Technology Development & Design for 22 nm InGaAs/InP-channel MOSFETs

Transcription

1 2008 Indium Phosphide and Related Materials Conference, May, Versailles, France Technology Development & Design for 22 nm InGaAs/InP-channel MOSFETs M. Rodwell University of California, Santa Barbara M. Wistey, U. Singisetti, G. Burek, A. Gossard, S. Stemmer, R. Engel-Herbert, Y. Hwang, Y. Zheng, C. Van de Walle University of California Santa Barbara P. Asbeck, Y. Taur, A. Kummel, B. Yu, D. Wang, Y. Yuan, University of California San Diego C. Palmstrøm, E. Arkun, P. Simmonds University of Minnesota P. McIntyre, J. Harris, Stanford University M. V. Fischetti, C. Sachs University of Massachusetts Amherst , fax

2 Specific Acknowledgements ( Device Team ) Dr. Mark Wistey Lead: Device Development MBE Regrowth Uttam Singisetti Prof. Chris Palmstrøm CBE Regrowth Greg Burek Dr. Erdem Arkun

3 Why III-V CMOS?

4 Why Develop III-V MOSFETs? Silicon MOSFETs continue to scale nm is feasible in production ( or so the Silicon industry tells us...)...16 nm? -- it is not yet clear If we can't make MOSFETs yet smaller, instead move the electrons faster: I d / W g = qn s v I d / Q transit = v / L g III-V materials lower m* higher velocities Serious challenges: High-K dielectrics on InGaAs channels, InGaAs growth on Si True MOSFET fabrication processes Designing small FETs which use big (low m*) electrons

5 Simple FET Scaling Goal: double transistor bandwidth when used in any circuit reduce 2:1 all capacitances and all transport delays keep constant all resistances, voltages, currents All lengths, widths, thicknesses reduced 2:1 S/D contact resistivity reduced 4:1 gs m C gd / W g / g ~ ε g / W ~ vε / T C / W ~ ε L / T C gs W gs, f / g g ~ ε g ox ox If T ox cannot scale with gate length, C parasitic / C gs increases, g m / W g does not increase hence C parasitic /g m does not scale C / W ~ ε L / T sb g c sub

6 FET scaling: Output Conductance & DIBL ( gs C expression neglects the effect of a finite density of states ) Cgs ~ εwε Wg Lg / Tox C d ch ~ ε Wg I d = Q / τ where δ Q = C δ V + gs gs C d ch δ V ds transconductance output conductance Keep L g / T ox constant as we scale L g

7 Well-Known: Si FETs No Longer Scale Perfectly Effective oxide thickness is no longer scaling in proportion to L g (ITRS roadmap copied from Larry Larson's files) High - K gate dielectrics :SiO2 interlayer limits acheivable ( capacitance / area ) Gate capacitance density is not increasing rapidly Output conductance is degrading ( C / I ) is improving only slowly...soon will dominate gate delay parasitic d

8 Highly Scaled MOSFETs: What Are Our Goals? Low off-state current (10 na/μm) for low static dissipation minimum subthreshold slope minimum L g / T ox low gate tunneling, low band-band tunneling Low delay C FET ΔV/I d in gates where transistor capacitances dominate. Parasitic capacitances are ff/μm while low C gs is good, high I d is much better d Low delay C wire ΔV/I d in gates where wiring capacitances dominate. large FET footprint long wires between gates need high I d / W g ; target ~6 ma/μm short transit time alone (low C gs, int ΔV gs /ΔI d ) is not sufficient

9 III-V MOSFETs: Drive Current and CV/I delay

10 III-V CMOS: The Benefit Is Low Mass, Not High Mobility Simple dif drift - diffusion i theory, nondegenerate, far above threshold h : I D c ox W v g injection ( V V ΔV ) gs th where vinjection ~ vthermal = ( kt / * m ) 1/2 I d V th ΔV = v injection L g / μ Ensure that ΔV V << ( V th ) gs V ~ 700 mv V gs low effective mass high currents mobilities above ~ 1000 cm 2 /V-s of little benefit at 22 nm Lg

11 Low Effective Mass Impairs Vertical Scaling Shallow electron distribution needed for high g m / G ds ratio, low drain-induced i d d lowering. Energy of L th well state 2 * 2 ~ L / m Twell For thin wells, only 1st state can be populated. For very thin wells, 1st state approaches L-valley.. Only one vertical state in well. Mimimum ~ 5 nm well thickness. Hard to scale below 22 nm L g.

