New Generation Nuclear Microprobe Systems: A new look at old problems

1
New Generation Nuclear Microprobe SystemsA new
look at old problems
7th International Conference on Nuclear
Microprobe Technology and Applications, Cité
Mondiale, Bordeaux, France, September 11 2000

• ByDavid N. Jamieson
• Microanalytical Research Centre
• School of Physics
• University of Melbourne
• Parkville, 3010
• AUSTRALIA

2
Electron Emission from Surfaces

• CVD B-doped diamond films are electrically
  conductive
• Diamond has a negative electron affinity
• Potential applications as a cold cathode electron
  emitter
• Measure g number of electrons emitted from
  surface per ion impact
• Measure g 15 to 30 (metals g 1.5)

Incident ion
electrons
H
H
H


I
RBS
3
Filiform Corrosion in Aluminium

• Filiforms grow under breaches in the
  anticorrosion coating on Al
• 3 MeV H PIXE data confirms role of Cl in
  catalysing growth of the filiform

Anticorrosion layer removed
Filiform growth
C- RBS
Al- RBS
O - RBS
Cl - PIXE
Al - RBS He
4
Menkes Syndrome revisited

• Menkes Syndrome is a Cu deficency genetic
  disorder.
• The gene responsible for the disorder has now
  been mapped.
• Pathways for Cu metabolism within cells can now
  be controlled and studied with unprecedented
  precsion.
• But can the nuclear microprobe cope?
• Need to resolve Cu distributions within single
  cells to a spatial resolution of sub-micron.
• Images here are by indirect immunofluorescence
  from anti-body labelled Menkes protein.
• Cells are less than 10 micons in width

5
Outline

• The quest for superior spatial resolution in the
  Nuclear Microprobe Why has the probe resolution
  stalled at 1 micron for 2 decades?
• Some new insights provide possible pathways to
  future progress
• Introduction to elementary ion optics
• Chromatic aberration - not a problem?
• Spherical aberration - not too much of a problem?
• Stray magnetic fields - definitely a problem
• Demagnification - the way forward
• Ion source brightness - small advances to be
  welcomed
• A review of the next generation systems
• Conclusion
• (Topics not addressed
• High efficiency detectors, fast DAQs to handle
  high intensity beams,
• specimen damage,channeling convergence angle)

6
Chip feature size and NMP resolution
Size (micron)
Year
7
Spatial Resolution RequiredApplications
published at the Last Conference 1998
1 mm wall
Pile up
8
Introductory Ion Optics
ImagePlane
Aperture Plane
Object Plane
(xo , ?o , yo , ?o , ?o )
Lens System
(xi , yi )
xi (x/x)xo (x/?)?o (x/?? )?o?o (x/?)?o3
(x/? ? 2)?o ?o2 yi (y/y)yo (y/?)? o (y/??
)?o?o (y/?)?o3 (y/? 2?)?o2?o 2
plus higher order terms
9
How to calculate probe resolution?

• Steps to evaluate lens system design
• 1. Calculate magnification and coefficients from
  ion optics computer codes
• 2. Measure
• Beam Brightness
• Chromatic momentum spread from the accelerator
  (use nuclear resonance)
• 3. Set object size so that demagnified image is
  equal to desired probe resolution
• 4. Set aperture size so that beam current is
  equal to desired beam current
• 5. Calculate aberration contribution from maximum
  divergence and energy spread
• 6. Add contributions to probe size in quadrature
  (or similar)
• 7. Spot size is now greater than desired spot
  size so go back to 3 and choose a smaller object
  size
• Repeat 4-7 until done.

dm 2(x/x)xomaxdc 2(x/?? )?o max ?o maxds
2(x/?)?o3max (x/? ? 2)?o ?o2max
di2 dm2dc2ds2
Wrong!!
10
Chromatic Aberration, A closer look
High excitation systems

• Singapore system achieves sub-micron probes with
  15o switcher magnet that has low energy
  dispersion
• Yet chromatic aberrations of this system should
  be large
• Skilled tuning of system is part of the answer,
  but not all!
• Maximum dc depends on getting maximum ? and ? in
  the same beam particle

dc 2(x/?? )?o max ?o max
11
Chromatic Aberration, A closer look

• Are ? and ? correlated?
• Use MULE to find out.
• Here is a slice of object plane phase space taken
  along ? and ?
• System was the HIAF accelerator in Sydney (From
  the work of Chris Ryan)
• Not much beam in the danger zone
• Beam intensity is peaked in the paraxial zone

• Conclusions
• Not much beam at edge of phase space
• Chromatic aberration is not a severe problem

Thank you G.W. Grime
12
Spherical Aberration, A closer look

• Traditionally, spherical aberration is computed
  from the rectangular model (RM)
• Rectangular model
• B(z) 0 z lt 0
• B(z) B0 0 lt z lt L
• B(z) 0 z gt L
• Results from this model agree with ray tracing
  codes that use B(r0 , z) measured at r r0
• Detailed studies have been done by Glenn Moloney
• Measured field profiles B(r , z) at several r
• Provides 3-D profile of True Fringe Field (TFF)
• Numerical raytracing from measured B(r , z)
  reveals different spherical aberration
  coefficients!

z
L
0
Coefficient RM TFFM (x/? 2)
-130 -130 (x/?? 2) -390
10 (y/? 3) -220 -190 (y/? 2?)
-390 2
13
Spherical Aberration, A closer look

• Coefficients calculated from the TFF model give
  aberration figures of different shapes compared
  to the rectangular model
• The figure is more intense in the paraxial region
  - good!

