S.r. Mohanty Centre Of Plasma Physics, Assam

S.R. Mohanty Centre of Plasma Physics, Assam

•Work carried out at Hotta Lab., Tokyo Institute of Technology, Yokohama under EUVA project of Japan 1

Presentation Outline 1. 2. 3. 4. 5.

EUV radiation Lithography Discharge Produced Plasma (DPP) EUV Sources EUV diagnostics developed for DPP Sources Capillary plasma

2

EUV Radiation

Extreme Ultra Violet

EUV radiation

Space Science Plasma Science

Electronics industry EUV radiation in Space Science • for studying specific groups of astronomical objects, including white dwarf stars, stellar coronae and interstellar medium. • for studying the temperature and density structure of the solar corona, the solar flares and other coronal disturbances. • for studying the invisible universe.

3

EUV Radiation

EUV radiation in Plasma Physics Tokamak researchers use EUV radiation emitted by highly ionized impurity ions to diagnose transport and other physical properties… Example lines: Li-II

13.5 nm

C-V

24.8 nm

He-II

30.4 nm

EUV radiation of Laser Produced Plasma and Discharge Produced Plasma are being studied for different application purposes.

EUV radiation in Semiconductor Industries EUV radiation is of significant interest for the Semiconductor Industries EUV Lithography is well on its way to commercialization

4

Lithography

What’s Lithography ? • A method to print text or figure on a plain surface (paper or other materials) Semiconductor Industries imitated the lithography process to fabricate ICs only in 1960…. IC (1962)

Basic steps for IC fabrication •

Wafer preparation – How Oxidation, Cleaning • Wafer fabrication Optical lithography and Etching Diffusion and Ion Implantation Metallization • Die testing and cutting • IC packaging and testing

Current generation IC

does optical lithography work ? Light source Lenses Mask (Circuit pattern) Lenses 5

Si Wafer coated with photoresist

Lithography

Fundamental Relations k1 Re solution  NA

k2 Depth of Focus = (NA)2 ±

k1 & k2 : application dependent coefficients (a constant between 0.25 to 1 depending upon optics, resist etc)  : wavelength of the radiation used for imaging NA : numerical aperture of the imaging system

Optical Lithography Road Map

6

Lithography

Why EUV lithography??? Semiconductor industry wants to produce smaller circuits 30 nm structures are needed around 2014 Not possible with “current” optical lithography EUV lithography (13.5 nm or 92 eV) is the solution

The light source is one of the main problems!!! Mo/Si multilayer mirror with high reflectivity at around 13.5 nm Mo/Si ML

7

Lithography

EUV Lithography System Discharge Produced Plasma (DPP) Light Source

kV

kHz

Intermediate Focal Point

Light source specifications ITRS*   = 13.5 nm ( = 2 %)  115 W at IF (100 wafers/hour)  > 7 ~ 10 kHz  Debris-free (lifetime > 30,000 hours)

* International Technology Roadmap for Semiconductor

From EUVA Official website

8

DPP Source Schematic of DPP EUV Light Sources Gas jet Z-pinch

Subsystems of DPP EUV Light Source •Electrical pulse power driver •Discharge head •Auxiliary systems like gas feeder and pumping system

Diagnostics •EUV photon detector •EUV energy monitor •EUV pinhole camera •EUV spectrometers •Time resolved visible light imaging 9 •Current and Voltage probe

DPP Source

HV

Pulse Power Driver-I

8 6

K

4 2

A

SI Thy

0 0

403 nF

PT 1:3

2 Time [ m s]

3

4

Magnetic Switch

15

10

K

SI Thy

40 nF

z 5

Current [kA]

Pulse Power Driver-II

HV

1

A 0 -0.2

0

Driver I

Driver II

dI/dt

20 A/ns

57 A/ns

Pulse width

3ms

350 ns

Input energy

5.6 J

4.7 J

0.2 Time [ m s]

0.4

10

Current [kA]

10

403 nF Unsat. inductor

Capillary Plasma

Capillary head Cold water Cathode : Mo Inner diameter : 5 mm Outer diameter : 38 mm Length : 13 mm

