1

Chemical Bonding Transformation Mapping to Optimize Low-k Dielectric Nanostructure Fabrication and Post-etch

Residue Clean

Oliver Chyan*, Sirish Rimal, Tamal Mukherjee,

Muthappan Asokan, Seare A. Berhe

Interfacial Electrochemistry and Materials Research Lab, University of North Texas

Present to SPCC-2017, *chyan@unt.edu

Kanwaljit Singh

Components Research, Intel Corporation

Ian Brown, Jacques Faguet

US-Technology Development Center, TEL Technology Center

2

Goal: Explore the underlying fundamental science that leads to real engineering solutions for important interfacial issues in

microelectronic fabrication.

Group members:

Alex Lambert, Issac Goutham, Muthappan Asokan, Seare Berhe

Alumni :

Dr. Sirish Rimal (JSR-Micro)

Dr. Tamal Mukherjee (LAM)

Dr. Nick Ross (Intel)

Dr. Kyle Yu (TEL)

Dr. Karthik Pillai (TEL)

Dr. Simon Koskey (Intel) 3

Interfacial Electrochemistry & Materials Research Lab

Front end (FEOL): Silicon surface preparation, H-passivation, SC1/SC2 cleaning, Silicon oxide etching, metal (Cu) and organic contamination control/detection.

"An Apparatus and Method for Detecting Impurities in Wet Chemicals " U.S. Patent 6145372, 2000.

Back end (BEOL): Cu ECD, Ru-based liner/barrier, Cu cleaning chemistry, Cu bimetallic corrosion mechanism, Post etch residue, Porous Low-k damage control.

"Method of Making Integrated Circuits Using Ruthenium and Its Oxides as Cu Diffusion Barrier", U.S. Patent 7247554, 2007.

IC Packaging: Cu vs. Au wire bonded device, Al pad corrosion prevention, molding compounds, Subtractive Cu Etching.

“Systems and Methods for Copper Etch Monitoring and Control", U.S. Patent Application, 2016.

“Method for Wafer Characterization", U.S. Patent 9366601, 2016.

(1992-2002)

(2000 - Date)

(2015 - Date)

Kanwal J. Singh, “Process and Environmental Challenges Facing BEOL Semiconductor Cleans Beyond the 32-nm node”, Electrochemical Society 220th Meeting, Oct, 2011

C. Weng, Materials Science in Semiconductor Processing, 13, 376 (2010)

BEOL Patterning & Damascene Processing

(Cu plating ): Cu+2

(aq) + 2e− → Cu (s)

Cu Cu

Low-k

dielectrics

CMP, PCMP

PR

Ru, TaN, Co,Mn

Micro-pattern Corrosion Screening Cu/Ru, Co/Cu

Co on Cu without inhibitor

Co on Cu with inhibitor

J. Electrochem. Soc., 2003, 150, 347. J. Electrochem. Solid-State Lett., 2004, 7, G154. J. Electrochem. Soc., 2005, 152, G808 Applied Physics Letter , 2005, 86, 083104 U.S. Patent, 7,247,554, 2007.

Invent Ru Barrier/liner

C. Weng, Materials Science in Semiconductor Processing, 13, 376 (2010) 8

BEOL Patterning & Damascene Processing

Etch/Clean Process

optimization

→MIR-IR Chemical Bonding Mapping

Attenuated total reflection (ATR) Infrared Spectroscopy (ATR-IR)

External ATR-IR

Not- reproducible • sample elasticity • flatness, position • fragile ILD damage Low sensitivity • miss max evanescent wave interaction

Patterned Si wafer

• No sample-contact problem

• Reproducible

• Max Evanescent Wave Interaction

• Superior Sensitivity (<10% monolayer)

Multiple Internal Reflection Infrared Spectroscopy (MIR-IR)

MIR-IR Advantages Patterned Si wafer = IR waveguide

“Method for Wafer Characterization", U.S. Patent 9366601, 2016. MIR-IR

External ATR-IR

Not- reproducible • sample elasticity • flatness, position • fragile ILD damage Low sensitivity • miss max evanescent wave interaction

11

MIR-IR Application: Ultra-clean Silicon Wafer Surface Preparation

MIR-R

O-SiHx

Surface Bonding Transformation on Si(100) by Trace Cu+2 impurity

12 MIR-IR

50 ppb Cu2+

J. Electrochem. Soc., 1996, 143, 92 J. Electrochem. Soc., 1996, 143, L235 J. Electrochem. Soc., 1997, 144, L17 U.S. Patent, 6,145,372, 2000

H

Cu-Si (100)

H H H H Cu2+

O

Trace Cu2+ induced early oxide and pitting

AFM

Cu2+

Optimize PECVD

Hydrogen Dilution,

BCl3 dosage,

Temperature, etc

𝒂𝒂-Si:H Microbolometer Pixel

IR Camera

MIR-IR : Optimize 𝒂𝒂-Si:H Microbolometer for Night Vision Camera

PECVD

Fabricate into IR waveguide

MIR-IR Characterization

Data Analyses L3 Wafer

PECVD Production of 𝒂𝒂-Si:H wafer.

