lecture 7 optical lithography

7/30/2019 Lecture 7 Optical Lithography

1/40

4/23/20

ECE416/516ICTechnologies

ProfessorJamesE.MorrisSpring2012

4/23/2012 ECE416/516ICTechnologies Spring2011

Chapter7OpticalLithography

2


2/40

4/23/20


Figure7.1SimplifiedICdesignprocessflowdiagram.3

Lecture topics:

A. Light sources

B. Wafer exposure systems

1. Optics

2. Projection

3. Proximity/contact

4. Immersion

C. Photoresist (also lecture 8) D. Simulation

E. Sub-wavelength lithography

F. Interference effects

G. Minimum feature sizes

LectureObjectives:

Can calculate resolution limits

and line width errors due to

optical limitations

Can calculate PR exposure

variations due to standing wave

effects

Lecture emphasis:

Optical Limits

o Proximity Printing

o Projection Printing

Technology Comparisons


3/40

4/23/20

- Process of image transfer to wafer

- Origin pattern drawn N x

- Images control diffusion, oxidation, & metallization sequences

- expose parts, mask others

-Masking levels refer to each mask used

-Lithographic Sequence:

1. Draw mask 100-2000x final size

2. Photographically reduce to 10x final size (glass)

3. Step & repeat -> [1 x final size] x [ matrix of images ] (glass)

4. Spin coat substrate with photoresist

- thickness (spin rate)-1/2

5. Expose PR through mask & develop to dissolve unwanted PR, etc.

Registration tolerance --> design rules

f1

f2 Mask 1ideal

pos iti on

Mask 2 ideal

po sit ion


4/40

4/23/20

Estimated final tolerance:

- T 3[( f1/2 )2 + (f2/2 )2 + r2]1/2

where r is registration errors Gaussian distribution if independent

For f1 ~ f2 ~ r~ 0.15m- T ~ 0.6 m- limits how densely features can be packed

Ideal feature separation mask 1 to mask 2> 1/2 (f1 + f2)


Figure7.2Excerptoftypicaldesignruleset.Thisportiondealswithfirstmetalrulesforaparticulartechnology.

8


5/40

4/23/20


Figure7.3 Typicalphotomasksincluding(fromleft)a1 plateforcontactorprojectionprinting,a10 plateforareductionstepper,anda10 platewithpellicles.

9


Figure7.4 Projectedlithographyrequirementsshowingoverlayaccuracy(rightaxis)andresolutionrequirements(leftaxis).Datatakenfrom2005InternationalTechnology RoadmapforSemiconductors.

10


6/40

4/23/20

11

LITHOGRAPHY

0.7X in linear dimension every 3 years. Placement accuracy 1/3 of feature size. 35% of wafer manufacturing costs for lithography. Note the ??? - single biggest uncertainty about the future of the roadmap.

Year of Production 1998 2000 2002 2004 2007 2010 2013 2016 2018

Technolo gy Node (half pitch) 250

nm

180 nm 130 nm 90 nm 65 nm 45 nm 32 nm 22 nm 18 nm

MPU Printed Gate Length 100 nm 70 nm 53 nm 35 nm 25 nm 18 nm 13 nm 10 nm

DRAM Bits/ Chip (Sampl ing) 256M 512M 1G 4G 16G 32G 64G 128G 128G

MPU Transistors/C hip (x106) 550 1100 2200 4400 8800 14,000

Gate CD Control 3 (nm) 3.3 2.2 1.6 1.16 0.8 0.6

Overlay (nm) 32 23 18 12.8 8.8 7.2

Field Size (mm) 22x32 22x32 22x32 22x32 22x32 22x32 22x32 22x32 22x32

Exposure Tech nology 248

nm

248 nm 248 nm

+ RET

193nm

+ RET

193nm +

RET

193nm

+ RET

+ H2O

193nm

+ RET

+ H2O

157nm??

??? ???

Data Volume/Mas k lev el (GB) 216 729 1644 3700 8326 12490


12

Historical Development and Basic Concepts

Patterning process consistsof mask design, maskfabrication and waferprinting.

