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

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

Mark A. Wistey

Now at University of Notre Damemwistey@nd.edu

mwistey@nd.edu

A.K. Baraskar, U. Singisetti, G.J. Burek, M.J.W. Rodwell, A.C. GossardUniversity of California Santa Barbara

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

Funding: SRC

2Wistey, NAMBE 2009

Outline: Regrown III-V FET Contacts

•Motivation for Self-Aligned Regrowth

•Facets, Gaps, Arsenic Flux and MEE

•MOSFET Results•Conclusion

3Wistey, NAMBE 2009

Motivation for Regrowth: Scalable III-V FETs

Classic III-V FET (details vary):

ChannelBottom BarrierInAlAs Barrier

Top Barrier or Oxide

Gate

Low doping

GapSource Drain

Large Rc {

Large Area Contacts{III-V FET with Self-Aligned Regrowth:

ChannelIn(Ga)P Etch Stop

Bottom Barrier

High-k

Gate

n+ Regrowth

High Velocity Channel

Implant: straggle,short channel effects

• Advantages of III-V’s

• Disadvantages of III-V’s

Small Raccess

Small RcSelf-aligned, no gaps

High doping: 1013 cm-2 avoids source exhaustion

2D injection avoids source starvation

Dopants active as-grown

High mobilityaccess regions

High barrier

Ultrathin 5nm doping layer}

4Wistey, NAMBE 2009

MBE Regrowth: Bad at any Temperature?

• Low growth temperature (<400°C):

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

Gate

200nm Gap

Source-DrainRegrowth

SEMs: Uttam Singisetti

Metals

Channel

SiO2

high-k Gate

Source-DrainRegrowth

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

•High growth temperature (>490°C):– Selective/preferential epi on InGaAs– No gaps near gate

– Rough far field – High resistance

5Wistey, NAMBE 2009

560C

490C460C

SiO2 dummy

gate

SiO2 dummy

gate

540C

No gaps, but

faceting next to gates

Gap

High Temperature MEE: Smooth & No Gaps

Smooth

regrowth

Note faceting: surface kinetics, not shadowing.

In=9.7E-8, Ga=5.1E-8 Torr

Wistey NAMBE 2009 6

Shadowing and Facet Competition

[111]

[100]

Fast surface diffusion = slow facet

growth

Slow diffusion = rapid facet growth

Shen & Nishinaga, JCG 1995

[111] faster[100] faster

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

SiO

2

[111]

[100]

Slow diffusion =fast growth

Fast surface diffusion =

slow facet growth

SiO

2

Good fill next to gate.

• But gap persists

Wistey NAMBE 2009 7

Gate Changes Local Kinetics

[100]

2. Local enrichment of III/V ratio

4. Low-angle planes grow insteadsidew

all

1. Excess In & Ga don’t stick to SiO2

• Diffusion of Group III’s away from gate

3. Increased surface mobility

Gate

SiO2 or SiNx

8Wistey, NAMBE 2009

Change of Faceting by Arsenic Flux

SiO2WCr

Increasing As flux

Original Interface

InAlAsmarkers

InGaAs

0.5x10-6

1x10-6

2x10-6

5x10-6

• Lowest arsenic flux → “rising tide fill”

• No gaps near gate or SiO2/SiNx

• Tunable facet competition

(Torr)

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

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

9Wistey, NAMBE 2009

Control of Facets by Arsenic Flux

SiO2WCr

Increasing As flux

Original Interface

InAlAsmarkers

InGaAs

0.5x10-6

1x10-6

2x10-6

5x10-6

• Lowest arsenic flux → “rising tide fill”

• No gaps near gate or SiO2/SiNx

• Tunable facet competition

(Torr)

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

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

[111]

[100]

SiO

2

Filling[11

1]

[100]

SiO

2

[111]

[100]

SiO

2

Conformal

Faceting

10Wistey, NAMBE 2009

Low-As Regrowth of InGaAs and InAs

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

SEMs: Uttam Singisetti

InGaAs regrowth (top view)

• Low As flux good for InAs too.

• InAs native defects are donors. Bhargava et al , APL 1997

• Reduces surface depletion.

InAsregrowth

InGaAs InAs

• No faceting near gate• Smooth far-field too

11Wistey, NAMBE 2009

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

• Upper limit: Rs,max = Rd,max = 370 Ω−μm.• Intrinsic gmi = 0.53 mS/μm• gm << 1/Rs ~ 3.3 mS/μm (source-limited case)

➡ Ohmic contacts no longer limit MOSFET performance.

740 Ω-µm

12Wistey, NAMBE 2009

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• InGaAs and relaxed InAs regrown contacts

–Not limited by source resistance @ 1 mA/µm–Results comparable to other III-V FETs... but now scalable

13Wistey, NAMBE 2009

Acknowledgements

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

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

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

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

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