SiNx

𝒂𝒂 -Si:H

Si

MIR

𝒂𝒂-Si:H

Comparing MIR vs Transmission (TIR) Sensitivity on 𝒂𝒂-Si:H

N. Ross, K. Shrestha, O. Chyan, C. L. Littler, V. C. Lopes, and A. J. Syllaios, Mater. Res. Soc. Symp. Proc., 1536, 127 (2013).

𝒂𝒂-Si:H Microbolometer Pixel

MIR-R

TIR

weak links

Si-H

MIR-IR : Optimizes 𝒂𝒂-Si:H Thin Film Quality

Bonding to Hydrogen is the key

Chemical Bonding Data Accelerates PECVD Process Optimization

MIR-IR

16

MIR-IR : Apply to BEOL Processing

Objective: Utilize MIR-IR to optimize RIE etching and post RIE cleaning with minimum dielectric damages

SEM and TEM • Electron beam based metrology • Cross sectional profiling • Destructive analysis, limited chemical bonding info • Laborious and time consuming XPS and TOF-SIMS • Evaluate depth of carbon depletion • Provides elemental composition • Based on x-ray, ion bombardment • Can’t reach deep-sidewalls and bottoms Spectroscopic Ellipsometry • For unpatterned films • Technique based on refractive index • Requires modeling

IR – traditional • Provide chemical bonding structure • Weaker signal and limited resolution 17

Trench bottom

XPS, TOF-SIMS, SEM

New MIR-IR Metrology

Current Metrology for Trench ILD Nano-structures

MIR-IR is a Powerful Substantive Too -- Isolate Low-k Film Stack only Spectra

Si sub

Porous Low-k (300nm) Dense Low-k (50nm)

Oxide (50nm)

Low-k Stack

Sample Spectrum

Background Spectrum

Si substrate

Vs.

( )

Low-k stack only Spectrum

18

Abso

rban

ce

Wavenumber cm-1

Si-OH

a-CH3

CH2

SiHx O-SiHx

Si-CH3 (bend)

Si-O cage Si-O

network

Si-CH3 (rock

Low-k Damage

Carbon Doping

Carbon Doping

Low-k

Low-k damage

Residue removal

Chyan, et al, ECS Solid State Lett., 3, N1 (2014)

Chemical Bonding Transformation Map for Porous Low-k NanoStructure

MIR-IR: Evaluating Plasma-induced Damages on Trench ILD

Chyan, et al, ECS Solid State Lett., 3, N1 (2014)

Post Strip

Over-Etch

Si-CH3

MIR-IR

Post Etch

Pristine Confirmed

by DHF etch

Post-Strip (O2 plasma)

more ILD damages

2400 2600 2800 3000 3200 3400 3600 3800 4000

Wavenumber (cm-1)

0.02

0.

04

0.06

0.

08

0.10

0.

12

0.14

Abso

rban

ce

Strip 1 Strip 2 Strip 3 Strip 4

a-CH3

Si-OH

MIR-IR: Minimize Low-k Damage of Manufacture Strip Processes

Four oxidative plasma strip processes screened by either adjusting process gases or by modulating chamber pressure for reduced O radical content

Strip process 1 induces minimal damage (lowest silanol) retaining maximal C-doping (highest CH3)

Strip 1 Strip 2 Strip 3 Strip 4

110

115

120

125

130

Peak

Hei

ghts

(mab

s)

Si-OH

Strip 1 Lowest Water Sensitivity

Highest Carbon Dopping

Abso

rban

ce

Wavenumber cm-1

Si-OH

a-CH3

CH2

SiHx O-SiHx

Si-CH3 (bend)

Si-O cage Si-O

network

Si-CH3 (rock

Low-k Damage

Carbon Doping

Carbon Doping

FC=CF

C=CF

C=C

CH3, CH2, CH (def.)

