finfet doping : fabrication and metrology challenges · finfet doping : fabrication and metrology...

FINFET doping : fabrication and metrology challenges

W.Vandervorst

Imec, Belgium

Also : Inst. Kern- en stralingsfysika, KULeuven

Surname + NameIMEC confidential 2009 2

Why FINFET?

Chem. oxide

HfS

iO+

TiN

c-Si10 nm

Chem. oxide

HfS

iO+

TiN

c-Si10 nm)

c)

-0.5-0.4-0.3-0.2-0.1

00.10.20.30.40.5

20 40 60 80 100Lg (nm)

V T_s

at (V

)

Wfin = 17 nmWfin = 12 nmWfin = 8 nmWfin = 6 nm

pMOS

nMOS

Ioff

Ion

Gate

S D

Ioff

Ion

Gate

S D

Ioff

IonIoff

Ion

Gate

S D

Gate

S D

Ioff

Ion

Ioff

Ion

Gate

S DGate

S DTrigate

(Intel)

J. Kavalieros

et al., Symp. VLSI Tech. Dig. 2006.

FinFET (IBM, AMD, Freescale, NXP)

Presented at Frontie

rs on metro

logy for n

ano-electronics


Conformal junctions in FINFET’s : 3D-profiling

buried oxide

Gate

uniform

Gate

Conformal(plasma, CVD?)

chan

nel

Gate

conformaltop rich

(tilted implants)

chan

nel

Gate

Top only(implant 0°)

chan

nel

Gate

uniform

Gate

uniform

Gate


chan

nel

Gate


chan

nel

Gate

conformaltop rich

(tilted implants)

chan

nel

Gate

conformaltop rich

(tilted implants)

chan

nel

Gate


chan

nel

Gate


chan

nel

Hardmask

0.0 0.2 0.4 0.6 0.8 1.01E-10

1E-9

1E-8

1E-7

1E-6

1E-5

1E-4

1E-3

I (dr

ain)

(Am

ps)

V(gate) (Volts)

Conformal Top Only

B

B’Tilted top view

Gate

Fin

Tilted top view

Gate

Fin

40nm gate16nm

fin

60 nm60 nm60 nm

Si

NiSi

NiSi NiSi

High-k + MG

poly-Si

Buried oxide

20 nm

S/D

Gate


rs on metro

logy for n

ano-electronics


S/D Junction formation requirements

• Junction conformality/uniformityGate

Source Drain

extensio

ns

extensio

ns

spacer spacerTop:high dose

Side:low dose

currentcurrent

SCE controlSCE control

B

Silicon BOX

Gate+ HMSpacer

Silicide

B’

B

Silicon BOX

Gate+ HMSpacer

Silicide

B’

FIN resistance : high doping, overlap

0° 10° 45°0° 10° 45°

Ionn

Ioff

Increased sidewall doping


rs on metro

logy for n

ano-electronics


0

0.2

0.4

0.6

0.8

1

1.2

0 10 20 30 40 50 60 70 80 90

implant angle

rela

tive

dose

As 5 kev

P 3 keV

B 1 keV

500 eV B

200 eV B

(non-conformal) Doping by I/I :Tilt angle effects

Θ°backscattering and sputtering (SRIM 2003)

side

top

Implanted dose reduction

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10 20 30 40 50 60 70 80 90Implant tilt angle (degrees)

Frac

tion

of th

e re

tain

ed d

ose

Arsenic 5keVBoron 1keVPhosphorus 3keVAs dose tilted:B dose tilted:P dose tilted:

tilt-relateddose reduction

Taperededge

Irregularedge

Reflection at surface

Back-

scatteri

ng

Target sputtering

BOX

Tilt angle

top

side

i.c.w. R.Duffy, B.J.Pawlak, NXP

Effect of wafer rotation still to be added (:2)

