CATHI Technical Report

psithurismaccountantΠολεοδομικά Έργα

29 Νοε 2013 (πριν από 3 χρόνια και 7 μήνες)

196 εμφανίσεις


Cryogenics, Accelerators and Targets at HIE
-
ISOLDE





CATHI

Technical

R
eport


Design study of an upgraded charge breeder for ISOLDE:

Report on breeder options

(M39)



A.

Shornikov, F.
Wenander


This work is part of
CATHI Work Package 8
:
Radioactive Ion Beam Quality Improvement
.













The research leading to these results has received funding from the European Commission under the
FP7
-
PEOPLE
-
2010
-
ITN

project
CATHI

(
Marie Curie
Actions
-

ITN
).

Grant agreement no
PITN
-
GA
-
2010
-
264330
.

Report on Charge Breeder Options



M39


28/02/2013

2


Content


Introduction

................................
................................
................................
................................
.

3

Parametric study

................................
................................
................................
..........................

4

Physics governing the breeder parameters

................................
................................
........................

4

Accuracy of the parametric study

................................
................................
................................
.......

7

HEC
2

early design

................................
................................
................................
........................

10

General overview and definition of the key points

................................
................................
...........

10

Electron beam design

: Introduction

................................
................................
................................
.

14

HEC
2

gun design

................................
................................
................................
................................
.

15

HEC
2

electron gun program

................................
................................
................................
.........

17

Goals and experimental program with the HEC
2

gun
................................
................................
........

17

Status of the HEC
2

gun project and current activities

................................
................................
.......

18

References

................................
................................
................................
................................
.

20





Report on Charge Breeder Options



M39


28/02/2013

3


Introduction


In this work we present a design study of a new Electron Beam Ion Source (EBIS) to be installed as a
charge breeder for reacceleration of rare ions at ISOLDE. The study has been triggered by the High
Intensity and Energy upgrade at ISOLDE (HIE
-
ISOLDE)
(Fraser, et al., 2012)

and the planned
TSR@ISOLDE project

(al. M. G., 2012)
. To fulfil the requests of the user community the new EBIS
should reach an electron beam density of 10
4

A/cm
2

with electron energies up to 150 keV and
provide UHV environment and ion cooling in the breeding region to ensure confinement of the ions
long enough to reach the requested charge states. We report on the established design parameters
and first prototypin
g steps towards production and testing of suitable equipment

The REX
-
ISOLDE facility was a pioneering machine

(al, 2003)

to use an EBIS as a charge breeder
before the ions are sent to a LINAC for subsequent reacceleration to th
e nuclear physics
experiments. Over the past decade, over 100 radioactive isotopes of 33 elements have been
reaccelerated using the REXEBIS charge breeder
(Duppen & Riisager, 2011)
. To extend the range of
experiments at ISOLDE a
n upgrade of the facility has been proposed and approved by the CERN
research board in 2009. The upgrade is driven by increased intensity of the primary proton beam at
ISOLDE targets and, but also includes an upgrade of the ISOLDE LINAC into a superconduct
ing
version. The LINAC energy upgrade is divided into three phases (2015, 2016, 2017) with 2, 4 and 6
cryomodules installed providing acceleration energies of 5.5/8.8, 9.3/14.5 and 10/16.8 MeV/u for
A/q 4.5/2.5 respectively

(Voulo
t, et al., 2012)
.

As reported in
(Shornikov, Pikin, Scrivens, & Wenander, 2013)

the user community has so far
submitted 31 dedicated proposals for the first stage of the energy upgrade. The main implication of
these propos
als for the charge breeder is the request for a reduced breeding time, not only to limit
the decay losses of short
-
lived ions but also to minimize the inevitable dead time of the detector
system due to the pulsed beam structure. In order to increase the ef
ficiency of the data acquisition it
is important to provide ion pulses with highest possible repetition rate, up to the 100 Hz limit of the
LINAC. For the requested isotopes, such as for example
132
Sn
(al. P. R., 2012)

in

charge state A/q<4.5
(stipulated by the LINAC), the existing REXEBIS can only provide pulses with a maximum of 8 Hz
repetition rate. An increase of the repetition rate requires a corresponding increase in the current
density from 150 to about 2000 A/cm
2

a
nd hence major changes in the EBIS design.

In parallel with the HIE
-
ISOLDE upgrade the integration of the TSR storage ring at ISOLDE
(al. M. G.,
2012)

is being studied. In order to reduce the risk for electron stripping inside the ring highly charged
ions are favoured, and to match the ring rigidity limitation of 1.57 Tm the injected A/q
-
values should
be lower than 3.4 for a stored ion
-
energy of 10 MeV/
u, which is beyond the REXEBIS reach for the
heaviest ions. For specific experiments such as p
-
capture process one needs bare ions of 30<Z<70

(al.
M. G., 2012)
. Few
-
electron configurations such as Li
-

and Na
-
like of the heavies
t species U/Th are
requested for dielectronic recombination tests on exotic nuclei

(al. M. G., 2012)
. The experiments
with the storage ring require lower repetition rates limited to 1
-
2 s by a combination of the electron
coolin
g and the actual data taking process. In the following section we will analyse the design
parameters for a charge breeder providing ion pulses matching

the requests of both HIE
-
ISOLDE and
TSR@ISOLDE.

Report on Charge Breeder Options



M39


28/02/2013

4


Parametric study


Physics governing the breeder paramete
rs

This entire section cites the results previously submitted by the authors to NIMB
(Shornikov, Pikin,
Scrivens, & Wenander, 2013)
.

The parameters to be determined for the future charge breeder are the electron beam energy, th
e
total electron current and its density, and vacuum and holding field conditions to confine the ions
during the breeding process. As study cases we consider fast breeding of Ba to an A/q
-
value between
4.5 and 3 to be used at HIE
-
ISOLDE, and bare Ba and Li
-
like breeding of U as the ultimate
performance limit for TSR@ISOLDE.

