R&D OF A GAS-FILLED RF BEAM PROFILE MONITOR FOR INTENSENEUTRINO BEAM EXPERIMENTS∗

K. Yonehara†, M. Backfish, A. Moretti, A. V. Tollestrup, A. Watts, R. M. Zwaska,Fermilab, Batavia, IL 60510, USA

R. Abrams, M.A. Cummings, A. Dudas, R. P. Johnson, G. Kazakevich, M. Neubauer,Muons, Inc., Batavia, IL 60510, USA

Q. Liu, Case Western Reserve University, Cleveland, OH 44106, USA

AbstractWe report the R&D of a novel radiation-robust hadron

beam profile monitor based on a gas-filled RF cavity forintense neutrino beam experiments. An equivalent RF circuitmodel was made and simulated to optimize the RF parameterin a wide beam intensity range. As a result, the maximumacceptable beam intensity in the monitor is significantlyincreased by using a low-quality factor RF cavity. The planfor the demonstration test is set up to prepare for futureneutrino beam experiments.

INTRODUCTIONIntense neutrino beam is a unique probe to research

physics beyond the standard model. Fermilab is the centerinstitution to create the most powerful and wide-spectrumneutrino beam. Fermilab recently achieved consistent 700kW proton beam delivery by using the Main Injector ring(NuMI) [1], but will need to almost triple this power toachieve the goals of future Neutrino experiments. Two ma-jor R&D activities have been going on to realize a multi-MWpower proton driver [2] and target [3]. On the other hand, theradiation robust beam diagnostic system is critical to main-tain the quality of neutrino beam required for future physicsexperiments. To this end, a novel radiation-resistive beamprofile monitor based on a gas-filled RF cavity is proposed.Gas in the cavity serves an ionization media by interact-

ing with incident charged particles. The amount of beam-induced plasma is proportional to the number of incident par-ticles. The plasma consumes a RF power, so called plasmaloading [4, 5]. It is interpreted as an imaginary part of theplasma permittivity [6] or a resistance of plasma which in-duces the RF power dissipation in plasma [7]. The beamprofile is reconstructed by observing the amount of plasmaloading from an individual cell of a multi-RF cavity assem-bly which forms a hodoscope structure.The gas-filled RF monitor is potentially radiation robust

and reliable by comparing to the present hadron monitorbased on an ion chamber [8]. Several remarks are listed asfollowing.

• A main component of the RF cavity is a simple metalchamber. A low power RF (∼mW) and near atmo-spheric pressurized gas are applied in the cavity, hence,

∗ Work supported by Fermilab Research Alliance, LLC under Contract No.DE-AC02-07CH11359 and DOE STTR Grant, No. DE-SC0013795.

† yonehara@fnal.gov

a simple RF window is used. On contrary, the struc-ture of present ion chamber is complicated, especiallyan electric feedthrough, which limits lifetime of thedetector.

• The RF signal can be remotely calibrated by measur-ing the quality (Q) factor when the beam is turned off.Impedance shifts due to aging and radiation damageare calibrated in the Q factor measurement. Remotecalibration is not possible for the present ion chamber.

• The quality of RF monitor signal is independent fromthe gas pressure and gas impurity while the ion chambersignal strongly relies on the fraction of those parame-ters.

The main goal in the R&D process of the gas-filled RFmonitor is validating the concept and developing the technol-ogy to apply the monitor for intense neutrino experiments.

TUNING PARAMETERA complete model of plasma production and plasma load-

ing mechanisms in a gas-filled RF cavity is given in ref. [4,5].Here, simplified formulae are recalled from the model todemonstrate how to tune the beam sensitivity in the RF mon-itor. The plasma loading takes place when ionized particles(electron-ion pairs) in the cavity gain a kinetic energy fromthe RF field and lost the energy via the Coulomb interactionswith the gas. The net result is that the gas is heated by theRF power. Especially, electrons are a good heat transfer me-dia which is two orders of magnitude higher efficiency thanheavy ions. Therefore, controlling the number of electronsin the plasma is the key to tune the plasma loading.

The number of ion pairs, Ne produced by crossing incidentcharged particles in the cavity is given,

Ne =dEdx

ρNlcW

Pg, (1)

where dEdx is a mean energy loss, ρ is a gas density at the

STP, N is the number of incident charged particles, lc is amean path length of incident charged particles, W is an ionpair production energy, and Pg is a gas pressure. On theother hand, the simplified plasma loading formula is,

Pplasma = eNeµE2

Pg≡

V 2

2Rg, (2)

FERMILAB-CONF-17-139-AD

where e is a unit charge, µ is a net reduced mobility of anion pair, and E is the RF gradient. V and Rg are a RF volt-age and a plasma resistance in an equivalent circuit model,respectively. The reduced mobility in eq. (2) is a constant ina low E/Pg condition. Thus, eq. (2) suggests that Pplasma

and Rg are independent from the gas pressure.The equivalent circuit diagramwhich involves a RF source

is shown in fig. 1. The time-domain RF power equation isgiven,

(V0 − V )2R

V =V 2

2Rg+

V 2

2Rc+ CV

dVdt, (3)

where V0 is the initial RF voltage. The shunt impedance ofthe cavity, Rc is given by the observed Q factor, Rc = QZ0,where Z0 is an intrinsic impedance, Z0 =

√L/C; That is

determined by the geometry of cavity.