Buried Oxide

SiSiSiSiSi

Resist

NMOSNMOS PMOSPMOS

10-20°

70-80°to side

Buried Oxide

SiSiSiSiSi

Resist

NMOSNMOS PMOSPMOS

10-20°

70-80°to side

1E+18

1E+19

1E+20

1E+21

1E+22

0 10 20 30 40Depth (nm)

Con

cent

ratio

n (a

t/cm

3)

5°

30°

60°

80°

As 1e15 5keV variable tiltImplanted


rs on metro

logy for n

ano-electronics


Junction parameters : Planar vs FINFET

•Rs

•Xj

•(Vertical)

•Lateral

•Steepness

•Lateral

•Conformality

•Metrology

•2D

•3D

•Rs

•Xj

•Vertical

•Lateral

•Steepness

•Lateral

•Metrology

•1D (Rs, SIMS)

•2D (SSRM)

PLANAR FINFET

buried oxide


rs on metro

logy for n

ano-electronics


FINFET’s conformal doping and its metrology

Conformal doping• Implantation

– Tilt angle and incorporation efficiency

– Shadowing in dense structures (< 10-20° tilt)

– Amorphization and recrystallization

• Plasma immersion– Conformality ??

– Incorporation versus erosion

• VPD– Integration

– Outdiffusion

Properties Rs vs Xj has no meaning! Lateral profiles, sidewall dose, conformality

Metrology•SIMS through FINs•Resistors•S/D area’s : X-SSRM•3D-SSRM•3D-Atomprobe


rs on metro

logy for n

ano-electronics


SIMS through FINs

SIMS conc.

Dep

th

Poly-Si

Si

Si Si Si Si Si Si

Poly-Si

Si

Si Si Si Si Si Si

total SIMS craterN Concentration H −= ×Δtotal SIMS crater SIMS crater SIMS craterN Concentration H B W− − −= ×Δ × ×

total fin SIMS crater SIMS craterN No. of fins in SIMS crater D H B− −= × ×Δ ×

fin finD Concentration P= ×Presented at Frontie

rs on metro

logy for n

ano-electronics


SIMS results

Sidewall Dose retention

Implants at 5keV 45° tilt 10° tiltSidewall dose

retention Ratio(45° vs. 10 °)

Arsenic 1 x 1015 /cm2 1.06 x 1015 /cm2 2.20 x 1014 /cm2 4.83

BF2 8 x 1014 /cm2 5.74 x 1014 /cm2 1.38 x 1015 /cm2 4.18

• BF2 5keV 8e14 at 45° and 10°• As 5keV 1e15 at 45° and 10°

• RTA annealed at 1050°C

Array of 10 fins

Deposited Si

200nm

80nm 200nm

Array of 10 fins

Deposited Si

200nm

80nm 200nm


rs on metro

logy for n

ano-electronics


Sidewall doping by I/I :SIMS vs theory

W.Vandervorst et al. , J. Vac. Sci. Technol. B 26 (1), Jan/Feb 2008, 396-401

Includes effect of 2 quad. implant


rs on metro

logy for n

ano-electronics


Scanning Spreading Resistance Microscopy : 2D-profiling with sub-nm resolution

SSRM

R=ρ/4a

Conducting probe

bias(50-500 mV)

to ADC

(10 -10-12 -4Α)

Log Amp

metal

diamond

metal

diamond

Back contact

R=ρ/4a

Conducting probe

bias(50-500 mV)

to ADC

(10 -10-12 -4Α)

Log Amp

bias(50-500 mV)

bias(50-500 mV)

to ADC

(10 -10-12 -4Α)(10 -10-12 -4Α)-12 -4Α)

Log Amp

metal

diamond

metal

diamond

Back contactBack contact

1014 -1021

1015 -1021

1014 -1021

1014 -1021

Dynamic range (at/cm3)

2%3-5%2%2%Concentration precision (%)

10.5 - 122Lateral/depth resolution (nm)

1.61-1.52.74.25Lateral/vertical steepness (nm/dec)