An efficient charge breeding requires significantly higher electron energy than just corresponding
ionization potential to counterbalance the competing radiative recombination (RR) proce
ss. In the
equilibrium state, the distribution of the ions over the highest charge states (N
q
, N
q
-
1
) is defined by
the ratio of corresponding impact ionization (





) and RR




cross sections at a given electron
energy
E
e













(


)



(


)

Using Lotz
(Lotz, 1970)

formula for the II cross
-
sections and
Stoehlker

(Becker, Kester, & Stoehlker,
2007)

for RR we find that the optimum electron energy is about 3 times higher than corresponding
ionization potential. For the above mentioned model cases of Li
-
like U and bare Ba we get 100 and
150 keV electron energies to reach 35% and 50% abundance in the st
atic II/RR equilibrium.

The current density required to provide few
-
electron U for TSR@ISOLDE and fast breeding of Ba for
HIE
-
ISOLDE can be calculated for a given time structure as follows:



where
t
q

is the breeding time
and J
e

the electron current density. As is shown in fig. 1 both cases yield
a design value of J
e
= 10
4

A/cm
2
.

While the breeding time is a function of J
e
, the trap capacity is defined by the total electron current
I
e
. In the
present experimental scheme, which will remain after the upgrade, the 1+ ions from the
ISOLDE targets are accumulated in the REXTRAP Penning trap and are then injected as pulses into the
breeder. The throughput of the entire system is limited by REXTRAP, w
hich can store about 10
8
ions.
Keeping the neutralisation factor
f

in the breeder to 0.1, and assuming a reasonable trap length of
L
trap
=1 m, the current
I
e

required to confine
N
=10
8

U
89+

ions at an electron energy of 150 keV is at
least 2 A as given by eq
uation:


where
v
e

is the electron velocity.












(


)

















Report on Charge Breeder Options



M39


28/02/2013

5


During the charge breeding the ions also gain kinetic energy due to small
-
angle scattering with the
electrons and undergo charge exchange in collisions with the neutral atoms of the residual gas. The
heating proc
ess is inevitable and proportional to the same ionization factor

J
e
τ

as the ionization
process. Therefore, it is independent of current density and breeding time and depends only upon
the reached charge state. Once the gained energy exceeds the space
-
charg
e potential of the electron
beam, the ion will radially leave the ionization region and will be effectively lost for further breeding
(i.e. will reach the target charge state not in time, or not reach it at all due to shifted II/CX balance).
As one can see

in fig. 1 the holding voltage
U
=
E
q
/q
required to confine the ions for the highest charge
states exceeds the radial space
-
charge potential of the electron beam, thus such charge states are
reachable only if the evaporative ion
-
ion cooling with low charge s
tate ions
(Currell & Fussmann,
2005)

will be effectively implemented in the breeder.


Figure
1
: Breeding of VHCI and fast breeding


The influence of the CX process due to the residual atmosphere
is governed by the ratio of II and CX
probabilities. Thus, the higher

current density will give a two orders of magnitude advantage
compared to REXEBIS to mitigate this process. Using the formulae by Lotz, Stoehlker and Selberg
(Selberg, Biedermann, & Cederquist, Semiempirical scaling laws for electron capture at low energies,
1996)

for II, RR and CX cross
-
sections, we calculate at which pressure CX will counterbalance the II
process and overtake RR as a major recombinatio
n process. Assuming a current density of 10
4

A/cm
2

and an ion temperature of 100 eV (effective ion
-
ion cooling implemented) we find that a pressure in
the breeding region of about 10
-
11

mbar is sufficient to make CX negligible compared to RR.

Report on Charge Breeder Options



M39


28/02/2013

6



Figure
2
:

Influence of the vacuum conditions on the charge breeding process


The summary of the electron beam design parameters is given in table 1. As we see, the new breeder
compared to REXEBIS features High electron Energy, Current and
Compression of the electron beam,
therefore hereafter referred to as the HEC
2

EBIS.

Parameter

HEC
2

EBIS

Compared to REXEBIS

Electron energy

150 keV

X30

Electron current

2
-
5 A

X10
-
20

Current density

10 000 A/cm
2

X100

Magnetic field

6 T

X3


Assuming these parameters, one can study the breeding dynamics as shown in fig. 3. With a
repetition rate of 1 Hz for bare Ba and 0.5 Hz for Li
-
like U, breeding efficiencies of 25% and 30%
respectively can be expected.

Report on Charge Breeder Options



M39


28/02/2013

7



Figure
3
:

Dynamics of VHCI breeding at HEC
2

EBIS


Accuracy of the parametric study

The accuracy limits known for the basic processes must be taken into account for the figures
appearing in the parametric study. According to Lotz
(Lotz, 1970)

the empirical approximation gives
an error of the impact ionization cross
-
section of
+40/
-
30%
. The approximation is derived for rather
low energy
-
states and becomes less precise at higher energies, especially due to relativistic
correction effects fo
r the electrons. Experiments by Marrs et al. show that the actual cross
-
section
was underestimated by a factor of 2
(R. E. Marrs, 1994)
.

Charge exchange cross
-
sections used in simulation programs such as CBSIM are accurate only

within
a factor of 2. Available data is mostly based on an approximation of CX cross
-
sections of light ions
Z~10, thus the application of these approximations to higher charge states must be done with
significant safety margins for the design
-
relevant par
ameters of the future HEC
2

charge breeder.

The widely used (CBSIM included)
Salzborn
-
Müller semi
-
empirical formula

for CX was

obtained by
parametric fitting of about 100 cross
-
sections, all below
charge state
8+, most
with a

projectile
energy
in the range
of 10
-
25 keV/amu. Below 25 keV/amu the cross
-
sections
are
believed to be
independent of the energy

(Müller & Salzborn, 1977)
.

The
Salzborn
-
Müller (1977) approximation

of the CX cross
-
section gives















where

A=1.43 (0.76) x10
-
12

cm
2
, α=1.17(0.09), β=2.76 (0.19)

and I=13.6 eV ionization potential of Hydrogen
as the primary component of the residual gas at UHV pressures.

Report on Charge Breeder Options



M39


28/02/2013

8


There is variety of other approximations similar to
Salzborn
-
Müller
, which are a
ll part of the
generalised
Classical absorbing scheme CASM by Janev and Presnyakov

(Janev & Presnyakov, 1981)
:












(
(







)

)


A further example of such fittings is the Kusakabe approximation
(Kusakabe, Horiuchi, Nagai, Hanaki,
Konomi, & Sakisaka, 1986)

based on experimental data for E=0.286 keV/u and 4<q<11:















with A=9.5 (3.8) x10
-
14

cm
2
, α=1.3 (0.1), β=2 (0.1) and I=13.6 eV ionization potential.

