1. ITRS Spring Conference 2009 Brussels, Belgium 1 Work in Progress: Not for Distribution 2009 ITRS Emerging Research Materials [ERM] March 18-20, 2009 Michael Garner – Intel Daniel Herr – SRC

2. ITRS Spring Conference 2009 Brussels, Belgium 2 Work in Progress: Not for Distribution ERM Outline Scope Introduction Difficult Challenges Challenges for Multi-application ERM (Back-up?) Materials for Alternate Channel CMOS (PIDS & ERD) –Critical Assessment ERM for Beyond CMOS Logic (ERD) Materials for Memory Devices –Critical Assessment ERM for Lithography –Resist (pixilated, Multi photon resist, novel) –Self Assembled Materials –Transition Table (Molecular glasses, evolutionary resist macromolecular design, etc.) ERM for FEP & PIDS –Deterministic Doping –Self Assembly for Selective Deposition & Etch ERM for Interconnects ERM for Assembly & Package ERM ESH Research Needs ERM Metrology Needs ERM Modeling Needs

3. ITRS Spring Conference 2009 Brussels, Belgium 3 Work in Progress: Not for Distribution X-cutting Challenges LDM –Control of placement & direction –Control of nanostructure, properties & macro properties Contact & Interface issues Self Assembled Materials –Control of placement, defects, and registration Complex metal oxides –Control of properties, interfaces, defects, and moisture degradation

4. ITRS Spring Conference 2009 Brussels, Belgium 4 Work in Progress: Not for Distribution III-V Ge Alternate Channel Partition Proposal ERM Materials, Interfaces & Process Issues & Challenges Critical Assessment of Materials & Integration Capabilities ERD Integrated Device Performance Assessment & Challenges (For different structures surface, buried channel, etc.) Critical Assessment of Device Performance PIDS III-V & Ge Potential Solution SiGe P-FET with Si N-FET Collaborate with ERD on device Readiness FEP Potential Solution: SiGe P-FET with Si N-FET III-V & Ge Potential Solution Track III-V & Ge Issues

5. ITRS Spring Conference 2009 Brussels, Belgium 5 Work in Progress: Not for Distribution ERM Device Materials Outline Emerging Logic Materials Alternate Channel Materials for Equivalent Scaling III-V Epi Materials Ge Epi Materials Graphite and Graphitic Materials Nanowires Carbon Nanotubes Critical Assessment Contact Materials (?) Beyond CMOS Logic Materials Spin Materials Ferromagnetic Semiconductors (III-V & Oxides) Magnetoelectric Materials (Alloys, Heterostructures, superlattices) Spin Injection Materials (Low barrier ferromagnetic metals, half metals, etc) Spin Tunnel Barriers (MgO, etc) Semiconductor & nanostructure Spin Transport properties (Si, Ge, Graphene, CNT, etc), Strongly Correlated Electron State Materials (Metal-Insulator) Molecular Devices (?) Emerging Memory Materials Molecular Devices (?) Complex Metal Oxides Critical Assessment (?)

6. ITRS Spring Conference 2009 Brussels, Belgium 6 Work in Progress: Not for Distribution Materials for Alternate Channel Logic Alternate Channel Materials for Equivalent Scaling III-V Epi Materials (John Carruthers) Ge Epi Materials (John Carruthers) Graphite and Graphitic Materials (Jeff Peterson & Daniel Bensahel) Nanowires (Ted Kamins) Carbon Nanotubes (Jean Dijon, Miyamoto-san, Awano-san) Critical Assessment Contact Materials (?)

