51 Kaabi Master Titanium Nitride For Rf Windows Enzo Palmieri

For this and many more thesis, visit the free download area on: http://www.surfacetreatments.it/ http://www.slideshare.net/PalmieriProfEnzo

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

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

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

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

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

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

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

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

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

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

TiN sputtering system Sample holder motion Master thesis presentation 12 February 16 th 2009

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

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

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

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

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

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

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

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

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

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

For this and many more thesis, visit the free download area on: http://www.surfacetreatments.it/ http://www.slideshare.net/PalmieriProfEnzo

51 Kaabi Master Titanium Nitride For Rf Windows Enzo Palmieri

by ISTITUTO NAZIONALE FISICA NUCLEARE & University of Padua

Accessibility

Upload Details

Usage Rights

Statistics

51 Kaabi Master Titanium Nitride For Rf Windows Enzo Palmieri Presentation Transcript