1. 11/8/99 SFR Workshop - Plasma 1 Small Feature Reproducibility A Focus on Plasma Etching UC-SMART Major Program Award E. Aydil, N. Cheung, D. Graves, E. Haller, and M. Lieberman, Second Annual Workshop 11/8/99

2. SFR Workshop - Plasma 2 Agenda 8:30 – 9:00 Introductions, Overview / Spanos 9:00 – 10:15 Lithography / Spanos, Neureuther, Bokor 10:15 – 10:45 Break 10:45 – 12:00 Sensor Integration / Poolla, Smith, Solgaard, Dunn 12:00 – 1:00 lunch, poster session begins 1:00 – 2:15 Plasma, TED / Graves, Lieberman, Cheung, Aydil, Haller 2:15 – 2:45 CMP / Dornfeld 2:45 – 3:30 Education / Graves, King, Spanos 3:30 – 3:45 Break 3:45 – 5:30 Steering Committee Meeting in room 775A / Lozes 5:30 – 7:30 Reception, Dinner / Heynes rm, Men’s Faculty Club

3. 11/8/99 SFR Workshop - Plasma 3 Plasma Milestones, Year 1  Develop spatially resolved Langmuir probe for measurements of electron and ion density, electron temperature, and plasma and floating potential.  Initiate measurements of plasma and process uniformity in LAPS using spatially resolved diagnostics.  Simulations of Cu+ and Ar+ on Co surfaces at 55 eV and 175 eV, at angles 0, 30, 45, 60, 75, 85 degrees. Collections of ion reflection and sputter product energy and angular distributions.  Using MD information, simulations of Cu seed layer deposition in high aspect ratio feature (5/1) under conditions relevant to ionized metal physical vapor deposition (IMPVD) tools.  Development of a working inter-atomic potential for the C-F-H system.  Test with trajectory simulations on carbon surfaces.  Refinement of low energy reactive ion source based on helicon technology.

4. 11/8/99 SFR Workshop - Plasma 4 Plasma Milestones, Year 2  Experimental study of plasma and process uniformity using Langmuir probe, OES, and etch-rate metrology.  Design matching networks for real-time control of standing wave ratio on LAPS and conventional tools.  Measurements of H/F radical-surface reactivity at photoresist surfaces.  Measurement of photoresist etch kinetics with Ar+/F/H  Studies of the structure and kinetics of the mixed CxFyHz layer formed during fluorocarbon plasma etching.  First experimental determination of Si TED in isotope heterostructures.  Secondary Ion Mass Spectrometry (SIMS) study of the Si interstitial "wind" (has never been detected directly!)  Effects of n- and p-type doping on Si TED.  Control of TED of Si and dopants with carbon implantation.

5. 11/8/99 SFR Workshop - Plasma 5 SFR Workshop November 8, 1999 Nathan W. Cheung Plasma Processing Effects Research Students: Yonah Cho Adam Wengrow, Changhan Yun Website: www.plasmalab@vivante.eecs.berkeley.edu

6. 11/8/99 SFR Workshop - Plasma 6 Goals 1) Layer Transfer Using Plasma Processing for SMART-Wafer - New applications of plasma processing for system integration of photonics, MEMS, and electronics 2) Modeling and experiments on surface charge accumulation during plasma processing -Multiple-species model developed to investigate effect of high density plasmas (HDP) on wafer charging

7. 11/8/99 SFR Workshop - Plasma 7 Plasma Assisted Materials Processing Lab General purpose HPD tools for plasma research (charging studies, etching, implantation, materials modification) Large selection of species available (BF 3, O 2, H 2, Ar, N 2, F, He, CO 2, H 2 O, plus others). Substrate bias 0-80kV DC, and 0-20 kV pulsed-AC. Plasma Immersion Ion Implanter Large Area Plasma Source Transformer Coupled Plasma Source

8. 11/8/99 SFR Workshop - Plasma 8 (1) Progress vs Milestones Year 1 Single-crystal Si membrane on buried cavities with thickness good uniformity (<0.3%) and surface micro-roughness (<10nm) Plasma surface-activated Si-Si direct bonding and anodic Si-Glass bonding demonstrated. Year 2 Demonstrate GaN Blue-LED layer transfer Integration of GaN LED array and c-Si resonator array in progress

