ma k i n g i n n o v a t i o n h a p p e n , t o g e t h e r
SAES experience in NEG coating of narrow gap insertion devices and
small diameter chambers for accelerators
Gero Bongiorno*, Paolo Manini, Enrico Maccallini, Fabrizio Siviero, Tommaso Porcelli, Stefano Raimondi
SAES Getters S.p.A. Viale Italia 77, 20020 Lainate (Milan) Italy
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r2
Outline
Narrow gap insertion devices : status of the activity at SAES
The new challenges : < 10 mm diameter NEG coated tubes
Status of the R&D activityResults of the characterizazion of the coated chambersCritical issues and challenges
Next steps
Conclusions
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r3
Narrow gap IDs : SAES experience
SAES getters has been pioneering NEG technology since the 60s,developing a variety of NEG alloys, getter related technologies andNEG solutions which have found wide use in a variety of industrial andresearch applications (X‐ray tubes, lamps, sealed off devices,accelerators, purifiers for the semiconductor industry ect).
About 15 years ago SAES added, through a licence agreement withCERN, the NEG coating technology to its product portfolio as this wasnicely complementing the product family based on lumped NEGpumps (e.g. Capacitorr pumps).
Since then a substantial R&D and NEG coating activity has been donein our company to :
industrialize the processexplore the feasibility of NEG coating of complex shape chambers,including narrow gap insertion devices.
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r4
Trends in narrow gap chambers…
The first large synchrotron project using relatively small section chambers was theSOLEIL ring, which was the first light source to adopt NEG coating extensively (> 55%of the ring circumference).For SOLEIL ring chambers , cathode wires could be suitably positioned inside thecross section of the chambers as this was reasonably large.
Depending on the required film specifications,2 or even 3 cathodes were used in theassembly.Spacing among cathodes was modeled toensure that the thickess distribution was asclose as possible to user’s requirements.Using three cathodes was for examplepossible to deposit more film at the edges ofthe chamber (e.g 1 micron) and less in thecentral part (e.g 0,5 micron), thus minimizingpossible impedance effects.
Courtesy of SOLEILCopyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r5
Trends in narrow gap chambers…In parallel with SOLEIL, several otherfacilities (DLS, MAX, NSRRC, PSI…) started to ask for NEG coating of insertion deviceswith gaps in the range between 12 and 9 mm.In the following years gaps have decreasedin size down to 7 mm. Also the sectionlength was accordingly reduced (from 70 mm to 35 mm, typically)
The combination of both effects make the coating much more difficult as
Inserting and mounting the cathodes ismore critical
the plasma volume is pretty much reducedwhich may cause plasma instabilitiesand/or dis‐uniformities unless properlyaddressed.
Plasma adjustments may require higherworking pressures
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r6
Typical thickness variation in a narrow IDA typical thickness distribution is showed in the chart below for the ID section given in thepicture.Average film thickness is about 0,6 micron, however thickness can move from a 0,3 micronminimum to a 1,5 micron high, depending on the area.This is unavoidable as there is no way to prevent sputtered atoms to deposit and grow a filmcloser to the emitting source (the cathode).Use of three cathodes can mitigate the issue and smooth somewhat the thicknessdistribution.
0
0,25
0,5
0,75
1
1,25
1,5
1,75
0 1 2 3 4 5 6
thic
knes
s (µ
m)
half perimeter lenght (cm)
Cathode Cathode
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r
Typical thickness variation in a narrow ID
100%30%
Coating thickness
3 Cathode wires
Copyright SAES Getters SpA7
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r
Typical thickness variation in a narrow ID
Cathode Cathode Cathode
Copyright SAES Getters SpA8
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r9
What is achievable, main limitationsIDs down to 8mm gap and up to 6m lenght have been coated in our facility.IDs 7mm up to 2,5 m long have been also successfuly processed. Thesputtering setup is well established so we do not expect major issues evenin coating longer chambers.
However, the following important points have to be considered :The use of three cathodes is almost impossible when the section is smallerthan 50 mm. Space for inserting them in a controlled and reliable way is verysmall.Cathodes at the edges may be just too close to the chamber, making thesputtering process unstable. This is especially true for long IDs ( e.g. 4‐6 m)where the possibility for the cathodes assemblies to move a few mm along thewhole lenght cannot be ruled out.With two cathodes you have to be aware that you have more thicknessvariation along the section profile.Cleaning of long IDs is also more critical as the flowing of reactants andchemicals is more dfficult
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r10Acronym 09/10/2015
What is achievable, main limitationsBased on this, we can consider as a rule of thumb that a practical limit for NEG coating with the present cathode technology may be an elliptical ID with section of 7mmx30 mm Lenght of 6 m
It is probably still possible to go a bit further down, however the users have to accept a more drastic thickness variation along the section ( e.g. 0,2 micron‐2,5 micron)
Elliptical sections: same perimeter, same ammount of deposited material
Elliptical section 60mm x 11,5mm Elliptical section 62mm x 6mm
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r11
Small diameter pipes
Since a few years now, pipes with internal diameter as small as 4 mm areincreasingly considered as insertion devices for high emittance lightsources. There is therefore a growing interest in the scientific acceleratorcommunity in understanding the coating process and feasibility.
As of now we have coated and characterized stainless steel pipes withdiameter of
10 mm8 mm6 mm
Pipes with 4 mm diameter have beenprepared andare being tested in these weeks.
