LAL-Orsay is developing an important effort on R&D and technology studies on RF power couplers for superconductive cavities. These are complex and high technology devices due to their basic functions: RF power matching between source and cavity, vacuum and temperature separation from the environment to the cavity. One of the most critical components of high power couplers is the RF ceramic window that allows the power flux to be injected in the coaxial line. The presence of a dielectric window on an RF power line has in fact a strong influence on the multipactor phenomena, a resonant electron discharge that is strongly limiting for the RF components performances. The most important method to reduce the multipactor is to decrease the secondary emission yield of the ceramic window. Due to its low secondary electron emission coefficient, TiN thin film is used as a multipactor suppressor coating on ceramic coupler windows. In addition, TiN permits to drain away electric charges on the surfaces to avoid material break down.
In this framework, a sputtering machine was developed allowing thin layer titanium nitride coating on ceramic. The coupler operating conditions, the physical properties of TiN layer and alumina substrate in addition to windows geometries have defined the strict constraints that have been taken into account in the definition of the coating bench design. In this study, a full description of sputtering machine will be given.
By maintaining a constant bias and a fixed ionisation gas flow, an optimisation of reactive gas flow will be necessary for stoichiometric deposit obtaining. Once these parameters determined, a study of deposition rate variation for different process pressure value will be done. The influence of substrate to target distance on deposit stoichiometry and thickness will be also presented.
As TiN deposit thickness is a very important parameter to control, a correlation between quartz crystal microbalance given value and the real deposition thickness is determined for a better in-situ monitoring.
25 Frigo Magnetron Sputtering Into S C R F Cavities Enzo Palmieri
51 Kaabi Master Titanium Nitride For Rf Windows Enzo Palmieri
1. For this and many more thesis, visit the
free download area on:
http://www.surfacetreatments.it/
http://www.slideshare.net/PalmieriProfEnzo
2. Master in Surface Treatments for Industrial Applications
Titanium Nitride Coating by Reactive DC Magnetron
Sputtering as a Multipactor Suppressor on Coupler
RF Ceramic Windows
Candidate: Dr. Walid KAABI
Thesis Advisors: Prof. V. Palmieri
Dr. A. Variola
Laboratoire de l’Accélérateur Linéaire
3. Power coupler for superconductive cavities
Principle and Functions
Power coupler have 3 functions to fulfil:
Electromagnetic role
Vacuum barrier
Thermal interface
Master thesis presentation 3 February 16 th 2009
4. Power coupler for superconductive cavities
Coupler Design: example of TTF-III Model
RF Source
Cryogenic module
Cavities
TTF-III is the baseline solution for the future XFEL linac in DESY. In this framework, our
laboratory is responsible of the RF conditioning task of 800 TTF-III couplers.
Master thesis presentation 4 February 16 th 2009
5. Power coupler for superconductive cavities
Coupler Ceramic Windows
Cold windows
(Ø = 47 mm, h= 48 mm)
Warm windows
(Ø = 75 mm, h= 57 mm)
Windows made of Alumina ceramic (97.6% Al2O3):
Mechanical and dielectric strength
High thermal stability
Low out-gazing rate
High Secondary Electron Emission Yield (SEY)
Master thesis presentation 5 February 16 th 2009
6. Multipactor phenomenon
Conditions of occurring
RF Field + + High vacuum
Best conditions for Multipactor phenomenon occurring:
Electron-avalanche discharge
Master thesis presentation 6 February 16 th 2009
7. Multipactor phenomenon
Principle
(A) Accelerated primary electron strikes the surface material with an impact energy in the
range corresponding to SEY>1, and extract secondary electrons,
(B) Field inversed and accelerated secondary electron hurt the opposite surface and eject
other electrons,
(C) The phenomenon still going on multiplying electrons.
Master thesis presentation 7 February 16 th 2009
8. Multipactor phenomenon
Damages caused in ceramic windows
Multipactor phenomenon may cause damages in ceramic windows:
Creation of damaging arcs,
Surface overheating that may lead to surface evaporation,
Load mismatch causing dangerous power reflection to the RF source.