9. 11/8/99 SFR Workshop - Plasma 9 Motivation: Layer Transfer for System Integration Electronic Devices Membrane by Ion-cut Integration of Optics, MEMS, and Electronics GaN LED array by Laser Liftoff GaN 1m1m 100  m Si LED/LASER Oxide Membrane

10. 11/8/99 SFR Workshop - Plasma 10 Ion-Cut Layer Transfer Process H+H+ 1. H + Plasma Implantation 2 Plasma activated Wafer Bonding Handle wafer Si donor Bonding interface 3. Donor wafer cleavage with heat treatment Hydrogen induced Si layer cleavage Hydrogen peak SiO 2 Transferred Si overlayer Handle wafer

11. 11/8/99 SFR Workshop - Plasma 11 Results: SOI fabrication by Plasma Implantation 1m1m 29nm 0nm 2m2m 1m1m 0m0m 2m2m 0m0m AFM scan over 2  m  2  m area of transferred silicon layer surface V bias = -20kV, Dose = 10 17 H/cm 2, T cut = 550  C Transferred silicon layer, 120nm Buried oxide, 200nm RMS roughness ~ 4.1nm (as-cut)

12. 11/8/99 SFR Workshop - Plasma 12 Wafer Bonding with Plasma Surface Treatment Wafer direct bonding 1) Chemical (piranha, HF, RCA) cleaning 2) Plasma surface treatment to increase bond strength L 2y Wafert Razor Bond strength = 8 3 Et 3 y 2 L4L4 Bond Strength Measurement SiO 2 Si XPS Result before and after Plasma Treatment 112 92 Binding Energy (eV) 10810410096 Counts (a.u.) 0 5000 1000 before after

13. 11/8/99 SFR Workshop - Plasma 13 Discussion * Membrane thickness uniformity across 100mm wafer is less than 0.3%. *AFM scan over membrane surface. The as-cut surface micro-roughness is 6nm. Excellent resonator Q-factor expected from uniform, smooth membranes

14. 11/8/99 SFR Workshop - Plasma 14 Results: Oxide Membrane Fabrication Etched Cavity Handle wafer H + 1. H + Implantation & Etching Cavity 2. Wafer Bonding at room temp. 3. Si Cleavage at ~500 O C 4. XeF 2 Etching of Si overlayer Silicon on Oxide Si donor 20  m 100  m 1m1m Optical micrographic top view SEM Cross-section Oxide Membrane

15. 11/8/99 SFR Workshop - Plasma 15 Laser Liffoff and Bonding 1  m 1 m1 m Si Pd-In GaN Pd-In bond h Develop methodology for integration of GaN with Si and GaAs sapphire Approach: 1. Low-T In-Pd bond of GaN to Si or GaAs 2. Laser Liftoff and transfer of GaN Si GaN GaAs GaN sapphire supporting substrate sapphire supporting substrate GaN successfully bonded and transferred from sapphire onto Si Next step Integration of GaN-based LEDs with Si-based ICs for emitter-detector arrays

16. 11/8/99 SFR Workshop - Plasma 16 Future Work (2000-2002) Prototyping of integrated optical, micro-mechanical resonator and active IC devices on a SMART wafer for real-time processing diagnostics LED Thermal Sensor resonator

17. 11/8/99 SFR Workshop - Plasma 17 (2) Progress vs. Milestones Year 1 Develop an ion flux model for multiple-species plasma Preliminary results demonstrating mass attenuation effect with different source apertures Year 2 Work in progress to verify effective mass separation using mass attenuation concept Verification of multiple-species charging model with various antenna ratios and thin dielectrics

18. 11/8/99 SFR Workshop - Plasma 18 Multiple-Species Effects on Charging Plasma composition will effect charging flux (e.g. molecular gas sources CF x +, carrier gas ionization, and plasma instability) A

