Status of high pressure RF cavity project K. Yonehara Accelerator Physics Center, Fermilab 10/08/20101MAP Friday meeting

805 MHz High Pressure RF cavity in B field Maximum electric field in HPRF cavity Schematic view of HPRF cavity High Pressure RF (HPRF) cavity has been successfully operated in strong magnetic fields Metallic breakdown Gas breakdown: Linear dependence Governed by electron mean free path Metallic breakdown: Plateau Depend on electrode material No detail study have been made yet Gas breakdown Operation range in cooling channel (10 to 30 MV/m) P. Hanlet et al., Proceedings of EPAC’06, TUPCH147 10/08/20102MAP Friday meeting, K. Yonehara

Apply HPRF cavity in front end channel Dense hydrogen gas can be used as an ideal buffer to suppress breakdown and also be used as an ionization cooling absorber GH2 cools down RF windows Schematic drawing of HPRF cavity in frontend pre- cooler channel Simulation of muon emittance in hybrid front end channel Hybrid: LiH (various widths (6~10 mm) in simulation) + 10 atm GH 2 Be pressure safety window is included J.C. Gallardo & M.S. Zisman et al., Proceedings of IPAC’10, WEPE074 Results are comparable with vacuum front end channel 10/08/20103MAP Friday meeting, K. Yonehara One example:

Breakdown probability in HPRF cavity Breakdown probability around boundary The data was systematically taken with copper electrodes K. Yonehara et al., Proceedings of IPAC’10, WEPE069. Operation gradient in cooling channel 10 to 30 MV/m E/p=14 E/p=10 E/p=7 old result E/p is a key parameter in DC breakdown model 10/08/20104MAP Friday meeting, K. Yonehara E/p is also a key parameter in RF breakdown!! It seems that gas breakdown mechanism is independent on RF frequency

Simulate beam loading effect in HPRF cavity with the MTA proton beam M. Chung et al., Proceedings of IPAC’10, WEPE067 Beam loading effect: Beam-induced ionized-electrons are produced and accumulated They are shaken by RF field and consume large amount of RF power A loading effect was estimated as a function of beam intensity Assumptions: - Only ionization electrons (~1000 e/p/cc) are taken into account - Recombination rate is cm 3 /s - Ignore thermal diffusion - Ignore dynamics of heavy ions 10/08/20105MAP Friday meeting, K. Yonehara

Criticism in HPRF cavity for down selection RF field must be recovered in few nano seconds 1. DC to 800 MHz, Hydrogen breaks down at E/P = 14. It indicates we can use DC data as a framework to explain results. Need to verify frequency independence by testing different frequency HP cavity 2. Electrons move with a velocity,. Current. Power dissipation due to electrons in phase with RF and dissipate energy through inelastic collisions = Measurements with beam verify mobility numbers and verify our loss calculation 3. Electrons recombine with positive ions and removed. If this is very fast they don’t load cavity, if slow cause trouble Beam measurement will give the recombination rate 4. Solution: use electronegative gas(es) to capture electrons and form negative ions Beam measurement will verify attachment rate 5. A+e →A - heavy negative ions. How long do these hang around and do they cause the breakdown voltage of the cavity to be lowered Beam measurement will give necessary answers Feasibility, including a hydrogen safety analysis, also must be assessed 10/08/20106MAP Friday meeting, K. Yonehara

MTA beam line 10/08/2010 MAP Friday meeting, K. Yonehara7 MTA experimental hall MTA solenoid magnet Final beam absorber 400 MeV H - beam Beam profile: Deliver 400 MeV H - beam in the MTA exp. hall to H - /pulse Tune beam intensity by collimator and triplet (reduce factor 1/10) MTA exp hall 400 MeV H - beam