Xe gas Anode : Mo Inner diameter : 8 mm Max. diameter : 32 mm Length : 13 mm

Capillary : Al2O3 Cold water

Inner diameter : 1, 2, 3 mm Outer diameter : 40 mm Length : 10 mm

11

DPP Source Auxiliary systems Gas Supplying Part

Magnetic Switch

Vacuum Chamber

Baratron Control Valve

EUV Light

Magnetic switch

Vacuum chamber

Valve

Discharge Part

Xe Gas

Pumping System

Pumping Part PFN (Capacitor Bank)

Rotary Pump+TMP

•Flexible enough to allow quick change of discharge head, gases, driver energy etc. •To detect EUV emission simultaneously using many diagnostics GREMI EUV device

12

EUV Diagnostics

EUV photon detector

Simple and Indispensable !!!

Time evolution of EUV photon output in the range of 5 to 18 nm

+ Zr/C filter

IRD AXUV-10

EUV energy monitor Absolute in band energy measurement at 13.5 nm ±2% BW EUV energy monitor

Zr filter

Multilayer Mirror

Calibrated by JENOPTIK Energy monitor (E-MON)

13

EUV Diagnostics EUV Pinhole camera

Magnification at axis: 1.7 at 45°: 2.1 at 90°: 3.1

Filter: 150 nm Zr having transparency of ~ 13% EUV light in between 5 to 18 nm Pinhole: 50 μm diameter X-ray CCD: Andor Techno. Ltd. DO434 (1024 x 1024 pixels, 13 x 13 μm2)

14

EUV Diagnostics

EUV spectrographs Grazing incidence spectrometer McPerhson 248/310G detector chamber

vacuum chamber grating box

Specification

aperture

Value

Grating radius (R)

998.8 mm

Angle of incidence ()

87 degree

Groove density (d)

600

Observable range

1 ~ 70 nm

Resolution

0.36 nm

15

EUV Diagnostics Transmission grating spectrometer

Specification

Value

Pinhole diameter S

50 mm

Aperture diameter of grating A

50 mm

Total length D

750 mm

Grove density

1000 lines/mm

Resolution

0.45 nm

16

Capillary Plasma

Block diagram of experimental setup with diagnostics

Xe Gas

Angular Distribution Measurement tool EUV Pinhole Camera

Gas Flow Controller

Visible Light Spectrometer & Camera

Discharge Head Detection Chamber

Pulse Power Driver I or II

Cooling Arrangement

P u m p

EUV Spectrometer

EUV Energy monitor EUV Photodiode

17

Capillary Plasma Results of EUV photo detector

10

0

10

5

0 0

1

2 3 Time [ms]

Photodiode (5~18 nm) Discharge current

20

10

10

5

0

0 0.6

4

0

0.2 0.4 Time [ms]

Photon intensity: 22 V

Photon intensity: 10V

Dependence of photon intensity on supply pressure

15

Current [kA]

Discharge current Photodiode (5~18 nm)

Photodiode [V]

20

30

15

Photodiode intensity [a.u.]

Photodiode [V]

30

High dI/dt pulse

Current [kA]

Low dI/dt pulse

Low dI/dt pulse High dI/dt pulse

1

0.5

0 2

4

6 Pressure [Torr]

8

18

Capillary Plasma

Absolute in-band EUV energy measurement At 10 degree EUV energy [mJ/sr/2%BW/pulse]

Low dI/dt pulse

High dI/dt pulse

2.5

3.3

0 To scope

EUV energy monitor

Capillary

EUV energy [mJ/sr/2% BW]

Angular variation of in-band energy 5 Low dI/dt pulse High dI/dt pulse

4 3 2 1 0

0

10

20 30 40 Angle [degree]

5019

60

Capillary Plasma Effect of dI/dt and pressure on EUV pinhole images Low dI/dt pulse Pressure

0

45

High dI/dt pulse 0

45

4 Torr Low dI/dt pulse High dI/dt pulse

1

3

2 0.5 1

0 2

4

6 Pressure [Torr]

8

Integrated intensity [a.u.]