CAD System Layout Simulation Design Rule Checking

ElectronGun

Focus

Deflection

Mask

Mask Making Wafer Exposure

LightSource

Mask

Wafer

CondenserLens

ReductionLens

AerialImage

(Surface)

N Well P Well

P+ N+ N+P+

P

LatentImage

in Photoresist

TiN LocalInterconnect Level

(See Chapter 2)

It is convenient to divide thewafer printing process into threeparts A: Light source, B. Waferexposure system, C. Resist.

Aerial image is the pattern ofoptical radiation striking the topof the resist.

Latent image is the 3D replicaproduced by chemical processesin the resist.



7/40

4/23/20

13

A. Light Sources

Decreasing feature sizes require the use of shorter .

Traditionally Hg vapor lamps have been used which generate many spectrallines from a high intensity plasma inside a glass lamp.

(Electrons are excited to higher energy levels by collisions in the plasma.Photons are emitted when the energy is released.)

g line - i line - (used for 0.5 m, 0.35 m)

Brightest sources in deep UV are excimer lasers

436 nm 365 nm

Kr NF3 energy KrF photon emission (1)

KrF - (used for 0.25 m, 0.18m, 0.13 m) ArF - (used for 0.13m, 0.09m, . . . ) FF - (used for ??)

248 nm 193 nm 157 nm

Issues include finding suitable resists and transparent optical components atthese wavelengths.



Figure7.14 Linespectraoftypicalmercuryarclampshowingthepositionsofthetwolinesmostcommonlyusedinlithography.

Figure7.13Typicalhighpressure,shortarcmercurylamp(courtesyOsramSylvania).HighTelectronsgray body radiation, (absorbed by silica glass)

Hgatoms

14


8/40

4/23/20


Figure7.15Schematicofatypicalsourceassemblyforacontact/proximityprinter(afterJain).15


Figure7.17Opticaltrainforanexcimerlaserstepper(afterJain).16


9/40

4/23/20

17

B. Wafer Exposure Systems

Three types ofexposure systemshave been used.

MaskPhotoresist

Si Wafer

LightSource

OpticalSystem

Contact Printing Proximity Printing Projection Printing

Gap

1:1 Exposure Systems Usually 4X or 5XReduction

Contact printing is capable of high resolution but has unacceptable defectdensities.

Proximity printing cannot easily print features below a few m (except forx-ray systems).

Projection printing provides high resolution and low defect densities anddominates today.

Typical projection systems use reduction optics (2X - 5X), step and repeat or stepand scan mechanical systems, print 50 wafers/hour and cost $10 - 25M.


Usually x, y, & control to align mask to substrate

- alignment marks

mask

PR

wafersContact

Proximity

Projection

Simultaneous XY scan

of mask & wafer


10/40

4/23/20


Figure 7.5 Schematic of a simple lithographic exposure system.

19


Figure7.6 Huygens'sprincipleappliedtotheopticalsystemshowninFigure7.5.Apointsourceisusedtoexposeanapertureinadarkfieldmask.

dRR

eAjR

erEr

RRjk

rj

),(

),(

0

)(

)(),(

20


11/40

4/23/20

21

B1. Optics - Basics and Diffraction (Huygens-Fresnel principle)

Ray tracing (assuming light travels in straight lines) works well as long as thedimensions are large compared to .

At smaller dimensions, diffraction effects dominate.

LightSource

ApertureImagePlane

If the aperture is on the order of , the light spreads out after passing through

the aperture. (The smaller the aperture, the more it spreads out.)

a). b).

2121

2

1

2

1

0101

2

000

*

cos

sourcespoint2for0,)(

,)(

2121

EEEE

eEeEeEeEILWb

EeEeEILWa

jjjj

jj

4/23/2012

22

Diffracted LightCollected Light

ImagePlaneAperture

PointSource

CollimatingLens FocusingLens

f

d

A simple exampleis the image formedby a small circularaperture (Airy disk).

Note that a point imageis formed only if ,

or .

If we want to image the apertureon an image plane (resist), we cacollect the light using a lens andfocus it on the image plane.

But the finite diameter of the lensmeans some information is lost(higher spatial frequencycomponents).

