Potential and Lim its of Texture Measurem ent Techniques for I nlaid - - PowerPoint PPT Presentation

Potential and Lim its of Texture Measurem ent Techniques for I nlaid Copper Process Optim ization

Holm Geisler, Inka Zienert, Hartmut Prinz, Moritz-Andreas Meyer, Ehrenfried Zschech AMD Saxony LLC & Co. KG, Dresden, Germany

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 2

Outline

• Microstructure characterization of inlaid copper interconnects
• Texture measurement techniques

– X-ray micro-diffraction – OIM: EBSD & ACT

• Application

– Microstructure monitoring – ECD-filled inlaid structures with new ILDs, capping layers and barrier layers – Texture and stress – Orientation stereology, grain size, grain boundary distribution – Texture in ECD-filled via chains – Texture of barrier and seed layers before ECD filling

• Summary

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 3

Microstructure characterization of inlaid copper interconnects

• Aluminum vs. inlaid copper: What is different ?
• Texture, EM & defects
• Microstructure characterization: general concept
• Orientation distribution function (ODF)
• Quantification

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 4

+ sidewalls + twins + engaged

Texture Al vs. inlaid Cu Al Al I nlaid Cu I nlaid Cu { 111} { 111} { 111} { 511} twins

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 5

What is different ?

Inlaid Copper Interconnects Aluminum Interconnects Twins + Sidewall Bamboo + Columnar ???? Optimum EM Behaviour Large Anisotropy Small Anisotropy Electroplating Vapour Deposition Cu CMP Metal etch .... ....

E〈111〉 = 1 9 1 GPa E〈100〉 = 6 6 .7 GPa

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 6

Texture, EM & defects Prevent grain boundaries along the trench direction! = Fast diffusion pathways Electromigration:

• j

Sidewall-oriented grains? High-angle grain boundary?

Top view

void v > v

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 7

Microstructure Characterization: General Concept

• Microstructure Function:

). , , ( / ) (

2 1

ϕ ϕ Φ = = f dg V dV g f

g

x1 x2 x3 g(x) x

• Orientation Distribution

Function: strain lattice defects, ns

• rientatio

phase ) ( ) ( ) ( ) (      = x D x g x i x G

V

g

dg

KB KA

H.J. Bunge (1999, 2001)

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 8

Quantification: ODF approximation

[ K. Helming, http: / / www.texture.de/ ]

). , , ( / ) (

2 1

ϕ ϕ Φ = = f dg V dV g f

g

Euler angles ϕ1, Φ, ϕ2

KA: sample coordinates KB: cryst. coord. system

) , ( φ χ P

(111)+ (200)+ (220) e.g., ADC

χ

φ

( hkl) pole figure, P( χ,φ) OD, f( φ1, Φ, φ2)

dg f P hkl

∫

Φ =

π

φ φ π φ χ

2 2 1 ) (

) , , ( 2 1 ) , (

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 9

Evaluation of pole figures

• Computational algorithms for OD analysis

– Harmonic Methods: computation in Fourier space – Discrete (Direct) Methods: computation in orientation space:

• Commercial software: LaboTex

– based on ADC (Arbitrarily Defined Cells) – direct method, good for sharp textures – quantification of

• fibers
• engaged fibers
• tw ins

– uncertainty in determination of random texture component (background, low signal-to-noise ratio)

• K. Pawlik et al.

(1991) U.F. Kocks et al. (1998)

] ) , , ( ) , [( 1 ) , (

2 1 1 i N i h

f N P φ φ φ χ φ χ Φ ⇐ = ∑

=

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 10

Texture measurement techniques

• Overview
• X-ray micro-diffraction
• OIM (Orientation Imaging Microscopy)

– EBSD: Electron Backscatter Diffraction – ACT: Automated Crystallography for the TEM

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 11

Texture measurement techniques for inlaid Cu interconnects: Overview ACT µ-XRD EBSD

• Single Inlaid
• Dual Inlaid

ACT EBSD ?

• Barrier
• Seed

nano- cryst.

EBSD µ-XRD ?

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 12

Techniques: comparison

• µ-XRD

Classical Texture, ODF: f(g) = f(ϕ1, Φ, ϕ2) Phase: i Strain (Stress): Ds

• OIM

Orientation Stereology: g(x) Grain Size Grain-boundary distribution

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 13

Probed volume

• Beam diameter between 50µm and several 100µm

X-rays: Penetration depth > > µm

Met n Met n+ 1

γ

Vian ILD

Compared with EBSD: ≤ tens of nm

• X-rays „Bulk“ information, penetration of ILD

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 14

Statistics

• X-ray m icro-diffraction

– Beam diameter d = 100µm A = π r2 = 7854 µm 2 – Test pattern: parallel trenches, w = 180nm, p = 360nm – Assumption: mean grain diameter = w

(one grain extends over the whole line width and depth)