Necessity of suppression, or at least limitation of this effect
Solution: decrease Alumina’s SEY by surface coating with multipactor
suppressor thin layer: Titanium Nitride
Master thesis presentation 8 February 16 th 2009
9. Multipactor suppressor thin film
TiN is characterised by a low SEY that remain stable on RF
operational conditions
Master thesis presentation 9 February 16 th 2009
10. Multipactor suppressor thin film
DC Reactive Magnetron Sputtering of TiN
Vaccum chamber
Ionisation gas inlet (Ar)
Substrate Reactive gas inlet (N2)
N2
polarisation
Electrical
Ar +
- Target
Magnetron
Pumping
Master thesis presentation 10 February 16 th 2009
11. TiN sputtering system
Rotating magnet
pack
Titanium target
Rotating magnet
shape
=
Shield Plasma shape Rotating sample holder
Sputtering machine overview
Master thesis presentation 11 February 16 th 2009
12. TiN sputtering system
Sample holder motion
Master thesis presentation 12 February 16 th 2009
13. Results and discussions
Practical guidance’s
Ceramic window exposure to RF field induce some additional constraints when
studying TiN deposition:
Stoichiometry control: the layer can get the multipactor suppressor property,
Thickness control: deposit should be thick enough to reduce multipactor, but not
so much to prevent increasing RF power reflection in ceramic surface.
XRD analyses for stoichiometry control: hundred nanometers deposit thickness,
X-Ray reflectivity for thickness control (in addition to microbalance monitoring):
deposit of some nanometers,
Deposits are made on 10x10mm quartz substrates,
Some special sample holder will be used in some particular experiences.
Master thesis presentation 13 February 16 th 2009
14. Results and discussions
Parameters optimization for stoichiometric deposit
Ar flow rate maintained at 0.1 sccm
Imposed current: I=3A
N2 flow rate variation gives TiNx deposits
x = 1,022 x = 1,089 x = 0,977
1400
Multi-sample holder
1200
T iN (220)
1000
From XRD plots, we determine d220
T iNx (111)
according to Bragg’s law:
R elative Intens ity
800
T iNx (220) n
d 220
600 2 sin
400 TiNx Lattice parameter is then calculated:
200
aTiNx d 2 20 h 2 k 2 l 2
0
30 35 40 45 50 55 60 65 70 75 80 85 90 95
Ang le (2θ)
Master thesis presentation 14 February 16 th 2009
15. Results and discussions
Parameters optimization for stoichiometric deposit
From the obtained values of lattice parameters, we calculate x, the N/Ti ratio
according to the relation below, valid in the range 0.6 < x < 1 *:
aTiNx = 4.1925 + 0.0467x
Ar (s c c m) N2 (s c c m) I (A) 2θ d (220) (Å ) a (Å ) x(T iNx)
0,10 0,11 3 61,8629 1,50115 4,2459 1,1423
0,10 0,12 3 61,8430 1,50028 4,2434 1,0896
0,10 0,13 3 61,8938 1,49917 4,2403 1,0223
0,10 0,14 3 61,9277 1,49843 4,2382 0,9775
0,10 0,15 3 62,1270 1,49410 4,2260 0,7153
* S. Nagakura et al. Thin Solide Films, vol 158, issu 2, (1988), 225-232.
