Transistors for THz Systems Mark Rodwell, UCSB IMS Workshop: Technologies for THZ Integrated Systems (WMD) Monday, June 3, 2013, Seattle, Washington (8AM-5PM) [email protected]Co-Authors and Collaborators: UCSB HBT Team: J. Rode, H.W. Chiang, A. C. Gossard , B. J. Thibeault, W. Mitchell Recent Graduates: V. Jain, E. Lobisser, A. Baraskar, Teledyne HBT Team: M. Urteaga, R. Pierson, P. Rowell, B. Brar, Teledyne Scientific Company Teledyne IC Design Team: M. Seo, J. Hacker, Z. Griffith, A. Young, M. J. Choe, Teledyne Scientific Company UCSB IC Design Team: S. Danesgar, T. Reed, H-C Park, Eli Bloch WMD: Technologies for THZ Integrated Systems RFIC2013, Seattle, June 2-4, 2013 1
60
Embed
Transistors for THz Systems - UCSB...RF Data: 25 nm thick base, 75 nm Thick Collector 0 5 10 15 20 25 1010 1011 1012 B) Frequency (Hz) H 21 U f W = 530 GHz f max = 750 GHz Required
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Transistors for THz Systems
Mark Rodwell, UCSB
IMS Workshop: Technologies for THZ Integrated Systems (WMD)
Monday, June 3, 2013, Seattle, Washington (8AM-5PM)
Required dimensions obtained but poor base contacts on this run
E. Lobisser, ISCS 2012, August, Santa Barbara
29
DC, RF Data: 100 nm Thick Collector
0
4
8
12
16
20
24
28
32
109
1010
1011
1012
Ga
in (
dB
)
Frequency (Hz)
U
H21
f = 480 GHz
fmax
= 1.0 THz
Aje = 0.22 x 2.7 m
2
Ic = 12.1 mA
Je = 20.4 mA/m
2
P = 33.5 mW/m2
Vcb
= 0.7 V
0
5
10
15
20
25
30
0 1 2 3 4 5
Je (
mA
/m
2)
Vce
(V)
P = 30 mW/m2
Ib,step
= 200 A
BV
P = 20 mW/m2
Aje
= 0.22 x 2.7 m2
10-9
10-7
10-5
10-3
10-1
0
5
10
15
20
0 0.2 0.4 0.6 0.8 1I c
, I b
(A
)
Vbe
(V)
Solid line: Vcb
= 0.7V
Dashed: Vcb
= 0V
nc = 1.19
nb = 1.87
Ic
Ib
Jain et al IEEE DRC 2011
30
Chart 31
THz InP HBTs From Teledyne
Urteaga et al, DRC 2011, June 31
Towards & Beyond the 32 nm /2.8 THz Node
Base contact process: Present contacts too resistive (4Wm2) Present contacts sink too deep (5 nm) for target 15 nm base
→ refractory base contacts
Emitter Degeneracy: Target current density is almost 0.1 Amp/m2 (!) Injected electron density becomes degenerate. transconductance is reduced.
→ Increased electron mass in emitter
32
Base Ohmic Contact Penetration
~5 nm Pt contact penetration (into 25 nm base)
Base Ohmic Contact Penetration
33
Refractory Base Process (1)
low contact resistivity low penetration depth
low bulk access resistivity
base surface not exposed to photoresist chemistry: no contamination low contact resistivity, shallow contacts low penetration depth allows thin base, pulsed-doped base contacts
Blanket liftoff; refractory base metal Patterned liftoff; Thick Ti/Au
34
Refractory Base Process (2)
0 100
5 1019
1 1020
1.5 1020
2 1020
2.5 1020
0 5 10 15 20 25
do
pin
g, 1
/cm
3
depth, nm
2 nm doping pulse
1018
1019
1020
1021
P-InGaAs
10-10
10-9
10-8
10-7
10-6
10-5
Hole Concentration, cm-3
B=0.8 eV
0.6 eV0.4 eV0.2 eV
step-barrierLandauer
Co
nta
ct
Res
isti
vit
y,
W
cm
2
32 nm node requirement
Increased surface doping: reduced contact resistivity, but increased Auger recombination. → Surface doping spike at most 2-5 thick. Refractory contacts do not penetrate; compatible with pulse doping. 35
Refractory Base Ohmic Contacts Refractory Base Ohmic Contacts
<2 nm Ru contact penetration (surface removal during cleaning)
Ru / Ti / Au
36
Degenerate Injection→ Reduced Transconductance
10-3
10-2
10-1
100
101
102
-0.3 -0.2 -0.1 0 0.1 0.2
J(m
A/
m2)
Vbe
-
Boltzmann
(-Vbe
)>>kT/q