12 Density-Of-States Capacitance E f Ewell = n s /( nm / 2πh * 2 ) V V = ρ / f well s c dos where c dos = q 2 nm * / 2π h 2 and n is the # of band minima Two implications: - With N s >10 13 /cm 2, electrons populate satellite valleys Fischetti et al, IEDM Transconductance dominated by finite state density Solomon & Laux, IEDM2001

13 Drive Current in the Ballistic & Degenerate Limits More careful analyses by Taur & Asbeck Groups, UCSD; Fischetti Group: U-Mass: IEDM2007

14 Drive Current in the Ballistic & Degenerate Limits J K ma = K 84 μm n=1 V gs V 1V th 3/ 2 eot includes the electron wavefunction depth, where K = * 1/ 2 ( m mo ) * ( 1+ ( c / c ) n ( m / m )) 3/ 2 dos, o 0.6 nm n ox 0.4 nm EOT=0.2 nm o nm m*/m o Inclusive of non-parabolic band effects, which increase c dos, InGaAs & InP have near-optimum mass for nm EOT gate dielectrics

15 Rough Projections From Simple Ballistic Theory 22 nm gate length ff/μm parasitic capacitances Channel EOT drive current intrinsic (700 mv overdrive) gate capacitance InGaAs 1 nm 6 ma/μm ff/μm InGaAs 1/2 nm 8.5 ma/μm 0.25 ff/μm Si 1 nm ma/μm 0.7 ff/μm Si 1/2 nm 5-7 ma/μm 1.4 ff/μm InGaAs has much less gate capacitance 1 nm EOT InGaAs gives much more drive current 1/2 nm EOT InGaAs & Si have similar drive current InGaAs channel no benefit for sub-22-nm gate lengths

16 Device Structure & Process Flow

17 Device Fabrication: Goals & Challenges III-V HEMTs are built like this Source Gate Drain K Shinohara and most...and most III-V MOSFETs are built like this

18 Device Fabrication: Goals & Challenges Yet, we are developing, at great effort, a structure like this N+ source regrowth TiW InGaAs well InP well N+ drain regrowth Why? Source Gate Drain K Shinohara

19 Why not just build HEMTs? Gate Barrier is Low! Gate is low: ~0.6 ev Source Gate Drain K Shinohara Tunneling through sets minimum thickness Emission over limits 2D carrier density E c E F E c E F E well-γ E well-γ At N s = / cm 2, ( E f E ) ~ 0.6 ev c

20 Why not just build HEMTs? Gate also lies under source / drain contacts Source Gate Drain N+ layer widegap layer K Shinohara low leakage: need high under gate low resistance: need low under contacts E c E c E F E F E well-γ N+ cap layer E well-γ

21 The Structure We Need -- is Much Like a Si MOSFET source contact metal gate sidewall gate dielectric drain contact N+ source N+ drain quantum well / channel no gate under S/D contacts substrate high-k gate Overlap between gate and N+ source/drain How do we make this device?

22 Source/Drain Implantation Does Not Look Easy implantation source contact metal gate implantation drain contact InGaAs quantum well substrate Implantation will intermix InGaAs well & InAlAs Annealing can't fix this. Incommensurate sublimation of III vs. V elements during anneal Need ~ 5 nm implant depth & ~ 6*10 19 /cm 3 doping Implanted structures have not shown the necessary low contact resistivity.

23 So, We Are Forming the Source/Drain By Regrowth sidewall metal gate InGaAs quantum well substrate metal gate InGaAs quantum well substrate metal gate Process selected to meet 22 nm ITRS targets But... unlike HEMT process flows, fully established in 1980's......most process steps here are completely new N+ source InGaAs quantum well N+ drain substrate sidewall gate dielectric metal gate source contact drain contact The technology is aggressive and challenging N+ source N+ drain InGaAs quantum well substrate

24 The Required Performance is Formidable ~5 nm thick well 1 nm Insulator EOT Target ~7 700 mv gate overdrive source contact metal gate sidewall gate dielectric drain contact N+ source N+ drain quantum well / channel substrate For < 10% impact on drive current, I D R S R s < 70 mv. < 10 Ω μm ( 20 nm N + extension) (100 Ω / square) = 2 Ω μm 2 (0.25 Ω μm ) /(25 nm wide contact ) = 10 Ω μm

25 Process Development

26 Process Flow with MBE Source/Drain Regrowth

27 Process Flow with MBE Source/Drain Regrowth

28 Gate Dielectrics ALD Al 2 O 3 from IBM (D. Sedana) & Stanford (P. McIntyre) ALD ZrO 2 from Intel (S. Koveshnikov et al, DRC 2008) Al 2 O 3 is more robust in processing. initial process development Process modules being developed for ZrO 2.

