Berkeley Lab Generic Presentation

1
Effect of atomic hydrogen exposure on electron
beam polarization from strained GaAs photocathodes
M. Baylac, JLab baylac_at_jlab.org
P. Adderley, J. Brittian, J. Clark, A. Day, J.
Grames, J. Hansknecht, M. Poelker, M.
Stutzman
2
Plan

• Motivations
• Experimental setup
• Test
• QE
• Polarization measurements (2002 2003)
• Analyzing power (2003)
• Profilometry
• Results, Conclusions

3
Motivations

• Early days at JLab, wet chemical etching provided
  unreliable results sometimes good, sometimes bad
  QE
• Since 1995, atomic hydrogen cleaning provides
  high QE and reproducible results at JLab other
  labs have adopted this method since then (MAMI,
  Nagoya, Bates, SLAC)
• Polarization varied from wafer to wafer
  originating from the same manufacturer and across
  one single wafer (12 mm diameter)
• What was JLab doing differently from SLAC?

? Study effect of hydrogen cleaning on wafer
properties
4
Strained layer GaAs photocathode
Bandwidth Semiconductor (formerly SPIRE)
Strained GaAs

• MOCVD-grown epitaxial spin-polarizer wafer
• Lattice mismatch

0ltxlt0.29
5
Photocathode preparation

• 3 wafer cleaved (15.5 mm)
• No wet chemical treatment, ie no acid or base
  etching, no degreasing, no anodization

• Sample on stalk w/

Indium Ta cup
lucky
6
Test Plan

• Stalk installed in gun vacuum chamber,
• chamber evacuated bake (250 C)
• Wafer heated 2 hours at 570 C (estimated wafer
  temperature)
• NEA activation (CsNF3) in gun chamber, QE scan
  of wafer
• 100 keV beam
• QE, polarization vs. wavelength and vs. wafer
  location
• Break vacuum, remove wafer from gun
• Load in portable hydrogen cleaning chamber, pump
  down
• Expose wafer to atomic hydrogen

7
Hydrogen source
wafer
300C

• H2 dissociation via RF inductive discharge
• Parameters adjusted then, wafer exposed
• Dose measured with an ion counter at chamber
  bottom
• Conditions kept identical for all cleanings

15 cm
100 MHz
20 W
Mc.Alpine Schildknecht, Proceeding of IRE,
1959 (2099)
H2, or D2
8
Hydrogen cleanerhttp//www.jlab.org/accel/inj_gro
up/h2/portable_H2.html
9
Test gun

• 100 keV DC beam
• Wavelength tunable TiSapp
• (750-850 nm)
• 10 Hz helicity reversal
• Mott polarimeter P
• Beam dump QE

lenses
e
correctors
viewers
10
Test gun (contd)
11
Initial benchmark, untreated surface
Stat. only

• Syst. 10

M. Baylac et al, SPIN 2002, 15th International
Spin Physics Symposium proceedings, 1073 (2003).
12
QE at band-gap vs. Hydrogen exposure
At 840 nm
Significant drop in QE vs. H cleaning Confirmed
by 2 similar tests in different chamber w/o
breaking vacuum
13
Polarization vs. cumulative H dose (central
location)

• Depolarization as H dose is increased
• ?-dependent, strongest at band-gap
• Found systematic effect

High H dose, QE low ND filter removal ?
increase non-IR light
Depolarization due to low ?, high QE light
? Reject 60 min. data
14
Depolarization tests

• Confirm/infirm previous depolarization results
• Test H ions effect
• 3 independent tests, 3 photocathodes, only one
  change
• Usual exposure (no bias on photocathode during
  cleaning)
• Ion-enhanced exposure (negative bias)
• Ion-reduced exposure (positive bias)
• Same method as before with one single exposure of
  80 minutes for each test

15
Polarization with usual exposure

P(H)-P(0)-10 at bandgap
80 min dose
16
Polarization with ion-reduced exposure

P(H)-P(0)-10 at bandgap
80 min dose
17
Polarization with ion-enhanced exposure

P(H)-P(0)-20 at bandgap
80 min dose
18
Analyzing power

• Measure A.P. across strained-layer GaAs
• 7 consecutive H exposures up to 4 hours

860 nm diode laser
19
Analyzing power (cont.)
Analyzing power at 5 locations 5 and uniform
20
Analyzing power

• No variation with H exposure, or location

gt Effect unrelated to strain
21
Profilometry after H exposure

• A roughened surface could explain depolarization
  as effective angle of incidence of light onto
  wafer varies
• High resolution 3d profilometry _at_ Jlab (Andy Wu)
• area 80 mm X 20 mm
• vertical resolution 7.4 Angstrom
• Measurement of RMS roughness

22
Profilometry after H exposure
Bare surface
RMS 155 A
H cleaned
RMS 8500 A
23
Hydrogen exposure results

• Depolarization at band-gap induced by H exposure
  
• confirmed at 10 level
• systematics explains stronger effect seen on
  older data
• QE seems to be only lowered by heavy H dose
• Depolarization with/without H ions, unexplained
  enhancement
• Increased anneal cycle (12 h instead of 2) does
  not restore Pe
• Analyzing power independent of H
• Surface analysis shows roughened surface which
  can explain depolarization (underway)

24
Polarization comparison
25
Laser power systematic check