19. 11/8/99 SFR Workshop - Plasma 19 Single Ion Species Model Electron Current Secondary Electrons Displacement n i ion density u b ion Bohm velocity ssheath width Mion mass V s sheath voltage v e electron velocity V p plasma potential ksecondary electron constant V o applied voltage C s sheath capacitance Sheath Diff. Eq. Ion Current

20. 11/8/99 SFR Workshop - Plasma 20 Multiple Ion Species Model and Results Effective mass concept: Effective ion Bohm velocity concept: BF 3 Plasma Results

21. 11/8/99 SFR Workshop - Plasma 21 Plasma Mass Attenuator Set-up Wafer BiasProcessing Chamber Plasma Magnetic Coils Microwave & Gas Input ECR Plasma Source Sheath Biased Ion Shutter Mass/Energy Spectrometer 3/2 " diameter3/4 " diameter3/8 " diameter

22. 11/8/99 SFR Workshop - Plasma 22 Small Diameter Shutter (3/8”) 300W  -wave power 5 sccm of both Ar and H Multi-cusp magnet mode (240Amps, plasma confined) No wafer bias 300W  -wave power 10 sccm of Oxygen Multi-cusp magnet mode (220Amps, plasma not confined) No wafer bias

23. 11/8/99 SFR Workshop - Plasma 23 Future Work for 2000-2002 Development and verification of a unified model to predict particle flux and charge flux related to HDP processing.

24. 11/8/99 SFR Workshop - Plasma 24 Plasma-Surface Interactions: Vacuum Beam Experiments and Molecular Dynamics Simulations David Graves, Frank Greer, and Cam Abrams University of California Berkeley Department of Chemical Engineering Workshop on SFR Nov. 8, 1999 Berkeley, CA

25. 11/8/99 SFR Workshop - Plasma 25 Motivation Plasma-surface interactions key to controlling plasma effects at feature scale –Most poorly understood part of plasma processing –Complex coupled processes: physical and chemical processes –Events often take place over small length scale and short time scales (e.g. ion-surface) –Surface chemistry affects feature scale and tool scale phenomena How to model feature scale and tool scale processes? –Development of surface process rate expressions –How does PR etch rate depend on ion energy and neutral flux? –How to model processes with simultaneous etch and deposition?

26. 11/8/99 SFR Workshop - Plasma 26 Plasma-Surface Chemistry in Plasma Etch Tools Bulk Plasma Inlet Gas Flow Feature scale Reactor Scale Reactive neutral and ionic species Etch and other reaction product species Wall Interactions Reactive/Etching/ Depositing Species Atomic scale

27. 11/8/99 SFR Workshop - Plasma 27 Schematic of the Beam Apparatus in Cross-Section Surface Reaction Products High Energy Ion Source Microwave Atom Source ICP Atom Source Main Chamber Analysis Section Rotatable Carousel H + + F + Quadrupole Mass Spectrometer Facing Quadrupole Mass Spectrometer

28. 11/8/99 SFR Workshop - Plasma 28 Etch Yield Results for Olin i-line Resist

29. 11/8/99 SFR Workshop - Plasma 29 Effect of Hydrogen Atoms at Surface Etch Yield Effects –Does adding flux of H atoms to photoresist surface reduce the Ar + /F atom etch yield? Abstraction Chemical Kinetics –What are the abstraction probabilities for the following? Incident F abstracting adsorbed H from the photoresist Incident H abstracting adsorbed F from the photoresist

30. 11/8/99 SFR Workshop - Plasma 30 Effect of Large Hydrogen Fluxes During Etching  D = 0  D ~  F

31. 11/8/99 SFR Workshop - Plasma 31 Effect of Other Species on Photoresist Ion-Assisted Etching

32. 11/8/99 SFR Workshop - Plasma 32 Abstraction Kinetics Experiments 1. Expose virgin PR to F atoms 4. Pump out deuterium from system 2. Pump out fluorine from system 5. Expose PR to F atoms 3. Expose PR to D atoms 6. Return to step 2 D F 1. D D F 2. D F QMS QCM 3. F QMS QCM 5. D D 4. D F 6. DD F DD F