)./exp( kTqVJJ bes
Transconductance is high
)./exp( kTqVJJ bes
JAg Em /
Current varies exponentially with Vbe
37
)./exp( kTqVJJ bes
Degenerate Injection→ Reduced Transconductance
10-3
10-2
10-1
100
101
102
-0.3 -0.2 -0.1 0 0.1 0.2
J(m
A/
m2)
Vbe
-
Fermi-Dirac
Boltzmann
(-Vbe
)>>kT/q
Transconductance is reduced
Current varies exponentially with Vbe
38
Degenerate Injection→ Reduced Transconductance
10-3
10-2
10-1
100
101
102
-0.3 -0.2 -0.1 0 0.1 0.2
J(m
A/
m2)
Vbe
-
Fermi-Dirac
Highly degenerate
(Vbe
->>kT/q
Boltzmann
(-Vbe
)>>kT/q
Highly degenerate limit:
)(8
* 2
32
3
beVmq
J
2* )( beE VmJ
current varies as the square of bias
39
Degenerate Injection→ Reduced Transconductance
10-3
10-2
10-1
100
101
102
-0.3 -0.2 -0.1 0 0.1 0.2
J(m
A/
m2)
Vbe
-
Fermi-Dirac
Highly degenerate
(Vbe
->>kT/q
Boltzmann
(-Vbe
)<<kT/q
Highly degenerate limit:
)(8
* 2
32
3
beVmq
J
2* )( beE VmJ
Transconductance varies as J1/2
current varies as the square of bias
JmAg EEm
*/
At & beyond 32 nm, we must increase the emitter effective mass.
...and as (m*)1/2
40
Degenerate Injection→Solutions
At & beyond 32 nm, we must increase the emitter (transverse) effective mass.
Other emitter semiconductors: no obvious good choices (band offsets, etc.).
Emitter-base superlattice: increases transverse mass in junction evidence that InAlAs/InGaAs grades are beneficial
Extreme solution (10 years from now): partition the emitter into small sub-junctions, ~ 5 nm x 5 nm. parasitic resistivity is reduced progressively as sub-junction areas are reduced.
41
3-4 THz Bipolar Transistors are Feasible.
4 THz HBTs realized by:
Extremely low resistivity contacts
Extreme current densities
Processes scaled to 16 nm junctions
Impact: efficient power amplifiers and complex signal processing from 100-1000 GHz.
Teledyne Scientific: moving THz IC Technology towards aerospace applications 1.1 THz pilot IC process 204 GHz digital logic (M/S latch) 670 GHz amplifier 300 GHz fundamental phase-lock-loop
Vout
VEE VBB
Vtune
Vout
VEE VBB
Vtune614 GHz fundamental oscillator (VCO)
182 mW 188 mW, 33% PAE
THz InP HEMTs and
III-V MOSFETs
46
FET parameter change
gate length decrease 2:1
current density (mA/m), gm (mS/m) increase 2:1
transport effective mass constant
channel 2DEG electron density increase 2:1
gate-channel capacitance density increase 2:1
dielectric equivalent thickness decrease 2:1
channel thickness decrease 2:1
channel density of states increase 2:1
source & drain contact resistivities decrease 4:1
Changes required to double transistor bandwidth
GW widthgate
GL
HBT parameter change
emitter & collector junction widths decrease 4:1
current density (mA/m2) increase 4:1
current density (mA/m) constant
collector depletion thickness decrease 2:1
base thickness decrease 1.4:1
emitter & base contact resistivities decrease 4:1
eW
ELlength emitter
constant voltage, constant velocity scaling nearly constant junction temperature → linewidths vary as (1 / bandwidth)2
fringing capacitance does not scale → linewidths scale as (1 / bandwidth )
47
FET scaling challenges...and solutions
Gate barrier under S/D contacts → high S/D access resistance addressed by S/D regrowth
High gate leakage from thin barrier, high channel charge density (almost) eliminated by ALD high-K gate dielectric
Other scaling considerations: low InAs electron mass→ low state density capacitance → gm fails to scale increased m* , hence reduced velocity in thin channels minimum feasible thickness of gate dielectric (tunneling) and channel