Apparatus for beam test 10/08/2010MAP Friday meeting, K. Yonehara8 400 MeV H - beam Elastic scattered proton from vacuum window No energized device within 15 feet from HP cavity due to hydrogen safety issue Beam must be stopped in the magnet due to radiation safety Beam signal: There are two toroid coils on the beam line One will measure the primary beam intensity and other will pickup current after collimator One telescope beam counter for DAQ trigger signal CCD + Luminescence screen Polaroid film after cavity (Ti)

Diagnostic system for beam test Two RF pickup probes – E/B sense antennas Two F/R RF power directional couplers Six optical ports – Light intensity monitor – spectrometer – TPB (wavelength shifter) Two toroid coils – Monitor current before/after collimator One beam counter CCD + Viewer RF circulator + Damper – Remove RF reflected power 10/08/2010MAP Friday meeting, K. Yonehara9

Variables in beam test Beam intensity – 10 7 protons/bunch as default – 10 7, 10 8, 10 9 protons/bunch – Beam pulse length 5 to 20  s GH2 pressure – 800 psi as default – 400, 800, 1400 psi RF amplitude – 15 MV/m as default – 0, 5, 15, 30 MV/m, on Paschen E Gas species (option) – N2, He Dopant gas effect – SF6 (0.1 %) as default – SF6 (0.01, 0.1 %), – N2 9/22/1010HPRF Wednesday meeting

Time table of HPRF test SepOctNov RF modification, assemble & calibration Wave Guide & Circulator Beam line Detector & DAQ Timing Calibration RF test without beam RF beam test 9/29/1011 Total run duration will be 2.5 months 10/11/10 ~10/22/10 Interruption due to modification of Cryo transfer line & installation of RF switch HPRF Wednesday meeting When do we really start? Rad Shielding Assessment is the first requirement. Button test may go first if the assessment is delayed.

RF breakdown signal analysis 10/08/2010MAP Friday meeting, K. Yonehara12 RF pickup signal in breakdown process Electron density from RF pickup signal analysis A. Tollestrup et al., FERMILAB-TM-2430-APC, K. Yonehara et al., Proceedings of IPAC’10, WEPE069 Equivalent resonance circuit Resonance circuit of normal RF cavity consists of L and C Breakdown makes streamer It produces additional L and R Resonance frequency is shifted by them Current can be estimated from L, R and drift velocity of electrons in hydrogen plasma Probe seems to be sensitivity ~10 13 electrons Lessons in breakdown study

Spectroscopy in breakdown 10/08/2010MAP Friday meeting, K. Yonehara13 Spectroscopy in the high pressure RF cavity Spectroscopy at Balmer line K. Yonehara et al., Proceedings of IPAC’10, WEPE069 Thermal radiation model: Broken line is a least square fitting of thermal radiation formula by taking into account red points which is on neither any hydrogen nor copper resonance lines “0 ns” is a peak light intensity Plasma temperature is raised up to 18,000 K in 5 ns and down to 10,000 K in 50 ns Deexitation emission: Solid line is a least square fitting of Lorentz function by taking into account all points Timing delay due to lifetime of de-excitation Broadened Balmer line (τ ~ 16 ns) is observed Stark effect well-explains resonance broadening Plasma density ~10 19 electrons/cm 3 Using fast data acquisition system Lessons in breakdown study

Precise timing calibration 10/08/2010MAP Friday meeting, K. Yonehara14 RF pickup signal Spectroscopic light Signal delay RF breakdown event

Study delay and distortion of signal in cable 10/08/2010MAP Friday meeting, K. Yonehara15 Giulia Collura (Summer student) made a fast laser circuit by using avalanche trangister Realize 0.6 ns timing resolution system Laser diode was located in parallel to R3 Similar test is planed by using < 100 ps Laser system Initial RF pickup signal Final signal after 300 ft ½’’ Heliax cable She also investigated the distortion of signal in cable No big signal distortion is observed Phase shift due to dispersion was observed

