Record Ion (0.50 mA/μm at VDD = 0.5 V and Ioff = 100 nA/μm) 25 nm-Gate-Length ZrO2/InAs/InAlAs MOSFETs Sanghoon Lee1*, V. Chobpattana2,C.-Y. Huang1, B. J. Thibeault1, W. Mitchell1, S. Stemmer2, A. C. Gossard2, and M. J. W. Rodwell1 1ECE and 2Materials Departments University of California, Santa Barbara, CA 2014 Symposium on VLSI Technology Honolulu, Hawaii, USA 06/10/2014 *sanghoon_lee@ece.ucsb.edu 1 VLSI 2014 Why III-V MOSFETs in VLSI applications? Low m* in III-V material high vinj high transconductance More transconductance per gate width more current lower intrinsic delay -or- reduced VDD less power consumption -or- small FETs reduced IC size TW. Kim, IEDM 2012 Other advantages Wide range of available materials nm-precise growth 1-2 nm thick channel Larger ΔEc Better confinement, Small EOT 2 VLSI 2014 Key Design Considerations Dielectric: Source/Drain: Low ρc Small contact size Self-aligned Small contact pitch Shallow Scaling (electrostatics) N+Source Dielectric Channel Thin high Ion, better SS and DIBL Low Dit Better SS N+Drain Barrier Channel: Thin Electrostatics Thin and wide bandgap Small band-band tunneling Thick and narrow bandgap higher injection velocity 3 VLSI 2014 FET Structures N+Source Dielectric N+Drain N+Source Body - Self-aligned - Implant damage - Large Raccess (limited doping) N+Source Channel - Good short channel effect - Large device footprint - Large Raccess (Barrier) Dielectric Channel Etch stop Barrier MOS-HEMT N+Source (Regrown) N+Drain Dielectric Barrier Inversion layer Inversion mode MOSFETs N+Drain Dielectric Barrier Channel Trench-etch - Small footprint - Small Raccess - Limited Lg scaling (wet etch) N+Drain (Regrown) Barrier Regrown S/D with gate-Last Regrown S/D with gate-first - Small footprint and Lg - Small Raccess - Abrupt junction - Low damage (No dry etch) - High damage (gate-stack etch) 4 VLSI 2014 Gate-Last Process ( Simplified for Development ) Channel growth By MBE Dummy gate formation e-beam lithography Cap: 2 nm In0.53Ga0.48As (U.I.D) Channel : 3.5 nm InAs (Strained) Setback: In0.52Al0.48As Setback (U.I.D) Pulse Doping (Si 2X1012 /cm2) Back Barrier: In0.52Al0.48As (U.I.D) P-type Doped Barrier: In0.52Al0.48As (Be 1017/cm3) 50 nm N+ In0.53Ga0.47As HSQ Regrown S/D HSQ 10 nm In0.53Ga0.47As Al2O3 InGaAs Cap InAs Channel In0.52Al0.48As Setback Vertical Spacer Al2O3 InGaAs Cap InAs Channel In0.52Al0.48As Setback In Al As Back Barrier Pulse Doping In0.52Al0.48As Back Barrier Pulse Doping 0.52 0.48 P-type Doped Barrier P-type Doped Barrier Substrate: InP (Semi-insulating) Mesa-isolation Surface digital etching Vertical spacer and N+ S/D regrowth in MOCVD InP (Substrate) InP (Substrate) Gate stack formation S/D metal contact formation 0.7/3.0 nm Al2OxNy/ZrO2 0.7/3.0 nm Al2OxNy/ZrO2 Ti/Pd/ Au N+ In0.53Ga0.47As Regrown S/D 12 nm In0.53Ga0.47As Vertical Spacer 2.5 nm InAs Channel In0.52Al0.48As Setback In0.52Al0.48As Back Barrier Pulse Doping P-type Doped Barrier InP (Substrate) Ni/Au (gate metal) N+ In0.53Ga0.47As 12 nm In0.53Ga0.47As Regrown S/D Vertical Spacer 2.5 nm InAs Channel In0.52Al0.48As Setback In0.52Al0.48As Back Barrier Pulse Doping P-type Doped Barrier InP (Substrate) 5 (S/D metal) LSD Ni/Au (gate metal) N+ In0.53Ga0.47As 12 nm In0.53Ga0.47As Lg Regrown S/D Vertical Spacer 2.5 nm InAs Channel In0.52Al0.48As Setback In0.52Al0.48As Back Barrier Pulse Doping P-type Doped Barrier InP (Substrate) VLSI 2014 High-k : MOSCAP with 0.7/5.0 nm Al2OxNy/ZrO2 Ni/Au 54 nm ZrO ZrO22 ~0.7 nm Al2OxNy N-type In0.53Ga0.47As N-type InP (Substrate) Cr/Au MOSCAP structure dielectric constant for ZrO2 is 23; EOT is ~1 nm 3.5 µF/cm2 accumulation capacitance at 1MHz ~1X1012 /cm2-eV Dit near midgap. Gate leakage < 1 A/cm2 up VG=2 V (V. Chobpattana, et al., ‘Scaled ZrO2 dielectrics for InGaAs gate stack with low interface trap densities’, APL 2014) 6 VLSI 2014 Off-state leakage and S/D spacers MOS-HEMT with large contact pitch Small S/D contact pitch 0 10 -1 10 3.2 Lg = 18 nm 2.8 VDS = 0.1, 0.5 V 2.4 2.0 -2 10 1.6 -3 1.2 10 0.8 -4 10 0.4 -5 10 -0.2 Lg =35 nm gm (mS/m) Current Density (mA/m) Band-band tunneling impact ionization 0.0 0.2 0.4 0.0 Gate Bias (V) D-H. Kim, IEDM 2012 Large lateral spacer low leakage, good short channel immunity Large lateral spacer large S/D pitch 7 VLSI 2014 Vertical Spacers reduced off-state leakage 200 Spacer thickness 2 nm 7 nm 12 nm SSmin (mV/dec) 180 160 Open: VDS = 0.1 V 140 Solid: VDS = 0.5 V 120 100 80 60 0.01 Spacer SSmin (mV/dec) 180 160 Open: VDS = 0.1 V 140 Solid: VDS = 0.5 V 120 100 /m) Larger spacer 80 60 and long channels better short channel effect at short 0.01 0.1 1 10 10 0 -1 Gate Length (m) 2 nm 8 10 -1 10 -2 10 -3 10 -4 10 -5 2 nm 7 nm Spacer thickness 2 nm 7 nm 12 nm 250 DIBL (mV/V) Spacer thickness 2 nm 7 nm 12 nm 10 7 nm 1 Gate Length (m) 0 300 Current Density (mA/m) 200 0.1 200 150 100 50 -6 V = 0.1 to 0.7 V 10 at 1 DS A/m 0 -7 0.01 10 0.2 V increment 0.1 1 length0.4 (m) 0.0 0.0Gate0.2 Gate (S. Lee, et al., EDL, June 2014) 12 nm VLSI 2014 Cross-sectional STEM image Courtesy of S. Kraemer (UCSB) 9 *Heavy elements look brighter VLSI 2014 -2 10 Dot : Reverse Sweep Solid: Forward Sweep Lg = 1 m 10 1 10 0 10 -1 10 -2 10 -3 -3 10 -4 10 -5 10 -6 10 -7 10 -8 SSmin~ 61 mV/dec. (at V DS = 0.1 V) SSmin ~ 63 mV/dec. 10-4 (at V = 0.5 V) 10 10 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 DS -5 2 Current Density (mA/m) -1 10 |Gate Leakage| (A/cm ) I-V characteristics for long channel device (Lg = 1 µm) Gate Bias (V) 61 mV/dec Subthreshold swing at VDS=0.1 V Negligible hysteresis <1 A/cm2 gate leakage at