Cluster Ion Beam Modification of Material Surfaces

1
Cluster Ion Beam Modification of Material Surfaces

• Hui Chen
• Texas Center for Superconductivity at University
  of Houston
• Presented at University of North Texas on May 02,
  2006

2
Outline

• Motivations
• Experiments
• ? Damage buildup induced by small cluster ion
  beam irradiation
• ? Large Ar gas cluster ion surface modification
• ? Large Ar gas cluster ion beam assisted thin
  film deposition
• Theoretical modeling
• Conclusions
• Acknowledgement

3
Motivation I
? Ion implantation is an indispensable
technique in semiconductor device
fabrication.
? Continuing shrinking of junction depth has
reached the limit of ion implantation technique.
? Space charge effect J ? KV3/2(Child, 1911
Langmuir, 1913).
4
Cluster Ion Beam Ultra-shallow Junction Formation
? Cluster ion beam implantation is one promising
alternative solution (B10H14, Goto 1997).
? Low energy effect Veff Vacc/10 ? High
mass to charge ratio ? Minimize space
charge effect ? Minimize surface charge
effect
? Neither collision mechanism nor their
influence on the type or number of induced
defects is clear.
5
Source of Negative Ion Cesium Sputtering (SNICS)
6
Cu Cluster Mass Spectra
7
RBS/Channeling Spectra
8
Enhanced Defect Productionvs. Cluster Size
9
Enhanced Defect Production vs. Dose
10
Defect Production in Collision Cascades by
Monomer Ion
K. Nordlund et al., Phys. Rev. B 80, 4201 (1998).
11
Synergism Overlapping Partial Damage
Permanent Damage
Vd
Vd
Vcas
Vcas
12
Damage Evolution
13
Damage Induced by One Cluster of Size n
14
Simulation Result- Enhancement factor vs. Copper
cluster Size
15
Simulation Result- Enhancement factor vs. Carbon
cluster Size
16
Critical Cluster Size for Maximum Enhancement
Factor
17
Simulation Result- Enhancement factor vs. dose
18
Simulation Result- Enhancement factor vs. dose
19
Conclusion I

• The radiation damage induced by cluster ion
  bombardment is not the linear superposition of
  damages induced by constituent atoms.
• The overlap model agrees well with the
  experimental results.

20
Motivation II
? Mobility degradation due to impurity
scattering in sub-90 nm node CMOS device.
? Strained-Si MOSFETs on strain-relaxed SiGe
exhibit enhanced mobility of holes and electrons
(Ismail, 1991 1993 Mii, 1991).
? Large lattice mismatch (4.2) leads to
crosshatched pattern and rough surface.
? Post-growth chemical mechanical polishing
(CMP) yields better performance (Sugii 2002
Sawano 2003).
? CMP is a slow and tedious process with
surface contamination.
21
Schematic of Cluster ion solid Interaction
Z. Insepov, I. Yamada, Nucl. Instr. Meth. B
112, 16 (1996).
22
(No Transcript)
23
Cross-section of strain relaxed Si0.7Ge0.3
Virtual Substrate
24
As-grown Si0.7Ge0.3 Morphology
25
Normal Incidence dose 5 x 1015 clusters/cm2,
Rav2.8 nm
26
Normal Incidence dose 1 x 1016 clusters/cm2,
Rav 0.7 nm
27
Lattice Disorder Study
28
Lattice Damage Annealed at 1000oC/10s
29
Crosshatch Pattern Reappear after Annealing
30
Post-smoothing Glancing Angle Etch Removes
Surface Damage
31
Glancing Angle Etching Remains Surface Smoothness
32
Ripple Formation under Off-normal Incidence ?
10o
33
Ripple Formation under Off-normal Incidence ?
30o
34
Ripple Formation under Off-normal Incidence ?
50o
35
Ripple Formation under Off-normal Incidence ?
70o
36
Distribution of Sputtered Substrate Atoms under
Normal and Off-normal GCIB Irradiation
I. Yamada et al., Mater. Sci. Eng. R 34, 231
(2001).
37
Assumed Schematic Diagram of Cluster Solid
Interaction
38
Surface Erosion for a Uniform Hopping Length a?
Dispersed into Vacuum
Surface diffusion
Hopping and redeposition
Sputtered into vacuum
39
Surface Diffusion Contribution
40
Continuum Surface Erosion Equation
41
Generation of Mobile Substrate Atoms Nm(x)
Ninc(x) number of incident cluster ions per unit
area per unit time Y(x) number of mobile atoms
generated per impact
42
Small Slope Approximation
Surface diffusion
Hopping redeposition
Sputtered into vacuum
43
Non-uniform Distribution of Hopping Length
Define p(a?)da? is the probability the jumping
length between a?, a?da?
44
Surface Stability Analysis
45
Normal Incidence Smoothening Effect (? 0o)
46
Off-normal Incidence Ripple Formation (0o lt ? ltlt
90o)
47
Conclusion II

• Crosshatch patterned and rough Si0.7Ge0.3
  surfaces were polished to atomic smoothness under
  Ar-GCIB irradiation at normal incidence.
• Nano-ripple structured Si0.7Ge0.3 surface were
  obtained under off-normal Ar-GCIB irradiation.
• A mesoscopic model was developed to explain
  qualitatively the surface smoothing effect under
  normal incidence and ripple formation under
  off-normal incidence.

48
Roadmap of HDD
49
Deposition method
50
Effect of C60/Ar Cluster Ratio
51
Breit-Wigner-Fano Fitting
52
Have to Be Simultaneous Bombardment
53
Hardness Measurement
54
Surface Roughness
55
Thickness measurement
56
23 Å Carbon Characterization
57
Field Emission Property
58
Conclusion III

• Ultra-thin, smooth and very hard DLC films were
  formed using Ar GCIB-assisted C60 evaporation.
• New method using non-Rutherford back scattering
  was developed to measure thickness of ultra-thin
  carbon film.
• Flexible field emitters with high emission
  current density has been realized using DLC films
  deposited room temperature .

59
Acknowledgement

• Wei-Kan Chu
• Jiarui Liu
• Ki Ma
• Xuemei Wang
• Irene Rusakova
• Milko N. Iliev
• Nacer Badi

60
Total Damage vs. Dose
61
Review of Monomer Ion Solid Interaction (after
Robinson 1965)
62
Monomer Ion Solid Interaction
Here ? lt 1 and ?(E) is the amount of PKA energy
not lost to excitation (after Robinson 1965
Sigmund, 1969).