1. 1 Nb Sputtered Cu QWR: a 20 years experience from an idea to the routine beam acceleration A.M. Porcellato S. Stark V. Palmieri F. Stivanello

2. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 2 Once upon a time, 20 years ago...  LNL started to think about ALPI, a heavy ion SC linac equipped with QW, Pb on Cu resonators, less expensive and easier to build than Nb QWR.  Since the beginning we investigated the possibility to substitute in the future Pb with Nb.  A laboratory for Nb QWR sputtering was set up and a devoted research project was funded in 1987.

3. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 3 V T ~ 14 MV (~13 MV at present) From H to 197 Au, E = 30 ÷ 1.5 MeV/A CW or pulsed XTU-Tandem ECR on 350 kV platform SC-RFQs and QWRs Veq ~ 8 MV PI Injector PIAVE SC Booster ALPI 70 SC Quarter Wave Resonators (Nb, Nb/Cu) V eq ~ 40 MV, species from 12 C to 197 Au Injected by Tandem or PIAVE TANDEM PIAVE ALPI COMPLEX

4. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 4 ALPI resonators , 80 MHz, full Nb=0.13, 160 MHz, Nb/Cu , 160 MHz,Nb/Cu , 160 MHz,Pb/Cu

5. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 5 QWR Sputtering development strategy  Sputtering of a standard medium  ALPI resonator (1998)  Upgrading of the medium  ALPI resonators (2003)  1987: Funding a research project on QWR Nb sputtering  1988-1999: DC biased sputtering choice and system set-up  1990: Obtaining good SC performance on samples  1991: Sputtering on a simplified prototype  1994: Design of a ALPI high  resonator suitable for sputtering production and compatible with existing cryostats  1995/1998 Production and installation of 4 high  cavities in ALPI  Sputtering on an Al substrate (2004)  New design of medium  substrate (2005)

6. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 6 LNL QWR sputtering milestones  1991: first Nb sputtered prototype produced;  1993: three prototypes reach 6 MV/m @ 7 W  1995: Four high  QWRs installed in ALPI (4 MV/m @ 7W); 7 MV/m @ 7W reached in laboratory  1998: Installation of four high  QWRs, operating at 6 MV/m @ 7W in ALPI  1998: First medium  ALPI QWR reaches 4 MV/m @ 7W  1999: The upgrading of medium  resonators begin  2003 : All the old 44 accelerating cavities have their Pb superconducting layer replaced by Nb

7. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 7 ALPI high  QWRs  CR20 resonators are in ALPI since 1988  The average operational accelerating field is 6 MV/m; the cavity CR20-3 is routinely amplitude and phase locked at 7.3 MV/m  CR19, housing cavities similar in shape, but having a brazed substrate, was installed ALPI in 2001  No other cryostats are available for installation of further cavities  =0.13, 160 MHz  Drilled by a billet of OFHC Cu, 99.95% certificate grade  No brazed joints, beam ports jointed by indium gaskets  Rounded shorting plate  Capacitive coupler

8. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 8  Brazed joints (especially the ones in the outer resonator surface)  Flat shorting plate  Beam ports shape  Inductive coupler (hole in high current region) The possibility of an effective improvement of the medium  resonators by Nb/Pb replacement was shown in 1998 ALPI  =0.11, 160 MHz Limited the reached performance to 4.6MV/m @7W, a lower level than the ones of high  resonators. Upgrading of ALPI medium  QWRs

9. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 9 ALPI upgrading results  46 Pb /Cu QWRs were upgraded by Nb sputtering.  In spite of operating at about twice higher accelerating fields than before, the cavities maintain the same cryostat, the same control system, the same rf hardware and software.  The performance is lower than that obtained by sputtering on new substrates, but the gap both in Q and Ea has been improving with time. Q 0 -value of 7x10 8 and Ea of 6 MV/m at 7 W were obtained in the last produced resonators.  The average E a in ALPI is however limited to 4.6 MV/m at 7 W, due to the lower E a of resonators produced in between 2001 and 2002, when only bad substrates were available and the production schedule was very tight.  The upgrading of medium  ALPI resonators gave a substantial increase in ALPI performance being the average Ea value of previously installed Pb/Cu resonators limited to 2.4 MV/m.

10. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 10 Substrate influence The cavity shape is important: high  series (HB), having rounded shorting plate, no beam ports or brazed joints, have the best performance. We have reduced performance in medium  (FC) and also in high  (hb) cavities in case of impurities delivered during the sputtering process from brazed joints. The PP series has joints in the outer cavity surface and an extra hole in the high current region, which both contribute to spoil the performance.

11. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 11 Production Experience Most of the ALPI cavities were produced in between June 00 and December 03 after an initial prototyping phase. The performance has been improving with time mainly because of: - Better substrate adjustment - HPW resonator rinsing before mounting the cavity in the line cryostat - Use of a new cathode In the production stage every cavity reaching 4 MV/m was installed, but sometimes, because of lack of time, we had to install also resonators which had inconvenience during the chemical or sputtering process and resulted in lower performance. Ea=freq.dependent PP cavity series Feed trough leak Q degradation after venting first Cu prototypes New cathode No In joint, but stronger end plate contact

12. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 12 During CR14-4 sputtering there were discharges in a feed-through, which forced to vent the sputtering chamber for maintenance. The process was continued later on without performing any chemical treatment. The resonator performance was reduced, however the cavity had to be accepted because of the necessity of installing the cryostat on line. The repeating of the chemical treatment and a new sputtering process would have allowed to reach standard results. Example of lack of time constraints Q- curve of resonators in the last two installed cryostats, CR14 and CR18

13. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 13 Cavities at work  Since 2001 most of the ALPI equivalent voltage is provided by Nb/Cu cavities, medium and high  types:  They do not need of continuous or fast tuning  They can operate at the accelerating field determined by the available cryogenic power.  There is no degradation with time after installation.  The average fields is still improving.  The low  resonators, which can reach average field of 6 MV/m at 7 W, have to be presently locked at an average field lower than 3 MV/m in the ALPI environment.  A plan is now funded to raise the operational accelerating field by increasing the rf driving power.  That will ask for new rf cryostat lines and a cooled coupler.

14. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 14 Possible improvements maintaining old substrates  More accurate “Lifting” the old Cu substrate (the opening of all the trapped enclaves was not well performed in the first produced resonators)  Better tuning plate contact to avoid Indium joint (modifying the end plate and doubling its fixing screws)  Cleaner assembling condition before and after sputtering  Longer high power rf conditioning (not enough devoted time up to now),

15. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 15  We do not have a clean room for cryostat assembling.  We have to align the resonators with their beam port open to air.  We have to close the cryogenic circuits after the resonator assembling  It would surely help: –Avoid cryostat venting to air (because of cryogenic circuits leaks) –Perform high pressure rinsing after resonator alignment (possible if we have not the In joint) –Longer rf and He conditioning (5 MV/m reachable @ 7W in old substrate) Low cost ways to improve the performance further

16. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 16 Further possible improvement in low  QWR performance  We are building 4 new substrates, using spare Cu parts with: –New beam port design –A rounded shorting plate –A capacitive coupler –Without holes in high current regions –Without brazing in the outer resonator body  First test foreseen by the end of 2006

17. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 17 Nb/Cu end plates  Two S.C. Nb SRFQs are operating at Legnaro in PIAVE, the new ALPI injector. Each of them is provided with a couple of sputtered Nb/Cu plates.  They demonstrate that by enough strength, the junction between the Nb cavity body and the end plates gives a sufficient good rf joint.  The ALPI and PIAVE QWR Nb cavities are also provided with sputtered Nb plates.

18. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 18 Nb/Cu sputtering advantages  Mechanical stability (mechanical vibrations are not an issue)  Frequency not affected by changes He bath  p (<0.01Hz/mbar)  Reduced over-coupling (smaller amplifier, coupler do not need cooling, rf lines have reduced size and limited rf dissipation)  High thermal stability (less prone to hot spots, conditioning easier)  Stiffness (in case of loss of isolation vacuum leak…)  Absence of Q-disease (less demand on cryogenic system cooling velocity and reliability)  Insensitivity to small magnetic fields (no magnetic shielding)  High Q of the N.C. cavity (easier coupling in N.C state)  Absence of In vacuum joints (vacuum leaks less probable)  Price (both material and construction) The lower performance of Nb/Cu cavities at high fields, due to the more pronounced Q-slope of Nb/Cu resonators, is not an issue in QWRs as it is in  cavities, because beam dynamic constraints require to limit the accelerating gradient in the low  section of linacs to values well reachable by Nb sputtered resonators.

19. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 19 Surface finishing and chemical treatments.  Electropolishing (20  m, 2 hours, phosphoric acid+butanol, computer controlled)  Rinsing (water, ultrasonic water, HPR)  Chemical polishing (10  m, 4 min, SUBU5)  Passivation (sulphamic acid)  Rinsing (water, ultrasonic water, HPR)  Drying (ethanol, nitrogen)

20. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 20 The sputtering process. Good vacuum No discharges High substrate temperature Cathode: Nb tube Cu base Sputtering chamber

21. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 21 The sputtering process. Operation sequence  Mounting the resonator in the sputtering chamber  Pumping the vacuum chamber  Resonator bake-out, at about 500°C for a couple of days  The sputtering process: in 12 steps of about 15 minutes each  Cooling the resonator at room temperature in vacuum  The sputtering cycle of a QWR requires 9 days The sputtering parameters  Argon pressure: 0.2 mbar  Substrate temperature: 300-500°C  Cathode voltage: 1KV  Power sustained by discharge: about 5 KW  Bias: - 120 V Average film thickness: about 2  m The cavity end plate is also Nb sputtered in a devoted chamber, set up for producing PIAVE SRFQs end plates

22. L.N.L, Thin film... Workshp, Porcellato 9 Oct 2006 22 Conclusion  The Nb sputtering technology shows to be very effective in producing reliable resonators, which have high performance, are very steadily phase locked and are easy to put into operation.  Even better results can be obtained using suitable substrates.  The high number of produced and operational resonators and the reliability of the sputtering process (rejection rate less than 10%) demonstrate that the technology is mature and very competitive and could be industrially applied. Thanks for your attention!