Etch residue Low-k

Low-k damage

Residue removal

Chemical Bonding Transformation Map for Porous Low-k NanoStructure

MIR-IR: Evaluating Plasma Damages and Etch Residue Removal

Pristine

Post Etch

Post Strip

60 sec over-etch

Post-Strip more

damages

Evaluate Post-etch Residue Removal Efficiency

Chyan, et al, ECS Solid State Lett., 3, N1 (2014)

60 sec over-etch

Post Strip

Post Etch

1X POLYMER 270 Pitch

Conformal Polymer Coating

1X POLYMER 180 Pitch

5X POLYMER 270 Pitch

Kanwaljit Singh Intel

Alan M Myers Intel

24

MIR-IR : Identify Chemical Bonding Structure for Post-Etch Residues

XPS Data: Difficult to Assign Chemical Bonding

542 540 538 536 534 532 530 528500

1000

1500

2000

2500

3000

3500

C=O533.7 eV

Coun

ts

Binding Energy (eV)

No Clean 1 Min 2 Min

300 296 292 288 284 280 276

0

500

1000

1500

2000

2500

3000CF3293.5 eV

CF2

291.1 eVCF/C=O288.8 eV C-CF/ C-O

286.8C-C, C-H284.8 eV

Coun

ts

Binding Energy (eV)

C 1s Curve Fitting

O 1s

302 300 298 296 294 292 290 288 286 284 282 280 278 276

0

500

1000

1500

2000

2500

3000

3500

CF3

CF2

CF/C=OC-CF/C-O

C-C/C-H

Coun

ts

Binding Energy (eV)

ECS Solid State Lett., 2, N11 (2013)

XPS: Elemental Info

(C, F, O, H)

C-F, C-H, C-O, C-C

But, no reliable chemical bonding

structure info

2000 1600 1200 800

Abso

rban

ce (a

.u.)

Wavenumber (cm-1)

1x 5x

. (CF2)

def (am)

. . (CF2)s

(CF2)as

. CF3i . . . .

C=CF F2C=CF C=O

FC=CF

. CF3

t

• Mainly fluoropolymer backbone

• Significant branching/cross-linking

• Olefinic unsaturation (fluorinated)

• Carbonyl functional groups

Polytetrafluoroethylene (Teflon)

6nm 28nm

Chemical Bonding Structure of Etch Residues

ECS Solid State Lett., 2, N11 (2013)

Teflon

FC

CF2

FC

CF

CF2CF2

CFC

F2F2C

CF2

FCF2C

CF

F2C CF2

FC

F2C CF2

CF

C

F2C

F2C

CCF2

O

H3C

C

F2C CF2 FC CF2

CF2

FC

CF2

CF2

F2C

n

n

CF2

FC F

C

F2CFC

F2C

F2C CF2

CFCF2

CF2

n

COO

n

n

n

n

nn

CF3F2C

CF2

F2C

CF3

F2C

n

F2Cn

Identify Chemical Bonding Structure by Functional Group Specific Surface Derivatization

Etch Polymer Residue

C

O NHH2N

NO2

NO2

NHN

NO2

NO2

C

H2O

Brady's test

Carbonyl Group

1710

C

O NHH2N

NO2

NO2

NHN

NO2

NO2

C

H2OFC

CF2

FC

CF

CF2CF2

CFC

F2F2C

CF2

FCF2C

CF

F2C CF2

FC

F2C CF2

CF

C

F2C

F2C

CCF2

O

H3C

C

F2C CF2 FC CF2

CF2

FC

CF2

CF2

F2C

n

n

CF2

FC F

C

F2CFC

F2C

F2C CF2

CFCF2

CF2

n

COO

n

n

n

n

nn

CF3F2C

CF2

F2C

CF3

F2C

n

F2Cn

1500 2200 1600 1700 2100 1800 1900 2000 cm-1

AU

-0.015

-0.010

-0.005

0.000

0.005 before

after

differential

1619

Confirmation of C=O by DNPH Reaction

DNPH 2,4-Dinitrophenyl Hydrazine

C=O C=N

≈ - 2 mabs Etch Polymer Residue

Hydrazine Hydrazone

29

2 x 10-3 mabs

Confirm all 24 spectroscopic features of Hydrazone modified Post-Etch Residues.

