4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 1
NEG thin film coatings: from the origin to the
next-generation synchrotron-light sources
Paolo Chiggiato
CERN, Technology Department
Vacuum, Surfaces and Coatings
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 2
Outline
Historical perspective
TiZrV thin film coating for the LHC
MAX IV
Future developments
1
2
3
4
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 3
From NEG strips and SC-RF cavities to NEG thin-film coatings
• Cris Benvenuti and the very first Nb
coating team in 1983.
• The study was triggered by Emilio
Picasso, the former leader of the SC
cavities development then LEP
project leader.
• At that time only bulk Nb was
considered for SC-RF cavities.
• Cris Benvenuti first technical note about
linear pumping by NEG strips in June 1977.
• At that time, the study was not directly
motivated by the LEP.
• At that time, linear ion pumps had being
used.
Nadia
Circelli
1
Max
Hauer
Jacques
Genest Cris
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 4
NEG linear pumping and Nb coated SC-RF cavities: two technological
achievements for the LEP
Installation of a LEP2 cryomodule with Nb
coated RF cavities (1995)
First prototype of the LEP dipole beam pipe
with the St101 NEG strip in the antechamber
(1983).
1
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 5
The merging: first indirect trials
Cris Benvenuti, private communication:
• Cris had the first idea of coating the inner wall of vacuum chamber by NEG materials in
the early Eighties.
• He tried to sputter a St-101 strip in a LEP chamber (Al alloy AA 6060).
• The ultimate pressure after sputtering was excellent, but the re-activation impossible.
Cris Benvenuti et al. LEP2-Note-94-21
Copper coatings for the main couplers of the LEP2 superconducting RF cavities
Jordi Fraxedas (EU fellow) performed ESD measurements of Cu and Ti coated stainless-
steel RF coupler extensions. For Ti coating, he measured an abrupt decrease of the
desorption yields after heating at 350°C (2 hours).
This measurement was not considered as unexpected since the ‘activation’ of Ti bulk was
already known from previous experience in the ISR. The ‘activation’ had been recorded at
350°C (24h) for Ti experimental vacuum chambers.
1
4/6/2014 Document reference 6
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 7
Intertwisted Ti and Zr elemental wires
It became clear that NEG thin-film
coatings would have been used in the
LHC only if materials with lower activation
temperature could be found.
1
Stainless steel ≤ 350°C
Copper alloys ≤ 250°C
Aluminium alloys ≤ 200°C
Beryllium ≤ 200°C
In the 1962 edition of the Lafferty,
it is written that Ti-Zr alloys are known to
activate after heating at 300°C.
It was a common practice in vacuum
technology.
We looked for wires or rods made of Ti-Zr
alloys.
1995: Proposal for intertwisted Ti and Zr
wires.
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 8
January 1996: faster activation with Ti and Zr, one wire each 1
10-11
10-10
10-9
10-8
10-7
10-6
100 150 200 250 300 350 400 450
st.steel
Ti-Zr
Hf-Ti
Hf-Zr
P
[Tor
r N
2 e
q.]
2 Hours Heating Temperature [°C]
10-11
10-10
10-9
10-8
10-7
10-6
100 150 200 250 300 350 400 450
st.steelTiZrHf
P
[Tor
r N
2 e
q.]
2 Hours Heating Temperature [°C]
10-11
10-10
10-9
10-8
10-7
10-6
100 150 200 250 300 350 400 450
st.steel
V
Nb
Ta
P
[Tor
r N
2 e
q.]
2 Hours Heating Temperature [°C]
10-11
10-10
10-9
10-8
10-7
10-6
100 150 200 250 300 350 400 450
st.steel
Ti-Zr
Zr-V2
Ti-Nb
P
[Tor
r N
2 e
q.]
2 Hours Heating Temperature [°C]
ESD
1 mA
500 V
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 9
December 1997: sometime wrong ideas give great results, TiZrV 1
01-1996: Ti-Zr
09-1997 Pd/TiZr
12-1997 TiZrV
During an informal meeting (V.
Rouzinov, J-M. Cazeneuve, F. Cicoira,
and myself) we decided to add a wire
of an element of the 5th group in Ti-Zr
braid. If positive, the result would have
been our Christmas present for Cris.
V was selected because it has the
largest O diffusivity, so it could have
increased the mobility of O in TiZr…
10-11
10-10
10-9
10-8
10-7
10-6
100 150 200 250 300 350 400 450
st.steel
Ti-Zr
Ti-Zr-V
P
[Tor
r N
2 e
q.]