The approximation best
-
suited for the charge
-
breeder case is done by Selberg. Based on over 300
experimental cross
-
sections for He, Ar, Xe targets and ions with 15<q<43
(Selberg, Biedermann, &
Cederquist, Absolute charge
-
exchange cr
oss sections for the interaction...., 1997)

the experimental
fitting by Selberg
(Selberg, Biedermann, & Cederquist, Semiempirical scaling laws for electron
capture at low energies, 1996)

gives:















with A
=2.70x10
-
13

cm
2
, α=0.98 (0.06), β=1.96 (0.04) and I=13.6 eV ionization potential.

Schlachter’s

widely used formula
(Schlachter, 1984)

is irrelevant to our conditions. According to the
author the formula gives a factor of two ag
reement if the reduced energy is 10<

̃
<1000
. The

reduced
energy is in keV/amu units scaled according to the rule

̃











,

so if Selberg data of 3.8 kev/q
for
132
Xe
would to be used,
than

̃

would range

from 0.03 to 0.1.
In an a
ctual EBIS
the
con
ditions will
differ even more prominently.

Primary experimental data on molecular Hydrogen targets with I and Ar ions

with a projectile energy
range relevant to a charge breeder can be taken from Mann experiments
(Mann, 1986)
. When
comparing the available data together with its uncertainties as shown in
Figure
4

one can see that
the most precise data is given by Mann. The Mann data report
ed for molecular Hydrogen is
consistent with the Selberg approximation within the error bars if one takes into account the
reported difference of about 30% in CX cross
-
sections for atomic and molecular Hydrogen, with
atomic cross
-
section being larger
(Phaneuf, 1983)
. From the given comparison, one can state that
amongst the semi
-
empirical scaling rules, the Selberg approximation is the most suitable and it was
used in the calculations for parametric studies.

Report on Charge Breeder Options



M39


28/02/2013

9



Figure
4

: CX models and their accuracy compared to each other and experimental data. Compilation from
the original sources
(Müller & Salzborn, 1977)
,
(Selberg, Biedermann, & Cederquist, Semiempir
ical scaling
laws for electron capture at low energies, 1996)
,
(Selberg, Biedermann, & Cederquist, Absolute charge
-
exchange cross sections for the interaction...., 1997)
,
(Kusakabe, Horiuchi, Nagai
, Hanaki, Konomi, &
Sakisaka, 1986)
,
(Schlachter, 1984)
,
(Mann, 1986)
,
(Phaneuf, 1983)
, by A. Shornikov.

A more accurate calculation of CX cross
-
sections involves a

quantum mechanical approach and yields
cross
-
sections, which oscillate (not monotonously increase/decrease). Such calculations are
implemented in open atomic codes such as FAC by Gu

(Gu, 2008)

and planned to be used in the nex
t
generation charge breeding simulation code based on TRIPSIM by R. Mertzig

(Mertzig, 2011)
; to be
implemented by himself upon joining the EBIS group at CERN.

The main result of the accuracy analysis regarding CX is the followi
ng. Even though many CX cross
-
section scaling rules exist, all of them still have large uncertainties. Therefore, the value of tolerable
pressure in the breeding region should be handled with care. Although the graphs show that a
pressure in lower 10
-
10

mbar is tolerable, one should set the design value to the mid 10
-
11

mbar
region to account for uncertainties in the CX values, especially for q>43 where the primary
experimental data for the relevant energy range is scarce.


0
2E-14
4E-14
6E-14
8E-14
1E-13
1.2E-13
1.4E-13
0
10
20
30
40
50
60
CX cross
-
section models, cross
-
section cm
2

vs charge state q

SM cross-sections
Selberg
Kusakabe
Schlachter
Mann_Ar+_on_H2
Mann_I_on_H2
Phaneuf_Fe_on_H
Phaneuf_Fe_on_H2
Report on Charge Breeder Options



M39


28/02/2013

10


HEC
2

early design


General
overview and definition of the key points

As mentioned above, the main changes in the EBIS design required to proceed from REXEBIS to HEC
2

EBIS are related to the HEC electron beam. Apart from the production of such a beam, there are
challenges related to
the beam dumping and maintaining sufficient vacuum conditions in the
presence of such high electron beam.



Figure
5

: Conceptual view of HEC
2
EBIS.

The beam
-
dump challenge consists of two main aspects: the energy deposition at th
e collector and
massive outgassing in the collector. The energy deposition problem originates from the limitations of
minimum electron energy at which the current can be dumped at the collector without sensitive
losses. At RHIC EBIS, the energy dumped at t
he collector is normally about 100 kW with a design limit
of 300 kW
(A., A., & L., 2006)
. The higher value is close to the technological limit and for the highest
energies at HEC
2

EBIS one should think of at least partial energy

recuperation (i.e. beam energy
deceleration) at the collector. For that reason, the commissioning of MIS
-
1
(Abdulmanov & Dikansky,
2010)

which is relying on a very advanced recuperation is of great interest. In general, the co
llector
for the HEC
2

should be designed following the BNL high
-
energy design with an option to lift the
collector potential and recuperate the energy. The BNL high compression electron gun is designed for
a 50 kV anode voltage. If the design parameters reg
arding the current and its density can be achieved
with the gun biased to 50 kV, then the operation at an electron beam energy of 150 keV in the
trapping region with recuperation of the energy at the collector back to 50 kV should be feasible. In
this case
, a few
-
ampere beam with 150 keV energy in the breeding region will deposit a power
similar to the design
-
value of the BNL
-
type collector.

Report on Charge Breeder Options



M39


28/02/2013

11


The challenge of UHV vacuum in the presence of a high
-
current beam is related to the electron
stimulated desorption (
ESD) phenomenon. For a well
-
prepared electron collector surface, after
weeks of electron bombardment, an ESD coefficient of 10
-
7
-
10
-
6

(Parkhomchuk, 2006)

can be
achieved. Assuming a 5 A electron beam and an ESD coefficient of 10
-
6

we get Q=5/(1.6x10
-
19
)x10
-
6
=3.1x10
13

molecules per second.