1.22 f/d 0f 0 d

Diffraction is usually described in terms of two limiting cases: Fresnel diffraction - near field. Fraunhofer diffraction - far field.

wheref=focallength,d=focusinglensdiameterDiameterofcentralmaximum=1.22f/d



12/40

4/23/20


Figure 7.8 Typical far field (Fraunhofer) image.

ndiffractioFraunhoferIf222 rgW

g

yL

g

yL

I

g

xW

g

xW

IIIg

LWoIyxI yxyxe

2

2sin

,2

2sin

,)2)(2(

)(),( 222

B2

Projection

printing


Figure 7.21 Schematic for the optical train of a simple projection printer.

Projection

NA=nsinRefractiveindex n=1forairRayleighsresolutioncriterion Wmin =k1/NA

24


13/40

4/23/20

25

B2. Projection Systems (Fraunhofer Diffraction)

These are the dominant systems in use today. Performance is usually describedin terms of: resolution

depth of focus field of view modulation transfer function alignment accuracy throughput

ImagePlane

PointSources d

A

BA'

B'

EntranceAperture

Consider this basic optical projectionsystem.

Rayleigh suggested that a reasonablecriterion for resolution is that thecentral maximum of each point sourcelie at the first minimum of the Airypattern.

With this definition,

sin

61.0

sin2

22.122.1

nfn

f

d

fR (2)

The numerical aperture of the lens is by definition, NA represents the ability of the lens to collect diffracted light.

NA n sin (3)4/23/2012 ECE416/516ICTechnologies Spring2011

26

R 0.61 NA k 1 NA (4) k1 is an experimental parameter which depends on the lithography system

and resist properties ( 0.4 - 0.8).

Obviously resolution can be increased by: decreasing k1 decreasing increasing NA (bigger lenses)

However, higher NA lenses also decrease the depth of focus. (See next slidefor derivation.)

DOF 2 NA

2 k2

NA2

(5)

k2 is usually experimentally determined.

Thus a 248nm (KrF) exposure system with a NA = 0.6 would have a resolutionof 0.3 m (k1 = 0.75) and a DOF of 0.35 m (k2 = 0.5).

R=Wmin &DOF= inCampbell4/23/2012 ECE416/516ICTechnologies Spring2011


14/40

4/23/20

4/23/2012 ECE416/516ICTechnologiesSpring2011 27

d

Entranceaperture

Imageplane

Limitoffocus

f

RayleighcriterionforDOF:

222

22

)()(2

2/sin

2211

4:smallFor

4/cos

NAk

NADOF

NAf

d

28

Another useful concept is the modulation transfer function or MTF, defined

as shown below.

MTF IMAX IMIN

IMAX IMIN (6)

Intensityat Mask

Intensityon Wafer

0

1

0

1IMAX

IMIN

Aperture

LightSource

CondenserLens

Objective orProjection

Lens

Photoresiston Wafer

Mask

Position Position

Note that MTF will be afunction of feature size



15/40

4/23/20


Figure 7.9 Far field image for a diffraction grating.

3

2

15

15

shownpatterngrating

ndiffractiofore.g.

image,anfor

:Define

minmax

minmax

MTF

II

II

MTF

29


Figure7.10

Plot

of

dose

versus

position

on

the

wafer.

Dose

is

given

by

the

intensity

of

thelightintheaerealimagemultipliedbytheexposuretime.TypicalunitsaremJ/cm2.

DD100 PRwillcompletelydissolveindeveloperD0


16/40

4/23/20

IdealPRdevelopsforD>Dcr only ActualPRpartiallydevelopsforD0>D>D100

>Partiallydissolves

1

0

1

0

Dose

Dose

Ideal PR

Actual PR

DCr

D100%

D0%

Rayleighresolutionlimit:ResolvableseparationSr =0.6 /NA

Examplesforobjectsequalsize&spacing

Notesquarepatternfromsineintensityallowsopticalprinttofinerlinesthanoriginallypredicted(duetoclippingactionofnonlinearPR)

Notealsocoherenceeffects(CampbellFig.7.18)

Image

intensities

Developed

patterns

optically

resolvable


17/40

4/23/20

33

Finally, another basic concept is thespatial coherence of the light source.

LightSource

MaskCondensorLens

s d

Practical light sources are not pointsources.the light striking the mask will not beplane waves.

The spatial coherence of the system isdefined as

S light source diametercondenser lens diameter

sd

(7)

or often as S NA condenser

NA projection optics (8)

Typically, S 0.5 to 0.7 in modern systems.

