1

1

The Effect of Ash Processes on Inorganic Porous Low-k Materials

Lei Jin, Anil Bhanap, Anna Camarena, Ananth Naman

Honeywell Electronic Materials

Carlo WaldfriedAxcelis Technologies, Inc.

2

OutlineOutline

• Background/Motivation

• Effects of Downstream Ash

• Effects of In-Situ RIE Ash

• Evaluation of New Ash Chemistries

• SLM (Single Layer Metal) Integration

• Conclusions

• Background/Motivation

3

Why the Need for LowWhy the Need for Low--kk

As transistor size shrinks, the interconnect dominates propagation delays

It’s all about speed

4

Deposition Process Candidate Materials k value VendorBlack Diamond™ < 3.0 Applied Materials

Coral® 2.5-3 NovellusAurora™ 2.0-2.7 ASMIOrion™ 2.0-2.5 Trikon

LKD 5109 2.2 JSRHOSP™ 2.5 Honeywell

NANOGLASS® 1.5-2.6 HoneywellSiLK* 2.1-2.6 Dow Chemical

GX-3™ 1.9-2.7 HoneywellZircon™ 1.9-2.7 Shipley

Spin-On

CVD

LowLow--k Materials Choicesk Materials Choices

Preservation of Porosity and Composition of Si-C-OH Materials Dictate Process Window

Carbon based materials Si-C-OH Materials

SiLK is a trademark of the Dow Chemical CompanyCoral is a registered trademark of NovellusZircon is a trademark of Shipley CompanyAurora is a trademark of ASMIOrion is a trademark of TrikonBlack Diamond is a trademark of Applied Materials

5

LowLow--k Materials Integration Challengesk Materials Integration Challenges

Patterning

• Photo resist poisoning

• Etch profile control

• Ash damage (Surface limiting or Bulk)

• Wet clean compatibility

Metallization

• CTE mismatch between Cu and Low-k materials

• Step coverage / Side wall roughness

CMP

• Adhesion of Low-k materials

• CMP delamination

6

OutlineOutline

• Background/Motivation

• Effect of Downstream Ash

• Effects of In-Situ RIE Ash

• Evaluation of New Ash Chemistries

• SLM (Single Layer Metal) Integration

• Conclusions

2

7

Effects of Conventional Downstream Ash Effects of Conventional Downstream Ash -- DimensionDimension

DownstreamO2 Ash

Side-wall bowing was observed after conventional downstream O2 ashing

Post Downstream-AshPre-Ash

1100W, 1200mT, 2000sccm O2, 250°C

8

Effect of Conventional Downstream Ash Effect of Conventional Downstream Ash -- CompositionComposition

All organic species are lost after conventional downstream ash process as indicated by the disappearance of SiC and CH peaks in FTIR spectra

-0.10

-0.08

-0.06

-0.04

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

Abso

rban

ce

1000 1500 2000 2500 3000 3500 Wavenumbers (cm-1)

CH: 2978 cm-1

SiC: 1277 cm-1

SiO

Post Cure

Post Etch

Post down-stream O2 Ash

Sheet Film Nanoglass – porous MSQ type material

9

OutlineOutline

• Background/Motivation

• Effect of Downstream Ash

• Effects of In-Situ RIE Ash

• Evaluation of New Ash Chemistries

• SLM (Single Layer Metal) Integration

• Conclusions

10

InIn--situ RIE Ash situ RIE Ash -- DimensionDimension

Pre-Ash

Post O2 RIE Ash

Use of an in-situ RIE ash preserves CD

In-situ RIE Ash

Post N2/H2 RIE Ash

11

Effect of Etch & InEffect of Etch & In--situsitu Ash Ash -- CompositionComposition

Significant carbon depletion in the bulk is measured after etch

and ash processes

Chg% SiC/SiO

-11.71-48.80-39.17

Post CurePost etch

Post Etch+O2 AshPost Etch+N2/H2 Ash

SiC/SiO0.03030.02670.01550.0183

-0.10

-0.08

-0.06

-0.04

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

Abso

rban

ce

1000 1500 2000 2500 3000 3500 Wavenumbers (cm-1)

CH: 2978 cm-1

SiC: 1277 cm-1

SiO

Post Cure

Post Etch

Post RIE O2Ash-0.10-0.08

-0.06

-0.04

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

0.22

Abso

rban

ce

1000 1500 2000 2500 3000 3500 Wavenumbers (cm-1)

Post Cure

Post Etch

Post RIE N2/H2Ash

Sheet Film Nanoglass – porous MSQ type material Sheet Film Nanoglass – porous MSQ type material

12

Effect of Etch & InEffect of Etch & In--situ Ash on Surface Carbon Depletionsitu Ash on Surface Carbon DepletionBare Si wafers

NGE coat/bake/cure (3000A)

RIE etch(1700A)

N2/H2 Ash(1400A)

Wet clean(DI Water)

XPS

Effect of Oxidative Ash

Effect of Reductive Ash

XPS

XPS

XPSXPS O2 Ash(1500A)

Wet clean(DI Water)

XPS

Control 16.1Etch 11.9Etch/O 2 ash 2.7

Etch/O2

ash/DI2.7

%C ((C-C,H)

Control 16.1Etch 11.9Etch/N2-H2ash

2.6

Etch/N2-H2ash/DI

3.3

%C ((C-C,H)

• Etch results in 30% carbon depletion @ surface

• Etch+Ash results in 80% carbon depletion @ surface

3

13

Consequence of CConsequence of C--depletion in Porous depletion in Porous SiCOH SiCOH Materials Materials