An update of this activity is providedin the following
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r12
Small diameters pipes: 10mm
Pipes lenght 2m, internal diameter 10mm, stainless steelPlasma could be ignited without major issues and coating processwas smooth with coating parameter recorded and stableThe pipes were cut and opened and visually and SEM inspected tocheck for uniform presence of the film Covered area: 100%Thickness measurement were carried out by EDS analysis
Average thickness about1 micron (± 30%)
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r13
Small diameters pipes (10mm): chemical composition uniformity
Ti V ZrNominal
composition30 %at 40 %at 30 %at
ExperimentalAverage
29,8 %at 40,0 %at 30,2 %at
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r14
The coated film is quite conformal and follows the surface roughness of the substrate.
Clean substrate surface Few microns thick NEG film
Small diameters pipes (10mm): SEM analysis
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r15
Tipical morphology on Si substrates (reference)
Si substrateNEG film on Si substrate
The coated film is quite conformal and follows the surface roughness of the substrate. As a reference this is the morfology of a NEG film (1,5 micron thick) deposited on a flat silicon wafer.
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r16
Small diameters pipes: 8mm
Pipes lenght 2m, internal diameter 8mm, stainless steelAlso in this case plasma ignited without issues and the coatingprocess was smooth with coating parameter recorded and stableThe pipes were cut and opened and visually and SEM inspected tocheck for uniform presence of the film Covered area: 100%Thickness measurement were carried out by EDS analysis.
Average thickness about1 micron (± 40%)
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r17
Small diameters pipes (8mm): chemical composition uniformity
Ti V ZrNominal
composition30 %at 40 %at 30 %at
ExperimentalAverage
25,3 %at 46,3 %at 28,4 %at
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r18
Small diameters pipes (8mm): SEM analysis
Film thickness: 1,5m
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r19
Small diameters pipes (8mm): ASTM tests
ASTM* tests have been carried out on 20 cm long coated pipes. These tests were notaimed at measuring the pumping speed (due to the pipe conductance limitation) butrather to assess the getter capacity for CO. Measured capacity values are in line with previous SAES measurement (see next slides). More tests are in progress.
*ASTM F798‐97Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r20
SAES historical data on NEG coating capacity for CO
SAES internal technical report (2006)
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r21
Small diameters pipes: 6mm
Initial tests on 6 mm stainless steel pipes, 2m lenght have showed a number ofissues related to the mounting of the cathodes and the stabilization anduniformity of plasma conditions.
To segregate the different factors we have reduced the test pipe length to 1 m.Plasma was ignited succesfully, even though higher plasma pressure wererequired. First ignition with the fresh cathode was more tricky than thefollowing ones. It seems that cathode ageing (removal of the oxide layer) helpsin ignite the plasma.As in previous cases also these chambers were tested for :
Film uniformity across the whole chamber length Covered area: 100%Film chemical composition and thickness uniformity.
Thickness was comprised between 0,8 and 1,8 micron (average value 1,3micron ± 30% )Chemical composition was close to the nominal values
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r22
Small diameters pipes (6mm): chemical composition uniformity
Average thickness about1,3 micron (± 30%)
Ti V ZrNominal
composition30 %at 40 %at 30 %at
ExperimentalAverage
30,5 %at 40,3 %at 29,2 %at
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r23
Small diameters pipes (6mm): SEM analysis
Stainless steelsubstrate
1.8m thickNEG film
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r24
Small diameters pipes: 4mm (work in progress)
Pipes have been prepared and are now ready for coating.
We will start with 1 m long samples and we will then increasethe length.
On these pipes in addition to the standard characterization wealso plan to measure the Kr outgassing properties. We expectin fact that due to the higher gas pressure needed, Kr may betrapped in the film and released later during the machineoperation, which would be detrimental to the electron beam.
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r25
Summary and conclusions
Narrow gap IDs have become quite popular in the past 10 years in light sources.Even though far from trivial, reliable and consistent NEG coating of IDs up to 6 m andwith gaps in the order of ≈7 mm is currently achieved in our facility.Almost 100 IDs have been delivered by SAES in the past 10 years which are currentlybeing used in most of the new generation machines worldwide.Chemical composition uniformity and thickness distribution is ensured by a stringentmanufacturing protocol and process control.
R&D activity on small diameter pipes is also in progress. First results with pipesdiameter of 6 mm indicate that homogenous coating through out the whole pipelength is possible at least up to 1 m length.NEG coating thickness can vary by within ± 40%, however there may be room fromimprovements as the process will be made more robust.In all cases, chemical composition was close to the nominal composition and withinthe acceptable range.Tests on 4mm pipes are in progress and we expect first results shortly.
Copyright SAES Getters SpA
m a k i n g i n n o v a t i o n h a p p e n , t o g e t h e r26
Summary and conclusions (2)
Coating 4mm pipes is not an easy walk as many issues have to be optimized. Inparticlular :
Adjust the plasma parameters and pressure to the given geometry (length, curvature…)Use high applied magnetic fields to ensure the plasma discharge is kept stable andrunning during the whole coating processDesign the cathode assembly in the proper way, to avoid shorts allowing at the sametime enough space for the pumping and plasma discharge volume. Use of very thincathodes is key with this regard.
Even though twisted multiple cathodes have been so far working, the use of asingle alloy wire might be beneficial. More considerations on this aspect areongoing.
Finally, as the pressure of the sputtering discharge has to be necessarily increased,a special focus must be given to Kr outgassing. Kr outgassing measurement as wellas pumpig speed tests on small diameter pipes are planned in the forthcomingmonths
Copyright SAES Getters SpA
www. s a e s g r o u p . c om
Thank you for your attention