECH

† This work was performed under the auspices of the U.S. Department of Energy by University of California Lawrence Livermore National Laboratory under contract No. W-7405-Eng-48

Support for this work provided in part by SEMATECH LITH160


Degradation of EUV projection optics

Industry ProblemDevelopment of novel capping layers capable of extending the lifetime of multi-layer optics in production environmentRequired: ΔR<1-2% over 30,000 hr lifetime

Conditions: EUV intensity, IEUV ≤ 10 mW/mm2

Unbaked vacuum, PH2O ≤ 10-7 Torr, PHCx~10-10 Torr

NIST Goals• Provide rapid feedback to capping-layer-development community

by establishing well characterized endurance testing facilities.• Develop accelerated testing methods to predict optic lifetime in

tool (~3.5 yrs) from ~102 hrs of exposure at aggressive IEUV and PH2O, HCx

conditions.


EUV HCxH2O

Demonstrably reversiblesurface carbon growth

Potentially Irreversible(sub)surface oxidation

Capping layer

Present assumption: two competing mechanisms


Existing EUV endurance testing facility at NIST

• Spot size ~ 650 μm × 650 μm (FWHM) at sample• Average, in-band (13.1-13.6nm) intensity ~5 mW/mm2

• Be window permits high water vapor pressures: PH2O ≤10-5 Torr• Automation upgrades increased exposure duty cycle: ~15 hrs/day

Input beam from SURF III

Mo/Si ML FocusingMirror

H2O atmosphere

Be windowMo/Si ML sample

Detector


EUVSample

13.5nm reflectivity map (ALS)

Post exposure analysis

0.660

0.655

0.650

0.645

0.640

0.635

Line profile of reflectivity map

0.64

0.65

0.66

0 0.5 1 1.5 2 2.5 3

Ref

lect

ivity

Distance along line profile (mm)


EUVSample

0.64

0.65

0.66

0

1

2

3

4

0 0.5 1 1.5 2 2.5 3

Ref

lect

ivity

Calculated Intensity (m

W/m

m2)

Distance along line profile (mm)

Line profile of reflectivity map

13.5nm reflectivity map (ALS)

Post exposure analysis

0.660

0.655

0.650

0.645

0.640

0.635

& EUV intensity distribution

Intensity dependence of damage inferred from a single exposure!


•Damage varies inversely with water partial pressure•Pressure dependence decreases with intensity•Suggests intensity saturation

-2.0

-1.5

-1.0

-0.5

0.0

0 1 2 3 4 5 6

2x10-7 Torr

5x10-7 Torr

1x10-6 Torr

5x10-6 Torr

Ref

lect

ivity

Cha

nge

(x10

0)

Calculated Intensity (mW/mm2)

Reflectivity loss as a function of incident intensity


2x10-6 Torr H2O4.6±1.5 mW/mm2

10 hrs (2C-S3)

0.67

0.66

0.65

0.64

151050

0.67

0.66

0.65

0.64

151050

5x10-6 Torr H2O5.5±1.8 mW/mm2

10 hrs (2C-S4)

0.67

0.66

0.65

0.64

151050

5x10-7 Torr H2O5.8±2.1 mW/mm2

10 hrs (2C-S5)

1x10-6 Torr H2O5.8±1.9 mW/mm2

10 hrs (1A-S1)

0.66

0.65

0.64

0.633020100

2x10-7 Torr H2O6.2±1.9 mW/mm2

10 hrs (1A-S5)

0.66

0.65

0.64

0.6320100

10hr Exposures at Maximum Intensity (~6 mW/mm2)

Reflectivity loss decreases with increasing water pressure

Increasing water vapor pressure

Unexpected damage dependence on H2O pressure


1x10-6 Torr H2O3.0±0.9 mW/mm2

10 hrs (1A-S2)

0.66

0.65

0.64

0.6320100

5x10-6 Torr H2O2.8±0.9 mW/mm2

10 hrs (1A-S3)

0.66

0.65

0.64

0.63151050

5x10-7 Torr H2O3.0±0.9 mW/mm2

10 hrs (1A-S4)

0.66

0.65

0.64

0.63151050

2x10-6 Torr H2O2.9±1.0 mW/mm2

10 hrs (2C-S2)

0.67

0.66

0.65

0.64

151050

Unexpected damage dependence on H2O pressure

10hr Exposures at Half Max Intensity (~3 mW/mm2)