Source diameter FWHM [mm]

3 Torr

5 Torr 6 Torr

7 Torr

0

8 Torr

20

Capillary Plasma

Effect of dI/dt on positional stability Low dI/dt pulse

High dI/dt pulse

2 mm

X

X：±0.16 mm Y：±0.13 mm

Y

Y

Center position of individual shots

Standard deviation σ

2 mm

X

X：±0.018mm Y：±0.013mm

 Peak intensity positions of 20 pinhole image shots were superimposed.  Positional stability of fast pulse is better than that of slow one.

21

Capillary Plasma

Stability at high dI/dt pulse

 Single pulse : 0.718 mm (FWHM)  20 pulses : 0.748 mm (FWHM) Plasma size appears about 5 % larger in 20 pulses than in single pulse experiments due to the fluctuations in positional stability. 22

Capillary Plasma

Grazing incidence spectra 200 ns

150

150

500 ns

100

Xe11+

Xe10+ Xe9+ Xe8+

50

Time 2 Intensity [a.u.]

Intensity [a.u.]

Time 1

0 10 11 12 13 14 15 16 17 18 19 Wavelength [nm]

150

Xe9+ 100

50

0 10 11 12 13 14 15 16 17 18 19 Wavelength [nm]

700 ns

150

1200 ns

100

Xe9+ Xe8+ 50

0 10 11 12 13 14 15 16 17 18 19 Wavelength [nm]

Time 4 Intensity [a.u.]

Peaks at 11, 12.5, 13.5 and 15 nm

Intensity [a.u.]

Time 3 O4+

O4+

100

O4+ 50

0 10 11 12 13 14 15 16 17 18 19 Wavelength [nm]

Time-resolved spectral analysis indicates different ionization states of Xe  After the max. of discharge current only O impurity lines dominate 23

GREMI EUV spectra

Capillary Plasma

Transmission grating spectra  Vcharge : 9 kV  200 Hz operation and 200 pulses averaged signal

3Torr 4Torr 5Torr 6Torr

1500

7Torr 8Torr 9Torr 10Torr

1500

1000

1000 10

2000

intensity[a.u.]

intensity[a.u.]

2000

12

14 16 wavelength[nm]

18

10

12

14 16 wavelength[nm] 24

18

Capillary Plasma

Summary •A high dI/dt discharge current is better for producing smaller and stable EUV emitting plasma • Maximum in-band energy obtained at source :3.3 mJ/sr/2%BW/pulse •EUV spectra show the evidence of emission at 13.5 nm

25

Survey: No EUV until 2015 or later By Ann Steffora Mutschler, Senior Editor -- Electronic News, 9/7/2007 According to a semiconductor manufacturing survey conducted by WeSRCH, the social networking site of market research firm VLSI Research Inc., 85 percent of respondents said that extreme ultraviolet (EUV) lithography will not make it to production until 2015 or later, with 74 percent of respondents saying that EUV will reach production between 2015 and 2024.

26

27

Gas Jet Pinch Plasma

28

Gas Jet Pinch Plasma

Background Capillary cathode

insulator

Gas jet pinch anode

cathode(nozzle)

anode(diffuser)

Xe

Xe

pump

Debris Demerits

Improvements



Debris from the insulator/electrodes



No insulator in the new design



Thermal load to the insulator





Narrow solid angle



Debris and EUV light in the same direction

Collection of EUV light from radial direction Large solid angle Less debris in EUV light 29

Gas Jet Pinch Plasma

Background Capillary cathode

insulator

Gas jet pinch anode

cathode(nozzle)

anode(diffuser)

He gas curtain Xe

Xe

pump

He gas curtain

Debris Demerits

Improvements



Debris from the insulator/electrodes



No insulator in the new design



Thermal load to the insulator





Narrow solid angle



Debris and EUV light in the same direction

Collection of EUV light from radial direction Large solid angle Less debris in EUV light 30

