Ashish Baraskar 1IPRM 2010
High Doping Effects on in-situ Ohmic Contacts to n-InAs
Ashish Baraskar, Vibhor Jain, Uttam Singisetti, Brian Thibeault, Arthur Gossard and Mark J. W. Rodwell
ECE, University of California, Santa Barbara, CA,USA
Mark A. WisteyECE, University of Notre Dame, IN, USA
Yong J. LeeIntel Corporation, Technology Manufacturing Group, Santa Clara, CA, USA
Ashish Baraskar 2IPRM 2010
• Motivation– Low resistance contacts for high speed HBTs– Approach
• Experimental details– Contact formation– Fabrication of Transmission Line Model structures
• Results– InAs doping characteristics– Effect of doping on contact resistivity– Effect of annealing
• Conclusion
Outline
Ashish Baraskar 3IPRM 2010
effcbbb CRff
8maxRC
f in 21
To double device bandwidth*:
• Cut transit time 1:2
• Cut RC delay 1:2
Device Bandwidth Scaling Laws for HBT
bcWcTbT
eW
*M.J.W. Rodwell, IEEE Trans. Electron. Dev., 2001
Scale contact resistivities by 1:4
Ashish Baraskar 4IPRM 2010
Emitter: 512 256 128 64 32 width (nm) 16 8 4 2 1 access ρ, (m2)
Base: 300 175 120 60 30 contact width (nm) 20 10 5 2.5 1.25 contact ρ (m2)
ft: 370 520 730 1000 1400 GHzfmax: 490 850 1300 2000 2800 GHz
- Contact resistivity serious barrier to THz technology
Less than 2 Ω-µm2 contact resistivity required for simultaneous THz ft and fmax*
*M.J.W. Rodwell, CSICS 2008
InP Bipolar Transistor Scaling Roadmap
Ashish Baraskar 5IPRM 2010
Emitter Ohmics-IMetal contact to narrow band gap material 1. Fermi level pinned in the band-gap
2. Fermi level pinned in the conduction band
Better ohmic contacts with narrow band gap materials1,2
Narrow band gap material:
• lower Schottky barrier height
• lower m* easier tunneling across Schottky barrier
InGaAs InP
GaAsInAs
1.Peng et. al., J. Appl. Phys., 64, 1, 429–431, (1988).2.Shiraishi et. al., J. Appl. Phys., 76, 5099 (1994).
Ashish Baraskar 6IPRM 2010
Choice of material:• In0.53Ga0.47As: lattice matched to InP
- Ef pinned 0.2 eV below conduction band[1]
• Relaxed InAs on In0.53Ga0.47As
- Ef pinned 0.2 eV above conduction band[2]
1. J. Tersoff, Phys. Rev. B 32, 6968 (1985) 2. S. Bhargava et. al., Appl. Phys. Lett., 70, 759 (1997)
Emitter Ohmics-II
Other considerations:• Better surface preparation techniques - For efficient removal of oxides/impurities• Refractory metal for thermal stability
Ashish Baraskar 7IPRM 2010
Semi-insulating InP Substrate
150 nm In0.52Al0.48As: NID buffer 100 nm InAs: Si (n-type)
Semiconductor layer growth by Solid Source Molecular Beam Epitaxy (SS-MBE): n-InAs/InAlAs - Semi insulating InP (100) substrate - Un-doped InAlAs buffer - Electron concentration determined by Hall measurements
Thin Film Growth
Ashish Baraskar 8IPRM 2010
In-situ molybdenum (Mo) deposition - E-beam chamber connected to MBE chamber- No air exposure after film growth
Why Mo? - Refractory metal (melting point ~ 2620 oC)- Easy to deposit by e-beam technique- Easy to process and integrate in HBT process flow
150 nm In0.52Al0.48As: NID buffer 100 nm InAs: Si (n-type)
20 nm Mo
Semi-insulating InP Substrate
In-situ Metal Contacts
Ashish Baraskar 9IPRM 2010
• E-beam deposition of Ti, Au and Ni layers
• Samples processed into TLM structures by photolithography and liftoff
• Mo was dry etched in SF6/Ar with Ni as etch mask, isolated by wet etch
150 nm In0.52Al0.48As: NID buffer 100 nm InAs: Si (n-type)
20 nm Mo20 nm ex-situ Ti
50 nm ex-situ Ni500 nm ex-situ Au
Semi-insulating InP Substrate
TLM (Transmission Line Model) Fabrication
Ashish Baraskar 10IPRM 2010
• Resistance measured by Agilent 4155C semiconductor parameter analyzer
• TLM pad spacing (Lgap) varied from 0.5-26 µm; verified from scanning electron microscope (SEM)
• TLM Width ~ 25 µm
Resistance Measurement
WR
R ShCC
22
0
1
2
3
4
5
0 0.5 1 1.5 2 2.5 3 3.5Gap Spacing (m)
Res
ista
nce
()
Ashish Baraskar 11IPRM 2010
• Extrapolation errors:– 4-point probe resistance
measurements on Agilent 4155C– Resolution error in SEM
0
0.5
1
1.5
2
2.5
3
3.5
0 1 2 3 4 5 6
Res
ista
nce
()
Gap Spacing (m)
R
d
Rc
Error Analysis• Processing errors:
Lgap
W
Variable gap along width (W)1.10 µm 1.04 µm
– Variable gap spacing along width (W)
Overlap Resistance
– Overlap resistance
Ashish Baraskar 12IPRM 2010
1018
1019