L/S = 0.5/0.24

L/S = 10.0/0.24

L/S = 0.5/1.0

Voids are observed in porous dielectric layer

after pre-CMP Cu anneal

Process Flow:

Etch (C4F8/CO/Ar/N2)

Ash (O2)

SiO2\SiN\NANOGLASS®E\SiO2

Trench litho

PVD Barrier SeedTaN/Ta/Cu

ECP Cu

Pre CMP Anneal200°C/60min

FIB

Voiding

CAP

14

Possible MechanismPossible Mechanism

Trench etch

Ash

Wet clean

Degas/Barrier dep/Cu seed dep

Cu electroplating (15000A)

Anneal (200C, 60 min)

MOST LIKELY SUSPECTS MECHANISM

Carbon depletion

May interact with Ash

No direct impact(Can affect Stress in annealed Cu)

Stress

Material is now structurally altered

TENSILE Stress pulls material apart.

Major factors in porous in-organic dielectric voiding:

C-depletion, Stress

No direct impact(Can affect Stress in annealed Cu)

15

OutlineOutline

• Background/Motivation

• Effect of Downstream Ash

• Effects of In-Situ RIE Ash

• Evaluation of New Ash Chemistries

• SLM (Single Layer Metal) Integration

• Conclusions

16

A New Ash Process A New Ash Process -- HH22/He/HeH2-He ash (300oC, 120s) – Process run by Axcelis Technologies

Chg% SiC/SiOPost CurePost etch -12.84%Post etch + ash -17.44%

-0.10

-0.08

-0.06

-0.04

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

Abso

rban

ce

1000 1500 2000 2500 3000 3500 Wavenumbers (cm-1)

Post Cure

Post Etch

Post RIE H2/He Ash

CH: 2978 cm-1

SiC: 1277 cm-1

SiO

• H2/He ash can effectively remove resist with little carbon depletion in porous Low-k materials

• Carbon depletion during ash may depend on the carbon depletion during etch. A modification of the etch recipe could help to further reduce carbon depletion during both etch and ash

17

DUV Resist Ashrate Calculated by Time to Clear

0100002000030000400005000060000700008000090000

240 270 300 330 360 390 420 450

Temperature [C]

Ash

Rat

e [A

/min

]

DUV Resist ash rateDUV Resist ash rate

18

Process Optimization of HProcess Optimization of H22/He Ash/He AshEffect of Temperature and Time

on SiC/SiO Ratio

05

101520253035404550

250C

60s

250C

180s

300C

120s

350C

60s

350C

180s

RIE N

2/H2 a

sh

RIE O

2 ash

Ch

g%

of

SiC

/SiO

Effect of Temperature and Time on Dielectric Constant k (Hg Probe)

2.20

2.30

2.40

2.50

2.60

2.70

2.80

2.90

Control

250C

60s

250C

180s

300C

120s

350C

60s

350C

180s

RIE N2/H

2 ash

RIE O2 a

sh

Process Condition

Die

lect

ric C

onst

ant k

• H2/He ash is less destructive than Std. Ash

• Selection of process conditions (ash temp/time) depends on resist removal rate and minimal damage to Low-k material

Process Condition

4

19

OutlineOutline

• Background/Motivation

• Effect of Downstream Ash

• Effects of In-Situ RIE Ash

• Evaluation of New Ash Chemistries

• SLM (Single Layer Metal) Integration

• Conclusions

20

SEM Images of SLM StructuresSEM Images of SLM Structures

Voids in porous Low-k materials can be significantly reduced by using H2/He plasma strip instead of in-situ RIE ash

Control ( RIE O2 ash)

H2/He plasma ash

Control ( RIE N2 / H2 ash)

Voiding

H2-He ash(350oC, 210s)

CAPNGE

Cu

Cu

Cu

CAPNGE

21

SLM Electrical Test DataSLM Electrical Test Data

300200100

99

95

90

80706050403020

10

5

1

Per

cent

pro

babi

lity Std NGE-Pre

Std NGE-Post

H2-He ash-Pre H2-He ash -Post

Comb capacitance (pF) L/S = 0.5/0.5

•Comb capacitance for O2 ash is significantly higher than H2/He ash before degas

•Post degas capacitance is similar for both. This indicates significantly lower moisture absorption in the case of H2/He ash

•RC product is similar for the two conditions

1000007500050000250000

99

95908070605040302010 5

1

Per

cent

pro

babi

lity

Snake resistance (ohm) L/S = 0.24/0.24

Std NGE-Pre Std NGE-Post

H2-He ash-Pre H2-He ash -Post

H2-He ashStd NGE

1.00.80.60.40.20.0

99

95908070605040302010 5

1

RC

Per

cent

pro

babi

lity

arbitrary units

Pre-Degas

Post Degas

22

SummarySummary

• Conventional down stream O2 ash results in both dimensional change (bowed profile) and compositional change (carbon depletion) in porous Si-C-OH dielectrics

• In-situ RIE ash does not affect etch profile, but results in significant carbon depletion(40-50%) which will lead to voiding in porous Si-C-OH dielectrics

• Modified H2/He ash process can minimize carbon depletion and therefore significantly reduce voiding in porous Low-k materials

• Electrical data indicates usage of H2/He ash results in improved dielectric performance (capacitance) to conventional ash processes

23

AcknowledgementsAcknowledgements

The authors would like to thank Orlando Escorcia and Dr. QingyuanHan from Axcelis Technologies for their assistance in completing this work.