Reflectivity loss decreases with increasing water pressure

Increasing water vapor pressure


-2.0

-1.5

-1.0

-0.5

0.0

0.5

0 1 10-8 2 10-8 3 10-8 4 10-8 5 10-8

1e-6 Torr H2O

Peak

Ref

lect

ivity

Cha

nge

(x10

0)

Methanol Partial Pressure (Torr)

Methanol significantly reduces damage

•Just 2x10-8 Torr methanol prevents measurable damage from water

10hr, ~6mW/mm2


-2.0

-1.5

-1.0

-0.5

0.0

0.5

0 1 10-8 2 10-8 3 10-8 4 10-8 5 10-8

1e-6 Torr H2O

5e-7 Torr H2O

Peak

Ref

lect

ivity

Cha

nge

(x10

0)


•Just 2x10-8 Torr methanol prevents measurable damage from water


10hr, ~6mW/mm2


-2.0

-1.5

-1.0

-0.5

0.0

0.5

0 1 10-8 2 10-8 3 10-8 4 10-8 5 10-8

1e-6 Torr H2O5e-7 Torr H2ONo Added H2O

Peak

Ref

lect

ivity

Cha

nge

(x10

0)


•Just 2x10-8 Torr methanol prevents measurable damage from water•Very little damage from low levels of methanol alone•Trace levels of carbon-containing compounds in ambient

background may similarly mitigate damage by water


10hr, ~6mW/mm2


10-13

10-12

10-11

10-10

10-9

10-8

10-7

Par

tial P

ress

ure

(Tor

r)

605040302010Mass (amu)

Base, 7e-10 Torr

Identify potential mitigating species in vacuum

•No long-chain (m>50) hydrocarbons present

H2OCO CO2

No peaks m>50


10-13

10-12

10-11

10-10

10-9

10-8

10-7

Par

tial P

ress

ure

(Tor

r)

605040302010Mass (amu)

Base, 7e-10 Torr Water, 2e-6 Torr

H2O

CO CO2

•CO and CO2 increase with water (no significant m>50)


No sig peaks m>50


10-13

10-12

10-11

10-10

10-9

10-8

10-7

Par

tial P

ress

ure

(Tor

r)

605040302010Mass (amu)

Base, 7e-10 Torr Water, 2e-6 Torr Methanol, 1e-8 Torr


•CO and CO2 contribution from water similar to methanol

H2O

CO CO2

No sig peaks m>50


2.0

1.5

1.0

0.5

0.0(CO

+CO

2):H

2O P

P R

atio

x10

-2

2.52.01.51.00.50.0Inverse Pumping Speed x10-2 (s/L)

Ancillary tests to find source of CO & CO2

Is water contaminated?

WWCC PLPS βα +=

WWW LPS =

Mass throughput equations

⎟⎟⎠

⎞⎜⎜⎝

⎛+∝

WW

C

SPP βα

•CO & CO2 NOT from contaminated water supply•Produced by water interaction with chamber

C-species displaced from unbaked chamber walls by waterReaction of water with C in hot tungsten/iridium filaments


10-13

10-12

10-11

10-10

10-9

10-8

Par

tial P

ress

ure

(Tor

r)

605040302010Mass (amu)

Base, 8e-10 Torr H2O, 5e-6 Torr

Improved mass spectrum after bake & cleaning

No sig peaks m>50

H2O

CO CO2


• C-species generated at rate proportional to water partial pressure

• Chamber contains two filaments: IG, RGA

• Is rate of C-species production different with only RGA on?

Ancillary tests to find source(s) of CO & CO2

1.2

0.8

0.4

0.0

(CO

+CO

2) P

artia

l Pre

ssur

e x1

0-9 T

orr

43210Water Partial Pressure x10-7 Torr

Ion Gauge ONs=(2.8 ± 0.1)x10-3


• C-species generated at rate proportional to water partial pressure

• Chamber contains two filaments: IG, RGA

• Is rate of C-species production different with only RGA on?

Ancillary tests to find source(s) of CO & CO2

• CO & CO2 3x smaller with IG off

• Likely reduced further if RGA also off

YES!

1.2

0.8

0.4

0.0

(CO

+CO

2) P

artia

l Pre

ssur

e x1

0-9 T

orr

43210Water Partial Pressure x10-7 Torr

Ion Gauge ONs=(2.8 ± 0.1)x10-3

Ion Gauge OFFs=(1.01 ± 0.02)x10-3

How will this affect EUV damage to ML optics?




