Superconducting Materials Testing With a High-Q Copper RF Cavity Sami Tantawi, Valery Dolgashev, Gordon Bowden, James Lewandowski, Christopher Nantista SLAC Ricky Campisi *1, Tsuyoshi Tajima *2, Alberto Canabal *2 *1 ORNL, *2 LANL Work supported by the U.S. Department of Energy under contract DE-AC02-76SF SRF Materials Workshop, FNAL, May 2007

Superconducting rf is of increasing importance in particle accelerators. We have developed a resonant cavity with high quality factor and an interchangeable wall for testing of superconducting materials. A compact TE 01 mode launcher attached to the coupling iris selectively excites the azimuthally symmetric cavity mode, which allows a gap at the detachable wall and is free of surface electric fields that could cause field emission, multipactor, and rf breakdown. The shape of the cavity is tailored to focus magnetic field on the test sample. We describe cryogenic experiments conducted with this cavity. An initial experiment with copper benchmarked our apparatus. This was followed by tests with Nb and MgB 2. In addition to characterizing the onset of superconductivity with temperature, our cavity can be resonated with a high power klystron to determine the surface magnetic field level sustainable by the material in the superconducting state. A feedback code is used to make the low level RF drive track the resonant frequency. Abstract 2SRF Materials Workshop, FNAL, May 2007

Finding the Right Cavity Shape TE 01n pillbox (field same on top and bottom) 3 SRF Materials Workshop, FNAL, May 2007

The Mushroom Cavity 57.1% peak 54.5% peak |H s | contour (inches) bottom plate r = 0.98” |E| TE 013 -like mode |H| Q 0 = ~44,000 (Cu, room temp.) material sample Features: No surface electric fields (no multipactor) Magnetic field concentrated on bottom (sample) face (75% higher than anywhere else) Purely azimuthal currents allow demountable bottom face (gap). Why X-band (~ GHz)?: high power & rf components available fits in cryogenic dewar small (3”) samples required axis peak 4 SRF Materials Workshop, FNAL, May 2007

6.8% mode separation for axisymm. TE modes. 4.0% mode separation from axisymm. TM mode. 1.4% mode separation from non-axisymm. mode. Other Nearby Resonance Modes: TE 131 TM 32? WC150 (3.81 cm diam.) Couple on axis with pure axisymmetric tranverse electric mode TE 01 in circular waveguide. 5 SRF Materials Workshop, FNAL, May 2007

copper plug temperature sensor WC150 Cu Mechanical Design Sample Dimensions ss Conflat flanges G. Bowden 6 SRF Materials Workshop, FNAL, May 2007

WR90-WC150 Compact High-Purity TE 01 Mode Launcher height taper (.400” .800”) planar TE 10  TE 20 converter rect.-to-circ. TE 20  TE 01 mode converter WC150 WR90 V. Dolgashev TE 10 TE 01 7SRF Materials Workshop, FNAL, May 2007

mode converter cavity Cryogenic Dewar different cavity vacuum can removed (cavity immersed in liquid He) inner diameter ~1 ft. Assembly WR90 waveguide 8 SRF Materials Workshop, FNAL, May 2007

HFSS (Cu)CopperNiobium frfr QLQL 30,99129,96120,128  Q0Q0 44,57543,77525,619 QeQe 101,69494,94493,906 “Cold” Tests (Room Temperature) HP 8510C Network Analyzer Nb sample mounted in bottom flange room temp. measurements 9 SRF Materials Workshop, FNAL, May 2007

Phase slew due to input waveguide, determined from a 50 MHz measurement is subtracted from 1 MHz data. A “Q circle” is fit to the corrected S 11 data in the complex plane to determine f r, , and Q L. From these Q 0 and Q e are derived. Processing Cold Test Data Temperature measured with a carbon-glass resistor (low end) inserted into a hole in bottom of cavity and from frequency shift (higher temperatures). Complex S 11 is measured with 1601 points in 1 megaherz around resonance. 10SRF Materials Workshop, FNAL, May 2007

Q 0 : 44,000  190,300 Q 0 : 25,600  330,700 Copper Sample Niobium Sample QLQL QeQe Q0Q0 11

9.3 K Transition of Cavity Q 0 During Warmup from Liquid He Temperature (reactor grade Nb) 12 SRF Materials Workshop, FNAL, May 2007

MgB 2 film (~500 nm) by STI 4.2K 13SRF Materials Workshop, FNAL, May 2007

MgB 2 surface resistance calculated from the difference betw. Cu and MgB 2 tests (preliminary) 14SRF Materials Workshop, FNAL, May 2007

Needed Power for Nb RF critical field H field along radius of bottom face 56.76% U= × J |H max | 2 = ×10 -6 A 2 /m 2 |H max | 2 = AU  A = ×10 10 A 2 /m 2 /J If |B c |=180 mT, then |H c |= kA/m,  U c  J = 8.754×10 -7 s P c for a 1.5  s flat input pulse.  P c  498 kW Vertical axis is stored energy divided by incident power 15 SRF Materials Workshop, FNAL, May 2007

New test setup for inexpensive accurate characterization of high-field RF properties of materials and processing techniques High Power Test of Niobium at 4.2K (reactor grade Nb) 16 SRF Materials Workshop, FNAL, May 2007

High Power Testing An initial high power experiment was performed, but data has not yet been fully analyzed. A feedback code is used to make the low level RF drive track the resonant frequency by flattening the phase during cavity discharge. Quenching of Nb sample seen as a roll off of Q as power was raised. Our experimental setup is still maturing. We will add a high-power circulator to isolate the klystron from the cavity reflection. We will also add a silicon diode and a Cernox temperature sensor. We will soon test a sample of MgB 2 from STI. Some samples from STI were low- power tested at LANL and has shown lower surface resistance than Nb at 4K. 17SRF Materials Workshop, FNAL, May 2007

Conclusions We have designed and fabricated a compact, high-Q rf cavity optimized for economically testing the rf properties of material samples and their dependence on temperature and field by means of frequency and Q monitoring. We’ve performed low power cryogenic tests with copper, niobium and MgB 2 samples. High power tests are under way. A similar cavity will be used for pulsed heating material testing. 18SRF Materials Workshop, FNAL, May 2007