7. ITRS Spring Conference 2009 Brussels, Belgium 7 Work in Progress: Not for Distribution III-V & Ge Key Messages Gate Dielectric Growth techniques are being developed –Current Approaches (III-V): MBE Growth of III-V/Ga2O3/GdGaO Stack (Freescale) As Cap/ In situ As decap +ALD HfO2 (Stanford) NH4OH-ALD Al2O3 or HfO2 on III-V (Purdue) InAlAs Barrier (MIT) –Current Approaches (Ge): GeOxNy Nitridation (Stanford) Ozone Oxidized Ge + ALD High κ dielectric HfO2 (Stanford) LaGeOx-ZrO2(Ge) High K (Dual Logic) Controlling surface oxide formation is critical for control of interface states –Control of interface stochiometry, structure and defects is critical –GeOx stochiometry control affected by growth temperature

8. ITRS Spring Conference 2009 Brussels, Belgium 8 Work in Progress: Not for Distribution III-V & Ge Key Messages Ge dopant activation requires high temperature –Incompatible with III-V process temperatures S/D Contact Formation Current Approaches: –Ge P-MOS: Boron with many ohmic metal contact options N-MOS: Dopants have high diffusivity & metals schottky barriers –III-V W contact/InGaAs cap/InAlAs (MIT) Are barriers needed to keep dislocations out of the channel?

9. ITRS Spring Conference 2009 Brussels, Belgium 9 Work in Progress: Not for Distribution III-V Ge Heteroepitaxy Challenges Reduction of dislocation densities Control of stress in III-V & Ge integrated on Si –Ultrathin films –Heterostructures to reduce defects Effect of antiphase domains on carrier transport Identify a crystal orientation that favors epitaxy and interface states.

10. ITRS Spring Conference 2009 Brussels, Belgium 10 Work in Progress: Not for Distribution Graphene Challenges & Status Ability to deposit graphene on appropriate substrates Producing a bandgap –Fabricating Narrow Graphene Lines –Applying a high electric field to bi-graphene Achieving high mobility in an integrated structure Achieving a high on-off conduction ratio

11. ITRS Spring Conference 2009 Brussels, Belgium 11 Work in Progress: Not for Distribution Graphene Deposition CVD of Graphene on Ni, Pt, and Ir –Graphene is strongly bonded to Ni, but has a lattice match –Graphene deposited on Pt is not distorted, is not lattice matched, but is weakly bonded SiC decomposition –Issue: High process temperature (>1100C) Exfoliation Techniques –Graphene Oxide Decomposition (Mobility <1000cm2/V-sec) Oxidation process produced islands of graphene surrounded by disordered material (hoping conduction) –Try less aggressive oxidation process –Solvent exfoliation Solvents capable of separating graphene sheets are difficult to evaporate (high boiling point) –Tape exfoliation

12. ITRS Spring Conference 2009 Brussels, Belgium 12 Work in Progress: Not for Distribution Producing a Graphene Bandgap Fabricating Narrow Graphene Lines –Requires patterning sub 20nm lines –Edge defect control is challenging (Eg & Mobility) Applying a high electric field to bi-graphene –Field ~1E7 V/cm

13. ITRS Spring Conference 2009 Brussels, Belgium 13 Work in Progress: Not for Distribution Graphene Mobility Mobility on substrates is reduced Graphene Oxide Mobility –Degraded by disordered regions

14. ITRS Spring Conference 2009 Brussels, Belgium 14 Work in Progress: Not for Distribution Nanowire Challenges Based on 2007 ERM the key challenges were: Position the nanowires during growth or reposition them after growth at the desired location and with the desired direction Provide performance exceeding patterned materials CMOS compatible catalysts. Additional –Integration of dopants –Gate Dielectric interface passivation

15. ITRS Spring Conference 2009 Brussels, Belgium 15 Work in Progress: Not for Distribution Nanotube Challenges Control of: –Location –Direction –Bandgap (Chirality & Diameter) –Carrier type & concentration Gate Dielectric Deposition Contact Resistance

16. ITRS Spring Conference 2009 Brussels, Belgium 16 Work in Progress: Not for Distribution Nanowire 2009 Potential Technology Advantages Status of demonstration Key Challenges & Status Position the nanowires during growth or reposition them after growth at the desired location and with the desired direction Provide performance exceeding patterned materials CMOS compatible catalysts. Additional –Integration of dopants –Gate Dielectric interface passivation