measured bias range 10 VLSI 2014 Current Density (mA/m) 2.8 Lg = 25 nm 2.4 0.8 Ion= 500 A/m (at 0.6 I =100 nA/m, V =0.5 V) off DD VDS = 0.1 to 0.7 V 0.2 V increment 2.0 1.6 1.2 0.4 0.8 0.2 0.4 0.0 0.0 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 10 1 10 0 -1 10 VDS = 0.1 to 0.7 V 0.2 V increment -3 10 -4 DIBL = 76 mV/V -5 VT = -85 mV at 1 A/m -6 SSmin~ 72 mV/dec. (at VDS = 0.1 V) -7 SSmin~ 77 mV/dec. (at VDS = 0.5 V) 10 1.0 VGS = -0.4 V to 0.7 V 0.1 V increment Ron = 303 Ohm-m 0.8 at V = 0.7 V GS LSD = 140 nm Lg = 25 nm 0.6 0.4 0.2 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Drain Bias (V) Lg = 25 nm -2 10 1.2 Gate Bias (V) 10 10 Current Density (mA/m) 1.0 gm (mS/m) Current Density (mA/m) I-V characteristics for short channel devices ( Lg = 25 nm) ~2.4 mS/μm Peak gm at VDS=0.5 V ~300 Ohm-µm on-resistance at VGS=0.7 V 77 mV/dec Subthreshold Swing at VDS=0.5 V, 76 mV/V DIBL at 1 µA/µm 0.5 mA/µm Ion at Ioff=100 nA/µm and VDD=0.5 V 10 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 Gate Bias (V) 11 VLSI 2014 N+S/D RRaccess Rspacer Rballistic 5000 800 Ron = 168 -m (+/- 25 -m) 4000 Resistance (Ohm-m) Rcontact On-Resistance (Ohm-m) Source/drain series resistance (extrpolated at zero Lg) 3000 2000 1000 at VGS = 0.7 V 0 0.0 0.2 0.4 0.6 0.8 600 Gate Length (m) RN+SD= 85 -m 500 400 300 Gap Ti/Pd/Au 200 60 nm N+ InGaAs (regrown contact layer) 100 0 1.0 Y = 24 + 25*X 2 c = 5.3 -m 700 5 nm Intrinsic InGaAs (Capping layer) Channel 0 5 10 15 Gap (m) 20 25 Rcontact [Ohm-µm] RN+ S/D [Ohm-µm] Rspacer [Ohm-µm] Rballistic [Ohm-µm] Ron at zero Lg[Ohm-µm] 25 60 35 50 170 for both source and drain sides From TLM measurement for N+S/D, RN+S/D sheet = 25 ohm/sq, ρc= ~5.3 ohm-μm2 Rspacer is estimated to be ~35 ohm-μm for both sides 12 VLSI 2014 Performance comparison: 2.5 nm VS 5.0 nm-thick channel 2.5 nm InAs 5.0 nm InAs 2.5 nm InAs 100 Open: VDS = 0.1 V 90 Solid: VDS = 0.5 V 80 70 0.1 Gate Length (m) 1 Current Density (mA/m) SSmin (mV/dec) 110 60 0.01 5.0 nm InAs 0 10 SS ~ 60 mV/dec -1 10 at V =0.1 V -2 DS 10 Lg =500 nm -3 10 ID -4 10 SS ~ 64 mV/dec at VDS=0.1 V Lg =500 nm ID -5 10 -6 10 -7 10 -8 10 IG IG -9 10 -10 10 -0.2 0.0 0.2 0.4 -0.2 0.0 Gate Bias (V) 0.2 0.4 Gate Bias (V) Better SS at all gate length scale: Better electrostatics, reduced BTBT ~1:10 reduction in minimum off-state leakage ~5:1 increase in gate leakage increased eigenstate 13 VLSI 2014 Performance comparison: 2.5 nm VS 5.0 nm-thick channel 2.8 1200 InAs channel thickness 2.5 nm 5.0 nm 1.2 0.8 400 200 VDS = 0.5 V 0 7 2.0 W/L= 25m/21 m 2 Cox = 4.2 F/cm 6 5 EOT= 0.8 nm 1.5 4 1.0 3 2 0.5 2.5 nm InAs 5.0 nm InAs 0.0 0.2 0.4 Gate Bias (V) 0.6 1 Energy (eV) Freq: 200kHz Carrier Density (/cm ) Gate Length (m) 0 1 12 0.1 2.5 0.0 -0.2 600 x 10 0.4 800 2 1.6 0.0 0.01 Gate