XTREME Technologies First Tin Beta SoCoMo ready for Wafer … · 2016-11-07 · Presentation at...

Taking light to new dimensions…

XTREME Technologies

First Tin Beta SoCoMo ready for Wafer Exposure

Marc Corthout, Yusuke Teramoto, Masaki Yoshiokaand the whole XTREME and EUVA Gotemba team

Presentation at EUVL Symposium @ Kobe, Japan

October 19, 2010

October 19, 2010 EUVL Symposium @ Kobe, Japan


Bringing together all DPP knowledge under 1 strong roof

2001 Founding of– PHILIPS Extreme UV– XTREME Technologies GmbH (XT)

PHILIPSExtreme UV

2001 2001

2005 Ushio buys Lambda Physik’s50% share in XT

2007 Joint development of Beta platformby XT and PHILIPS EUV starts

2008 Ushio buys Jenoptik share in XT

2010 Ushio buys PHILIPS Extreme UV

2010

100% Daughter



Ushio: Lighting Edge Technologies -> EUV at the core

Established in 1964

Headquarters Tokyo, Japan

Consolidated financial figures ( 2010/03)- Net sales 119 billion yen- Net profit 7 billion yen- Net asset 157 billion yen- Number of employees 4,700

Core business segments:- Electronics ( UV lamps for photolithography, Excimer, EUV )- Visual image equipment ( Cinema projector, non-cinema projectors)- Office automation equipment ( Light sources for copiers )- Illumination

Mr. UshioChairman

USHIO Group

Mr. SugataPres. & CEOUSHIO Inc.

CEO of XTREME

i-line lamp PCB Stepper Xe excimer dry cleaner


Taking light to new dimensions…4

• 24 DPP sources in use for wafer exposures and EUV R&D

• Many years of runtime:more than 100 billion DPP pulses used for exposures and source testing

• Delivering all the photons for all 12” wafer results presented to date• Continuous Learning and Improvement

Managing all commercial DPP sources world-wide



An organization with all competences on board to support you…

Eind/Veldhoven

Leuven Aachen

Growing fast at 1 hour’s drive from ASML and IMEC

XTREME: 180 Staff+ Admi: 22

R&D: 85 Operations: 72

Supply Chain:18Engineering:14 Production: 17 CS: 22

USHIO Korea Inc

12.08.2010



Building several NXE 3100 sources

• Several full products for ASML to be delivered in coming months:

• Internal support platforms: Availability– Source for power scaling R&D April 2008– Full SoCoMo* R&D September 2009– SoCoMo* for module qualification Ops September 2010– Source + FT fully for reliability eng Ops January 2011

*SoCoMo = Source Collector Module (complete source upto IF)

ASML will qualify the first productwith the scanner in coming weeks.

Product close to beingSEMI S2 qualified.



Building the infrastructure to support this and future business

• New operations site since June 2010:former Mitsubishi plant in Alsdorf (15 min from ILT)

• Brand new 1100 m2 cleanroom(ready October 2010)

• Expansion for NXE 3300 source business secured for 2011

NXE 3300source

expansionarea

Warehouse



Tin bath

capacitor bank

Tin Film

cooling

Building an NXE 3100 SoCoMo…

Laser~10 mJ

~100 mJEUV

in 2

Taking light to new dimensions…confidential



Modular SoCoMo architecture:allowing typically a module swap in 1 shift:

SF: Swap Flange incl. Collector and FoiltrapSH: Source HeadTHB: Tin Handling Box

Power and reliability upgrades based on on-site module swaps of SH, THB, SF

Modular architecture easy exchange of modules

SF

SHTHB



Debris mitigation: principle

foil-trap

buffergas

particles are trappedradiation can pass

collector

thin metallamellae

Source emits around 0.5 kg/day in HVM phaseParticles are stopped by collisions with buffer Ar gas and deflected to foilsThese foils are transparent for the light


The source emits fast ions with energies above 10 keVThese ions can not be stopped by the FT and end up at the collector

which is of the grazing incidence typeThere they will sputter the optical coating

Consequences:• no growth of tin on the optical coating• thickness of coating determines

the life-time of collector,see next slides

Debris mitigation: fast ions

0 50 100 150 200 250 300 350104

105

106

107

108

109

1010

d = 1.36 m d = 2.12 m

(intensity 1/r2 corrected)

n(E

) / a

.u.

energy / keV


Taking light to new dimensions…0%

20%

40%

60%

80%

100%

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0Normalized Ru thickness

Nor

mal

ized

Ru

refle

ctiv

ity

Exposed collector was cut in samples and analysed:• hardly any tin deposition found• thickness of ruthenium coating was reduced due to sputtering• no increase in roughness found• reflectivity was independent of removed material

Collector: constant optical performance over lifetime

ruthenium coatingcan be used as

sacrificial materialwith constant

optical performanceof the collector



Foil-trap Test-stand: integrated experiment

Debris mitigation tests performed with collector samples during 8-hour runs:• Very low levels of Sn deposition: 0.1 nm• Very low Ru sputter rates: 2 nm/Gshot (about 2 nm/day for 3300)