2 Hours Heating Temperature [°C]
10-7
10-6
10-5
10-4
10-3
10-2
10-1
100 150 200 250 300 350 400 450
H2
CH4
COCO
2
Eff
ect
ive
Deso
rpti
on Y
ield
[mol
ecu
les/
elect
ron]
2 Hours Heating Temperature [°C]
R. Kersevan
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 10
IVC-1998 in Birmingham 1
+
= Presented at
EVC-1999
Lyon (France)
ESRF stainless steel undulator
vacuum chambers
&
change of EU regulation
in matter of RP
10-7
10-6
10-5
10-4
10-3
10-2
10-1
100
101
102
103
104
105
106
Eff
ecti
ve d
esorp
tion
yie
ld
[mole
cule
s p
hoto
n-1]
Dose [mA h]
Activation
= 0,38 D-1
= 4,2 x 10-5 D
-0,38
Dose [photons m-1]
1019
1020
1021
1022
1023
1024
(2.3x10-3)
(6.1x10-6)
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 11
An unexpected support after the IVC-1998 in Birmingham
‘The new alternative technology for flat screen displays, namely Field Emission
Displays require UHV to work. It could be possible to use NEGs to provide pumping
in the small space allowed in between the two panels.’
(see The Economist, September 12th-18th 1998, p.96-97)
1
A more systematic study started in 1998
4/6/2014 Document reference 12
Ti
Zr V
Quick activation: no + yes
(degrees)
30 40 50 60
TiZrV Substrate
2 (degrees)
30 40 50 60
TiZrV
Substrate
2
STM
100 nm
A. Prodromides, Ph.D Thesis No. 2652 (2002), Faculte´ Sciences de Base,
Section de Physique, EPFL; and references therein. CERN Library URL:
http://documents.cern.ch/cgibin/setlink?base=preprint&categ=cern&id=
cern-thesis-2002-042.
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 13
Optimization of the deposition parameters: the role of the substrate
temperature 1
300°C
250°C
200°C 1 m
100°C
150°C
350°C
10-3
10-2
10-1
100
101
10-7
10-6
10-5
10-4
10-3
1013
1014
1015
1016
Pum
ping
Spe
ed [
s-1 c
m-2
]
CO surface coverage [Torr cm2]
[molecules cm-2
]
coated at 300 °C
on rough Cu
coated at 300 °C
on smooth Cu
coated at 100 °C
on smooth Cu
C. Benvenuti, P. Chiggiato, P. Costa Pinto, A. Prodromides, V. Ruzinov,
Vacuum 71 (2003) 307.
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 14
1999-2000 After activation, TiZrV has a low SEY: a solution to ecloud in LHC
SEY after saturation with CO
1
B. Henrist, N. Hilleret, C. Scheuerlein, M. Taborelli, App. Surf. Sci. 172
(2001) 95.
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 15
Installation in accelerators: the production of undulator vacuum
chambers started in 1999 for the ESRF
1
EPAC 2000 and AVS-2000 Boston
ESRF: 10 chambers already inserted within two year
ELETTRA: 1 chamber installed (11/2000)
NSRRC: 1 prototype chamber coated, test in August 2002
DIAMOND: 1 prototype chamber coated (11/2000)
SOLEIL: under consideration By J
un
e 2
002
First chamber produced for other Institutes: PSI June 1999
4/6/2014 Document reference 16
Cross section of the TiZrV coated ESRF’s vacuum chamber (length=
5.5 m)
No effect on the machine global impedance was observed.
First ESRF Al-alloy vacuum chamber 1
4/6/2014 Document reference 17
SAES was the first industrial partner
(January 2001)
2000-2001: Licensing of ESRF and SAES Getters 1
ESRF was the first institute/firm to
be licensed by CERN (2000).
The support of Jean-Marc Filhol
was important.
Other licensees:
• DESY, GSI, LNLS
• VARIAN (now Agilent), FMB
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 18
First hypothetical application in the LHC: TiZrV in a beam
screen without holes
1
• The coating of a tight beam screen was inquired by L.
Evans, the LHC project leader.
• Heating would have been possible by 4 molybdenum
heaters isolated by alumina tubes.
• The main advantages would have been:
• No need for beam scrubbing in the LHC arcs.
• He leaks from the capillaries would have been less
critical.
• Possible heating of the beam pipe.
However:
- The pumping efficiency of TiZrV at 20 K was not confirmed.
- Pumping of CH4?
- The cryopumping of the cold bore would have been lost.
- It was too late to start a parallel study of integration.
- Additional costs for coating and cabling.