33. 11/8/99 SFR Workshop - Plasma 33 Time DF QMS Signal F 2 only Plasma on F and F 2 Time D 2 only Plasma on D and D 2 Surface gains mass as D is replaced by F F does not etch PR, so mass gain saturates Mass loss due to slow PR etching from D atoms D does not abstract F to form DF  DF ~ 0 F abstracts D and replaces it on surface DF signal declines as D is depleted from surface  FD = 0.06

34. 11/8/99 SFR Workshop - Plasma 34

35. 11/8/99 SFR Workshop - Plasma 35 Molecular, Reactive Ion-Surface Interactions: CF x + on Si Many plasma etching applications include molecular ions that can fragment on impact and that can both deposit and etch. CF x + Si Si x C y F z Mixed layer structure especially complicated with depositing species present in the ion (e.g. C or Si). What is thickness, composition, profile within layer? What controls etch rate? What determines transition between etching and deposition? Can we develop and apply a site balance model of the mixed layer? Si x C y F z mixed layer

36. 11/8/99 SFR Workshop - Plasma 36 Molecular Dynamics (MD) Simulation Simulates the motion of a collection of several hundred to several thousand coupled atoms Interatomic potential energy function (PEF) governs the forces atoms exert on each other Numerical integration of the atomic equations of motion until steady state is reached For C-F: Tersoff/Brenner/Tanaka PEF; accounts for neighboring atom influence on bonding Added Si for a Si-C-F PEF (Abrams)

37. 11/8/99 SFR Workshop - Plasma 37 Interatomic Potential (Tersoff/Brenner Formalism) Simulation Scheme Introduce ion above surface Integrate using MD Desorb weakly bound clusters, cool to T s Repeat x2000 i j Molecular Dynamics Simulations of CF x + /Si

38. 11/8/99 SFR Workshop - Plasma 38 Steady-State Si Etching Through a Fluorocarbon Overlayer: 100 eV CF 3 +

39. 11/8/99 SFR Workshop - Plasma 39 Mass Balance Model Si Si x C y F z C,FC,F Si,C,F  z 01.0 Si C F MD Model

40. 11/8/99 SFR Workshop - Plasma 40 Mass Balance Model, continued = Sites/cm 3 ; L = depth; J + = ions/(Acm 2 )

41. 11/8/99 SFR Workshop - Plasma 41 Mass Balance Model, continued Comparison of Model to MD Simulation Basis: J + = 5 mA/cm 2

42. 11/8/99 SFR Workshop - Plasma 42 Second Year Plans Measure effects of surface temperature and surface type on photoresist etch/abstraction chemistry in beam experiments Add neutral radical (F, CF x ) impact to MD simulations Further development of phenomenological models for fluorocarbon chemistry

43. 11/8/99 SFR Workshop - Plasma 43 TRANSIENT ENHANCED DIFFUSION (TED) IN ISOTOPICALLY ENGINEERED SILICON Hartmut A. Bracht, Cynthia B. Nelson and Eugene E. Haller University of California at Berkeley and Lawrence Berkeley National Laboratory SFR-UCB SMART, Nov. 8, 1999

44. 11/8/99 SFR Workshop - Plasma 44 Motivation Transient Enhanced Diffusion (TED) in Silicon Semiconductor Isotope Heterostructures Future Work

45. 11/8/99 SFR Workshop - Plasma 45 COLLABORATORS Joel W. Ager III, LBNL Steven Burden, ISONICS Corp., Golden, CO Manuel Cardona, MPI Stuttgart, Germany Nick Cowern, Phillips Eindhoven, Holland Hans Gossmann, Bell Labs, Lucent Techn., Murray Hill, NJ William Hansen, LBNL

46. 11/8/99 SFR Workshop - Plasma 46 MOTIVATION Shrinking device dimensions: the SIA roadmap Doping techniques: ion implantation, diffusion A new tool: isotopically enriched Si

47. 11/8/99 SFR Workshop - Plasma 47 6515 J. Appl. Phys., Vol. 81, No. 10, 15 May 1997Appl. Phys. Rev.: Chason et al. The in-plane dimension reductions (e.g. gate length) demand equivalently shallower implantation and diffusion depths.