1020
0
1000
2000
3000
4000
5000
6000
1018 1019 1020
Active carrierconcentration
Mobility
Mob
ility
(cm
2 /Vs)
Elec
tron
conc
entra
tion,
n (c
m-3
)
Si atom concentration (cm-3)
Tsub: 380 oC
Results: Doping Characteristics
n saturates at high dopant concentration
5 1019
1 1020
1.5 1020
380 400 420 440 460Substrate Temperature (oC)
Elec
tron
conc
entra
tion,
n (c
m-3
)- Enhanced n for colder growths- hypothesis: As-rich surface drives Si onto group-III sites
Ashish Baraskar 13IPRM 2010
Metal Contact ρc (Ω-µm2) ρh (Ω-µm)
In-situ Mo 0.6 ± 0.4 2.0 ± 1.5
WR
R ShCC
22
0
1
2
3
4
5
0 0.5 1 1.5 2 2.5 3 3.5Gap Spacing (m)
Res
ista
nce
()
Lowest ρc reported to date for n-type InAs
• Electron concentration, n = 1×1020 cm-3
• Mobility, µ = 484 cm2/Vs• Sheet resistance, Rsh = 11 ohm/ (100 nm thick film)
Results: Contact Resistivity - I
Ashish Baraskar 14IPRM 2010
Results: Contact Resistivity - II
• ρc measured at various n - ρc decreases with increase in n
• Shiraishi et. al.[1] reported ρc = 2 Ω-µm2 for ex-situ Ti/Au/Ni contacts to n-InAs
• Singisetti et. al.[2] reported ρc = 1.4 Ω-µm2 for in-situ Mo/n-InAs/n-InGaAs 0
1
2
3
4
0 2 4 6 8 10 12Electron Concentration, n (x 1019cm-3)
Con
tact
Res
istiv
ity,
c (
m2 )
Shiraishi et. al.[1] (ρc = 2 Ω-µm2)
Singisetti et. al.[2] (ρc = 1.4 Ω-µm2)
Present work
1Shiraishi et. al., J. Appl. Phys., 76, 5099 (1994). 2Singisetti et. al., Appl. Phys. Lett., 93, 183502 (2008).
Extreme Si doping improves contact resistance
Ashish Baraskar 15IPRM 2010
• Contacts annealed under N2 flow at 250 oC for 60 minutes
(replicating the thermal cycle experienced by a transistor during fabrication)
• Observed variation in ρc less than the margin of error
Thermal Stability
Contacts are thermally stable
Results: Contact Resistivity - III
Ashish Baraskar 16IPRM 2010
• Optimize n-InAs/n-InGaAs interface resistance• Mo contacts to n-InGaAs*: ρc = 1.1±0.6 Ω-µm2
n-InAs n-InGaAs
Application in transistors !
*Baraskar et. al., J. Vac. Sci. Tech. B, 27, 2036 ( 2009).
Ashish Baraskar 17IPRM 2010
• Extreme Si doping improves contact resistance
• ρc= 0.6±0.4 Ω-µm2 for in-situ Mo contacts to n-InAs with 1×1020 cm-3 electron concentration
• Need to optimize n-InAs/n-InGaAs interface resistance for transistor application
Conclusions:
Ashish Baraskar 18IPRM 2010
Thank You !
Questions?
AcknowledgementsONR, DARPA-TFAST, DARPA-FLARE
[email protected] of California, Santa Barbara, CA, USA
Ashish Baraskar 19IPRM 2010
Extra Slides
Ashish Baraskar 20IPRM 2010
1. Doping CharacteristicsResults
6 1019
7 1019
8 1019
9 1019
1020
5 10 15 20 25 30As flux (x10-7 torr)
Act
ive
carr
ier c
once
ntra
tion,
n (c
m-3
)
Ashish Baraskar 21IPRM 2010
0.1
1
10
0 5 10-10 1 10-9
n-1/2 (cm-3/2)
Con
tact
Res
istiv
ity,
c (
m2 )
Ashish Baraskar 22IPRM 2010
n-InAs p-InAs
– : simulation[1]
x : experimental data[2]
Results: Contact Resistivity - III
Possible reasons for decrease in contact resistivity with increase in electron concentration• Band gap narrowing• Strain due to heavy doping• Variation of effective mass with doping
Ashish Baraskar 23IPRM 2010
Accuracy Limits
• Error Calculations– dR = 50 m (Safe estimate)– dW = 0.2 m– dGap = 20 nm
• Error in c ~ 50% at 1 -m2
Ashish Baraskar 24IPRM 2010
Random and Offset Error in 4155C
0.6642
0.6643
0.6644
0.6645
0.6646
0.6647
0 5 10 15 20 25
Res
ista
nce
()
Current (mA)
• Random Error in resistance measurement ~ 0.5 m
• Offset Error < 5 m
*4155C datasheet
Ashish Baraskar 25IPRM 2010
Correction for Metal Resistance in 4-Point Test Structure
WLsheet /metalR Wcontactsheet /)( 2/1
Error term (Rmetal/x) from metal resistance
L
WI
I
V
V
xRWLW metalsheetcontactsheet ///)( 2/1
Ashish Baraskar 26IPRM 2010
Overlap Resistance
Variable gap along width (W)1.10 µm 1.04 µm
W
Ashish Baraskar 27IPRM 2010
• Variation of effective mass with doping• Non-parabolicity• Thickness dependence• SXPS (x-ray photoemission spectroscopy)• BEEM (ballistic electron emission microscopy)• Band gap narrowing• Strain due to heavy doping
Ashish Baraskar 28IPRM 2010
For the same number of active carriers (in the degenerate regime) Ef - Ec > Ef - Ec (narrow gap, Eg1) (wide gap, Eg2)
m*e(Eg1)< m*e(Eg2)
Metal contact to narrow band gap material
Emitter Ohmics-I