ITRS 2016

SSRMITRS 2010

ITRS 2005

1014 -1021

1015 -1021

1014 -1021

1014 -1021

Dynamic range (at/cm3)

2%3-5%2%2%Concentration precision (%)

10.5 - 122Lateral/depth resolution (nm)

1.61-1.52.74.25Lateral/vertical steepness (nm/dec)

ITRS 2016

SSRMITRS 2010

ITRS 2005

P.Eyben, W.Vandervorst, D.Alvarez,, M.Xu. and M.Fouchier, in “Scanning Probe Microscopy, Electrical and Electromechanical Phenomena at the Nanoscale”, edited by S. Kalinin and A. Gruverman (Springer, New York, 2007), Vol. 1, Chapter 2, p. 31

96nm

8nm72nm

30nm16nm

96nm

8nm72nm

30nm16nm

96nm

8nm72nm

30nm16nmsilicide

source

silicide

drain

silicide

gateextension

halo

We can see it, can you make it?


rs on metro

logy for n

ano-electronics


SSRM (FIN-like) applicationsΩ

ONO layer

Si FinPoly

1010 Ω

1012 Ω

Inversion layer, corner effect

1010 Ω

1012 Ω

1010 Ω

1012 Ω

1010 Ω

1012 Ω

Inversion layer, corner effect

M.Specht, D.Alvarez

S.Kluth et al.


rs on metro

logy for n

ano-electronics


SSRM on FIN : As I/I

Arsenic ImplantImplant

angleTop Sidewall Theoretical

45° 28nm 24nm 28nm

10° 35nm 18nm 6nm

( )sidewall top surfaceDepth tan Depthα −= ×Array of 10 fins

Deposited Si

200nm

80nm 200nm

Array of 10 fins

Deposited Si

200nm

80nm 200nm


rs on metro

logy for n

ano-electronics

J.Mody et al, Insight-2009


Additional information using SSRM

• Ability to obtain 1D carrier profile with sub-nm resolution.

• Resolve the center of the fin which cannot be resolved using SIMS.Presented at Frontie

rs on metro

logy for n

ano-electronics


SSRM on FIN : BF2 I/I

BF2 Implant

Implantangle

Top Sidewall Theoretical

45° 35nm 31nm 35nm

10° 36nm 15nm 7nm

( )sidewall top surfaceDepth tan Depthα −= ×


rs on metro

logy for n

ano-electronics


Comparison SIMS-SSRM

ImplantSIMS Sidewall Ratio

(45° vs. 10°)SSRM Sidewall Ratio

(45° vs. 10°)

Arsenic 4.83 3.42

ImplantSIMS/SSRM Sidewall

Ratio 10°

SIMS/SSRM Sidewall Ratio 45°

Arsenic 0.95 1.33Presented at F

rontiers on m

etrology fo

r nano-electro

nics


Summary of sidewall doping : As

Single sidewall dose Implant Dose sin( ) cos( ) incorporation efficiency ( )α α α= × ×

J. Vac. Sci. Technol. B 26 (1), Jan/Feb 2008, 396-401

fin finD Concentration P( ) / 2= ×


rs on metro

logy for n

ano-electronics


Comparison SIMS-SSRM

ImplantSIMS Sidewall Ratio

(45° vs. 10°)SSRM Sidewall Ratio

(45° vs. 10°)

BF2 4.18 9.79

ImplantSIMS/SSRM Sidewall

Ratio 10°

SIMS/SSRM Sidewall Ratio 45°

BF2 1.52 0.63Presented at F

rontiers on m

etrology fo

r nano-electro

nics


Summary of sidewall doping : BF2

Single sidewall dose Implant Dose sin( ) cos( ) incorporation efficiency ( )α α α= × ×