29 Gate Definition: Challenges Must scale to 22 nm Dielectric cap on gate for source/drain regrowth well InP well Metal & Dielectric etch must stop in 5 nm channel Semiconductor etch must not etch through 5 nm InP subchannel Process must leave surfaces ready for S/D regrowth

30 Gate Stack: Multiple Layers & Selective Etches Key: stop etch before reaching dielectric, then gentle low-power etch to stop on dielectric

31 Sidewall Formation well well well (starting material) PECVD Si3N4 ICP RIE etch PECVD SiN sidewall deposition, CF 4 / O 2 low power anisotropic RIE etch...sidewall etch must not damage the channel

32 Clean, Undamaged Surface Before Regrowth undamaged InP subchannel (after Al 2 O 3 dielectric etch & InGaAs recess etch )

33 Two Source/Drain Regrowth Processes non-selective area S/D regrowth by Molecular Beam Epitaxy: Wistey InGaAs well InP well well well well well InP well InP well InP well InP well well InP well (starting material) recess etch nonselective regrowth in-situ S/D metal planarize etch strip planarization material electroplate S/D metal selective area S/D regrowth by Chemical Beam Epitaxy: Palmstrøm / Arkun well well InGaAs well InP well InP well InP well well InP well well InP well well InP well (starting material) CBE in-situ etch CBE selective-area regrowth in-situ S/D metal deposition planarize & etch electroplate S/D metal

34 Recess Etch & Regrowth: Inter-Relationships non-selective regrowth selective CBE regrowth InGaAs well InP well well InP well or well InP well well InP well InGaAs/InP composite channel permits selective InGaAs wet-etch, stopping on InP regrowth initiated iti t on InP (desirable?) If regrowth can extend laterally under sidewall, sidewall can be thicker

35 Regrowth interface resistance In addition to the contact & link resistances source contact metal gate sidewall gate dielectric i drain contact N+ source N+ drain InGaAs channel InP sub-channel substrate resistance at the regrowth interfaces is also of concern...

36 Contact & Regrowth Interface Resistance Selective -Arkun Nonselective -Wistey Mo contact N+ RG N+ InGaAs Mo contact well InP well well InP well Re esistance (Ω) SI InP 0.8 Ω - μm TLM B 2 Mo contact 0.5 Ω - μm TLM A N+ RG N+ InGaAs SI InP 2 Mo contact Pad Spacing (μm)

37 TEM of Regrowth: InGaAs on InGaAs (Mark Wistey) InGaAs n+ regrowth HRTEM Interface HAADF-STEM` InGaAs n+ Interface 2 nm

38 Images of MBE Regrowth (dummy sample) Regrowth process is still being de-bugged growth on sidewall: bad! to suppress, grow hotter lateral regrowth under gate: good! Oxide TiW

39 Planarization / Etch-Back Process Ashed-back PR covers S/D contacts Mo etched in SF 6 /Ar dry etch PR strip removes polymers. Mo protects semiconductor from descum plasma.

40 Images of Completed Device

41 Results & Status 1 st working devices: (ISCS submission) MBE S/D regrowth incomplete S/D growth under gate high access resistance low drive current (1 μa/μm!) Cause of Problems related to regrowth on InP subchannel has impacted both MBE & CBE regrowth...and is now resolved New devices now in fabrication...stay tuned

42 InGaAs/InP Channel MOSFETs for VLSI Low-m* materials are beneficial only if EOT cannot scale below ~1/2 nm Devices cannot scale much below 22 nm Lg limits IC density Little CV/I benefit in gate lengths below 22 nm Lg Need device structure with very low access resistance radical re-work of device structure & process flow Gate dielectrics, III-V growth on Si: also under intensive development