Improved Regrowth of SelfImproved Regrowth of Self-Aligned ... · Improved Regrowth of SelfImproved Regrowth of Self-Aligned ... A.C. Gossard P. McIntyre, B. Shin, ... NAMBE 2009

North American Molecular Beam Epitaxy Conference (NAMBE),8-11-2009

Improved Regrowth of Self-AlignedImproved Regrowth of Self Aligned Ohmic Contacts for III-V FETs

Mark A. Wisteyy

Now at University of Notre [email protected]

A.K. Baraskar, U. Singisetti, G.J. Burek, M.J.W. Rodwell, A.C. Gossard

P. McIntyre, B. Shin, E. KimStanford University

[email protected]

M.J.W. Rodwell, A.C. GossardUniversity of California Santa Barbara

Stanford University

Funding: SRC

Outline: Regrown III-V FET Contacts

•Motivation for Self-Aligned Regrowthg

•Facets, Gaps, Arsenic Flux and MEEMEE

•MOSFET Results•Conclusion

2Wistey, NAMBE 2009

Motivation for Regrowth: Scalable III-V FETs

Classic III-V FET (details vary):

GateGapSource Drain

Large Area Contacts

• Advantages

ChannelBottom BarrierInAlAs Barrier

Top Barrier or Oxide

Low

Large Rc

Implant: straggle

gof III-V’s

• Disadvantages of III-V’s

InAlAs BarrierLow doping

III-V FET with Self-Aligned Regrowth:

Implant: straggle,short channel effects

Hi h b i

Gate

High Velocity ChannelSmall Raccess

Small RcSelf-aligned, no gaps

High barrier

ChannelIn(Ga)P Etch Stop

Bottom Barrier

High-kn+ Regrowth

Small RcHigh mobility

access regionsUltrathin 5nm doping layer

3Wistey, NAMBE 2009

High doping: 1013 cm-2

avoids source exhaustion 2D injection avoids source starvation

Dopants active as-grown

MBE Regrowth: Bad at any Temperature?200nm Gap

• Low growth temperature (<400°C):–Smooth in far field GateGate

p

Smooth in far field–Gap near gate (“shadowing”)–No contact to channel (bad) Source-Drain

RegrowthSource-Drain

RegrowthRegrowthRegrowth

Metals

SiO2

Metals

Channelhigh-k GateGate

• High growth temperature (>490°C):

Source-DrainRegrowth

Source-DrainRegrowth

–Selective/preferential epi on InGaAs–No gaps near gate

–Rough far field

4Wistey, NAMBE 2009 SEMs: Uttam Singisetti

RegrowthRegrowth

Regrowth: 50nm InGaAs:Si, 5nm InAs:Si. Si=8E19/cm3, 20nm Mo, V/III=35, 0.5 µm/hr.

Rough far field –High resistance

490C460C

High Temperature MEE: Smooth & No Gaps

490C460C

Gap

560C540C 560C

SiO2 dummy gate

540C Smooth regrowth

SiO2 dummy gate

gateNo gaps, but faceting next

to gates

5Wistey, NAMBE 2009 Note faceting: surface kinetics, not shadowing.In=9.7E-8, Ga=5.1E-8 Torr

Shadowing and Facet CompetitionShen & Nishinaga, JCG 1995

Fast surface diffusion = slow facet

growth

Slow diffusion = rapid facet growth

O2

[100]

growth

SiO

Fast surface

• Shen JCG 1995 says:Increased As favors [111] growth

[100]

Slow diffusion =fast growth

Fast surface diffusion =

slow facet growth

SiO

2

[100]

Wistey NAMBE 2009 6

Good fill next to gate.

• But gap persists

Gate Changes Local Kinetics

1. Excess In & Ga d ’t ti k t SiO

2. Local i h t

4. Low-angle planes grow

don’t stick to SiO2

[100]

enrichment of III/V ratio

g p ginsteadsidew

SiO2 or SiNx

[100]all

3. Increased surface mobility

Gate

Diff i f G III’ f t

Wistey NAMBE 2009 7

• Diffusion of Group III’s away from gate

Change of Faceting by Arsenic Flux

• InGaAs layers with increasing As fluxes, separated by InAlAs.

Increasing As fluxInAlAs

markers InGaAs

SiO2WCr 1x10-6

2x10-65x10-6

W

0.5x10-6

(Torr)

• Lowest arsenic flux → “rising tide fill”

8Wistey, NAMBE 2009

• No gaps near gate or SiO2/SiNx• Tunable facet competition

Growth conditions: MEE, 540*C, Ga+In BEP=1.5x10-7 Torr, InAlAs 500-540°C MBE.

Control of Facets by Arsenic Flux

• InGaAs:Si layers with increasing As F ti

InAlAs

• InGaAs:Si layers with increasing As fluxes, separated by InAlAs.

[100]O2

Faceting

Increasing As flux

InAlAsmarkers

InGaAs

SiO

SiO2WCr 1x10-6

2x10-65x10-6

[100]O2

ConformalW

0.5x10-6

(Torr)

SiO

• Lowest arsenic flux → “rising tide fill” [100]iO

2

Filling

9Wistey, NAMBE 2009

• No gaps near gate or SiO2/SiNx• Tunable facet competition

Growth conditions: MEE, 540*C, Ga+In BEP=1.5x10-7 Torr, InAlAs 500-540°C MBE.

S

Low-As Regrowth of InGaAs and InAs

InGaAs InAs

InAsth

InGaAs regrowth (top view)

regrowth

(top view)

• Low As flux good for InAs too. • InAs native defects are donors.

Bhargava et al , APL 1997

R d f d l ti

• No faceting near gate• Smooth far-field too

10Wistey, NAMBE 2009

4.7 nm Al203, 5×1012 cm-2 pulse dopingIn=9.7E-8, Ga=5.1E-8 Torr

SEMs: Uttam Singisetti

• Reduces surface depletion.

InAs Source-Drain Access Resistance4.7 nm Al203, InAs S/D E-FET.

740 Ω-µm

• Upper limit: Rs,max = Rd,max = 370 Ω−μm.• Intrinsic gmi = 0.53 mS/⎧m

1/R 3 3 S/ ( li it d )• gm << 1/Rs ~ 3.3 mS/μm (source-limited case)

11Wistey, NAMBE 2009

Ohmic contacts no longer limit MOSFET performance.

Conclusions

• Reducing As flux improves filling near gate• Self-aligned regrowth: a roadmap for scalable III-V FETs–Provides III-V’s with a salicide equivalent

I G A d l d I A t t• InGaAs and relaxed InAs regrown contacts–Not limited by source resistance @ 1 mA/µmResults comparable to other III V FETs but now–Results comparable to other III-V FETs... but now scalable

12Wistey, NAMBE 2009

Acknowledgements

• Rodwell & Gossard Groups (UCSB): Uttam Singisetti, Greg Burek, Ashish Baraskar, Vibhor Jain...

• McIntyre Group (Stanford): Eunji Kim Byungha Shin• McIntyre Group (Stanford): Eunji Kim, Byungha Shin, Paul McIntyre

• Stemmer Group (UCSB): Joël Cagnon, Susanne StStemmer

• Palmstrøm Group (UCSB): Erdem Arkun, Chris Palmstrøm

• SRC/GRC funding• UCSB Nanofab: Brian Thibeault, NSF

13Wistey, NAMBE 2009