– n = (L / w) / (2Lw) = 1 / (2w2) L: length of the line – n ∼ 15 grains / µm 2 – N = n A ∼ 118000 grains

• EBSD

– A = 3µm x 10µm = 30 µm 2 – N = n A ∼ 450 grains

X-ray EBSD { 1 1 1 } pole figures

RD TD

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 15

X-ray

X-ray micro-diffraction

• Arrays of ECD-filled inlaid copper lines
• Arrays of ECD-filled inlaid line segments
• Arrays of ECD-filled vias (?)
• Process monitoring
• In-line application

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 16

Texture and stress measurements at inlaid test structures using X-ray micro-diffraction

• Tool perform ance:
• large detector area with high detector

sensitivity (80% quantum efficiency)

• small area beam focus with high

intensity

• Test structures:
• blanket or structured thin film samples

from 120 x 120 µm 2 up to 10 x 10 mm 2 Bruker AXS D8 micro-diffraction tool Huber goniometer with ¼ Eulerian cradle, PolyCap and area detector (GADDS) Video + laser for accurate height adjustment

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 17

X-ray micro-diffraction on arrays of inlaid Cu lines

• r

Copper lines Barrier Silicon Si(F)O, SiCOH SiN, SiCN

Narrow inlaid Cu lines Wide inlaid Cu lines

Inlaid structure > 120µm X-ray beam 80µm Ø

Metal 1

Video camera + laser beam alignment

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 18

X-ray µ-diffraction on arrays of Cu lines and line segments geometry effects

Narrow Cu lines and line segments – width = 180nm

4.3µm x 4.3µm 10µm x 10µm 10µm x 10µm 10µm x 10µm 10µm x 10µm

.......

{ 111} { 111} { 111} { 111}

X-ray beam ~ 8 0 - 1 0 0 µm Ø

good good poor bad

I t should w ork on dense via arrays

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 19

X-ray microstructure monitoring: Arrays of inlaid copper lines { 111} - narrow copper lines (180nm) { 111} - wide copper lines (1.8µm) Week

• ILD = Si(F)O
• SiN etch stop
• Metal 1
• sharp { 111} fiber
• engaged component
• { 511} twins
• sidewall-oriented grains

negligible

• Stability of process of

record TD RD RD TD

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 20

Texture components in Cu lines

Si

SiO2

Cu

( 1 1 1 ) lattice planes Sym m etry equivalent, 7 0 .5 ° Sym m etry equivalent, 7 0 .5 °

Blanked

{111} {112} {110} {111} (-110) {111} (0-11) (-1-12)

Sketch of (111) Pole Figure

Trenches

Engaged ( 1 1 1 ) Fiber Texture

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 21

Texture Components in copper lines

{111} {111} {111} {111} (-211) (-110) {111} (0-11) (-1-12) {111} {111} (-110) (-1-12) (-211) (0-11)

Sketch of { 1 1 1 } pole figure

Tilted sidewalls

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 22

Texture Components in copper lines

{111} {111}

Superposition sidewall + fiber

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 23

1st generation twins

{ 1 1 1 } pole figure { 111} center peak

Additional circles @ 38.9°, 56.2° and 70.5° in { 111} pole figure

{ 5 1 1 } : 1 st generation tw ins

{ 511} { 111} Twin

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 24

2nd generation twins

Additional circles @ 22.19°, 56.25° and 65.95° in { 111} pole figure

{ 111} centre peak { 1 1 1 } pole figure

{ 5 7 1 3 } : 2 nd generation tw ins

{ 5 7 13} { 111} Twin

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 25

Effect of different ILDs and etch stop layers on texture in copper lines

• Broadening transverse to the metal line direction
• 100 MPa
• 100 MPa

SiCOH + SiCN Si(F)O + SiN

• Narrow lines (180nm)

{ 111} { 111}

TD RD RD TD RD TD

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 26

Orientation spread: FWHM

1 2 3 4 5 6 7 8 9 10

1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0

Week

b)

w = 1.8µm

TD RD TD RD

ILD #2 ILD #1

TD RD

FWHM (111) [Deg.]

1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0

a)

w = 180nm

TD RD TD RD

ILD #2 ILD #1

TD RD

FWHM (111) [Deg.]

Deviation: changed ILD + capping layer Process of record stable Wide lines: no influence

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 27

Monitoring: Quantification

10 20 30 40 50 10 9 8

{111} total {111} {111}[0 1 -1] {111}[1 0 -1] Volume Fraction [%]

Week

{511} {611} {5 7 13}

• Narrow lines

(180nm)

• Weeks 8-9:

Si(F)O / SiN

• Week 10:

SiCOH / SiCN

• Uncertainty in volume fraction of random component !