Capacitance (F/cm2) eff (cm /s-V) 2.0 1000 2 Peak gm (mS/m) 2.4 0.8 0.6 0.4 0.2 0.0 -0.2 -0.4 -0.6 2.5 nm InAs 5.0 nm InAs 1 2 3 4 5 6 Carrier Density (/cm2) ~1.25 nm EF E1 2.5 nm thick 10 20 30 Position (nm) 0 7 12 x 10 ~2.5 nm EF E1 5 nm thick 10 20 30 Position (nm) 1D-Possion Schrodinger solver (coded by W. Frensley, UT Dallas) 14 VLSI 2014 SS and DIBL vs. Lg (Benchmarking) [2] [4] [3] [7] [1] [6] [5] SSmin (mV/dec.) Open: VDS = 0.1 V 110 100 90 80 Planar [6] [3] [1] Tri-gate or GAA [2] [4] [7] 140 120 DIBL (mV/V) Solid: VDS = 0.5 V 120 100 80 60 40 This work 20 70 This work 60 0.01 0 0.1 1 0 40 80 120 160 Gate Length (nm) Gate Length (m) [1] Lin IEDM 2013,[2] T.-W. Kim IEDM 2013,[3] Chang IEDM 2013,[4] Kim IEDM 2013 [5] Lee APL 2013 (UCSB), [6] D. H. Kim IEDM 2012,[7] Gu IEDM 2012,[8] Radosavljevic IEDM 2009 <80 mV/dec at sub-30 nm Lg and VDS=0.5 V Record low subtheshold swing among any reported III-V FETs. Lowest DIBL among planar-type III-V FETs. 15 VLSI 2014 Peak gm and Ion at fixed Ioff vs. Lg (Benchmarking) 0.50 VDS = 0.5 V 0.45 2.4 2.0 Ion (mA/m) Peak gm (mS/m) 2.8 1.6 [5] [6] [7] [3] [4] [1] 1.2 0.8 0.4 0.0 0.01 VDD = 0.5 V This work 0.40 0.35 0.30 This work J. Lin, IEDM2013 T. Kim, IEDM2013 Intel, IEDM2011 Intel, IEDM2009 J. Gu, IEDM2012 D. Kim, IEDM2012 0.25 0.20 0.15 This work 0.1 Gate Length (m) Ioff = 100 nA/m 0.10 1 100 Gate Length (nm) [1] Lin IEDM 2013,[2] T.-W. Kim IEDM 2013,[3] Chang IEDM 2013,[4] Kim IEDM 2013 [5] Lee APL 2013 (UCSB), [6] D. H. Kim IEDM 2012,[7] Gu IEDM 2012,[8] Radosavljevic IEDM 2009 >2.4 mS/µm peak gm at VDS=0.5 V and sub-30 nm Lg. Highest Ion at Ioff=100 nA/µm and VDD=0.5 V 0.5 mA/µm Ion at sub-30 nm Lg 16 VLSI 2014 Benchmark with 22 nm node Si Fin- and nanowire FET Intel 22 nm Si FinFET IBM 22 nm nanowire FET S. Bangsaruntip et al., IEDM 2013 Jan, IEDM 2012 Intel 22 nm FinFETs (HP) : ~0.5 mA/µm (?) @ VGS=0.5 V, VDS=0.75 V IBM 22 nm nanowire : ~0.4 mA/µm @ VGS=0.5 V, VDS=0.5 V Comparable performance with state-of-the-art Si-FinFETs (nanowire). 17 VLSI 2014 Conclusion Developed vertical spacer to reduce off-state leakage and to improve short channel effect. Integrated sub-1 nm EOT ZrO2 high-k with low Dit Obtained 61 mV/dec at VDS =0.1 V and 1 μm-Lg. Obtained 0.5 mA/μm at Ioff =100 nA/μm and VDD =0.5 V (best reported Ion among any reported III-V MOSFETs) Achieved comparable Ion to state-of-art multi-gate Si-FETs 18 VLSI 2014 Acknowledgment Thanks for your attention! Questions? This research was supported by the SRC Non-classical CMOS Research Center (Task 1437.006). A portion of this work was done in the UCSB nanofabrication facility, part of NSF funded NNIN network and MRL Central Facilities supported by the MRSEC Program of the NSF under award No. MR05-20415. *sanghoon_lee@ece.ucsb.edu 19 VLSI 2014