Optical Lithography Simulation and PhotoresistOptimization for Photomask Fabrication

Benjamen M. Rathsack1, Cyrus E. Tabery1, Steven A. Scheer1, MikePochkowski2, Cece Philbin3, Franklin Kalk3, Clifford L. Henderson4, Peter D.

Buck5 and C. Grant Willson1

March 15, 1999

1Department of Chemical Engineering, The University of Texas at Austin

2 ETEC Systems Inc., 9100 S. W. Gemini Dr., Beaverton, OR 970083 DPI Reticle Technology Center LLC, 2011 Greenhill Dr., Round Rock, TX 78664

4 Georgia Institute of Technology, 7778 Atlantic Dr., Atlanta, GA 30332 5 Dupont Photomask Inc., 1955 Division St. Gresham, OR 97030

Photomask Fabrication OptimizationGoal: Improve resolution and process latitude for photomask fabrication using laser pattern generators

Method: Line edge optimizationof exposure image and resistdevelopment response

Results: Sub 0.30 µm resist features on photomasksubstrates

UV Light

Reduction Lens

Exposed ResistPattern

Silicon Substrate

Photomask

Photomask Fabrication Process

• Laser pattern generatorexposes a pattern in theresist

• Developer dissolves awaythe exposed regions of resist

• Etch transfers the reliefimage into the chromium

Chromium

Quartz

Photoresist

Laser exposure

Development

Etch

Exposed resist

Novolac/DNQ Resist Chemistry

Photoactive Compound Photoreaction

Photoresist polymer matrix synthesis

N2

Diazonaphthoquinone Carbene Ketene Intermediate

O

N2

S

R

O

hν

O

S

R

OO

O

C

S OO

R

O

OH

O

S R

O

O

Indenecarboxylic Acid

H2O

H+

Novolac Resin

OH

CH3

+ H2C=O

CH2

OH

CH3

( )n

Photomask Resist Characterization

0204060

80100120

0 200 400 600 800 1000 1200 1400Time (s)

Tem

pera

ture

(ºC)

Commercial

Univ. Texas

Chromium 70 nm

Resist 570 nmChromium Oxide 35 nm

Quartz 6.35 mm

Coated Photomask Substrate

Photomask Post Application Bake

• I-line photoresists are coated onantireflective (AR3) substrates

• A hotplate was built to mimicthe post application bake (PAB)of production photomasks

• Photokinetics measured through exposure parameters (Dill’s A, B and C)

• Generate development ratefunction (R(m)) for resists

•Methodology–Bleach resist with light at 365 nm

–Photoactive compound undergoes chemical change–Photoproducts do not absorb light– Measure transmittance of resist film as a function of exposure dose

TransimpedanceAmplifiers

500 W HgArc Lamp Source Electronic

Shutter

Water Filter

Bandpass Filter

Beam Splitter

Resist Sample

UV-Enhanced Silicon Photodetectors

Neutral DensityFilter

Automated DataAcquisition System

Photokinetics Experiment

SpatialFilter

Photoresist Reaction Kinetics

• Dill parameters A, B and CA = bleachable component in resist

B = non-bleachable

component in resist

C = rate of photochemical

reaction in resist

• Rigorous A,B and C ExtractionSimultaneous solution to the thin filmoptics and the Beer-Lambert equations

A = 0.808 1/µmB = 0.086 1/µmC = 0.010 cm2/mJ

0.50

0.60

0.70

0.80

0.90

0 200 400 600

Exposure Dose (mJ/cm2)

Tra

nsm

ittan

ce (

%)

Experimental

Calculated

Mult-Wavelength DissolutionRate Monitor

Broadband Light Source

PC Card Spectrometers

PC DataAcquisitionSystem

Ref

lect

ance

WavelengthT

hick

ness

Time

Data Analysis I(λ,t)

Puddle Developer Cell withDeveloping Resist Film

Lithography Simulation

m’(x,z)

I(x)

z

x

E(x,z)

m(x,z)