[Note:SketchMTFvs (featuresize)1andvariationwithSnext]4/23/2012 ECE416/516ICTechnologies Spring2011


Figure7.22 Modulationtransferfunctionasafunctionofthenormalizedspatialfrequencyforaprojectionlithographysystemwithspatialcoherenceasaparameter.

S=(Sourceimagediameter)/(Pupildiameter)

34


18/40

4/23/20

35

Illumination System EngineeringAperture Projection

Lens

"Captured"Diffracted

Light

"Lost"Diffracted

Light

"Captured"Diffracted

Light

"Lost"Diffracted

Light

Wafer

Off-axis illumination also allowssome of the higher order diffractedlight to be captured and hence canimprove resolution.

Aperture

PointSource

CollimatingLens

ProjectionLens

Photoresiston Wafer

Kohler illumination systems focus thelight at the entrance pupil of the objectivelens. This captures diffracted lightequally well from all positions onthe mask.

Advanced optical systems using Kohlerillumination and/or off axis illumination

are commonly used today.



Figure7.23Schematicfortheoperationofascanningmirrorprojectionlithographysystem(courtesyofCanonU.S.A.).

36


19/40

4/23/20


Figure7.26ConfigurationofStepandScan193nmsystem.Thelaserisattheleft.Thereticleisattheupperright,whilethewaferisatthelowerright(photocourtesyASML).

37


Figure7.7 Typicalnearfield(Fresnel)diffractionpattern.

ndiffractioFresnelIf 222 rgW

ForWverylarge ray tracing

Image width on waferW=W(g/D)

38

B3. Contact/proximity systems


20/40

4/23/20

39

B3. Contact and Proximity Systems (Fresnel Diffraction)

Contact printing systems operate in the near field or Fresnel diffraction regime. There is always some gap g between the mask and resist.

WaferResistMaskAperture

IncidentPlaneWave

Light Intensityat Resist Surface

g

W

The aerial image can be constructed by imagining point sources within theaperture, each radiating spherical waves (Huygens wavelets). Interference effects and diffraction result in ringing and spreading outside

the aperture.


40

Fresnel diffraction applies when g W2 (9)

Within this range, the minimum resolvable feature size is Wmin g (10) Thus if g = 10 m and an i-line light source is used, Wmin 2 m.

Summary of wafer printing systems:

WaferResistMaskAperture

IncidentPlaneWave

Light Intensityat Resist Surface

W

Contact

Proximity Projection

Separation Dependson Type of System



21/40

4/23/20

Maskclosetowafer/resistminimizediffraction/divergence.

Separation ~m Fresnel(nearfield)diffraction Minimumlinewidth:

Wm =(d )1/2

(d=maskwafer

separation)

10

1

.11 10 100

Mask-wafer spacing (m)

Line width (m) Wm

= 36 nmuv = 249 nm

Deep uv

1 nmx-ray

D g

ForW2>> (g2+r2)1/2

W=W(g/D)

r

mask

pt source

2 W

2 L

wafer

wafer

position

Intensity


22/40

4/23/20

Penumbraeffectonlinewidth: W=2dtanwhere angleoffcollimation

Wellcollimatedlightminimizespenumbra. Wellcollimatedlightmaximizesdiffraction. ~ fewdegreesforoptimumbalance.

w

mask

wafer

d


Figure7.19Typicalcontactexposuresystem(courtesyofKarlSuss).

Contactexposure

44


23/40

4/23/20


Figure7.20 Intensityasafunctionofpositiononthewaferforaproximityprintingsystemwherethegapincreaseslinearlyfromg = 0tog = 15m(afterGeikasandAbles).

Contact proximity

Gapg


24/40

4/23/20


Figure7.24 Setupofanimmersionsystemusingsurfacetension(fromSwitkesetal.,reprintedfromtheMay2003editionofMicrolithographyWorld.Copyright2003byPennWell.)

B4.Immersion

gratinguniformafor2WPwhere.

andO,Hfor43.1

)/(1

)/(

DOF

DOFand

)2/(sin.4

P2

2

22

1(air)n

n(liq)

2

Pnn

P

Pn

nDOF

P

47


Figure7.25 NozzlesystemusedbyNikontoputthewaterdownandsuctionitupforeachstage(fromGeppert,reprintedbypermissionIEEE.)