48. 11/8/99 SFR Workshop - Plasma 48 Dominant Diffusion Mechanisms in Silicon -Interstitial assisted diffusion (kick-out mechanism) A S + I A I -Vacancy assisted diffusion (Frank-Turnbull or dissociative mechanism) A S A I + V (A = Si or impurity)  Self-diffusion coefficient D SD :

49. 11/8/99 SFR Workshop - Plasma 49 SELF-DIFFUSION IN SILICON Native point defects: self-interstitials (I) and vacancies (V) Si-tracer self-diffusion coefficient:

50. 11/8/99 SFR Workshop - Plasma 50 TRANSIENT ENHANCED DIFFUSION (TED) The origin of TED: Si interstitial “wind” Boron clustering The “+1” rule The effect of carbon

51. 11/8/99 SFR Workshop - Plasma 51 6516 J. Appl. Phys., Vol. 81, No. 10, 15 May 1997Appl. Phys. Rev.: Chason et al.

52. 11/8/99 SFR Workshop - Plasma 52 6522 J. Appl. Phys., Vol. 81, No. 10, 15 May 1997 Appl. Phys. Rev.: Chason et al.

53. 11/8/99 SFR Workshop - Plasma 53 6036 J. Appl. Phys., Vol. 81, No. 9, 1 May 1997 Stolk et al.

54. 11/8/99 SFR Workshop - Plasma 54 6037 J. Appl. Phys., Vol. 81, No. 9, 1 May 1997 Stolk et al.

55. 11/8/99 SFR Workshop - Plasma 55 6041 J. Appl. Phys., Vol. 81, No. 9, 1 May 1997 Stolk et al.

56. 11/8/99 SFR Workshop - Plasma 56 SEMICONDUCTOR ISOTOPE HETEROSTUCTURES Isotopically engineered semiconductors: the case for self-diffusion studies Self-diffusion in Si, Ge, GaAs, AlGaAs, GaP and GaSb

57. 11/8/99 SFR Workshop - Plasma 57 ISOTOPE MULTILAYER STRUCTURES natural Si = 92.23% 28 Si + 4.67% 29 Si + 3.10% 30 Si Post Cold War collaborations with Russian and Ukranian scientists have given us access to highly enriched 28 Si (99.95%) Lawrence Semiconductor Research Corporation in Tempe, AZ has grown undoped and doped multilayer structures: nat Si (2  m) : 28 Si (2  m) : nat Si (substrate wafer)

58. 11/8/99 SFR Workshop - Plasma 58 as-grown 30 Si profile diffused 30 Si profile (1322°C/30min) [ 30 Si ] (cm -3 ) nat. Si = 92.23% 28 Si + 4.67% 29 Si + 3.10% 30 Si

59. 11/8/99 SFR Workshop - Plasma 59 Si Self-Diffusion Coefficient H. Bracht, E.E. Haller and R. Clark-Phelps, Phys. Rev. Lett. 81(2), 393 (1998).

60. 11/8/99 SFR Workshop - Plasma 60 FUTURE WORK: TED RESEARCH MILESTONES YEAR 2 First experimental determination of Si TED in isotope heterostructure Secondary Ion Mass Spectroscopy (SIMS) studies of Si interstitial wind (has never been detected directly!) Effects of n-type and p-type doping on Si TED Control of TED of Si and dopants with carbon implantation

61. 11/8/99 SFR Workshop - Plasma 61 Plasma Sources for Small Feature Reproducibility A.J. Lichtenberg †, M.A. Lieberman †, A.M. Marakhtanov † and Yaoxi Wu * Departments of † Electrical Engineering and Computer Sciences * Materials Science and Engineering University of California, Berkeley

62. 11/8/99 SFR Workshop - Plasma 62 Motivation Scaling inductive sources to larger sizes (LAPS) Usual external coils give inherently nonuniform power deposition over large areas - use internal coil Antenna standing wave effects lead to plasma nonuniformities - employ traveling wave antenna Controlling instabilities in inductive sources (TCP) Instabilities are observed in commercial inductive discharges with electronegative gas feeds – determine stable/unstable operating parameter windows (power, pressure, gas feed mix, etc) – develop a theory of the instability to learn how to control it