Further improvement possible by optimisation of gas flows

1 year collector HVM lifetime can be achieved



Foil-trap and baffles separate plasma area from IF area

GI collector

plasma

FT

IFaperture

“clean” partof chamber

“dirty” partof chamber

vessel wallbaffles

All Sn isremoved via the

drainpicture before

installation of tincatch



00:00 00:10 00:20 00:30 00:40 00:50 01:00550

600

650

700

750

800

100% duty cycleno dose control

EU

V p

ower

at p

inch

[W/2

pi]

time [hh:mm]

Source power: 640 W/2 continuous

2.13% conversion efficiency at 30 kW input power100% duty cycle, no dose control50% etendue match -> 320W collectable light at pinch level



DPPLPP

100W

65W

15W

3W

ExposePower

60wph

25wph

4wph

<1wph*

Throughput

100W

40W

7W

1W

ExposePower

2wph**IntegrationQ3/2010

60wph

35wph

10wph

Throughput

Upgrade 2Mid 2011Upgrade 1Q1/2011IntegrationQ4/2010

Source ConfigTiming

NXE 3100 Power roadmap and status

Q4 ‘10 Q1 ‘11 Q2 ‘11

100% Duty cycleClean photons (no SPF needed)

Dose controlledAfter IF

2.13

320

15.2 20

38

14.7

54

2.3

30

345

Nov

See ASML’s pres. by Rard De Leeuw

Measured

Sour

ce M

odul

eC

oll.

Mod

ule



Dose Control well within spec

• Feed-back loop from measured IF power on the electrical pulse energyby changing charging voltage from pulse-to-pulse

• Same principle that is being used within ADT for years• No loss of EUV output• LLB system under development allows for operation > 40 kHz

one LOT(several wafers,incl. overhead)

±0.2%

time [min]

wafer overhead exposures

field #1 field #2

Wafer LOT100% duty cycle• Dose window 1 J (NXE:3100)• Duty cycle 100%• Measurement interval 30 min• Dose stability <±0.2%• Dose repeatability (3 ) 0.08%

0 102 864time [s]

0.5

-0.5

-0.2

0

0.2

devi

atio

n fro

m n

omin

al d

ose

[%]

7.5kHz 4J in <-> 35W@IF

Dos

e re

peat

abili

ty [%

]D

ose

repe

atab

ility

[%]

Measure time [min]

After 100 waferequivalent dose runat 100% Duty Cycle

Measured time [min]

10 hr run (100% duty cycle)



00:00 04:00 08:00 12:00 16:000.6

0.8

1.0

1.2

1.4

100% duty cycleno dose control

readjustmentby the operator

clea

n ph

oton

pow

er [a

.u.]

time [hh:mm]

Source maturity

18 hours continuous lightwithout any interruptrunning mostly without operator

MTBIp13 Performance S510 Sources

0.0

0.1

0.2

0.3

0.4

02-Jun-09 02-Jul -09 02-Aug-09 02-Sep-09 02-Oct-09 02-Nov-09 02-Dec-09

MTB

I [G

shot

s]

Alpha source MTBI4x improvement

944 952 1007 1015 1023 1031 10390

4

8

12

16

0

1000

2000

3000

4000

tota

l dos

e [M

J]

week number

equi

vale

nt w

afer

s

NXE 3100prototype

> 3600Eq. wafers



HVM roadmap: DPP scalability proven

• Evolutionary increase in efficiency and transmission• Significant increase in peak source input power but half in average power• Feasibility shown by proof-of-principle experiments

Peak0 1000 2000 3000 4000 5000

0

10

20

30

400.00 0.02 0.04 0.06 0.08 0.10 0.12

40 kHz

outp

ut e

nerg

y [m

J/2

]

pulse number

time [s]

0 1 2 3 4 5 6 70

40

80

120

160

2% conversionefficiency

EU

V p

er p

ulse

[mJ/

2]

electrical energy per pulse [J]

Measuredresults

Duty Cycle and Average power about halvesThus: Thermal challenges do not scale linearly with peak power

Sour

ce M

odul

eC

oll.

Mod

ule



Conclusions

• USHIO takes the lead in EUV business bringing all DPP knowledge underone roof and investing heavily in its future

• Several NXE 3100 source products being built for delivery in comingmonths

• Modular architecture allowing for easy field-upgrades in power andimproved reliability

• Debris mitigation as the corner stone of constant optical performanceand affordable cost of ownership

• Strong proof of power scalability as basis for designing NXE 3300products and expanding our business

• Building a sizable & experienced EUV service crewfor world-wide 24/7 support



Acknowledgements

The whole XTREME technologies’ and EUVA Gotemba teams

ASMLFraunhofer ILTNikonMedia LarioUSHIOZeiss

Part of this work is supported by EUVA, Japan

The financial support of the German Research Ministry under contract numbers13N8865 and 13N8866 is gratefully acknowledged

Work ongoing under the EXEPT project within CATRENE

XTREME Technologies First Tin Beta SoCoMo ready for Wafer … · 2016-11-07 · Presentation at...

Documents

Transcript of XTREME Technologies First Tin Beta SoCoMo ready for Wafer … · 2016-11-07 · Presentation at...