Assuming the collector is equipped with a 1000 l/s cryopump (conductance limited to S=300 l/s), we
obtain a pressure of

P=QRT/N
A
S=4.28x10
-
9

mbar.

To provide a vacuum pressure two orders of magnitude better in the breeding region than in the
collector, one needs a distributed pumping system. As mentioned above a suitable length of the
breeding region is about 1 m. Given the

other necessary requirements for space and the limited
magnet bore radius, any pumping relying on conductance connection to the outside of the magnet
will be very limited.


Figure
6

: Vacuum concept of HEC
2

EBIS with warm
-
bore cr
yogenically pumped breeding region.


Report on Charge Breeder Options



M39


28/02/2013

12



Figure
7

: Vacuum concept of HEC2 EBIS with warm
-
bore breeding region pumped mostly by NEG coatings
(red lines).

Thus, the only suitable means of pumping in the breeder region are NEGs and lon
g distributed
charcoal panels. The two vacuum concepts under consideration now are the NEG dominated fully
warm concept, and a cryogenically dominated cold
-
in
-
warm
-
bore design.

Both concepts provide comparable pumping speed. The decision has to be made bas
ed on
technological feasibility and tests. The main question for the cryogenic option is if the HEC
2

electron
beam will produce RF on a level above 10
-
3
% of the beam power. If yes, that will be enough to
counterbalance a powerful cryocooler at low temperat
ure and cause disruption in the vacuum
system.

Let us assume we have 500 cm
2
of pumping surfaces in the breeding region. In case of cryopumping
of H
2
, the combined coefficient c of sticking probability and penetration through the first stage baffle
is 0.21

(Day, 2007)
. Then assuming pumping of H
2

thermalized to 300/80 K according to














we obtain a specific pumping speed of
9.3
-
4.7 ls
-
1
cm
-
2

or a total pumping of 4650
-
2350 l/s. The
combined coefficient
c

is used as specified for a cryopump and depends upon exact geometry.
Nevertheless using the value for a round commercial cryopump gives a more cconservative value
than just the sticking coefficient, as it takes into account the reflection of Hydrogen from

non
-
pumping first stage baffle.

The specific pumping speed of H
2

using Ti
-
Zr
-
V NEGs is 0.35
-
1.3

(Chiggiato, 2006)

l/scm
2

depending
on roughness of the NEG surface. This would give some
S
2
=175
-
650 l/s total pumping.

Assuming in

both cases a pressure in the breeding region

P
b
= 1x10
-
11

mbar we calculate the allowed
gas load
Q

in both cases as















.


Report on Charge Breeder Options



M39


28/02/2013

13


Table
1

: Comparison of Cryopumping and NEG
-
dominated vacuum concepts

case

Pumping rate
l/s (S
1,2
)

Molecules/sec (Q
1,2
)

Cryopumping

2350

5.67x10
11

NEG pumping

175

4.23x10
10


If the collector is at
a pressure

P
c
=
5x10
-
9

mbar
,

it
will a set an upper limit for the conductance
between collector and breeding region as














=
0.35
/4.7

l/s for NEG/cryogenic design.
Assuming a single stage differential pumping insert of 20 cm length one can estimate the inner
diameter of the insert for a given conductance as follows. The c
onductance of an annular tube is















where
v

is the mean molecular velocity. For Hydrogen at room temperature






































1.81x10
3

m/s

For
a
4.7 l/s

conductance






































=12
.
6

m
m
.
To have a conductance

C
2
=
0.35 l/
s the diameter of the aperture should be







































=5
.
3

m
m.
While the former

i
s a
practical number, the latter

suggests insufficiency of 500 cm
2

of NEGs.

Having the NEG surface
substantially increased is well within pr
actical limits. The REXEBIS now has 6000 cm
2

of NEG pumping
surface integrated in the breeding region of comparable size.
If NEG coating of the entire chamber is
used instead of discrete NEG stripes
the
overall surface of

~ 2x3.14x5x200=6280

cm
2

can be used
,

so
the conductance
of the connection in the range 4
.
4

l/s is acceptable.

Let us now
compare the gas intake due

to the ESD outgassing in the collector to the thermal
outgassing in the breeder region. Assuming
a
100

mm diameter warm steel bore
of 2 m full length
baked to 150
-
200°C we get S=200x3.14x10=6280 cm
2

surface with outgassing 10
-
14

mbarl/scm
2

(Jousten, 1998)
,
so in total 6.28x10
-
11

mbarl/sec=6.28x10
-
12

Pam
3
/sec
, which corresponds to

1.5x10
9
molecules/s. There
fore, the gas intake from collector/gun dominates the thermal outgassing in the
breeder by one

to
two orders of magnitude.

To bring this intake down intermediate pumping chambers between the breeding region and the
regions of the electron collector and gun

as shown in fig
. 7 and 8

are highly recommended. Such
arrangement used at BNL allowed

reaching

a
differential pumping from 10
-
8

mbar in the collector to
10
-
10

mbar in the breeding region. To improve the pressure in the breeding region
further
as specified

above
,

the HEC
2

EBIS will include
a
significantly higher pumping rate in the breeding region.

Apart from increased pumping rate there is a suggestion to reduce the ESD in the collector by coating
its surface with NEG. Showing in general
a
positive impact
on desorption and ESD the stability of NEG
coating has never been studied under relevant electron radiation doses.



Report on Charge Breeder Options



M39


28/02/2013

14


Electron beam design

:
Introduction

Amongst the parameters studied above the total current, vacuum pressure, electron energy and
power
dissipation on the collector are within the range of state
-
of
-
the
-
art machines such as RHIC
EBIS, or Livermore SuperEBIT. The actual figure of merit for the HEC
2

EBIS defining its ultimate
performance is the current density, which exceeds the values reache
d in EBISes so far by several
times. Such current density is way beyond the reach of the most stable electron beams produced by
immersed electron guns in today’s best performing breeders such as REXEBIS
(Wenander, Charge
breeding of

radioactive ions with EBIS and EBIT, 2010)

or RHIC EBIS
(Pikin, et al., 2010)
.