Master thesis presentation 15 February 16 th 2009
16. Results and discussions
Deposition rate variation with process pressure
Once stoichiometric parameters reached (fAr = 0.1 sccm, fN2 = 0.14 sccm, I = 3A), we vary
process pressure by changing both gases flow rate keeping the same ratio obtained in
stoichiometric conditions (fAr / fN2 = 0.71):
5
4,5
Quiet linear dependency
Deposition rate (Å s-1)
4
3,5
3
2,5
2
1,5
1
0,5
Deposition rate decrease with increasing process pressure
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Process pressure (x10-3 mbar)
Continuous decrease of deposition rate: Target not totally poisoned
Deposition rate estimation is possible for a given process pressure
Master thesis presentation 16 February 16 th 2009
17. Results and discussions
Thickness control of stoichiometric deposit
In-situ deposition rate and film thickness monitored by quartz crystal microbalance
Due to its emplacement, quartz crystal of
the microbalance receive sputtered atom
from both target
While
Our samples receive deposition only from
the target in front of witch they are placed
Quartz cristal
location
Correlation must be found between
thickness given by microbalance and the
real one on samples
Master thesis presentation 17 February 16 th 2009
18. Results and discussions
Thickness control of stoichiometric deposit
Measured thickness: 16.2 nm Measured thickness: 37 nm
Measured thickness: 23 nm Measured thickness: 72 nm
Expected thickness (nm) Mesured thickness (nm) Density (g cm-3) Roughness (nm) Very good agreement between
the two thickness values:
16.2 16.2 4.4 1.9
23.4 23 3.48 2 No effect of sample orientation
36.0 37 3.4 3.9
(horizontal or vertical ),
according to sputtered atom flow,
71.1 72 3.09 7
on thickness of deposit received
Master thesis presentation 18 February 16 th 2009
19. Results and discussions
Influence of substrate-target distance on deposit thickness
Sputtering conditions:
fAr = 0.1 sccm; fN2 = 0.14 sccm; I = 3 A;
Pbase = 2.4 10-6 mbar; Pprocess = 5.33 10-3 mbar;
VDep = 3 Å s-1; tDep = 68 s;
5 samples at respectively: d1 = 6.5 cm, d2 = 11.5 cm, d3 = 16.5 cm, d4 = 22 cm, d5 = 25.5 cm.
Multi-sample holder at
different distances from
the target
Master thesis presentation 19 February 16 th 2009
20. Results and discussions
Influence of substrate-target distance on deposit thickness
60
Clear influence of target-substrate
50 distance on film thickness:
Deposit Thikness (nm)
40
More the sample is close to target,
30 the thicker will be the deposit
20
(regardless to the stoichiometry)
10 Influence of sputtered atom flow
0
2 4 6 8 10 12 14 16 18 20 22 24 26 28
Target-substrate distance (cm)
Deposit is not uniform along vacuum chamber axis. We should precise that this conclusion is
available when we consider only one target.
This result is not constraining for deposition on cylindrical ceramic windows, since it will receive
sputtered atoms from both targets. Thus, every target will contribute to film thickness according
to the distance that separate it from the substrate
Master thesis presentation 20 February 16 th 2009
21. Conclusions
XRD analyses of TiNx films shown that stoichiometry was reached for an N2 flow rate
of 0.14 sccm with a fixed Ar flow rate of 0.1 sccm and an imposed current bias of 3A,
Process pressure influence on deposition rate is quit linear (maintaining the fAr /fN2 as
for stoichiometric conditions). Deposition rate decrease with process pressure increase.
This tendency allows the prediction of deposition rate by a simple pressure lecture. This
can be useful in case on deposition of very thin films,
Good agreement between thickness measurements given by microbalance and the
one measured by Reflectivity. No effect of sample orientation (horizontal or vertical )
according to sputtered atom flow on thickness of deposit received,
Important influence of the target-substrate distance on deposited film thickness due to
sputtered titanium emission flux difference. However, this result is no constraining for the
deposition on cylindrical windows.
Master thesis presentation 21 February 16 th 2009
22. Acknowledgments
Thanks to A. Variola and Pr. Palmieri who allowed me to follow this master courses.
Thanks to all my master colleagues and all the staff of the LNL.
Thanks to my colleagues in LAL for their help each time that I asked for.
Master thesis presentation 22 February 16 th 2009
23. For this and many more thesis, visit the
free download area on:
http://www.surfacetreatments.it/
http://www.slideshare.net/PalmieriProfEnzo