>

Characterization of the CEBAF 100 kV DC GaAs photoelectron gun vacuum system
Nuclear Instruments and Methods in Physics Research A 574 (2007) 213220
Characterization of the CEBAF 100 kV DC GaAs photoelectron gun
vacuum system
M.L. Stutzman
Ã
, P. Adderley, J. Brittian, J. Clark, J. Grames, J. Hansknecht,
G.R. Myneni, M. Poelker
Thomas Jefferson National Accelerator Facility, Polarized Source, 12000 Jefferson Avenue, MS 5A, Newport News, VA 23606, USA
Received 22 January 2007; accepted 27 January 2007
Available online 12 February 2007
Abstract
A vacuum system with pressure in the low ultra-high vacuum (UHV) range is essential for long photocathode lifetimes in DC high
voltage GaAs photoguns. A discrepancy between predicted and measured base pressure in the CEBAF photoguns motivated this study
of outgassing rates of three 304 stainless steel chambers with different pretreatments and pump speed measurements of non-evaporable
getter (NEG) pumps. Outgassing rates were measured using two independent techniques. Lower outgassing rates were achieved by
electropolishing and vacuum ring the chamber. The second part of the paper describes NEG pump speed measurements as a function of
pressure through the lower part of the UHV range. Measured NEG pump speed is high at pressures above 5
10
11
Torr, but may
decrease at lower pressures depending on the interpretation of the data. The nal section investigates the pump speed of a locally
produced NEG coating applied to the vacuum chamber walls. These studies represent the rst detailed vacuum measurements of CEBAF
photogun vacuum chambers.
r
2007 Elsevier B.V. All rights reserved.
PACS: 29.25.Bx; 29.27.Hj; 07.30.Cy; 07.30.Kf
Keywords: Polarized electron source; Photoinjector; Electron gun; Non-evaporable getter (NEG); Pump speed; Outgassing
1. Introduction
Electron beams with polarization up to 85% are used to
probe nuclear structure at three independent experimental
halls at the Continuous Electron Beam Accelerator Facility
(CEBAF) at Thomas Jefferson National Accelerator
Facility (Jefferson Lab). The electron beams for each hall
originate from the same gallium arsenide (GaAs) photo-
cathode that sits in a 100 kV DC high voltage photogun
[1]
.
Residual gasses in the gun chamber are ionized and
degrade photoelectron yield (quantum efciency, QE)
when ions are accelerated into the photocathode, damaging
the crystal structure or sputtering away the chemicals used
to create the negative electron afnity condition.
Photocathode operational lifetime falls as average beam
current extracted from the photocathode rises. At lower
currents (
o100 mA), the photogun can operate uninter-
rupted for months before the QE and laser power are
insufcient to provide the required beam. At high currents,
the gun can provide uninterrupted beam for one to two
weeks before the focused laser beam must be moved to a
new location on the photocathode where the QE is still
high. After exhausting the usable laser spot locations, the
photocathode must be heated and reactivated in situ. This
process restores photocathode QE and beam delivery can
resume.
There are two identical photoguns at the CEBAF
injector, with one providing beam to the accelerator and
the other serving as a spare. The single-chamber design
(
Fig. 1a
) requires venting to atmospheric pressure to
replace the photocathode. After venting, the photogun
must be baked to achieve a suitable vacuum condition by
ARTICLE IN PRESS
www.elsevier.com/locate/nima
0168-9002/$ - see front matter r 2007 Elsevier B.V. All rights reserved.
doi:
10.1016/j.nima.2007.01.170
ÃCorresponding author. Tel.: +1 757 269 7073; fax: +1 757 269 7363.
E-mail address:
marcy@jlab.org (M.L. Stutzman). removing adsorbed water vapor and activating the NEG
pumps. Bakeouts are conducted using a thermally insu-
lated oven and forced hot air to achieve a bake temperature
of 250 1C with a typical duration of 30 h. Two types of
bakes are performed: an initial bake with an inexpensive
sacricial photocathode following extensive vacuum work
where the NEG modules have been exposed to atmosphere,
and routine bakes following venting to dry nitrogen and
installation of high polarization photocathode material.