Management decision: Forget beam-screen and be ready for the coating of
experimental and long-straight section beam pipes.
4/6/2014 Document reference 19
2 TiZrV thin film coating for the LHC
• 6 Km of beam pipe
• About 1400 vacuum chambers
• Experimental vacuum chambers
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 20
In 2003 the team focused on the facility for the LSS series production 2
• The development and the construction of the coating
facilities started at the end of 2002.
• Pre-series of LSS standard chamber: November 2003
2013
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 21
The series production started in February 2004 1 The new set-up for the coating of the LSS chambers
Manifold and chambers on the bench
The solenoids
The cathodes
Slide prepared
for the
Department
Head
4/6/2014 Document reference 22
3mm wires of Ti, Zr and
V
The coating system 2
4/6/2014 Document reference 23
316LN st. steel: allows high T treatments without softening
OFE copper: high conductivity,
reliable brazing with st. steel
OFS copper: high conductivity, better mechanical performance than OFE at high temperature
L max = 7 m ID = 80 mm thickness = 2 mm
The LSS standard vacuum chamber 2
Courtesy of S. Atieh
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 24
The adhesion of the TiZrV film on copper was probed by high-pressure water
rinsing and/or ultra-sound agitation in deionized water.
October-November 2003: modification of surface treatment for copper tubes 2
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 25
Flake with copper
on one side and
NEG on the other
October-November 2003: modification of surface treatment for copper tubes 2
4/6/2014 Document reference 26
October-November 2003: modification of surface treatment for copper tubes 2
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 27
The coating team in September 2004 2
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 28
2004: Ultrathin heaters for vacuum chambers inserted into magnets 2
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 29
Functional quality control: XPS and pumping speed 2
Coupon samples for SEM and XPS 25-cm long vacuum chambers for
pumping speed measurement
4/6/2014 Document reference 30
0 50 100 150 200 250
Heating Temperature [ºC] - 1 hour duration
O 1
s p
eak a
rea (
A.U
.)O 1s peak area
Functional quality control: XPS and pumping speed 2
Courtesy of M.Taborelli
4/6/2014 Document reference 31
Functional quality control: XPS and pumping speed 2
10-3
10-2
10-1
100
1010
1011
1012
1013
1014
1015
1016
Cap
ture p
rob
ab
ilit
y
Q [molecules cm-2
]
CO
4/6/2014 Document reference 32
NEG coated chambers for the LHC
Production weeks
Num
ber
of
cham
be
rs
0
200
400
600
800
1000
1200
0 10 20 30 40 50 60 70
Mar 2004 Oct 2004
454 7m long
standard LSS chambers
Jun 2005 Jan 2006
650 chambers
MBXW, MQW, MBW, MCBW,
short LSS, experiments, etc
Series production 2
Courtesy of P. Costa Pinto
4/6/2014 Document reference 33
Series production 2
The LSS chamber was installed in the LHC tunnel in the summer 2006
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 34
Claudio Raggi, Master Degree Thesis, 2005
Activation in the LHC tunnel 2
4/6/2014 Document reference 35
NEG activation monitoring.
Installation of the bakeout heater and
control
Activation in the LHC tunnel 2
4/6/2014 Document reference 36
Pressure profile during heating
180°C RGA + SVT
140°C Sec. Valves
250°C Flanges
150°C S.S.
230°C NEG
Activation in the LHC tunnel 2
Courtesy of G. Bregliozzi
4/6/2014 Document reference 37
End of bakeout and NEG activation: acceptance test.
0
2 10-10
4 10-10
6 10-10
8 10-10
1 10-9
200 300 400 500 600 700
Mass 2
RG
A s
ign
al [A
]
Time [s]
Valve closed
Valve opened
TMP
RGA
St. Steel module
Valve
Beam pipe
Beam pipe
Experiments in the lab and a dedicated Monte Carlo simulation indicate that a good activation is obtained when the RGA H2 signal decreases by a factor of at least 3 when opening the valve.
H2
Sector A7R5
Activation in the LHC tunnel 2
4/6/2014 Document reference 38
Average pressure, measured by BA gauges, 1 month after the end of bakeout: P ≈ 6x10-12 mbar N2 eq.
Ultimate pressures after activation and cooling 2
Courtesy of G. Bregliozzi
4/6/2014 Document reference 39
The ultimate pressure was limited by the outgassing of the uncoated st. steel module
With BA Gauge ON: QH2 ≈ 4.5 · 10-9 mbar /s
NEG Pumping Speed for H2 ≈260 /s
Degassing of the Stainless Steel Module:
P (N2 eq) ≈ 7 · 10-12 mbar N2 eq.