Gas Jet Pinch Plasma

Discharge head

cathode

Curtain gas

anode

cathode

anode

Curtain gas Xe Curtain gas 125 mm

anode Xe

cathode

Curtain gas

 Cathode (dual orifice nozzle) - Inner nozzle diameter: 2 mm - Outer nozzle diameter Inner diameter: 11.6 mm Outer diameter: 12 mm

 Anode (diffuser) Inner diameter: 6 mm Outer diameter: 20 mm

31

Gas Jet Pinch Plasma

Time resolved EUV photon measurement 5 -18 nm With He gas curtain and diffuser With He gas curtain Without gas curtain and diffuser

30

15

10

20 5 10 0 -100

0

100

200 300 Time [ns]

400

0 500

He

Discharge current [kA]

EUV signal [V]

40

pump Xe

cathode

anode

He

EUV photons appear after 70 ns of initiation of discharge current and reach a maximum nearly just before the maximum of current pulse The EUV intensity peak as well as EUV photon flux improves around 30 % 32 because of the presence of gas curtain

Gas Jet Pinch Plasma

Absolute in-band EUV energy measurement gap distance ：12 mm, Xe pressure ：20 Torr, with He gas curtain and diffuser

cathode

anode

EUV energy [mJ/sr/2% BW]

Angular distribution of EUV radiation 5 4 3 2 1 0 -20

-10

0 10 Angle [degree]

20

EUV energy = 4.66 mJ/sr/2％BW/pulse Energy monitor

33

Gas Jet Pinch Plasma gap distance ：12 mm

EUV pinhole images

15 Torr

25 Torr

1000

Intensity [a.u.]

800 600

20 Torr

30 Torr

4 mm 6 mm 8 mm 10 mm 12 mm 14 mm 16 mm

400 200 0 10

20 Xe pressure [Torr]

30

34

Gas Jet Pinch Plasma

Positional stability Single pulse

Radial

20 pulses

Axial

Axial

Length (FWHM)

Radial

1 pulse

20 pulses

Remark

Axial [mm]

0.34

0.96

Radial [mm]

0.07

0.16

- 20 pulses : 2.5 times larger

35

Gas Jet Pinch Plasma Effect of gas curtain on EUV emission gap distance ：12 mm，Xe pressure ：20 Torr Without gas curtain

2 mm

With He gas curtain

4.3 mm

Length [mm] （FWHM）

He：120 sccm

axial

0.92

radial

0.16

Intensity [a.u.]

967

Length [mm] （FWHM）

axial

0.80

radial

0.14

Intensity [a.u.]

1213

36

Gas Jet Pinch Plasma Dependence of EUV intensity on pressure

Intensity [a.u]

1500

Without gas curtain He gas curtain 120 sccm He gas curtain 240 sccm He gas curtain 280 sccm

1000

500

0 10

20 Xe pressure [Torr]

30

Nearly 25 to 50 % increase in EUV intensity is marked in 20 to 25 Torr Xe gas pressure because of the presence of gas curtain .

37

Gas Jet Pinch Plasma

Transmission Grating Spectrometer

11 nm 13.5 nm 15 nm 16.5 nm

38

Gas Jet Pinch Plasma

Summary  New gas jet pinch discharge system successfully demonstrated the concept of radial extraction of EUV light.  Highly intense and small size EUV emitting plasma was produced at Xe pressure of 20 Torr.  Total EUV energy at the optimum condition is 4.6 mJ/2％BW/pulse.  He gas curtain limits the expansion of Xe gas jet and increases the EUV intensity.

e

39

40

Source material Zn

Re Sn

Xe

 elements having an atomic number Za ~ 30 (Zn), Za ~ 50 (Sn) and Za ~ 75 (Re) are likely to produce the highest yield  The not-optimal EUV power level achieved using Xe, about a factor of 10 is yet missing for the lithography application, brought some of the research back to the spectrally much more favorable element of tin (Za = 50), in spite of the huge debris problem which is to be expected

Dependency of the conversion efficiency (CE) at  = 13.5 nm on the atomic number Za of the target material. The CE was measured on laser plasmas at a power density of 1012 W/cm2 41