The requested electron beam compression is possible only with Brillouin
-
like guns with combined
magnetic and electrostatic compression of the
electron beam. There are two major families of such
guns. The first relying on active cancelation of the stray magnetic field at the cathode originates from
Livermore SuperEBIT
(Marrs, 1996)

and has been widely used in EBIT dev
ices including
FreEBIT/HDEBIT
(López
-
Urrutia, Bapat, Draganic, Werdich, & Ullrich, 2001)
, TITAN
(Gallant, et al.,
2010)

and recently built MSU EBIT
(Schwarz, Bollen, Johnson, K
ester, & Kostin, 2010)
. Being rather
tuneable and thus less sensitive to misalignments, all these guns were however designed and used
for electron currents significantly lower than requested for HEC
2
. The reported current densities for
these guns are
below 5000 A/cm
2

for 200 mA currents
(Marrs, 1996)

which means that an up
-
scaling
of such a design will require significant R&D efforts.

The other family originates from the guns
(Baryshev, et al., 1994)

for high
-
power RF tubes such as
magnicons. These guns rely mostly on passive magnetic shielding of the cathode area using high
permeability screens. In such guns electron currents of about 200 A with a density up to 7000 A/cm
2

have been reported for 2

μs pulsed operation
(Nezhevenko, Yakovlev, Hirshfield, Gold, Fliflet, &
Kinkead, 2001)

at compression fields of only 2.2 T, compared to 7 T in the EBITs. The experimental
electron
-
beam density in these guns is reported to be on
ly 40% lower than that of the theoretical
Brillouin flow
(al. V. Y., 2001)
. The newer versions of that gun type have design values as high as 30
000 A/cm
2

(Shchelkunov, LaPointe, Jiang, Yakovlev, & Hirshfi
eld, 2012)
. Recently a similar gun design
has been proposed for the MIS
-
1 EBIS, currently under construction at BINP
(Abdulmanov &
Dikansky, 2010)
.








Report on Charge Breeder Options



M39


28/02/2013

15


HEC
2

gun design

A similar gun design (see
Figure
8
) was developed at BNL by Pikin et al.
(Pikin, Beebe, & Raparia,
Simulation and optimization of a 10A electron gun with electrostatic compression, 2013)

for future
applications at RHIC EBIS. The design features electron energies up to 50 kV, adjustable current in the
range 0.5
-
10 A and high compression ratio for the beam.



Figure
8
: Overview of the HEC
2

gun placed in TestEBIS.


The HEC
2

EBIS team recently joined the BNL high
-
compression gun project to build and test a gun of
the above
-
mentioned design in long pulse/continuous operation at an EBIS. According to numerical
simulations by Pikin the stray magnetic field on the catho
de surface is in the range of 3.8
-
4.5 Gauss
for 2
-
5 A optimised settings, which according to Hermann formula:


































(



























)








[

]



[

]



[

]



[

]



Report on Charge Breeder Options



M39


28/02/2013

16



Figure
9

: Estimation of the HEC
2

gun performance.

allows to reach

J
e
> 10
4

A/cm
2

in the breeding region at a full field of 6 T. Due to mounting of
electrostatic and magnetic elements on the same single unit a proper matching of electric a
nd
magnetic fields allows to lower the radial beam oscillations down to 4%.

A comprehensive
description of the simulations is given in
(Pikin, Beebe, & Raparia, Simulation and optimization of a
10A electron gun with electrostatic co
mpression, 2013)
.


Figure
10

: Dependence of relative amplitude of radial oscillations on the gun coil current for varying values
of the cathode shift. Picture by A. Pikin from his presentation during HIE
-
EBIS workshop 16
-
18.
10.2012 at
CERN.

0
5000
10000
15000
20000
25000
30000
35000
0
2
4
6
8
10
Current density in the ionization
region, A/cm
2

Field on the cathode, [G]

Current density in the interaction region

working region

I
e
=5A ,
B=6
T,
E
e
=50

keV

R
c
=10 mm,
Report on Charge Breeder Options



M39


28/02/2013

17


The bias limit of the cathode is about 50 kV to keep the cathode
-
anode gap safely from an electric
breakdown see
(Pikin, Beebe, & Raparia, Simulation and optimization of a 10A electron gun with
electrostatic compre
ssion, 2013)
. Therefore, electron energies above 50 keV have to be reached by
biasing the entire gun assembly together with the vacuum chamber by 100 kV relative to the ground.

The gun matches to the above
-
mentioned HEC
2

specification regarding achiev
able energy, current
density and the trap capacity. An additional aspect affecting the feasibility of the HEC
2

EBIS is if the
breeder equipped with such an electron beam will be able to capture the injected rare ions by
providing sufficient transverse acce
ptance. The transverse acceptance of an EBIS can be represented
as follows

(Wenander, Jonson, Liljeby, & Nyman, 1999)
:










(























)

where
r
b

is the electron beam radius,
m

is mass of the ion,
B

is the magnetic field,
U
ext

is the
extraction voltage and

l

is the linear density of charges in the electron beam. Setting
U
ext
=30 kV
,
assuming m=150 u (mass dependence is a minor effect), and an electron current of 5 A, with a
dens
ity according to Hermann radius for 4
-
6 G cathode field (see
Figure
9
), one obtains an
acceptance of 22
-
26/18
-
21 micrometers for 50/150 keV electron energy respectivel
y. This value is
lower than 34 micron acceptance calculated for the existing REXEBIS
(Wenander, Jonson, Liljeby, &
Nyman, 1999)
, which in turn is lower than emittance of 30πxmmxmrad (94 microns) of ISOLDE
targets, but exceeds t
he 9.4 microns emittance out of REXTRAP. Therefore, a buffer
-
gas cooled beam
from the REXTRAP can be effectively accepted into the trap, while direct injection will cause
significant losses due to low capture probability.

The gun is currently being built a
t CERN to be delivered and tested at the BNL TestEBIS in the first
half of 2013.


HEC
2

electron gun program

Goals and experimental program with the HEC
2

gun

Being
a rather successful design in general the above mentioned type of guns has a number of
parameters, which need to be tested before such a gun can be successfully implemented in an EBIS.



First of all, we need to make sure that the downscaling of the gun in terms of geometrical
dimensions, electron current and energy, still allows reaching the
desired value of the beam
compression. The main predicted risk arises from the fact that unknown thermal distribution
along the gun structure may cause a mismatch of the cathode emitting surface with respect
to the magnetic and electric fields. This proble
m is known

(Nezhevenko, Yakovlev, Hirshfield,
Gold, Fliflet, & Kinkead, 2001)

to compromise the performance of such guns. Thus, an
important step is the precise measurements of the cathode position with the cathode heater
stabil
ized at the working temperature. Such a measurement is planned using a short
-
range
laser tracker and a special chamber equipped with suitable viewport for the heated cathode.