Rig

orou

s P

hysi

cal

Rel

atio

nshi

psE

mpi

rica

lR

elat

ions

hips

ResistProfile

Input Parameters

PEB Time & Temp

Photokinetics Parameters

Development Rate Model

OutputSimulation

Optical Equations

Optical Equations

Photochemistry

Diffusion Equations

Dissolution Theory

Exposure Tool

Resist Optical Properties

m(x,z) à relative photoactive compound concentration

∇m à relative photoactive compound concentration gradient

E(x,z) à exposure energy distribution

Basis of Line Edge Optimization(0.5 µm Space in Resist)

-1000 -500 0 500 1000Horizontal Direction (nm)

Rel

ativ

e In

tens

ity

Develop Time Contours

Aerial Image

Output: Resist ProfileSidewall angle θ and feature size w

w

θ

PROLITH 2 v6.04

Relative PAC (m)Exposure Dose + Kinetics (A, B and C)

∇m

Lithographic Imaging Equation

dx

dm

dx

dR

xTH

*

γ=

dm

dRTH = γ

R = Dissolution ratex = Horizontal positionm = Relative PAC Concentrationγ TH = Theoretical resist contrast x* = Nominal edge of resist feature

–Maximize the change indevelopment rate at theedge of the resist feature

–Maximize ∇m = (dm/dx)through exposure dose andgiven aerial image

–Adjust location ofdissolution rate notch (γ ΤΗ)to the edge of the nominalresist edge (target m) withdeveloper concentration

Simulated Optimum Exposure Dose forPhotomask Lithography

2.0

2.4

2.8

3.2

100 150 200 250 300

Exposure Dose (mJ/cm2)

PA

C g

radi

ent (

m-1

)

IP3600PFI88A5

Determined exposure dose thatresulted in the maximum ∇∇m

Best DosesIP3600 ≅ 210 mJ/cm2

PFI88A5 ≅ 260 mJ/cm2

Determined m at the edge of theresist feature at the dose givingmaximum ∇m

Target mIP3600 = 0.3

PFI88A5 = 0.3

m(x,z) of IP3600 with a 210 mJ/ cm2 dose

Optimal Development Rate Function (R(m))

Note:Resists A, B, PFI88A, E and Fhave large dissolution ratenotches, however, the notchesoccur at high m. These resisthave been optimized for highthroughput on Si (high m ~low dose).

Optimal resist for masklithography has…

Large dissolution Notch

near Target m

R(m) response of six I-line resistsdeveloped with 0.26N TMAH

0

10

20

30

40

50

60

0.4 0.5 0.6 0.7 0.8

Relative PAC (m)Di

ssol

utio

n Ra

te (n

m/s

)

ABIP3600PFI88A5EF

Target m Notch

Optimal Dissolution Notch Location

IP3600 Notch Locationm ≅ 0.35

(developed w/ 0.20 N TMAH)

0

20

40

60

80

0 0.2 0.4 0.6 0.8Relative PAC (m)

Dis

s. R

ate

(nm

/s)

0.26NModel(0.26N)0.23NModel(0.23N)

0.20NModel(0.20N)

Target m

Lower developer concentration shifts notch toward the target m

PFI88A5 Notch Locationm ≅ 0.45

(developed w/ 0.23 N TMAH)

0

20

40

60

80

0 0.2 0.4 0.6

Relative PAC (m)

Dis

s. R

ate(

nm/s

)

0.26N

Model (0.26N)

0.23N

Model (0.23N)

Low Developer Conc.

Target m

Low Developer Conc.

Standing Waves in Photomask Resists

• Lower developer concentration amplifies influence of standing waves

• Hotplate built that mimics thebake profile of photomask surface

• Developed post-exposure baketo minimize standing waveson photomasks

0

2040

6080

100120

140

0 100 200 300Time (s)

Tem

pera

ture

(ºC

)

MIN

MAX

Photomask Post-Exposure Bake

0

5

10

15

20

25

0 200 400 600Thickness (nm)

Dis

solu

tion

Rat

e (n

m/s

)

IP3600 developed with0.20 N TMAH (NMD-W)