 Ein : Input energy of pulse power supply/pulse [J/pulse]  CE : Conversion efficiency  Tsystem : Transmission ratio of debris shield and purity filter Ein

CE

Tsystem

collection

Tgas

Rep-rate  collection : Collection efficiency

 Tgas : Gas absorption  Rep-rate : System frequency [Hz]

42

Collection efficiency - collection Discharge produced plasma (DPP)

Laser produced plasma (LPP)

DPP

LPP

Max. solid angle

2

4

Condenser mirror

Grazing incidence

Normal incidence

Collection efficiency

29% ( 50%)

40% ( 50%)

Discharge structure, mirror and plasma generation 43

LPP vs DPP

back

44

Source Specifications

ITRS*-2004 •EUV power at the intermediate focus •Repetition rate •Energy stability •Lifetime •Maximum etendue of source output •Maximum solid angle input to illuminator •Spectral purity: 10 – 40 nm 40 – 130 nm 130 – 400 nm (DUV/UV) > 400 nm •Pulse-to-pulse positional stability •Source emission volume •Vacuum before intermediate focus •Spatial distribution of power •Stability of repetition rate •Angular distribution of power •Rotational symmetry of power

115 W in 2% BW > 7 – 10 kHz  0.3%, in 3  over 50 pulses > 30,000 hours of operation < 3.3 mm2sr 0.03 – 0.2 sr TBD (To be declared) TBD < 3 – 7% TBD < 10% of source size   1.3  1.5 mm2 TBD TBD <  0.1% (long term) TBD TBD

* ITRS-International Technology Roadmap for Semiconductor

45

Factors that influence EUV Lithography tool

Source Power Spectral purity Rep rate

Optics Chamber life

Products

Contamination Illuminator Max field Collimation Stability size Mask Material damage Trans.

Wafer size Test wafer

Stage accel. Stage speed Acid reaction rate Transparency Wafer load/unload Photo resist sensitivity

Vibration isolation Alignment mark Overlay calculation

Alignment

back

46

47

back

GREMI EUV Source looks like…….. 0.8- 7.2 J Switch

Capillary

0.6 m* 0. 7 m* 1.7 m

25 cm

* Much compact * Simpler in design

back

* Cost-effective * Easy to operate

MEE (2003);PSST (2003) 48

EUV Spectrometer

Gazing incidence Jobin-Yovon Spectrometer (15-35 nm max. Efficiency, 0.1 nm spectral resolution, 10 nm temporal resolution): •800 g/mm platinum coated grating •Two stage MCP (Galileo 3040FM) •ICCD array detector (Princeton)

Pinhole imaging facility

1700 A° thick Zr filter allows only 10-16 nm EUV radiation

back

49

Results Xe

11

Time integrated xenon spectra at 0.46 mbar for different charging voltages

Xe10

Xe 9

7000

Oxygen Lines

15 kV 18 kV

6000

24 kV

• Three broad band peaks of Xe centered at 11, 13.5 &15 nm

Intensity [a.u.]

5000 4000 3000

• Oxygen lines from wall

2000 1000 0 10

12

14

16

18

20

Wavelength [nm]

22

24

26

28

• Xe/O lines intensity increases with the increase in voltage

• Observation of 11 nm peak indicates that plasma temp. of 45 eV

50

Results

Time resolved spectra at 0.46 mbar and 24 kV charging voltages

800

600 400 200

800

70 ns

600 400 200 0

0 8

10 12 14 16 18 20 22 24 26 28

W avelength [nm ]

Intensity [a.u.]

50 ns

Intensity [a.u.]

Intensity [a.u.]

800

115 ns

600 400 200 0

8

10 12 14 16 18 20 22 24 26 28

W avelength [nm ]

8

10 12 14 16 18 20 22 24 26 28

W avelength [nm ]

•11nm broad band peak appears first & it goes maximum at the maximum of compression. •At the instant of maximum of discharge current the 13.5 nm peak appears to be prominent. •Spectra trapped after first half cycle of current do not show any emission from broad bands.

back 51

Results Pinhole (50 µm)

Capillaire

7 cm

MCP

70 cm

ICCD

Time resolved pinhole images from 10 mm long 1.2 mm diameter alumina capillary at 1 torr Xenon and at 24 kV charging voltage...