17. ITRS Spring Conference 2009 Brussels, Belgium 17 Work in Progress: Not for Distribution Critical Assessment

18. ITRS Spring Conference 2009 Brussels, Belgium 18 Work in Progress: Not for Distribution Beyond CMOS M. Garner Molecular State (Alex Bratkovski & Curt Richter) Spin Materials (U-In Chung / Kang Wang, Nihey-san) –Ferromagnetic Semiconductors (III-V & Oxides) –Magnetoelectric Materials (Alloys, Heterostructures, superlattices) –Spin Injection Materials (Low barrier ferromagnetic metals, half metals, etc) –Spin Tunnel Barriers (MgO, etc) –Semiconductor & nanostructure Spin Transport properties (Si, Ge, Graphene, CNT, etc),

19. ITRS Spring Conference 2009 Brussels, Belgium 19 Work in Progress: Not for Distribution ERM Beyond CMOS Scope: 2009 2007Transition InTransition Out2009 Molecules & Interfaces Transition out? Inadequate progress Status FM Semiconductors Curie Temp Table Tc Graph FM Oxide Semiconductors Status, Table or Graph Spin Semiconductor Status Spin Tunnel Materials Status Magnetoelectric materials & structures Status Low barrier spin injection materials Status

20. ITRS Spring Conference 2009 Brussels, Belgium 20 Work in Progress: Not for Distribution Spin Materials Ferromagnetic III-V (Mn) semiconductors have verified Curie temperatures 100-200K –Carrier mediated exchange Nanowires of GeMn have reported ferromagnetic properties at 300K+, but carrier mediated exchange with gated structure is difficult to verify Oxides doped with transition metals have ferromagnetic properties –Ferromagnetism can be controlled with carrier doping, but it isn’t clear whether this can be modulated with electric fields –Ferromagnetism is proposed to be in an impurity band vs. the oxide bands. –It is not clear whether this is useful for device applications

21. ITRS Spring Conference 2009 Brussels, Belgium 21 Work in Progress: Not for Distribution Spin Materials (Cont.) Spin Tunnel Barrier Materials –MgO crystalline material is the best spin selective tunnel barrier to date May work with a limited number of materials due to lattice match requirement –Films must be ~9A thick –Al2O3 films work, but with much lower selectivity Multiferroics –Need higher coupling coefficient (Electrical to Magnetic) Nanostructures Heterostructures –BaFeO3 has ferroelectric & antiferromagnetic properties coupled Limited degrees of freedom & low coupling

22. ITRS Spring Conference 2009 Brussels, Belgium 22 Work in Progress: Not for Distribution Strongly Correlated Electron State Materials (For Spin Logic)(Kariya-san) Potential Physics of Interest –Competing Non-Ferromagnetic/ Ferromagnetic Phase Transitions Nanoscale phase segregation near phase transition compositions Magnetic fields can convert the phases (multi Tesla) –Insulator to Ferromagnetic Metallic state Carrier doping may be able to cause the transitions –Electric Field –Issues: Most phase transitions occur below room temperature Phase transitions may be first order “Pure” phases may not exist (Nanoscale phase segregation)

23. ITRS Spring Conference 2009 Brussels, Belgium 23 Work in Progress: Not for Distribution Strongly Correlated Electron State Heterointerfaces (For Spin Logic)(Kariya-san) Oxide heterointerfaces don’t appear to have interface pinning Interfacial reconstruction at charged interfaces –Charged interfaces result in interface reconstruction –Hole doped interfaces are “metallic”

24. ITRS Spring Conference 2009 Brussels, Belgium 24 Work in Progress: Not for Distribution Phase Competition

25. ITRS Spring Conference 2009 Brussels, Belgium 25 Work in Progress: Not for Distribution Nanoscale Phase Segregation

26. ITRS Spring Conference 2009 Brussels, Belgium 26 Work in Progress: Not for Distribution 1 st Order Phase Transitions Coexistence of competing phases