29

FC

CF2

FC

CF

CF2CF2

CFC

F2F2C

CF2

FCF2C

CF

F2C CF2

FC

F2C CF2

CF

C

F2C

F2C

CCF2

O

H3C

C

F2C CF2 FC CF2

CF2

FC

CF2

CF2

F2C

n

n

CF2

FC F

C

F2CFC

F2C

F2C CF2

CFCF2

CF2

n

COO

n

n

n

n

nn

CF3F2C

CF2

F2C

CF3

F2C

n

F2Cn

1500 1000

4000 3500 2000 1500

0.60

0.70

0.90

0.80

3500 3000 2500 2000 1500 1000

FC

CF2

FC

CF

CF2CF2

CFC

F2F2C

CF2

FCF2C

CF

F2C CF2

FC

F2C CF2

CF

C

F2C

F2C

CCF2

O

H3C

C

F2C CF2 FC CF2

CF2

FC

CF2

CF2

F2C

n

n

CF2

FC F

C

F2CFC

F2C

F2C CF2

CFCF2

CF2

n

COO

n

n

n

n

nn

CF3F2C

CF2

F2C

CF3

F2C

n

F2Cn

C C

F

FBr

2C C

F

F

Br

C C

F

Br

Br

F

C C

F

Br

OH

F

H2O

Br

Br2

Before

After

Differential

After

Identification of C=C by Br2 Reaction

-OH

Before

After

Differential -C=C- F-C-Br

Before

After

Differential

Mukherjee et. al, ECS Solid State Lett., 2, N11 (2013) 30

+ Na + NaEther

R F : Na + NaF + R

R FC CF2 n=_

Reductive Extraction of F

Naphthalenide Anion Radicals attack C-F bonds

FC CF2 n+

Na/Naphthalene in etherFC CF n

FC CFn

NaF

FC CF2 n

FC CF2 n

dimerization

C CFn

Na/Naphthalenein ether NaF

dimerization

C CFn

C CFn

so on

Generate reactive radicals on Fluoro-polymer

FC

CF2

FC

CF

CF2CF2

CFC

F2F2C

CF2

FCF2C

CF

F2C CF2

FC

F2C CF2

CF

C

F2C

F2C

CCF2

O

H3C

C

F2C CF2 FC CF2

CF2

FC

CF2

CF2

F2C

n

n

CF2

FC F

C

F2CFC

F2C

F2C CF2

CFCF2

CF2

n

COO

n

n

n

n

nn

CF3F2C

CF2

F2C

CF3

F2C

n

F2Cn

Reductive Defluorination

31

Sodium Naphthalenide

Identify Fluoropolymer Backbone By Reductive Defluorination on C-F bond

Mukherjee et. al, ECS Solid State Lett., 2, N11 (2013)

32

MIR-IR : UV-assisted Cleaning on Post-etch Residues

FC

CF2

FC

CF

CF2CF2

CFC

F2F2C

CF2

FCF2C

CF

F2C CF2

FC

F2C CF2

CF

C

F2C

F2C

CCF2

O

H3C

C

F2C CF2 FC CF2

CF2

FC

CF2

CF2

F2C

n

n

CF2

FC F

C

F2CFC

F2C

F2C CF2

CFCF2

CF2

n

COO

n

n

n

n

nn

CF3F2C

CF2

F2C

CF3

F2C

n

F2Cn

UV

UV

Kanwaljit Singh Intel

Ian Brown TEL

Jacques Faguet TEL

UV

10sec UV

10sec UV + Wet Clean

No UV

Ian Brown

SEM Images of UV & UV+Wet Clean Samples

#265_no Ash

Deposit 5X Polymer

No Wet Clean

UV induced Polymer Residue Removal

Complete removed by Wet Clean

Hydrophobic Hydrophilic

Kanwal Jit Singh

10sec UV

10sec UV + Wet Clean

No UV

Ian Brown

SEM Images of UV & UV+Wet Clean Samples

#265_no Ash

Deposit 5X Polymer

No Wet Clean

UV induced Polymer Residue Removal

Complete removed by Wet Clean

Hydrophobic Hydrophilic

Kanwal Jit Singh

36

C=O Chromophore Required for UV-assisted Clean

800 1000 1200 1400 1600 1800 Wavenumber (cm-1)

as-deposited

300s UV

CF2 (as)

C=O/ CFx=CFy

800 1000 1200 1400 1600 1800

CF2 (as)

Nafion

as- spin-coated

300s UV

Abso

rban

ce (A

.U.)