Report on Charge Breeder Options



M39


28/02/2013

18




The second point of the program is to make sure that the gun design, until now im
plemented
in short pulse
-
operation mode, can operate for a time period of about a second as required
to breed VHCI. The concern is magnetron discharges, which may take place around the
cathode arm. While in short pulse
-
operation mode this is not a big prob
lem as the discharge
will be extinguished between the 1 μs long pulses spaced with approximately 1 s, in
continuous mode such a discharge if persistent will make the vacuum conditions in the gun
chamber too bad for sustained operation of the device. To min
imize the influence of possible
magnetron discharges during the test phase it was suggested to lift the potential of the gun
chamber to the cathode arm potential. In this case, there is no radial electric field between
the cathode arm and the chamber, and
a radial field exists only between the anode and the
chamber during the actual electron pulse. A future solution of the discharge issue has been
suggested and will be numerically studied. It is based on a set of ribs on the cathode arm,
which should interc
ept drifting electrons.




The third point is to make sure that the electron beam for the given parameters is not
producing significant RF noise at the TestEBIS fundamental frequency of 11 GHz (given by the
internal diameter of the drift tubes). The vacuum c
onditions calculated in the parametric
study require intensive pumping of the breeding region to compensate for gas intake from
the collector, where massive non
-
thermal outgassing will take place due to electron
stimulated desorption. One of the possible w
ays to provide large pumping capacity in the
breeding region is to install cryogenically cooled panels. This puts strong limits on the
tolerable RF production, in fact not more than a fraction of Watt can be adsorbed by metal
surfaces, otherwise the RF ads
orption will counterbalance the cryocooling at the lowest
temperature, where disposable cooling power is a few Watt only. If found, the RF noise will
force us to switch to completely warm design with NEG dominated pumping in the breeding
region.




The final

point is to perform the charge breeding in the TestEBIS, as in the absence of other
diagnostic methods it is the only way to experimentally measure the current density in the
breeder. The charge breeding test also allows measuring the acceptance of the EB
IS and the
emittance of charge bred beam.

Status of the HEC
2

gun project and current activities

According to the agreement with BNL, the project is organized in the following way. BNL provides
CERN with drawings and 3D models of the electron gun (delivered

10.12.2012 ). CERN manufactures
the gun assembly and the gun chamber. BNL provides access to the TestEBIS where the gun will be
installed, including necessary experimental infrastructure and beam diagnostics such as a primary ion
source, Faraday cups, a p
epper
-
pot emittance meter, beam focusing elements and a time
-
of
-
flight
mass
-
spectrometer. CERN provides vacuum components, a complete turbo pump unit with backing
pump, and linear translation stages for the diagnostics Faraday cups as specified by BNL to m
ake the
test bench (see
Figure
11
) operational. In order to minimize the risk for magnetron discharges the
gun will be tested with the gun chamber under high voltage p
otential, in contrast to how tests have
been performed at the TestEBIS setup in the past. Thus, a set of ceramic HV breaks shown in the
layout will also be provided by CERN.

Report on Charge Breeder Options



M39


28/02/2013

19



Figure
11

: Experimental setup for test operation of HE
C
2

gun. Encircled areas show locations where some
adjustments have to be done in order to carry out the tests. These include the new gun, focusing elements
and Faraday cups. Picture by A. Pikin.

The purchasing of the specified equipment started in Dec 2012

and is summarized in
Table
2
. As of
end of January 2013, the drawings of the gun were transferred to ISOLDE and the main CERN
workshops, so the manufacturing is split

to limit the cost and to speed up the process. At the
moment (Feb 2013) the purchasing of materials not available in the CERN stock (large dimension
ARMCO steel) is taking place together with processing of the drawings (optimizing to standard EU
component
s), assigning the specific jobs to the staff and production of pieces from already available
materials.

Table
2

: Preparation to the experiments. Use of financial resources.

#

Description

Ref on
EDH

Contact
person

C
ompany

Price
CHF

Status

on 19.02.2013

1

Dry Pump Edwards
XDS

5156290

Giovanna
Vandoni

Edwards

7413

Delivered 28.01

2

Filter for the dry
pump

5156257

Giovanna
Vandoni

Agilent

283.5

in accounting

3

Turbo pump

+controller HiPace

5158278

Wim Maan

Pfeiffer

7221

in accounting

4

Linear positioners
Huntington

5150197

Andrey
Shornikov

ITL

2922

Delivered 19.02

5

Vac fittings, flanges
and feedthroughs

5157198

Andrey
Shornikov

Vaqtec

10781

Partly delivered 19.02

6

HV connections

5171584

Andrey
Shornikov

Lewvac

1625

Delivered 07.02.2013

7

Ceramic tube
isolators

5172694

Andrey
Shornikov

Almath

75

Delivered
, used


8

HV ceramic breaks

5184449


Andrey
Shornikov

Lewvac


5053


In
accounting

9

Raw materials for
the gun chamber

5185369

Emilien


RIgutto

CERN

3609

Delivered 05.02

A
Electron gun
Test EBIS
Pepper-pot
emittance meter
Faragay cup-2
Einzel lens
Y-chamber
Faraday cup-1
Wien filter
Ion source
Time-of-flight
mass-spectrometer
Faraday cup-3
A
View A
Report on Charge Breeder Options



M39


28/02/2013

20


References

[1]

M. Fraser et al., "Status of the HIE
-
ISOLDE project at CERN," in
Proceedings of HIAT 2012
,
Chicago, 2012, pp. 175
-
181. [Online].
http://accelconf.web.cern.ch/AccelConf/HIAT2012/papers/wec02.pdf

[2]

M. Grieser et al., "Storage ring at HIE
-
I
SOLDE,"
Eur. Phys. J. Special Topics
, vol. 207, pp. 1
-
117,
2012. [Online].
http://link.springer.com/article/10.1140%2Fepjst%2Fe2012
-
01599
-
9

[3]