27. ITRS Spring Conference 2009 Brussels, Belgium 27 Work in Progress: Not for Distribution Heterostructures

28. ITRS Spring Conference 2009 Brussels, Belgium 28 Work in Progress: Not for Distribution Heterostructures Surface reconstruction hole generation No polar discontinuity except at STO/LaAlO3 interface

29. ITRS Spring Conference 2009 Brussels, Belgium 29 Work in Progress: Not for Distribution ERM Beyond MOS Memory: 2009 2007Transition InTransition Out2009 Complex Metal Oxide Resistance Change Status Oxides & Interfaces FE Memory Status Nanotube for Nanomechanical memory Status Molecules & interfaces for Molecular Memory Transition out?Status MRAM MaterialsStatus Ionic Transport Materials

30. ITRS Spring Conference 2009 Brussels, Belgium 30 Work in Progress: Not for Distribution Oxide Memory Materials Multiple mechanisms proposed –Phase transformation –Change of polarization alignment (E or H) –Nanofilament formation –Cation migration –Anion Migration Oxygen Vacancies Should we assess the consequences of the different mechanisms? (Scaling & Reliability) –Resistance Change –Ferroelectric FET & Barrier –Mott FET

31. ITRS Spring Conference 2009 Brussels, Belgium 31 Work in Progress: Not for Distribution Mechanism Assessment Cation migration (Ag, Cu) Filament formation Anion migration Vacancy Migration Charge Trapping (Vacancies or defects) Electronic Phase Transition –Mott FET

32. ITRS Spring Conference 2009 Brussels, Belgium 32 Work in Progress: Not for Distribution C. Dubourdieu - LMGP-CNRS & D. Bensahel - STMicroelectronics - France 32 In December 2007, the journal Science considered the recent discoveries emerging from oxide interfaces as one of the 10 breakthrough of the year 2007 Oxides interfaces -New properties arise from surface, electronic or orbital reconstructions. (Stacking for example two insulating compounds such as LaAlO 3 and SrTiO 3 can lead to a metallic or superconducting LaAlO 3 /SrTiO 3 interface) -Interfaces in superlattices can change the nature of the coupling between competing instabilities and produce new properties. (superlattices combining the proper ferroelectric PbTiO 3 and the paraelectric SrTiO 3 compounds behave like a prototypical improper ferroelectric due to interface coupling based on rotational distortions).

33. ITRS Spring Conference 2009 Brussels, Belgium 33 Work in Progress: Not for Distribution Memory & Oxides DeviceMaterialsMaterial Mechanisms Interface Mechanisms Phase ChangeNiO, TiO2, IrO2-NiO, CuO 1.Nanofilament formation 2.Anion migration CuO electrode interaction ElectrochemicalCu & Ag with Oxides or sulfides Cu or Ag ion migration though the oxide Electrochemical Charge Trapping

34. ITRS Spring Conference 2009 Brussels, Belgium 34 Work in Progress: Not for Distribution Memory & Perovskites DeviceMaterialsMaterial Mechanisms Interface Mechanisms FE FETPZT, BFOFerroelectric Polarization Electrodes can degrade reliability Pt: Hydrogen: SRO preferred FE BarrierPZT, BFO, etc. Ferroelectric Polarization changes Schottky Barrier height or charge TBD Mott FETPCMO, LCMO, STO Carrier injection drives a metal insulator transition TBD

35. ITRS Spring Conference 2009 Brussels, Belgium 35 Work in Progress: Not for Distribution Perovskite Challenges Ferroelectrics: Electrode Interactions –Pt: Hydrogen ion generation degrades polarization –SRO: Increases resistance Strongly Correlated Electron Material Challenges (Mott M-I Transition) –Nanoscale phase segregation may suppress sharp transition –Materials are very sensitive to stress (Piezo effects) Selection of substrate & interface layers –“Disorder” can dramatically reduce critical temperatures

36. ITRS Spring Conference 2009 Brussels, Belgium 36 Work in Progress: Not for Distribution Molecular Devices Top contact formation is still a significant issue Determining that switching is due to the molecular energy levels is difficult