48


25/40

4/23/20

49

C. Optical Intensity Pattern in the Resist (Latent Image)

Aerial Image Ii(x,y)

Exposing Light

P

P WellN Well

PNP+ P+ N+ N+

Latent Image inResist I(x,y,z)

The second step in lithographysimulation is the calculation of thelatent image in the resist.

The light intensity duringexposure in the resist is a functionof time and position because of Light absorption and bleaching. Defocusing. Standing waves.

These are generally accounted for by modifying Eqn. (21) as follows:

I x, y, z I i x, yI r x ,y , z (22)

where Ir(x,y,z) models these effects.


50

Microns0 0.8 1.6 2.4

Microns

0

0.4

0.8

1.2

Example of calculation of light intensitydistribution in a photoresist layer duringexposure using the ATHENA simulator.A simple structure is defined with aphotoresist layer covering a silicon substrawhich has two flat regions and a slopedsidewall. The simulation shows the [PAC]calculated concentration after an exposureof 200 mJ cm-2. Lower [PAC] valuescorrespond to more exposure. The colorcontours thus correspond to the integratedlight intensity from the exposure.

C. Photoresist Exposure

Neglecting standing wave effects (for the moment), the light intensity in theresist falls off as

dI

dz I (23

(The probability of absorption is proportional to the light intensity and theabsorption coefficient.)



26/40

4/23/20

51

The absorption coefficient depends on the resist properties and on the [PAC].

resist Am B (24)where A and B are resist parameters (first two Dill parameters) and

m PACPAC

0

(25)

m is a function of time and is given by

dm

dt C Im (26)

Substituting (24) into (23), we have:

IBAmIdz

dI (27)

Eqns. (26) and (27) are coupled equations which are solved simultaneously byresist simulators.


[PAC] = photoactive compound concentration

C is 3rd Dill parameter

52

The Dill resist parameters (A, B and C) can be experimentally measuredfor a resist.

LightSource

CondenserLens

Photoresiston Transparent

SubstrateFilter to Select

Particular

LightDetector

TransmittedLight

D

Transmittance

Exposure Dose (mJ cm-2)

1

0.75

0.5

0.25

200 400 600

T

T0

By measuring T0 and T,

A, B and C can be extracted.

2

120

1200

0

1

11where|

)1(

ln1

ln1

resist

resistE

n

nT

dE

dT

TTAT

BAC

TD

B

T

T

D

A



27/40

4/23/20


Figure7.11Contourplot(left)and10slicethroughthecenterofprojectedintensityasafunctionofpositionforagratingwith1.6mpitch,exposedatR=436nmandNA= 0.43.

D.Simulation

53


Figure 7.12 Same 1D plot for a grating with a pitch of 0.8 m exposed on the same system.

54


28/40

4/23/20

55

Models and Simulation

Lithography simulation relies on models from two fields of science: Optics to model the formation of the aerial image.

Chemistry to model the formation of the latent image in the resist.

A. Wafer Exposure System Models

There are several commercially available simulation tools that calculate theaerial image - PROLITH, DEPICT, ATHENA. All use similar physical models.

We will consider only projection systems.

Light travels as an electromagnetic wave.

W, t Re U Wejt where U W C Wej P P, t C Wcos t (t)

or, in complex exponential notation,

(13)

(14)


56

Consider a genericprojection system:

Aperture

LightSource

Condenser

Lens


Lens

Photoresiston Wafer

Mask

z

x

y

x1y1 Planex'y' Plane

x y Plane

The mask is considered to havea digital transmission function:

After the light is diffracted, it is

described by the Fraunhoferdiffraction integral:

where fx and fy are the spatialfrequencies of the diffractionpattern, defined as

t x1 , y 1 1 in clear areas0 in opaque areas

(15)

x',y' t x1 , y1e2j fxxfy ydxdy (16)fx x'

z and fy y'z4/23/2012 ECE416/516ICTechnologies Spring2011


29/40

4/23/20

57

(x,y) is the Fourier transform of the mask pattern. fx ,fy F t x1 , y 1 (17)

The light intensity is simply the square of the magnitude of the field, so thatI fx ,fy fx ,fy 2 F t x1 , y 12 (18)

Example - consider a longrectangular slit. The Fouriertransform of t(x) is in standardtexts and is the sin(x)/x function.