48
-0.2 0.0 0.2 0.4 0.60.0
0.4
0.8
1.2
1.6
2.0
2.4
Gm
(mS
/m
)
Cu
rre
nt
De
ns
ity
(m
A/
m)
Gate Bias (V)
0.0
0.4
0.8
1.2
1.6
2.0
2.4 VDS
=0.05 V
VDS
=0.5 V
III-V MOS
Peak transconductance; VLSI-style FET: 2.5 mS/micron ~85% of best THz InAs HEMTs
0.1 10.0
0.4
0.8
1.2
1.6
2.0
2.4
2.8 90 [011]
0 [01-1]
at Vds
=0.5V
Gate length (m)
Pe
ak
Tra
ns
co
nd
uc
tan
ce
(m
S/
m)
Lg = 40 nm (SEM)
III-V MOS will soon surpass HEMTs in RF performance
Sanghoon Lee
40 nm devices are nearly ballistic
49
In ballistic limit, current and transconductance are set by: channel & dielectric thickness, transport mass, state density
2/3*
,
2/1*
1
2/3
1
)/()/(1 where,
V 1m
mA84
oequivodos
othgs
mmgcc
mmgK
VVKJ
FET Drain Current in the Ballistic Limit
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.01 0.1 1
no
rma
lized
drive
cu
rre
nt
K1
m*/mo
g=2
EET=1.0 nm
0.6 nm
0.4 nm
g=1
InGaAs <--> InP Si
0.3 nm
EET=Equivalent Electrostatic Thickness
=Tox
SiO2
/ox
+Tinversion
SiO2
/semiconductor
/EETε
)/c/c(c oxequiv
2SiO
1
depth
11
2
0
2
, 2/ mqc odos
minima band # g
50
Low m* gives lowest transit time, lowest Cgs at any EOT.
Transit delay versus mass, # valleys, and EOT
51
2/1*
,
2/1
0
*
2
2/1
72 1 whereVolt 1
cm/s 1052.2
oeq
odos
thgs
g
D
chch
m
mg
c
c
m
mK
VV
LK
I
Q
0
0.5
1
1.5
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
No
rma
lize
d t
ran
sit d
ela
y K
2
m*/mo
EOT=1.0 nm
EOT includes wavefunction depth term
(mean wavefunction depth*SiO2
/semiconductor
)
0.6 nm
0.4 nm
g=1, isotropic bands
g=2, isotropic bands
1 nm
0.4 nm
0.6 nm
/EOTε
)/c/c(c oxequiv
2SiO
1
semi
11
InGaAs
Si, GaN
FET Scaling: fixed vs. increasing state density
52
0
500
1000
1500
2000
2500
3000 canonical scaling
stepped # of bands
transport only
f , G
Hz
0
500
1000
1500
2000
2500
3000
3500
4000
f ma
x, G
Hz
0
0.5
1
1.5
2
2.5
0 10 20 30 40 50 60
dra
in c
urr
ent
density, m
A/
m
gate length, nm
200 mV gate overdrive
0
200
400
600
800
1000
0 10 20 30 40 50 60
SC
FL
sta
tic d
ivid
er
clo
ck r
ate
, G
Hz
gate length, nm
f fmax
mA/m→ VLSI metric SCFL divider speed
Need higher state density for ~10 nm node
2-3 THz Field-Effect Transistors are Feasible.
3 THz FETs realized by:
Ultra low resistivity source/drain
High operating current densities
Very thin barriers & dielectrics
Gates scaled to 9 nm junctions high-barrier HEMT MOSFET
Impact: Sensitive, low-noise receivers from 100-1000 GHz.
3 dB less noise → need 3 dB less transmit power.
or
53
4-nm / 5-THz FETs: Challenges
4 nm
Channel thickness 0.8 nm: UTB 1.6 nm: fin atomically flat
Gate dielectric 0.1 nm EOT: UTB 0.2 nm EOT: fin
Estimated (WKB) tunneling current is just acceptable at 0.2 nm EOT.
How can we make a 1.6 nm thick fin, or a 0.8 nm thick body ? 54
Thin wells have high scattering rate
Need single-atomic-layer control of thickness Need high quantization mass mq.
Sakaki APL 51, 1934 (1987).
./1y probabilit Scattering 62
wellqTm
55
III-V vs. CMOS: A false comparison ?
UTB Si MOS UTB III-V MOS III-V THz MOS/HEMT
III-V THz HEMT
III-V MOS has a reasonable chance of future use in VLSI
The real THz / VLSI distinction: Device geometry optimized for high-frequency gain vs. optimized for small footprint and high DC on/off ratio.
56
Conclusion
57
THz and Far-Infrared Electronics
IR today→ lasers & bolometers → generate & detect
It's all about classic scaling: contact and gate dielectrics...