H2O atmosphere


Detector

Use in situ reflectivity monitor for initial tests

•Only need relative measurement of real-time decay




H2O atmosphere


Detector


•Only need relative measurement of real-time decay




H2O atmosphere


Detector

EUV beam

ML sample

Photodiode


•Only need relative measurement of real-time decay•Suitable for monitoring initial decay and changes >0.5% (abs)•Not reliable for absolute measurements over long exposures due to

highly non-uniform spatial distribution of damage.


in situ measurement of reflectivity decay

-2

-1

0

% c

hang

e of

in s

itu re

flect

ivity

00:00 01:00 02:00 03:00 04:00 05:00Elapsed Exposure Time (hr:min)

5e-6 Torr H2O



-2

-1

0

% c

hang

e of

in s

itu re

flect

ivity


5e-6 Torr H2O 1e-6 Torr H2O

•Observe increased damage at lower H2O pressures



-2

-1

0

% c

hang

e of

in s

itu re

flect

ivity


5e-6 Torr H2O 1e-6 Torr H2O

•Observe increased damage at lower H2O pressures•Initial decay empirically described by exponential


Effect of filaments on reflectivity decay0.448

0.444

0.440

0.436

In S

itu R

efle

ctiv

ity

04:00 05:00 06:00 07:00 08:00 09:00 10:00Elapsed Exposure Time (hr:min)

Filaments ON

5 x 6-10


Effect of filaments on reflectivity decay0.448

0.444

0.440

0.436

In S

itu R

efle

ctiv

ity


Filaments ON


Effect of filaments on reflectivity decay

•Damage is measurably accelerated when filaments are switched off•Hot filaments play critical role in damage processes•Production of mitigating species: CO, CO2 and/or perhaps H

0.448

0.444

0.440

0.436

In S

itu R

efle

ctiv

ity


Filaments OFF

Filaments ON


Effect of filaments at different pressures

-3

-2

-1

0%

cha

nge

of in

situ

refle

ctiv

ity


Filaments ON 5e-6 Torr H2O 1e-6 Torr H2O


-3

-2

-1

0%

cha

nge

of in

situ

refle

ctiv

ity


Filaments OFF 5e-6 Torr H2O 1e-6 Torr H2O

Filaments ON 5e-6 Torr H2O 1e-6 Torr H2O

•Damage still greater for lower H2O partial pressure, though less pronounced without filaments

•Suggests trace levels of mitigating species still present and active

Effect of filaments at different pressures


• Mitigating C-species may result from water-induced desorption/displacement of C-compounds on chamber walls. Common concern in surface science

Designing new, UHV bakable chamberUse doser to reduce ambient pressure

Preliminary indications and next steps• In situ reflectivity can monitor initial evolution of damage

Verify with accurate reflectometry (next weeks)

• Observed mitigation attributed to low-mass species like CO, CO2. NOT due to m>45 amu hydrocarbons (e.g., m=55)

Direct admission of CO and/or CO2

• CO, CO2 & possibly H produced by H2O reacting with hot filaments significantly slows damage of Ru-cap MLs

Reflectometry of long exposure with and without filaments


Gassupply

Target chamber

EUV

Sample

<10-7 Torr

Flux equivalent to 10-5 Torr ambient

~ 1 Torr~ 10-

2 Torr Metering

aperture

Capillaryarray

EUV

Planned use of doser


New UHV exposure facility under construction

• Load lock: cleanliness and increased throughput• Bake to 150-200 C• Gold plate majority of chamber surface area• Can accommodate two independent dosers• Potential to heat and cool samples during exposure (~ 0-50 C)


• Continue to identify critical factors to establish well characterized exposure environment for benchmarking and capping layer development

• Investigate differences between pulsed and quasi-cwsynchrotron exposures

• Commission third high intensity exposure facility~150 mW/mm2 broadbandAlso UHV designInvestigate intensity dependence for accelerated testing

• Collaborate to identify useful surface analysis techniques:e.g., IR, elastic ion backscatter, XPS, TOF-SIMS,…

Future directions

Documents

EUV Endurance Testing of Ruthenium- Capped Multilayer …euvlsymposium.lbl.gov/pdf/2005/pres/43 3-CC-13 Hill.pdf2005 EUVL Symposium 12-2.0-1.5-1.0-0.5 0.0 0.5 0 1 01-8 2 10-8 3 10-8