14
Ion Source Brightness Flux Peaking

• Legge et al (1993) showed a 1 order of magnitude
  decrease in probe size required a 5 orders of
  magnitude increase in brightness for uniform
  model
• True situation more complicated 1 order of
  magnitude decrease in probe size requires 2
  orders of magnitude increase in brightness

For 5 nA divergence is 2.5 times less than
uniform model so spherical aberration is reduced
by a factor of 16
2 MeV He
Current (pA)
15
Stray DC Magnetic Fields Parasitic aberration
Without magnet
With Magnet

• Non-uniform stray DC fields are a problem
• Shadows of a line focus on a fine grid should be
  straight line
• Small bar magnet has severe effect
• See large sextupole field component aberrations
• Sources of stray DC fields in the MARC
  laboratory
• Iron gantry and stairway over the beam line
• Steel equipment racks
• Gas bottles
• Stainless steel beam tube itself!

16
Stray DC Magnetic Fields Aberrations of a beam
pipe

• Type 316 stainless steel beam pipe through
  quadrupole lenses
• 10 mm internal diameter
• Beam diameter 6 mm
• Grid shadow pattern reveals aberrations
• See strong effect from different deflections of
  the beam pipe!
• Effect here produced by a few cm length
• What effect does 8 m have?

17
Stray AC Magnetic Fields Beam spot jitter

• Stray AC field causes a shift in the virtual
  object position
• The beam spot is scanned by the stray field in a
  complex fashion

object
lens
http//www.meda.com/fm3page.htm
18
Stray AC Magnetic Fields Beam spot jitter

• Stray AC fields cause virtual movement of the
  object collimator
• Used a 2-D scanwith y-coilsdisconnected
• Gives position asa function of timein map of Cu
  x-rays

3 mm
19
Stray AC Magnetic Fields

• It is good to have
• High demagnification systems
• Short systems
• On the Melbourne system it is required that
• Bstray lt 20 nT for xi lt 0.1 mm

• Where
• M Magnification 1/Demagnification
• q beam particle charge
• L Length of beam line
• E beam energy
• m beam particle mass

20
Stray AC fields in MARC laboratory Where from?

• Field as a function of time tells the story
• Start 6pm April 18 2000
• Place MP2 beam line, MARC laboratory

To MARC lab 50 m
21
Modify RF Ion Source

• Beam from ion source emerges with low energy
• Gas leakage from ion source canal fills low
  energy end of accelerator
• Gas scattering degrades ion source brightness
• Solution Add recirculating turbopump

new
old
From the work of Roland Szymanski
22
Modify Accelerator Column

• Remove corona needles and replace with resistors
• (Have now increased brightness by a factor of 10)
• So need to design a system optimised for a flux
  peaked beam
• High demagnification!

23
Selected new quadrupole systems
24
CSIRO quintuplet system Leipzig two stage
system

• Strong demagnification in a short system, 80 mm
  WD
• Very intense beam spot into 1 mm

• Strong demagnification in a long system

25
Resolution Versus Beam Current CSIRO/MARC
quintuplet system
1.3 mm at 0.5 nA
3100 pA/mm2 !
2.0 mm at 10 nA
3 mm at 20 nA
Accelerator brightness 1.2 pA.mm-2.mrad-2.MeV-1
CSIRO-GEMOC Nuclear Microprobe
From the work of Chris Ryan
26
Future Developments

• Conclusion To break through the 1 micron wall
• Install heavier magnetic shielding! But be sure
  to clean off DC fields (10 nT).
• Dont worry about chromatic and spherical
  aberration, they are not a severe as first though
  because of flux peaking (lt0.1 mm)
• Make brighter ion sources by small tweaks, even a
  factor of 10 is helpful (x1/3)
• Install an optimised system for a strongly flux
  peaked accelerator, this will have a large
  demagnification (of necessity a high excitation
  system) (M-1 gt 200)
• Need more radical lens design to reduce working
  distance and increase fields (40 mm)
• Apply the new system to some interesting
  problems! (lt 0.1 mm resolution)

27
(No Transcript)
28
Spherical Aberration A closer look

• The TFF model also revealed the need for careful
  attention to the field overlap between adjacent
  lenses
• Must have a linear field gradient as a function
  of beam direction to minimise aberrations
• Need to shape pole ends to achieve this

Pole tip
N
S
?
z
S
N