[Company Name]

1
Progress in tungsten x-ray exposures on XAPPER
Presented by Jeff Latkowski XAPPER Team Ryan
Abbott and Brad Bell HAPL Program
Workshop Rochester Laboratory for Laser
Energetics November 8-9, 2005 Work performed
under the auspices of the U. S. Department of
Energy byLawrence Livermore National Laboratory
under Contract W-7405-Eng-48.
2
Outline

• General XAPPER updates
• Moved to a circular sample tray with all key
  instruments machinedto be at the same height
• CMOS camera activated as complement to CCD camera
• Scanning electron microscope delivered
  activated
• New understanding of XAPPERs time dependence
• Measurements, modeling and implications
• Future plans
• New exposures of single crystal powder met
  tungsten
• Thermometer results
• Future plans

3
We designed and installeda new sample /
instrument tray

• Provides a single platform for fielding all
  in-situ diagnostics
• Corrects our single largest sourceof
  experimental error ensuringthat fluence
  measurement and sampleirradiation are on the
  same plane,and thus, are at the same fluence
• Able to field both camerassimultaneously to
  cross-check againsteach other while gaining
  CMOSoperating experience and writingimage-proces
  sing software

4
NRL delivered and installeda CMOS camera for use
in XAPPER

• New camera is fast, small, and cheap
• 10 Hz imaging, 1024 x 1280 _at_ 5.2 mm pixels
• Dimensions 2 x 1.26
• 1500 with PCI card, software and cables
• June attempt to field CMOS failed due to
  electrical noise on XAPPER
• Glenn Holland returned in October / was able to
  get uncoated CMOS sensor to work
• Can operate camera in vacuum without cooling

Stars are repeatable and can be subtracted as
background May be due to micro-lens residue
5
Used SEM with digital imaging upgradeswas
delivered activated in September

• SEM includes Be-window Edax system chemical
  identification down to sodium (Z11)
• Digital imaging upgrades from 4pi, Inc.

6
Weve done quite a bit regarding the time
dependence of XAPPERs x-ray pulse

• We have been making incorrect assumptions
  regarding the x-ray pulselength
• We have operated using vendors statement that
  x-ray pulse risetime is 40-50ns
• Unfortunately, we misinterpreted this as a 40-50
  ns energy deposition time
• Further, modeling generally assumed a square
  rather than a gaussian pulse
• Result ? belief that we needed 0.7 J/cm2 to rise
  from 600 to 2500ºC
• In June/July, we implemented a highly-biased
  AXUV-100 photodiode
• Thanks 2 to Glenn Holland of NRL
• Diode is too slow to fully resolve the x-ray
  pulse, but the bias gives it a fast risetime
• Used to confirm 50ns risetime (10-90 of peak)

7
Modeling realizations

• Longer x-ray pulselength requires increased
  fluence 0.7 ? 1.0 J/cm2
• Using 0.7 J/cm2 only heats surface to 1800
  instead of 2500ºC goal
• Will confirm overall pulse with fast photodiode
  (waiting for suitable x-ray filters)

8
An increased target fluence makeslife more
difficult expensive

• Ellipsoidal focusing optics start at a high
  fluence and gradually degrade
• High target fluence means that less degradation
  is tolerable
• Increased change-out rate has materiel and labor
  cost (time for replacement and alignment)
• Experimenting with meshes to protect optic
• Reduces fluence slightly at beginning
• Protect optic and greatly slow rate of
  degradation
• Able to make this a net win?
• Have had two optics cleaned by group that
  fabricates them unclear if process restores
  x-ray reflectivity

9
Powder met and single crystal tungsten samples
have been exposed on XAPPER

• Each sample was hit with 2 spots of 50,000 pulses
• Looking for repeatability at least in single
  crystal sample (powder met might be different due
  to varying grain orientations)