During the initial bake, maximum pressure, deduced from
ion pump current, is
5
10
6
Torr and QE of the
sacricial GaAs wafer is typically less than optimal.
Routine changes of photocathode material take place by
venting with dry nitrogen and using a glove bag with a dry
nitrogen purge around the opened ange to prevent water
vapor from entering the vacuum chamber. During bake-
out, following a routine photocathode change, maximum
pressure is typically 5
10
9
Torr and QE is high (up to
1% QE at the bandgap of 780 nm for high polarization
superlattice material
[2]
).
Following a routine bakeout, the photocathode is
activated to a negative electron afnity condition using
Cs and an oxidant, NF
3
[3]
. Activation occurs within the
chamber, and the deposition of the chemicals introduces
temporary vacuum excursions in the gun. The vacuum
recovers within an hour following cathode activation and
beam delivery takes place with the gun at base pressure.
The photoguns are connected to the accelerator using
large bore (6.35 cm diameter) beampipes coated with NEG
material (
Fig. 1b
). A 151 bend eliminates line of sight
between the photogun and the accelerator beamline, and
provides means to illuminate the photocathode at normal
incidence, which is desirable for high polarization guns.
The rst 3 m of beamline downstream of the guns was
baked to
220 1C and a differential pump station isolates
the baked beamline from the unbaked portion of
the accelerator. These features of the injector ensure that
the dominant gas load at the photoguns originates from the
gun chamber outgassing rather than from the accelerator.
Ten NEG modules
[4]
surround the cathode/anode gap
in the gun vacuum chamber providing extensive pumping
of hydrogen gas, which is the dominant gas species in steel
UHV systems
[5]
. A differential ion (DI) pump
[6]
located
downstream of the anode pumps gasses that are poorly
pumped by the NEG pumps, such as noble gasses and
methane. There are also two GP100 NEG pumps
[7]
on
ports downstream of the anode plate. An extractor gauge
[8]
and residual gas analyzer (RGA)
[9]
can be used to
monitor pressure within the photogun. These hot lament
gauges produce light and gas when energized, so they
cannot be left powered during photogun operation since
the light could be a source of uncontrolled and unpolarized
photoemission
and
optimal
vacuum
conditions
are
achieved with the gauges off. Consequently, pressure is
measured only during long accelerator maintenance
periods. The best recorded pressures for the gun are shown
in
Fig. 2
, as well as pressure values from test chambers also
using NEG and DI pumps. The measured pressure in all
chambers is signicantly higher than predicted values,
assuming typical outgassing rate for 304 SS of 1
10
12
Torr L/(s cm
2
) and the vendors stated NEG and DI
pump speeds. These observations provided motivation for
the outgassing and NEG pump speed studies presented here.
ARTICLE IN PRESS
a
b
Fig. 1. (a) The CEBAF 100 kV DC high voltage GaAs polarized
photogun and (b) the CEBAF 100 kV photoinjector.
1E-12
1E-11
1E-10
0
0.2
0.4
0.6
0.8
1
1.2
Getter Surface Area (m
2
)
Pressure (Torr)
measured
predicted
Fig. 2. Measured versus predicted pressure inside CEBAF photoguns and
test stands. Closed and open triangles represent best pressure achieved in
test stands with two and four fully activated SAES WP950 getter modules.
Closed and open diamonds represent CEBAF photogun vacuum
chambers, each with ten WP950 modules passively activated. Circle
represents a new load-locked vacuum chamber with six fully activated
SAES WP1250 modules. Lower data set represents values predicted from
calculations based on chamber surface area, assuming an outgassing rate
of 1
10
12
Torr L/s cm
2
and using manufacturers quoted pump speed.
M.L. Stutzman et al. / Nuclear Instruments and Methods in Physics Research A 574 (2007) 213220
214 Section 2 describes outgassing rate measurements for
one of the original CEBAF photogun vacuum chambers, a
new chamber constructed for these tests, and a chamber
designed for a new load-locked gun. Section 3 describes
NEG pump speed measurements of commercial pumps and
in-house fabricated coatings.
2. Outgassing measurements
2.1.