Mask

t(x)

F{t(x)}

I(x')

w/2

x

z

Aperture

LightSource

CondenserLens


Lens

Photoresiston Wafer

Mask

z

x

y

x1y1 Planex'y' Plane

x y Plane


58

But only a portion of the light is collected. This is characterized by a pupil function:

P fx ,fy 1 if fx2 fy2 NA

0 if fx2 fy2 NA

(19)

The objective lens now performs the inverse Fourier transform.

x, y F1 fx ,fy P fx , fy F1 F t x1 , y1 P fx ,fy (20)

resulting in a light intensity at the resist surface (aerial image) given by

I i x, y x, y2 (21)

Summary: Lithography simulators performthese calculations, given a mask design and

the characteristics of an optical system.These simulators are quite powerful today.Math is well understood and fast algorithmshave been implemented in commercial tools.These simulators are widely used.

MaskTransmittance

t(x1,y1)

Far FieldFraunhofer

Diffraction Pattern Pupil FunctionP(fx,fy)

Lens Performs InverseFourier Transform

Light Intensity

(x,y) = F-1{(fx,fy)P(fx,fy)} Ii(x,y) = (x,y)2

(fx,fx) = F{t(x1,y1)}4/23/2012 ECE416/516ICTechnologies Spring2011


30/40

4/23/20

59

Microns

Microns

0

1

2

3

-1

-2

-3 0 1 2 3-1-2

-3

Microns0 1 2-1-2

Microns

0

1

2

-1

-2

Microns

0

1

2

-1

-2

Microns0 1 2-1-2

Exposure system: NA =0.43, partially coherentg-line illumination( = 436 nm). Noaberrations ordefocusing. Minimumfeature size is 1 m.

ATHENA simulator (Silvaco). Colors correspond to optical intensity in theaerial image.

Same example except thatthe feature size has beenreduced to 0.5 m. Notethe poorer image.

Same example except thatthe illuminationwavelength has now beenchanged to i-lineillumination ( = 365 nm)and the NA has beenincreased to 0.5. Note the

improved image.


60

E. SubWavelength Lithography

Beginning in 1998, chip manufacturers began to manufacture chips withfeature sizes smaller than the wavelength of the light used to expose thephotoresist.

This is possible because of the use of a variety of tricks- illumination system optimization- optical pattern correction (OPC), and- phase shift mask techniques.

E. Mask Engineering - OPC and Phase Shifting

Optical Proximity Correction (OPC) can be used to compensate somewhat for

diffraction effects. Sharp features are lost because higher spatial frequencies are lost due to

diffraction. These effects can be calculated and can be compensated for.



31/40

4/23/20


Figure 7.27 Basic concept of phase shift masksas described by Levenson et al. [49].

Creating phase shift masks involves massive numerical calculations and oftenthe implementation involves two exposures - a binary mask and a phase shiftmask - before the resist pattern is developed.

61

62

Another approach uses phase shifting to sharpen printed images.

Mask

Amplitudeof Mask

Amplitudeat Wafer

Intensityat Wafer

180 Phase Shift

A number of companies now provide OPC and phase shifting software services. The advanced masks which these make possible allow sharper resist images

and/or smaller feature sizes for a given exposure system.



32/40

4/23/20


Figure 7.28 Selfaligned method of phase shifting the radiation at the edges of a pattern.

63


Figure7.29 Lightfromtheexposedregionscanbereflectedbywafertopologyandbeabsorbedintheresistinnominallyunexposedregions.

Here,inthelaterstages,withsurfacetopography

64

Reflection & Scattering


33/40

4/23/20


Figure7.30 Resistlinesforpatterningthemetalrunningdiagonallyfromthelowerleft.OntherightisanexpandedviewofthegeometrydescribedinFigure7.31(afterListvanetal.).

65


Figure7.31Micrographsofresistlinesonreflectivemetalwithoutandwitha2700antireflectinglayerundertheresist(afterListvanetal.).