Images show powder met (left) and single crystal
(right) tungsten during irradiation on XAPPER at
10 Hz. The brighter spot on each image is the one
that is currently being exposed. Due to surface
damage, a dimmer spot lights up in the location
of the previous spot.
10
Using Veeco, the two powder metspots look
different
20x
100x
Spot 1 170nm Ra
Spot 1 110nm Ra
All scales-1000 ? 1000 nm
Spot 2 147nm Ra
Spot 2 57nm Ra
Unexposed 24nm Ra
11
SEM confirms that spot 1 is considerably rougher
than spot 2 on the powder met
PM Spot 1
Note that scale bars do not match up (sorry)
PM Spot 2
12
Due to an error in the fluence measurement
process for the single crystal tungsten, spot 1
was hit at a fluence of 0.8 J/cm2
10x
50x
All scales-500 ? 500 nm
Spot 1 23nm Ra
Spot 1 64nm Ra
0.8 J/cm2(2100ºC)
1.0 J/cm2(2500ºC)
Spot 2 48nm Ra
Spot 2 150nm Ra
Despite the lower peak temperature, both single
crystal tungsten spots roughened.
Unexposed 19nm Ra
Unexposed 14nm Ra
13
SEM on the single crystal sample
SingX Spot 1 (0.8 J/cm2 2100ºC)
SingX Spot 2 (1.0 J/cm2 2500ºC)
14
Observations from powder met single crystal
exposures (50K pulses, 0.8 1.0 J/cm2)

• Powder met spots look qualitatively different
  under Veeco
• Despite different fluences, single crystal spots
  look more alike than the powder met spots It
  seems like the damage on 1.0 J/cm2 spot is simply
  the logical extension of that seen on the 0.8
  J/cm2 spot
• Interesting that single crystal tungsten
  roughened at 0.8 J/cm2 ? 2100ºC
• Would it be worth generating a grain map of the
  powder met samples (either before or after
  exposure)?
• Would exposure of nanocrystalline tungsten
  possibly shed light on this?
• We are eager to repeat these exposures with more
  (5-10) spots per sample Can less than 50K pulses
  be used?

15
Options issues for startingXAPPER work on
optics

• Options for reducing x-ray fluence (WAY too high
  for optics!)
• Design, fabricate field new optic that provides
  larger, flat-top type spot instead of
  high-intensity spike (enables larger area studies
  too)
• Modify system design to provide reduced energy
  per pulse, yet conserve average power ? could
  support a significantly higher repetition rate
• Move to argon or nitrogen discharges
• Have an argon optic, but never tested no
  nitrogen optic exists
• Either gas gives slightly higher harder x-ray
  spectrum (250-400 eV) and significantly reduced
  energy per pulse
• Issue Do GIMMs need to be exposed at grazing
  angles? Cant do this with ellipsoidal optic!

16
Future plans

• Fast photodiode measurements to confirm x-ray
  pulse history
• Need effort to clean and/or protect ellipsoidal
  focusing optics
• Additional tungsten exposures
• Goal is many spots on same sample
• Grain size determination useful?
• Exposure of nanocrystalline tungsten samples
• Get ready to test optics on XAPPER
• Trade study for best way to implement
• Decision by end of year
• Complete implementation by summer 2006
• Finally get back to the thermometer

17
Back-up slides
18
The XAPPER experiment is used to studydamage
from rep-rated x-ray exposure

• Source designed / built by PLEX LLC
• Operates with xenon gas pinch to produce 80-150
  eV x-rays
• Operation possible at up to 10 Hz for millions of
  pulses

Materialsample
Condensingoptic
Plasmapinch
19
XAPPERs mission is to investigate cyclic fatigue
and other sub-threshold effects

• XAPPER is looking for sub-threshold (e.g.,
  without melting or ablation) effects such as
  roughening and thermomechanical fatigue.
• XAPPER cannot match the x-ray spectrum, but it
  can replicate a selected figure of merit (e.g.,
  peak surface temperature, dose, stress, etc.).
• XAPPER is used in the study of x-ray damage to
  chamber wall materials and will be applied to
  optics in the near future.

Our results to date show some roughening of
tungsten, but we do not see anything that would
suggest the first wall armor concept would not
work.
20
SEM images of unexposedsingle crystal tungsten