Hydrogen plasma model in breakdown (I) 10/08/2010MAP Friday meeting, K. Yonehara16 We are still working on the fine timing calibration. We do not know when the plasma light is emitted in the RF signal timing yet. But, we have already seen several remarkable phenomena in RF signal and spectroscopic light. The evidences are: We see the RF modulation from -74 ns, i.e. RF amplitude gets lower but no frequency shifts up to -67 ns. From the resonance circuit analysis, electrons are in the plasma and grows exponentially by next 7ns. It is clearly seen the frequency shift at -67 ns. On the other hand, the plasma temperature is extracted from thermal radiation model, where 0 ns corresponds to the maximum light intensity. It seems that the RF frequency modulation happens at the highest plasma temperature

Hydrogen plasma model in breakdown (II) 10/08/2010MAP Friday meeting, K. Yonehara17 On the other hand, we have observed the broadening of resonance line after 0 ns. This suggests that the plasma density suddenly grows after frequency modulation happens. Something takes place at t=-67 ns! Alvin’s model: The electron swarm touches down on the electrode and extremely high current flows between two electrodes. It makes a plasma pinch due to self inductance. Plasma lens! What we have learned from the breakdown analysis: The number of plasma seeding electrons is ~ It grows exponentially up to ~10 15 in 5~10 ns On the other hand, the electron density in breakdown plasma grows up to Assume that the plasma shape is cylinder then the radius of plasma is ~100 μm Plasma temperature grows up to 18,000 K (ave. Ke = 2.3 eV)

Hydrogen plasma model in breakdown (III) 10/08/2010MAP Friday meeting, K. Yonehara18 More thoughts: Initial electron kinetic energy seems not to be enough to initiate cascade process although some electron is accelerated and warm up the plasma temperature I guess that the main driver of initial plasma growth is surface emission electrons Electron is emitted from surface and decelerated immediately by hydrogen gas It stays in the cavity for a while since there is no dominant absorption process in the cavity (cross section of dissociative absorption of H2 is to cm 2 ) Once the density reaches the critical value, the plasma pinch takes place and change the resonant frequency RF power cannot go into the cavity due to mismatching while the plasma disappears due to thermal diffusion (τ ~ 1 μs) Breakdown plasma is too dense by comparing with beam-induced plasma Breakdown plasma has too many electrons to investigate recombination process

Recombination process in pure GH2 10/08/2010MAP Friday meeting, K. Yonehara19 A.Tollestrup et al., FERMILAB-TM-2430-APC Polyatomic hydrogen cluster: H n + are formed from H 2 + and H + in very short time Recombination of H n + is < 1 μs that has been observed in dilute condition No measurement has been done in dense hydrogen environment Careful RF Q reduction measurement with beam (as shown in previous slide) will indicate recombination rate with polyatomic hydrogen cluster Key physics

Electronegative dopant gas 10/08/2010MAP Friday meeting, K. Yonehara20 SF6 and other Fluorine base gas has a large attachment cross section However, SF6 is dissociated and forms F - that damages the cavity surface Besides, cavity cannot operate in cryogenic temperature lower than the boiling point of SF6 (209 K in STP) NH3 has 1/1000 lower attachment cross section than SF6 but it forms in H2 + N2 mixture gas due to beam energy deposition Question is how much amount of N2 need Damaged Cu electrode Key physics

High Pressure pillbox cavity 10/08/2010MAP Friday meeting, K. Yonehara21 Muons, Inc. built a new cavity: Confirm B field independence in realistic RF field condition Operate cavity in vacuum to high pressure gas condition Test from room temperature down to LN2 temperature Test various window materials; Be is the most interesting material Test RF grid window

Summary HPRF cavity can operate in B field (3 T) without any field degradation We understand large part of gas breakdown physics and realize it is not a good model to investigate the beam loading effect Beam loading effect must be measured by using intense beam as the next feasibility study Preparation for beam test is on going – Waiting for radiation shielding assessment to begin the second stage including with beam test 10/08/2010MAP Friday meeting, K. Yonehara22