2
Courtesy of G. Bregliozzi
4/6/2014 Document reference 40
TiZrV coating during the Long Shutdown 1: assorted geometries 2
Courtesy of P. Costa Pinto
4/6/2014 Document reference 41
TiZrV coating during the Long Shutdown 1: assorted geometries 2
Courtesy of P. Costa Pinto
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 42
MolFlow+: ray-tracing, colour-coded pressure profile, and 2000 l/s NEG lump pump
2 New project: ELENA. NEG coating needed in 2015.
ELENA ring:
Calculation of the pressure profile
seen by the antiproton
beam along one dipole magnet
Courtesy of R. Kersevan
Vacuum, Surfaces & Coatings Group
Technology Department
3 NEG coatings in synchrotron radiation facilities
CERN
ESRF ELECTRA
SOLEIL DESY DIAMOND
MaxLab
BNL
LNLS
KEK
NSRRC
AS
ISA
ALBA
users
SESAME
ILSF
ANL
, in design/study
4/6/2014 Document reference 44
BPM location
Ion pump location
Crotch absorber location
Sector valve location
VC10 – straight section for IDs
VC1
VC2 VC3
VC4
VC5
VC6
VC7
VC8
VC9
One achromat Length 26 m
In each achromat: -> 10 BPMs -> 3 pumping ports (with ion pumps), -> 1 crotch absorber, -> 3 gate valves.
3 MAX IV in Lund
Courtesy of M. Grabski (MAX IV)
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 45
Distributed
cooling
Ribs
Cooling for
corrector
area
Cooling for
corrector area
Welded
bellows
Welded
bellows
Chamber
body
Inside diameter: 22
mm,
Total length: 2.5 m,
Bent part:
Arc length 1 m,
Bending angle 30,
Bending radius 19 m.
Bent part
3 MAX IV in Lund: the ‘standard’ vacuum chamber
Courtesy of M. Grabski (MAX IV)
4/6/2014 Document reference 46
3 MAX IV in Lund: the light-e beam bifurcation chambers (VC1 and VC2)
Ø25
5
7 Ø22
3
Ch
am
ber
exit
Ch
am
ber
en
tran
ce
Courtesy of M. Grabski (MAX IV)
4/6/2014 Document reference 47
3rd coating – 3+1 cathodes (9e-1 mbar, 240V, 0.28A, 68W)
360 Gauss
7th coating - 1 cathode (3.1e-1 mbar, 700V, 0.03A, 26W)
260 Gauss, diode
9th coating - 2 cathodes (2.7e-1 mbar, 775V, 0.06A, 47W)
360 Gauss, magnetron
VC2 with antechamber:
360 -> 270 Gauss – 1 cathode
12th coating - 1 cathode (3.6e-1 mbar, 650V, 0.04A, 26W)
3 MAX IV in Lund: the light-e beam bifurcation chambers (VC1 and VC2)
Courtesy of M. Grabski (MAX IV) and P. Costa Pinto
4/6/2014 Document reference 48
Courtesy of S. Marques Dos Santos
The standard coating of very small and asymmetric cross-section beam pipes:
• requires a significant amount of time for the optimization of the coating
parameters;
• the quality and uniformity of the film and its performance are unmeasurable;
In addition, at present, vacuum chambers are not designed to facilitate surface
treatments and coating.
4/6/2014 Document reference 49
4 Future developments
Ju
ly 1
999
CE
RN
qu
ota
tio
n f
or
ES
RF
4/6/2014 OLAV IV Taiwan - Paolo Chiggiato 50
4 Future developments: the electroforming way
Courtesy of L. Marques Antunes Ferreira
4/6/2014 Document reference 51
4 Future developments: the electroforming way
Courtesy of L. Marques Antunes Ferreira
4/6/2014 Document reference 52
4 Future developments: the electroforming way
Courtesy of L. Marques Antunes Ferreira
4/6/2014 Document reference 53
4 Future developments: the electroforming way
Courtesy of L. Marques Antunes Ferreira
4/6/2014 Document reference 54
Acknowledgments
I warmly thank:
• Pedro Costa Pinto
• Giuseppe Bregliozzi
• Vincent Baglin
• Giulia Lanza
• Leonel Marques A. Ferreira
• Mauro Taborelli
• Marek Grabski
• Roberto Kersevan
for their help in the preparation of
this presentation;
• Cristoforo Benvenuti
• José Miguel Jimenez
for useful discussions.