O. Kester et al, "Accelerated radioactive beams from REX
-
ISOLDE,"
Nuclear Instruments and
Methods in Physics Research B
, vol. 204, pp. 20
-
30, 2003. [Online].
http://www.scie
ncedirect.com/science/article/pii/S0168583X02018864

[4]

P. Van Duppen and K. Riisager, "Physics with REX
-
ISOLDE: from experiment to,"
JOURNAL OF
PHYSICS G: NUCLEAR AND PARTICLE PHYSICS
, vol. 38, p. 024005, 2011. [Online].
http://iopscience.iop.org/0954
-
3899/38/2/024005/pdf/0954
-
3899_38_2_024005.pdf

[5]

D. Voulot et al., "HIE
-
ISOLDE SC LINAC: operational aspects and commissioning pre
paration," in
Proceedings of IPAC 2012
, New Orleans, 2012, pp. 3853
-
3855. [Online].
http://accelconf.web.cern.ch/AccelConf/IPAC2012/papers/thppp050.pdf

[6]

A. Shornikov, A
. Pikin, R. Scrivens, and F. Wenander, "Design study of an upgraded charge
breeder for ISOLDE,"
Nuclear Instruments and Methods Section B
, vol. submitted, 2013.

[7]

P. Reiter et al., "Coulomb excitation of doubly magic 132Sn with MINIBALL," CERN, Geneva,
Proposal to the ISOLDE and Neutron Time
-
of
-
Flight Committee INTC
-
P
-
345, 2012. [Online].
cds.cern.ch/record/1482395/files/INTC
-
P
-
345.pdf

[8]

W. Lotz, "Electron
-
impact ionization cross
-
sections

for atoms up to Z=108,"
Zeitschrift für Physik
,
vol. 232, no. 2, pp. 101
-
107, Apr 1970. [Online].
http://link.springer.com/article/10.1007%2FBF01393132

[9]

R. Becker, O. Kester, and Th
. Stoehlker, "Simulation of charge breeding for trapped ions,"
Journal
of Physics: Conference Series
, vol. 58, pp. 443
-
446, 2007. [Online].
http://iopscience.iop.org/1742
-
6596/58/1/102

[10]

F. Currell and G. Fussmann, "Physics of electron beam ion traps and sources,"
IEEE Transactions
on Plasma Science
, vol. 33, no. 6, pp. 1763
-
1777, 2005. [Online].
http://i
eeexplore.ieee.org/lpdocs/epic03/wrapper.htm?arnumber=1556668

[11]

N. Selberg, C. Biedermann, and H. Cederquist, "Semiempirical scaling laws for electron capture
at low energies,"
Physical Review A
, vol. 54, no. 5, pp. 4127

4135, 1996. [Online].
http://link.aps.org/doi/10.1103/PhysRevA.54.4127

[12]

S. R. Elliott, D. A. Knapp R. E. Marrs, "Production and Trapping of Hydrogenlike and Bare
Uranium Ious in an Electron Beam Ion Trap,"
Phys
ical review letters
, vol. 72, no. 26, pp. 4082
-
Report on Charge Breeder Options



M39


28/02/2013

21


4085, Jun 1994. [Online].
http://link.aps.org/doi/10.1103/PhysRevLett.72.4082

[13]

A. Müller and E. Salzborn, "Scaling of cross sections for

multiple electron transfer to highly
charged ions colliding with atoms and molecules,"
Physics Letters A
, vol. 62, no. 6, pp. 391
-
394,
September 1977. [Online].
http://dx.doi.org/10.1016/0375
-
9601(77)90672
-
7

[14]

R. K. Janev and L. P. Presnyakov, "Collision processes of multiply charged ions with atoms,"
Physics Reports
, vol. 70, no. 1, pp. 1
-
107, April 1981. [Online].
http://dx.do
i.org/10.1016/0370
-
1573(81)90161
-
7

[15]

T. Kusakabe et al., "Charge transfer of multiply charged slow argon, krypton and xenon ions on
atomic and molecular targets. Single
-
charge transfer cross sections,"
Journal of Physics B: Atomic
and Molecular Physics
, vol. 19, no. 14, pp. 2165
-
2175, 1986. [Online].
http://dx.doi.org/10.1088/0022
-
3700/19/14/011

[16]

N. Selberg, C. Biedermann, and H. Cederquist, "Absolute charge
-
exchange cross sections for t
he
interaction.,"
Phyical Review A
, vol. 56, no. 6, pp. 4623

4632Ⱐ1997⸠孏.汩n敝⸠
h瑴pJIIp牡r慰献V牧I慢獴牡捴IP剁Iv56Ii6Ip4623弱

嬱[]

N⸠卥汢敲本⁃⸠䉩敤敲e慮nⰠ慮d⁈ ⁃ d敲煵楳琬i≓"m楥mp楲i捡氠c捡汩ng慷猠fo爠敬e捴牯n 捡c瑵牥t
慴a汯眠敮敲杩敳Ⱒ
Physical Review A
, vol. 54, no. 5, pp. 4127

4135Ⱐ1996⸠.佮O楮敝⸠
h瑴pJII汩n欮
慰献V牧IdoiI10.1103IPhX獒Vv䄮54⸴127

嬱[]

䄮⁓捨 慣a瑥爬r≃桡牧e
-
䍨ang楮g⁣o汬楳ion猬∠楮
10th international conference on Cyclotrons
, East
Lancing, 1984. [Online].
http://www.osti.gov/bridge/servlets/purl/6939436
-
HdAJC8/6939436.pdf

[19]

R. Mann, "Total one
-
electron capture cross section for Arq+ and iq+ ions in slow collisions on H2
and He,"
Z. Phys. D
-

Atoms, Molecules and Clusters
, vol. 3, pp. 85
-
90, 1986. [Onlin
e].
http://link.springer.com/article/10.1007%2FBF01442351

[20]

R. A. Phaneuf, "Electron capture by slow Feq+ ions from hydrogen atoms and molecules,"
Physical review A
, vol. 28, no. 3,
pp. 1310
-
1314, Sep 1983. [Online].
http://link.aps.org/doi/10.1103/PhysRevA.28.1310

[21]

M. F. Gu, "The flexible atomic code,"
Canadian Journal of Physics
, vol. 86, no. 5, pp. 675
-
689,
2008.

[Online].
http://dx.doi.org/10.1139/p07
-
197

[22]

R. Mertzig,
Untersuchungen zur Physik von in hochdichten EBIS
-
Elektronenstrahlen gespeicherten
Eisenionen
. Dresden, 2011.