Intensity

6000 5000

45 ns

50 ns

60 ns

95 ns

110 ns

45 ns

50 ns

60 ns

95 ns

110 ns

4000 3000 2000 1000 0

-60-40-20 0 20 40 60

-60-40-20 0 20 40 60

-60-40-20 0 20 40 60

Pixels

-60-40-20 0 20 40 60

-60-40-20 0 20 40 60

•A second faint compression observed at the maximum of discharge current. •The beam intensity profile changed from single peak pattern to annular pattern. Annular pattern may be due to increased refraction of EUV beam caused by larger density in plasma column. 52

Magnetic Pulsed Compression circuit Magnetic switch

v0

MS1

MS2

v1

v2

B H

C0

i0

0

C1 i2

i1

C2

(a) circuit hysteresis loop

ferromagnetic material (Finemet)

    B  S 

V  t



Φ : flux reversal χ : rate of occupation B : magnetic flux density S : cross section of magnetic path V : magnetic switch voltage t : pulse width

V

v0

v1

t0

t1

v2

i2 i1 i0 (b) operation C0=C1=C2

t2 53

t

New Set-up of Hotta Lab EUV Source 2 stage MPC circuit

LC inversion circuit

1st Magnetic Switch (2turns)

Ceramics Capacitor

HV

Capacitor bank

Capacitor 4uF ×2

2nd Magnetic Switch 1:3 Transformer gas supply

1:1 Transformer

RF preionization RF power supply

Xe gas preionization

Rotary pump + TMP pump 54

PFN (Pulse Forming Network)

Pulse Power Driver I

・ 68 Ceramic capacitor (5.3 nF, 20kV) ・ Total Capacitance : 370 nF ・ Impedance : 0.44  HV PFN SI thyristor stack

SI Thyristor Stack ・3 thyristors connected in series

R

10

10

5

5

0

0 0

1

2

L

L

C

C

Magnetic switch

Load

3

Discharge current [kA]

Switch voltage [kV]

・Blocking voltage : 12.0 kV ・Conducting current : 400A

D

430 nF

Anode

Cathode

Reset circuit

Low dI/dt (20A/ns, 3μs) 55

Pulse Power Driver II

Primary capacitor bank (370 nF)

Magnetic Switch

Voltage probe

HV

Cathode Switch

Load

40 nF

Anode

High dI/dt (57 A/ns, 350ns)

Secondary capacitor bank

Driver I

10

Current [kA]

1：3

Current probe

Driver II

Current dI/dt 5

20 A/ns

57 A/ns

Pulse width 0 -100

0

100

200 300 Time [ns]

400

500

3ms

350 ns

Electrical input energy into capillary (per shot) 5.6 J

4.7 J

56

Effect of dI/dt on the etendue* Capillary edge

0

0

0

Radial Axial

45

Low dI/dt pulse Xe : 5 Torr

Etendue [mm2sr]

15

10

5

0 0

45

High dI/dt pulse Xe : 5 Torr

Low dI/dt pulse High dI/dt pulse 5 Torr High dI/dt pulse 10 Torr

15 30 45 60 75 Collection angle [degree]

45

High dI/dt pulse Xe : 10 Torr

*Etendu: A measure of the capacity of an optical system to transfer power 57

90

The singularity current probe signals is observed due to large change in the plasma impedance during the pinching. The rapid increase in plasma impedance can be understood principally in term of the rapid increase in plasma inductance. As the plasma inductance during radial phase is given as L=Zp*(µ/2* π )ln(a/rp) Where a is the initial radius, Zp and rp are length and radius of plasma column).

The intensity of Xe emission around 13.5 nm is mainly due to 4d8 - 4d75p transitions of the Xe+10 ion 58