Dry-Ice Cleaning on SRF-Cavities A. Brinkmann, J. Iversen, D ... - Desy

DRY-ICE CLEANING ON SRF-CAVITIESA.Brinkmann # , J.Iversen, D.Reschke, J.ZieglerDESY, D-22603 Hamburg,GermanyAbstractHigh pressure rinsing with ultra-pure water is the wellprovenstandard cleaning step after chemical orelectrochemical surface treatment of SRF cavities. <strong>Dry</strong>icecleaning (DIC) is a powerful additional cleaningoption which depends on the sublimation-impulsemethod. Particles and film contaminations, especiallyhydro-carbons, are removed without residues.Furthermore DIC offers the possibility of a finalhorizontal cleaning of a fully equipped cavity becausewater is not present in the cleaning process. Horizontalcleaning tests on single-cell cavities showed promisinghigh gradient, high Q-value performances, but fieldemission is still the limiting effect. On the basis of thesetests a new IR-heater module is installed to keep a hightemperature gradient between the CO2 jet and the cavitysurface.INTRODUCTIONAlthough many improvements of Cavity preparationprocedures had been done, enhanced field emission stilllimits the high gradient of superconducting cavities.[1]Advanced final cleaning steps and handling proceduresmust be applied to avoid surface contaminations likeparticles and hydrocarbons etc. High pressure rinsing withultra pure water is a powerful method to reduce enhancedfield emission, but dry-ice cleaning might have additionalcleaning potential[2]. <strong>Dry</strong>-ice cleaning avoids a wetcavity surface, removes carbonhydrates and is applicableto ceramics, so the possibility to clean a cavity withpower couplers is given. Improvements with a new IRheaterto keep the cavity surface warm lead to promisingtest-results on single-cell cavities.sublimation. Contaminations gets brittle and start to flakeoff from the surface. When snow particles hits the surfaceand melts at the point of impact, the chemical cleaningeffect occurs.Liquid CO 2 is a good solvent especially forhydrocarbons and silicons. Achieving an optimal cleaningprocess, it is necessary to reach a high thermal gradientbetween jet and surface. It is absolute essential to keep thecavity warm (20° – 30° C) during the cleaning process.Furthermore a sufficient exhaust system is needed to keepdown the CO 2 and N 2 rate in the cleanroom atmosphere.Basic cleaning parameters are shown in Table 1.CO 2 -pressureN 2 -pressureParticle filtrationTable 1: Basic cleaning parametersTemp. of liquid CO 2~ 50 bar12 – 18 bar< 0.05 μm-5° - -40° CEnviroment of cleaning Laminar flow class 10DRY ICE CLEANING (DIC)In comparison with high pressure water rinsing where amechanical effect is the major cleaning contribution DICadditionally offers thermal and chemical effects ascleaning effects. Relaxation of liquid CO 2 in a nozzle(Fig. 1), results in a snow/gas mixture with approximately45 % snow-rate with a temperature of 194 K. To ensurean accelaration and to focus the CO 2 -stream, a supersonicjet of N 2 surrounds the stream. At the same time the N 2prevents condensation of humidity on the cavity surface.The mechanical cleaning effect is based on shockfreezingof the contaminations, strong impact of the snowcrystals and a 500 times increasing volume afterFigure 1: Nozzle inside a cavityInstallationsDuring cleaning the cavity turns round driven by amotor and the nozzle-bar is moving in horizontaldirection. Compared to the first setup of the DIC devicewe changed the orientation of the nozzle to horizontaldirection to test DIC with cavities in mountingposition[3]. (Fig.: 2). Former heating installations were_________________# arne.brinkmann@desy.de

not efficient enough, more heater power was needed. Acompany made IR-heater system with 8 shortwave twintube heaters and a maximum power of 5.6 KW is installednow (Fig.:3 and 4) The wavelength is 1.0 μm – 1.4μm.With this heating power and wavelength the cavity is notfrosted on the outer surface during the cleaning process.The temperature is measured with a IR-Temp.Sensor.The exhaust system ensures a harmless CO 2 –concentration in the cleanroom and keeps particles awayfrom the cavity.CO 2 and O 2 concentration in the cleanroom atmosphere,cleanroom fans and exhaust are controlled by aninterlock-system, which automatically stops the gassupplyif one of these components or concentrations failsor gets too high.1,00E+11Qo1,00E+101,00E+09RF-TEST RESULTSQo/Eacc1DE7 BD1DE4 FE1DE7 FE1DE7 PWR1AC4 BD1,00E+081,00E+070 5 10 15 20 25 30 35 40 45Eacc [MV/m]Figure 2: horizontal orientation of nozzleFigure 3: IR-heaterFigure 5: Test resultsFig.5 shows some RF-test results of several 1-cellcavities. No previous cleaning steps were applied to thecavities before dry ice cleaning. The cavities were storedunder cleanroom-air or kept under vacuum and ventedwith particle-filtered N 2 . In all tests maximum gradientsabove or at 30 MV/m were achieved. The tests showedtypical Q-values above 10 10 at 2 K for superconductingcavities at 1.3 GHz. High Q-values indicate that DIC notcontaminates the inner cavity surface. An optimizedcleaning and handling procedure led to a RF-test with agradient up to 38 MV/m limited by breakdown. Fieldemission is still a limiting factor, but most likely causedby particle contamination during assembly of the cavities.Figure 4: IR-heaterSUMMARYIt was shown that no previous cleaning steps like highpressure water rinsing applied to cavites is necessary toobtain high performance in RF-tests. The new IR-heaterwith optimal wavelength and countoured alignment of thetube heaters provides a high thermal gradient betweenCO 2 -jet and Cavity-surface. A safety interlock is build tocontrol gas concentrations. Field emission is a limitingfactor, but surely caused by particle contamination duringfinal assembly of the cavity. Particle contaminationduring the cleaning procedure is almostnegligible,nevertheless particles must be monitored. Insome cases the particle counter showed a very high rate ofparticles, but this occurred due to a defect in the exhaustsystem.

ACKNOWLEDGEMENTWe acknowledge the support of the EuropeanCommunity Research Infrastructure Activity under FP6“Structuring the European Research Area” program(CARE, contract number RII-CT-2003-506395REFERENCES[1] D.Reschke, Proc. 10 th Workshop on RFSuperconductivity, Tsukuba, Japan, p.144 (2001)[2] D.Reschke et al., Proc. LINAC 2004, Lübeck,Germany, p.776 (2004)[3] A.Brinkmann et al., 12 th Workshop on RFSuperconductivity, Ithaca, USA