63. 11/8/99 SFR Workshop - Plasma 63 Large Area Plasma Source Antenna coil embedded in the plasma Eight quartz tubes threaded by copper antenna tubes Eight vertical gas feed lines with equally spaced pin- holes 71 cm x 61 cm plasma area I rf

64. 11/8/99 SFR Workshop - Plasma 64 Progress vs. Milestones Year 1 Develop spatially resolved Langmuir probe for measurements of plasma properties (done) Initiate measurements of plasma uniformity in LAPS using spatially resolved diagnostics (done) Year 2 Experimental study of plasma and process uniformity using Langmuir probe and OES (on-going) Study effect of alternative coil configurations on plasma density profile (on-going) LAPS

65. 11/8/99 SFR Workshop - Plasma 65 LAPS Driving Circuit (a) Power supply (b) Matching network (c) Tuning network (d) Antenna and plasma system Traveling waves are launched by a tuning network

66. 11/8/99 SFR Workshop - Plasma 66 Configurations Investigated 6 rods powered4 rods powered8 rods powered

67. 11/8/99 SFR Workshop - Plasma 67 Six Rods Powered – Experimental Results Density profile for p = 5.8 mTorr, varying power Density profile for P  1211 W, varying pressure

68. 11/8/99 SFR Workshop - Plasma 68 Eight Rods Powered Density profile - without tuning Density profile - with tuning

69. 11/8/99 SFR Workshop - Plasma 69 Capacitive 450 W Unstable 595 W Inductive 900 W End-On View of 5.4 mTorr SF 6 Discharge TCP Experiment Plasma InductiveCoil Langmuir Probe PMT Video

70. 11/8/99 SFR Workshop - Plasma 70 Progress vs Milestones TCP Source Year 1 Mapped inductive plasma instability regions inside TCP source. (Done) Observation of optical emission modulations due to instabilities. (Done) Develop theoretical model of instabilities (Done) Year 2 Add additional TCP plasma diagnostics. (Done) Further develop theory for electronegative instabilities in inductive discharges. (on-going) Bring high flow Lam oxide etch prototype on-line. (on-going)

71. 11/8/99 SFR Workshop - Plasma 71 Langmuir probe current time history p SF6 = 5.2 mTorr, 500 W PROBE CURRENT,  A TIME V=28 V V=14 V V=10 V V=0 V=-14 V V=-28 V -300 - 460 - 620 - 15 - 30 - 45 - 1 - 9 - 17 00.2 ms 0 0 0 00 14 10 6 29 25 21 44 40 36

72. 11/8/99 SFR Workshop - Plasma 72 Range of Instability in SF 6 Discharge Instability

73. 11/8/99 SFR Workshop - Plasma 73 Fluctuation of Light Emission vs. Power SF 6 10 mTorr 25 mTorr 90 mTorr 80 mTorr 65 mTorr 50 mTorr 40 mTorr

74. 11/8/99 SFR Workshop - Plasma 74 Theoretical Model of Instability 10 -4 10 -2 10 0 10 0 1 n e n - 0 0.4 n e 0 5 n - 0 10 T e 0 20  0 0.4  - 012345678910 0 1 t,ms optical emission p SF6 = 5 mTorr, I COIL =8.2 A dn e /dt=0 dn - /dt=0 (10 10 cm -3 )

75. 11/8/99 SFR Workshop - Plasma 75 Future Work LAPS Investigate material etching rate and uniformity; e.g., oxygen etching of photoresist, & correlate with plasma profiles (Year 1) Optimize etching uniformity using alternative tuning networks to control antenna standing wave ratios (Year 1) Investigate real time control of plasma uniformity using the tuning network. (Year 2) TCP Continue work, characterizing the influence of the matching network and power supply on the instability (Year 1) Initiate instability studies on a Lam high flow oxide etch prototype TCP reactor (Year 1) Incorporate matching network in instability theory (Year 2)