• Good correlation between theoretical model, SIMS and SSRM

J. Vac. Sci. Technol. B 26 (1), Jan/Feb 2008, 396-401

fin finD Concentration P( ) / 2= ×


rs on metro

logy for n

ano-electronics


1D-profiles (and 1D-tuning) are meaningless

1.E+18

1.E+19

1.E+20

1.E+21

1.E+22

1.E+23

0 10 20 30 40 50 60 70 80 90

depth (nm)

Con

cent

ratio

n(a

t/cm

3

1.2K/25 sec

1.7 K/25 sec

1.7 K/200W/25 sec

1.7K/150/50(Ar). 25 sec eq

1.7K/50/150 (Ar),25 sec eq

1.7K/5.8/100 sec

elec. Active

Dose(atm/cm2)

2.47E152.52E152.51E154.98E15

Rs(Ω/cm2)

457457

-446

Similar 1D-profiles give very different conformality!!!

d07

-b2h

6


rs on metro

logy for n

ano-electronics


Resistor methodology

Hard mask protecting top of FIN

Implant or Plasma

BOX

FIN

HM

Pseudo VDP- or 2PP measurement

Poly-Si Fin,high diffusion!

RFin ~ 1/N

Full silicidation of unprotected area.

W.Vandervorst et al. JVST (2008)

BF3 P3I

W03 W04 W05 W06 W07 W08 W09 W10

200

400

600

800

1000

1200

2_po

int_

resi

stan

ce/R

2vf

wafer

BF3 P3I

W03 W04 W05 W06 W07 W08 W09 W10

200

400

600

800

1000

1200

2_po

int_

resi

stan

ce/R

2vf

wafer


rs on metro

logy for n

ano-electronics


Metrology for conformality

• SIMS through FIN’s : – Dopants,not active carriers

– No details on lateral junctions depths

– No wafer mapping

• Scanning Spreading Resistance Microcopy (SSRM) on cross sections of S/D fin’s

– Active carriers, real Xlat numbers

– No wafer mapping

• Resistors (R ~ 1/Sidewall dose)

– Relative

– Wafer mapping

P.Eyben, Vandervorst W. et al. . “Scanning Probe Microscopy : Electrical and Electromechanical Phenomena at the Nanoscale. Chapter II:SSRM, pp.31-87 (2007) (Springer).

R.Duffy et al., MRS -2008, J.Mody, Insight- 2009

W.Vandervorst et al., Proc. INSIGHT-2007, JVST B (2008)


rs on metro

logy for n

ano-electronics


plasma doping :Concurrent doping and erosion.

1.E+02

1.E+03

1.E+04

1.E+05

1.E+06

1.E+07

1.E+08

0 50 100 150 200norm. BF3-dose

RFI

N

150/50

50/150

previous, 100/50, 100/100

100 Bf3

HM after RTA< Ar-dilution

W.Vandervorst, IIT-2008

6 .0x10 -7

7 .0x10 -7

8 .0x10 -7

9 .0x10 -7

4 .0x10 -7 5 .0x10 -7 6 .0x10 -7 7 .0x10 -7

SSRM : Conformal!

1.00E+03

1.00E+04

1.00E+05

1.00E+06

1.00E+07

0 50 100 150 200

Depth (nm)Sp

read

ing

Res

ista

nce

(ohm

s)

D06 vert inD06 vert out

D06 lat in

1.E+02

1.E+03

1.E+04

1.E+05

0 0.5 1 1.5 2Vbias (KV)

RFI

N (O

hm)

0

1

10

100

HM

-ero

sion

(nm

)

source 200 Wsource 1000 WBF3 + B2H6no HMErosionErosion 1000W

lack of conformality

1.E+02

1.E+03

1.E+04

1.E+05

0 0.5 1 1.5 2Vbias (KV)

RFI

N (O

hm)

0

1

10

100

HM

-ero

sion

(nm

)

source 200 Wsource 1000 WBF3 + B2H6no HMErosionErosion 1000W

lack of conformality


rs on metro

logy for n

ano-electronics


3D-metrology

•Carriers : 3D-SSRM : Slice and view

•DopantsTomographic atomprobe


rs on metro

logy for n

ano-electronics


3D-profile in FinFET : SSRM slice and view

j.mody, mrs 2008S/D

Gate


rs on metro

logy for n

ano-electronics


Practical Problems

G

B

B’

A

A’

C

C’

G

B

B’

A

A’

C

C’

• To obtain 3D-profile we must obtain successive 2D-spreading resistance maps in one of the planes with 1nm step.