66


34/40

4/23/20


Figure7.32Theformationofstandingwavesintheresist.67

F.InterferenceEffects

EI =E0 cos(t - z)ER =RE0 cos(t + z) =2 (n i k)/R=reflectioncoefficient=r2

=[(n1n2)/(n1+n2)]2

n2 =Sicomplexrefractiveindexn1 =PRcomplexrefractiveindexk 0forPRoroxide, 2 n/

EI +ER =E0[cos(t z)+Rcos(t + z)]>StandingWave

Es=Eso sint sin((2 nz)/ + )Maxima/Minimaseparations /2nMinimawhen(2 nz/ )+ =m

m=0,1,2,...Forr= 1(Si,GaAs,metals), =0.minimaatz=m/2n,m=0,1,2&exposureduetointensityI=I0 sin2 z

Notes:(1)Poordevelopmentatsurface(2)PRonoxide>opticalpropertiessimilar,soSWcontinuousacrossinterface;

adjustthicknessofoxideZ0 togetantinodeatPR/oxideinterface.

Z0 =(2m+1) /4n(3)EliminateSWformation:

(a)thinabsorbinglayerbetweenSi&PR

(b)dyeinPRforlightscatter(4)PlasmaetchremovesPR

residue


35/40

4/23/20

1988Technologies: Optical Proximity

Full Wafer Projection

Optical Stepper

Deep UV contact

X-Ray

Electron beam & ion beam

Minimum Feature Size

0.1 1 10 m

Light

Developed

resist

Focussed

image

edge

- 0 +

Exposure Energy

mJ/cm

W (m)0.1

-0.1

-0.2

-0.3

-0.4

0

20 807060504030


36/40

4/23/20

W (m)

Light

Developed

resist

Focussed

image

edge

- 0 +

Ratio of Water to Developers

1 65432

0

-0.8

-0.6

-0.4

-0.2

+0.2

Light

Developed

resist

Focussedimage

edge

- 0 +

1 5432 60

0.05

0.1

W (m)

Time Delay between exposure and Development


37/40

4/23/20

.8 1.0 1.2 1.4 1.6 m

Light

Developed

resist

Focussed

image

edge

- 0 +

-0.2

-0.1

0

W (m)

Resist Thickness

Increasingexposuretimedecreasesintensityforthreshold.

Light

Developed

resist

Focussed

image

edge

- 0 +

Light intensity

mask

Diffraction

No change

Defocus

Incr expos energy

-0.9 -0.6 -0.3 0 0.3 0.6 0.9 1.2 1.5

m0

0.2

0.4

0.6

0.8

1.0

1.2


38/40

4/23/20

Light

Developed

resist

Focussedimage

edge

- 0 +

W (m)

Defocus

Exposure times

8.3 sec

5.3 sec

0 5 10 15 20 m

-0.2

-0.1

0

+0.1

+0.2


Figure7.33Twotypicalregistrationerrors.Temperaturechanges

76


39/40

4/23/20

Reliesonfractureofmetalfilm

PR

metal

Less soluble PR

(surface treated)

Highly soluble PR

metal

metal

78

Some Final Thoughts

Lithography is the key pacing item for developing new technology generations.

Exposure tools today generally use projection optics with diffraction limitedperformance.

Lithography simulation tools are based on Fourier optics and do an excellent jobof simulating optical system performance. Thus aerial images can be accuratelycalculated.

A new approach to lithography may be required in the next 10 years.



40/40

4/23/20

Problems: 6.1 7.46.2 7.56.5 7.6/7

Midtermcourseevaluation:Summarizewhatyoulikeaboutthecourse,andwhatyoudontlike.

(Suggestion:Identify5each.)Considerlectures,textbook,assignments,notes,videos,etc.(Weightedas2problems)

4/23/2012 ECE416/516ICTechnologiesSpring2011 79

Phasemasking Specificdesignexample

Opticalproximitycorrection Specificdesignexample

Throughsiliconvias (TSVs) Solidcopper&barrel

Planarization Chemical/MechanicalPolishing

Onchip

resistors

Materials/fabrication/properties

Onchipinductors Spiralandferrite

Waferthinning Thinwaferapplications

Cryopumping Theoryandpractice

Unbalancedmagnetron Sources/deposition

Thinmetalfilmresistivity (Continuous)thicknessvariation

Stiction avoidance Includingsurfacestructuring

Electroless deposition Theoryandtechniques

Atomiclayerdeposition Theoryandtechniques

MEMSmotorfabrication Linearandrotational

4/23/2012 ECE416/516ICTechnologies Spring2011 80

lecture 7 optical lithography

Documents