With LaboTex (ADC)

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 28

Influence of barrier layers on final Cu texture in inlaid copper lines

Ta Std Ta resputter TaN resputter Ta/ TaN/ TiSiN/ Ta

T a S t d T a / T a N / T i S i N / T a T a N / T a T a / T a N / T a T a r e s p u t t e r T a N r e s p u t t e r T a T E O S

1 2 3 4 5 6 7 8 9 10 FWHM(RD) FWHM(TD)

FWHM {111} [Deg.]

Barrier Layer

5 10 15 20 25 30 35 40 45 50 55

{511} {611} {5 7 13}

Volume Fraction [% ]

{111} total {111} {111}[0 1 -1] {111}[1 0 -1]

b) a) Barrier Layer RD TD changed EM behaviour

!

ILD = SiCOH

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 29

Texture: explanation process (b)

{111}

RD TD

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 30

GADDS Texture & Stress

• GADDS + precise ¼ circle Eulerian cradle
• Pole figures + stress on patterned wafers
• n the same test structures
• Record higher order { hkl} , e.g., { 311}
• Study of possible influence of changed

texture on stress values

• Choose: (fiber + engaged) or fiber only
• Optimization of { χ; φ} for stress analysis
• 2D (triaxial) stress data analysis

Anisotropy & shear stresses

• Limit: intensity

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 31

GADDS 2D stress analysis (triaxial) Bruker AXS (311)

180nm inlaid copper lines

GADDS

shear

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 32

X-ray: in-line application

Why? Time critical issues (e.g., recrystallization) In-line process monitoring (also 300mm) How? Nondestructive, relatively fast X-ray See this conference:

• W E-10 : „Room Temperature Electroplated Copper Recrystallization: In-Situ

Mapping on 200/ 300mm Patterned Wafers“, K. J. Kozaczek, et al.

• W E-11 : „Metrology Tool for Microstructure Control on 300mm Wafers

During Damascene Copper Processing“, K. J. Kozaczek, et al.

• W E-17 : „Texture Evolution in Interconnects upon Annealing“, K. Mirpuri, et

al.

• W E-18 : „Microstructure Variations in Annealed Damascene Cu

Interconnects“, K. Mirpuri, et al.

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 33

X-ray summary:

X-ray micro-diffraction suitable for texture (and stress) analysis on arrays of ECD-filled inlaid copper lines, line segments, and vias (?) Nondestructive ( also in-line) X-ray lim its: Inlaid structures with barrier only or barrier and seed only (intensity!) Does not provide g(x), grain-boundary distribution and in- plane grain size

X-ray OIM

OIM: EBSD, ACT

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 34

EBSD: principle 70° sample tilt Kikuchi patterns Depth ∼ tens of nm SEM

Here: LEO 1 5 5 0 + therm . FEG, EBSD: TexSem Lab. ( TSL)

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 35

EBSD on copper lines & via chains

I nverse pole figure m aps ( a) 2 0 0 nm lines ( b) via1 , via3 , via5 twin boundary

a)

Grain size monitoring

g(x)

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 36

Limits of EBSD: passivation H= 27nm H= 34nm Narrow lines Cu Pad Narrow lines Cu Pad (200nm)

(Thickness of Passivation)

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 37

Limits of EBSD: passivation Lines Cu pad

IQ= 93 IQ= 143

EBSD Analysis

H= 3 0

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 38

EBSD: perspectives

• Future perspective

Sequential FIB + EBSD 3D orientation image (e.g., via cross-section)

• EBSD after EM test (difficult: before EM test !)

which grains & grain boundaries are critical ? locally resolved since g(x) is measured

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 39

Barrier & Seed: X-ray, EBSD ACT

• Lateral resolution of EBSD is of the order of grain sizes in

copper seed layers (∼30-50nm) EBSD not possible for nanocrystalline barriers and seed inside inlaid structures

• ACT

ACT needed for barriers and seed in inlaid structures

ACT – Automated Crystallography for the TEM

• Small, non-planar inlaid structures accessible by ACT
• But: time consuming
• Special sample preparation needed

[ TSL]

Multiple dark field images are collected by rotating the beam

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 40

ACT on Cu seed in inlaid structures

Courtesy of Holger Saage, Hans-Jürgen Engelm ann

AMD Saxony, Dresden, Germ any

Grain size distribution of Cu seed inside inlaid structures Grain orientation map not possible yet in this case Compare seed grain sizes inside the structures with seed grain sizes on top and in the ECD-filled metal layer underneath !!

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 41

Summary

• ECD-filled inlaid copper structures

X-ray micro-diffraction & EBSD process monitoring, texture, grain size, grain boundary distribution, trenches and vias

• Copper seed and nanocrystalline barrier layers in inlaid

structures

ACT needed

03/ 26/ 03 ICCM, March 24 - 28, 2003, Austin, Texas Holm Geisler 42

Trademark Attribution

AMD, the AMD Arrow Logo and combinations thereof are trademarks of Advanced Micro Devices, Inc. Other product names used in this presentation are for identification purposes

• nly and may be trademarks of their respective companies.