• Polishing ??? • Cleaving ???

• Cleaving and polishing with nanometer step in the planes ???

• Successive cleaving with 1nm step on the same transistor ???• Successive polishing with 1nm step on the same transistor ???


rs on metro

logy for n

ano-electronics


3D SSRM : slice and view

1 nm


rs on metro

logy for n

ano-electronics


3D-SSRM :proof of concept

100 150 200 250 300 350 400 450 500

105

106

Error bar = Average Reproducibility = 18%

Res

ista

nce

(Ohm

s)

Length (nm)

Average resistance (Fin 1 - Fin 8)

FinSubstrate

J.Mody et al., MRS-Spring 2008

0 100 200 300 4000

20

40

60

80

100

120

140

Width (nm)

Hei

ght (

nm)

100006.105E43.728E52.276E61.389E78.483E75.179E83.162E91.931E101.000E11


rs on metro

logy for n

ano-electronics


Fs Laser pulse

Vdc

Tof 2

Tof 1X1, Y1

X2, Y2

3D reconstruction

1⎟⎠⎞

nM

2⎟⎠⎞

nM

Chemical identification

Fs Laser pulse

Vdc

Tof 2

Tof 1X1, Y1

X2, Y2

3D reconstruction

1⎟⎠⎞

nM

2⎟⎠⎞

nM

Chemical identification

Tomographic Atomprobe


rs on metro

logy for n

ano-electronics


Atomic resolution with the Atomprobe: analysis of doped Si

“homogeneously” B-doped Si

3.1 Å

Si-atoms/ lattice planes in Si<111>Estimated depth resolution <0.2 nm Presented at F

rontiers on m

etrology fo

r nano-electro

nics

W.Vandervorst et al, AIP 2007


Lateral As-profile

• HRTEMK.Jones, IIT-2006, Univ. Florida


rs on metro

logy for n

ano-electronics


20 nm

20 nm

120 nm

3D-dopant profiling counting atoms

1 cts ~ 5 e19 at/cm3


rs on metro

logy for n

ano-electronics


FINFET underdiffusion : Counting single atoms with the atomprobe…

Gate

10 x 1 x1 = 500 Si atoms5e19 ~ 0.5 cts/pixel

Gate

Fin

1-2 nm

10 nm

60 nm

•Gradients : 1nm/dec

5e20 ~ 5 cts/pixel

5e19 ~ 0.5 cts/pixel

•Registration to gate (LER)Presented at F

rontiers on m

etrology fo

r nano-electro

nics


Conclusions

• FINdoping fabrication and metrology is a major challenge– I/I ( ) , plasma doping ( ), VPD ()

• Metrology– conformality :

• SIMS through FINs• Resistors• S/D area’s : X-SSRM

– 3D-profiles

• Pseudo 3D-SSRM : dedicated test structures• 3D-Atomprobe : statistics!!!


rs on metro

logy for n

ano-electronics


Acknowledgements

Device fabricationHighly automated volume

nm-scale characterizationSkillfull experts

•The art of many student-experts :

M.Meuris, P.De Wolf, D.Alvarez, T.Hantschel, T.Trenkler, M.Fouchier, N.Duhayon, W.Polspoel, J.Mody, T.Janssens, S.Koelling,M.Gilbert, H.Bender, O.Richard,

•Collaboration Cameca (LAWATAP)

finfet doping : fabrication and metrology challenges · finfet doping : fabrication and metrology...

Documents