Photomask post-exposure bake improvesprocess latitude

IP3600/ 0.23N TMAH (NMD-W)/90 s dev. time/ 110 mJ/cm2/ 0.5 µm space

No PEB PEBFocus

+0.4 µm

Best Focus

-0.4 µmScum

StandingWaves

Scum

Simulated Process LatitudeImprovements for IP3600

IP3600 (TOK)/ 0.26N TMAHNMD-W/ No PEB

IP3600 (TOK)/ 0.20N TMAHNMD-W/ PEB

PROLITH 2/ 6.04/ 0.5 µm isolated space/ 82º+ sidewall angle/ ± 5% linewidth latitude

Current Photomask Process Improved Photomask Process

Simulated Process Latitude Improvementsfor High Resolution Resists

PFI88A5 (Sumitomo)/ 0.26NTMAH NMD-W/ PEB

PFI88A5 (Sumitomo)/ 0.23NTMAH NMD-W/ PEB

PROLITH 2/ 6.04/ 0.5 µm isolated space/ 82º+ sidewall angle/ ± 5% linewidth latitude

High-Resolution Photomask Process

Improved High-ResolutionPhotomask Process

Focus-Exposure Process LatitudeImprovements (0.5 µm space)

0.35

0.4

0.45

0.5

0.55

0.6

0.65

-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8

Cri

tical

Dim

ensi

on ( µ

m)

90100111123

0.35

0.4

0.45

0.5

0.55

0.6

0.65

-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8Focus (µm)

Cri

tical

Dim

ensi

on ( µ

m) 122

132

144

Current Photomask Process

Improved Photomask Process

High Resolution Photomask Process

IP3600/ 0.26N TMAH (NMD-W)/60 s dev./ No PEB

IP3600/ 0.20N TMAH (NMD-W)/180 s dev./ PEB

PFI88A5/ 0.26N TMAH (NMD-W)/180 s dev./ PEB

Manufacturing trials at the DPI Reticle Technology Center

Focus (µm)

0.35

0.4

0.45

0.5

0.55

0.6

0.65

-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8Focus (µm)

Cri

tical

Dim

ensi

on ( µ

m)

144160177.78197.53219.48243.87

Focus-Exposure Process LatitudeImprovements (0.3 µm spaces)

Improved Photomask Process

Current Photomask Process

High Resolution Photomask Process

IP3600/ 0.26N TMAH (NMD-W)/60 s dev./ No PEB

IP3600/ 0.20N TMAH (NMD-W)/180 s dev./ PEB

PFI88A5/ 0.26N TMAH (NMD-W)/180 s dev./ PEB

Manufacturing trials at the DPI Reticle Technology Center

0.25

0.29

0.33

0.37

0.41

0.45

-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8

Cri

tical

Dim

ensi

on (

177197219243

0.25

0.29

0.33

0.37

0.41

0.45

-0.8-0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8

Focus(µm)

Cri

tical

Dim

ensi

on (

µm) 132

144156

0.25

0.29

0.33

0.37

0.41

0.45

-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8

Cri

tical

Dim

ensi

on(

132

137

SCUMMEDFEATURES

Focus(µm) Focus(µm)

µm)µm

)

High Resolution Optical PhotomaskLithography

PFI88A5/ 0.26N TMAH NMD-W/ 180 s dev. time/ PEB/ 220mJ/cm2

0.5 µm isolated resist space 0.3 µm isolated resist space

Optimal Photomask Process forSub 0.30 µm Resist Features

Process Conditions• High Resolution I-line resists

• Higher exposure energy

• Lower developer concentration

• Longer development time

• Post-exposure bake

Process Parameter Optimization• Sub-0.30 µm resist features

• Improved focus-exposure process latitude

• Reduce Scumming/ defects

• Reduced CD sensitivity to dose

• Reduce pattern density effects

0.25 µm isolated resist space

PFI88A5/ 0.26N TMAH NMD-W/ 180 s dev. time/ PEB/ 220mJ/cm2

Acknowledgements

• ETEC Systems Inc.• SRC/ National Semiconductor Fellowship• Sumitomo, Shipley, Clariant, Olin and TOK• FINLE Technologies• J. A. Woollam (Ron Synowicki)