Wavenumber (cm-1)

With -C−O, no UV clean Need -C=O, for UV clean

Etch Residue

O2 is required

Fluorocarbon polymer backbone -(CF2)n- steadily decrease as UV treatment time increases Gradual reduction of CFx=CFx band

Fluoro-carbonyl FxC=O forms by

photo-cleavage of CFx=CFx bonds

Chemical Bonding Transformation Mapping - UV Radiation

FC

CF2

FC

CF

CF2CF2

CFC

F2F2C

CF2

FCF2C

CF

F2C CF2

FC

F2C CF2

CF

C

F2C

F2C

CCF2

O

H3C

C

F2C CF2 FC CF2

CF2

FC

CF2

CF2

F2C

n

n

CF2

FC F

C

F2CFC

F2C

F2C CF2

CFCF2

CF2

n

COO

n

n

n

n

nn

CF3F2C

CF2

F2C

CF3

F2C

n

F2Cn

UVC

CFx=CFx

Provide Critical Process Evolution Info for Engineering Optimization

Summary of Chemical Bonding Transformation with UV Radiation

FC

CF2

FC

CF

CF2CF2

CFC

F2F2C

CF2

FCF2C

CF

F2C CF2

FC

F2C CF2

CF

C

F2C

F2C

CCF2

O

H3C

C

F2C CF2 FC CF2

CF2

FC

CF2

CF2

F2C

n

n

CF2

FC F

C

F2CFC

F2C

F2C CF2

CFCF2

CF2

n

COO

n

n

n

n

nn

CF3F2C

CF2

F2C

CF3

F2C

n

F2Cn

UVC

1O2

O2

1O2 1O2

[2+2] Cycloaddition

CF

CF

O O

CF

CF

O OCF

CF

Post-Etch Residues

2400 2800 3200 3600 4000 Wavenumber (cm-1)

Abso

rban

ce U

nits

60 sec UV/air

C-O-H str.

OH OH OH

UV/air

C OR2

R1C

R2

R1 OH

OHH2O+

as-deposited

R1 R2 Keq = [Hydrate]/[Carbonyl]

CH3 CH3 1.4 x 10-3

CF3 CH3 35 CF3 CF3 1.2 x 106

~103 times faster ~109 times

faster

Ref: Guthrie, J. P. Carbonyl Addition Reactions: Factors Affecting the Hydrate-Hemiacetal and Hemiacetal-Acetal Equilibrium Constants. Can. J. Chem. 1975, 53, 898-906

IR Spectroscopic Evidence for Hydroxyl Formation during UV/Air Treatment

Hydrophobic

Hydrophilic

MIR-IR

Chemical Bonding Transformation Insights – UV Clean Optimization More Efficient

Chemical Bonding Structure ACS Applied Materials and Interfaces, 2015

Effective Post-etch cleaning with minimum dielectric damages

Can Not Wet Clean

Clean !

單線態氧 Cleaning mechanism

BEOL Integration Challenges

BEOL Etch and Cleans Process Integration

ULK Materials Engineering • Low Keff • UV cured PECVD • HF resistant

Reactive Ion Etching • CD control • Clean friendly • Damage control

Post Etch Cleans

• Less damage • Fast clean • Keep CD, Keff

Physical

Electrical

Chemical

Metrology Tools

Chemical Bond Formation & Bond Breaking - Highly Selective in time & space !

Chemical bonding transformation map

PI: Oliver Chyan*, Chyan@unt.edu

Costly

Not complete

7 nm finFETs

MIR-IR Applications in Advanced IC Fabrication

FEOL (Ge, SiGe etc) , BEOL etching and cleaning, formulation

Monitor Post-etch residue removal & minimize Low-k damages

Optimize Plasma Etch/Strip/Clean Process integration

Monitor Low-k damages and optimize restoration process

UV curing on porous Low-k dielectrics materials

Evaluate TiN hard mask for Low-k pattern fabrication

Flowable low-k dielectrics for gap filling in nanostructure

Atomic Layer Deposition/Etching : provide critical interfacial

chemical bonding info for better atomic layer control.

PI: Oliver Chyan*, Chyan@unt.edu

Goal: Achieve better understanding of fundamental materials properties at the critical interfaces of practical applications.

Group members:

Alex Lambert, Issac Goutham, Muthappan Asokan, Seare Berhe

Alumni :

Dr. Sirish Rimal (JSR-Micro)

Dr. Tamal Mukherjee (LAM)

Dr. Nick Ross (Intel)

Dr. Kyle Yu (TEL)

Dr. Karthik Pillai (TEL)

Dr. Simon Koskey (Intel)

43

END

44