[23]

Pikin A., Kponou A., and Sn
ydstrup L., "Optical, Thermal and Stress Simulations of a,"
Brookhaven national laboratory, Upton, C
-
A/AP/#246, 2006. [Online].
http://www.bnl.gov/cad/eb
is/docs/Tech_Notes/pdf/ap_note_246%20Electron%20Collector.pdf

[24]

V. G. Abdulmanov and N. S. Dikansky, "MIS
-
1 electron
-
beam ion source,"
Physics of Particles and
Nuclei Letters
, vol. 7, no. 7, pp. 539
-
542, 2010. [Online].
Report on Charge Breeder Options



M39


28/02/2013

22


http://link.springer.com/article/10.1134%2FS1547477110070241

[25]

V. V. Parkhomchuk, "Vacuum pumping with an electron beam,"
INSTRUMENTS AND
EXPERIMENTAL TECHNIQUES
, vol. 49, no. 3
, pp. 446
-
448, 2006. [Online].
http://www.springerlink.com/content/d6u35p6717248m41/

[26]

Ch. Day, "Basics and applications of cryopumps," in
CERN accelerator school vacuum in
accelerat
ors
, Platja d'Aro, 2007, pp. 241
-
274. [Online].
http://cdsweb.cern.ch/record/923393/files/CERN
-
2007
-
003.pdf?version=1

[27]

P. Chiggiato, "Ti



嘠Von
-
敶慰o牡r汥⁧l瑴敲⁦楬 猺⁆ om⁤ v敬opm敮琠eo慲 e⁳捡汥⁰牯duc瑩tn
景爠rh攠L慲a攠H慤牯n 䍯汬楤敲Ⱒe
Thin solid films
, vol. 515, pp. 382
-
388, 2006. [Online].
http://www.sciencedirect.com/science/article/pii/S0040609005025496

[28]

K. Jousten, "Dependence of the outgassing rate of a “vacuum fired” 316LN stainless steel
捨cmb敲eon⁢慫e
-
ou琠t敭p敲慴畲攬e
Vacuum
, no. 4, pp. 359
-
360, 1998. [Online].
http://www.sciencedirect.com/science/article/pii/S0042207X98000025

[29]

F. Wenander, "Charge breeding of radioactive ions with EBIS and EBIT,"
Journal of
Instrumentation
, vol. 5
, no. 10, p. C10004, 2010. [Online].
http://iopscience.iop.org/1748
-
0221/5/10/C10004/

[30]

A. Pikin et al., "RHIC EBIS: basics of design and status ofcommissioning,"
Journal of
Instrumentation
, vol. 5, p. C09003, 2010. [Online].
http://iopscience.iop.org/1748
-
0221/5/09/C09003/

[31]

R. E. Marrs, "Production of U92+ with an electron beam ion trap (invited),"
Review o
f scientific
instruments
, vol. 67, pp. 941
-
944, 1996. [Online].
http://dx.doi.org/10.1063/1.1146646

[32]

J. R Crespo López
-
Urrutia, B. Bapat, I. Draganic, A. Werdich, and J. Ullrich, "First results from
the
Freiburg Electron Beam Ion Trap FreEBIT,"
Physica Scripta
, vol. T92, pp. 110
-
112, 2001. [Online].
http://iopscience.iop.org/1402
-
4896/2001/T92/023

[33]

A.T. Gallant et al., "TITAN
-
EBIT


捨c牧e⁢牥敤 ng o映f慤楯慣瑩v敩eotop敳efo爠桩gh⁰牥捩獩 na獳V
m敡獵e敭敮瑳Ⱒ
Journal of instrumentation
, vol. 5, p. C08009, 2010. [Online].
http://iopscience.iop.org/1748
-
0221/5/08/C08009/

[34]

S Schwarz, G Bollen, M Johnson, O Kester, and M Kostin, "The NSCL electron beam ion trap for
the reacceleration of rare isotopes coming to life : First extraction tests with a high
-
current
electron gun,"
Review of scientific instruments
, vol. 81, p.
02A503, 2010. [Online].
http://dx.doi.org/10.1063/1.3271246

[35]

Y.V. Baryshev et al., "A 100 MW electron source with extremely high beam area compression,"
Nuclear Instruments and Methods in Physics
Research Section A
, vol. 340, no. 2, pp. 241

258Ⱐ
䙥F⸠.994.⁛佮 楮敝⸠
h瑴pJIId砮xo椮o牧I10⸱.16I0168
-
9002⠹4)90100
-
7

Report on Charge Breeder Options



M39


28/02/2013

23


[36]

O.A. Nezhevenko et al., "STATUS OF X
-
BAND PULSED MAGNICON PROJECT," i
n
Proceedings of
the 2001 Particle Accelerator Conference
, Chicago, 2001, pp. 1023
-
1025. [Online].
http://accelconf.web.cern.ch/AccelConf/p01/PAPERS/MPPH107.PDF

[37]

V.P. Yakovlev et al., "100 MW electron gun for a 34.3 ghz magnicon," in
Proceedings of the 2001
particle accelerator conference
, Chicago, 2001, pp. 1041
-
1043. [Online].
http:
//ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=986570&tag=1

[38]

S.V Shchelkunov, M.A. LaPointe, Y. Jiang, V.P. Yakovlev, and J.L. Hirshfield, "MODIFIED
MAGNICON FOR HIGH
-
GRADIENT ACCELERATOR R&D," in
Proceedings of IPAC2012
, New
Orleans, 2012, pp. 3377
-
3379. [Online].
http://accelconf.web.cern.ch/AccelConf/IPAC2012/papers/thppc042.pdf

[39]

A. Pikin, E. N. Beebe, and B. Raparia, "Simulation and optimization of a 10A elect
ron gun with
electrostatic compression,"
Review of scientific instruments
, vol. accepted for publication, 2013.

[40]

F. Wenander, B. Jonson, L. Liljeby, and G. Nyman, "REXEBIS the electron beam ion source for the
REX
-
ISOLDE project design and simulations,
" CERN, Geneva, 1999. [Online].
http://cds.cern.ch/record/478399