Figure 1: Equivalent RF circuit diagram.

Control plasma population by electronegativesNe µ in eq. (2) is broken down to each component,

Ne µ = neµe + n+µ+ + n−µ−, (4)

where ne, n+, n−, µe, µ+, and µ− are a population and re-duced mobility of electron, positive ion, and negative ion,respectively. Contributions of multi-charge state ions to theplasma loading is negligible in past measurements. Thispermits us to set n+ ≈ ne + n−. By doping electronegativein the cavity, a great amount of electrons are captured by thedopant. Hence ne is a time-domain function as exp(−t/τ)where τ is an electron capture time. Fig. 2 shows the es-timated electron capture time in a O2 doped N2. The RFpower simulation suggests that the required τ for LBNF ap-plication is shorter than 10 ns while τ in an atmospheric dryair is ∼ 6 ns. Because the electron capture is the three-bodyprocess τ can be sub-ns in 3 atm dry air. Other electronattachment processes, e.g. a charge recombination or escapefast electrons from the cavity are negligible.Fig. 3 shows the simulated 2.45 GHz RF envelope by

solving eq. (3). The beam parameters are taken from theLBNF application. A sawtooth in the RF envelope is thetiming of injected bunched beam into the cavity. Becauseelectrons are quickly captured by the dopant the RF voltageis partially recovered in a ∼20 ns bunch-gap.An averaged RF envelope reaches equilibrium after 4∼5

bunches due to balancing between the plasma loading andthe feeding RF power. The equilibrium RF voltage is given

0.01 0.02 0.05 0.10 0.20 0.50 1

10-10

10-9

10-8

10-7

10-6

O2 mixing ratio

Electroncapturetime[s]

Figure 2: Electron capture time in O2 doped N2 gas (blue:Pg = 1 atm, orange: Pg = 2 atm, and green: Pg = 3 atm).

Figure 3: Blue line is a beam loaded RF envelope in atmo-spheric dry air filled RF cavity. The Q factor is 500. Greenand orange correspond to the RF envelope with no-beamand a moving averaged of beam-loaded signal, respectively.

from eq. (2),VV0=

Rg

R Rc

Rg +R+Rg

R Rc

. (5)

Because R and Rc are a constant the equilibrium voltage isvaried by Rg which is roughly proportional to the intensityof incident charged particles in the cavity. Fig. 4 showsthe simulated equilibrium voltage as a function of Rg bysolving eq. (5). A half of RF current goes to Rc so that thevoltage ratio V/V0 is a half without plasma (Rg = ∞). Thelowest acceptable designed value V/V0 is set 0.05 to holdthe reasonable signal-to-noise ratio. In case of LBNF, therange of Rg is 1,800-18,000 Ω so that we fixed the designedQ factor 500.

DEMONSTRATION TESTThe concept will be verified in the demonstration test.

The plasma loading will be measured as functions of beamintensity and the concentration of electronegative dopant.One of the technical challenges is measuring a low power RFsignal in a low Q factor RF cavity. Two step demonstrationsare considered to establish the RF monitor technology.

1000 104 105 106 107

0.005

0.010

0.050

0.100

0.500

Rg [Ω]

V/V0

Figure 4: Calibration of the beam loaded RF signal withvarious Q factors (green: Q = 50, orange: Q = 500, and blue:Q = 5,000). Assume that the input impedance is matched tothe cavity (R = Rc).

Table-top testA main goal of the table-top test is developing the low

power RF signal measurement in a low Q factor RF cavity.Fig. 5 shows the conceptual design of a tunable Q-factorpillbox RF cavity. Two coupling loops load the cavity. Bychanging the coupling strength of loops, the Q factor willbe tuned 100-1,000. A fast RF peak power detector will beused in the RF data acquisition system. A dummy dielectricmaterial will be loaded in the cavity to check the sensitivityof the DAQ system.

Figure 5: Conceptual design of a Q-factor variable pillboxRF cavity.

Beam testThe tunable Q-factor pillbox RF cavity will be used for

the first beam test. The response function of the gas-filledRF cavity will be measured by using a known-intensity andknown-size beam. The electron capture rate will be mea-sured by tuning the Q factor. Then, the optimum Q-factorand ideal electronegative gas concentration are identified inthis test.

Re-entrant shape RF cavity with the fixed Q factor isdesgined for real neutrino beam applications. Fig. 6 showsa superfish simulation of the re-entrant RF cavity for theRF monitor. The size of coupling loop in the re-entrantcavity is smaller than the tunable Q cavity. Therefore, there-entrant cavity has a small perturbation on the RF signaldue to plasma which is created near the loop. Electric fieldin the re-entrant cavity is concentrated and radially flat in agap of two electrodes. The plasma loading can be localizedin the gap. It increases a sensitivity and hence reduces anuncertainty of the RF signal.

Figure 6: Conceptual design of a re-entrant RF cavity.

Time lineThe table-top test will be finished by Summer 2017 and

the first beam test will be taken place in Fall 2017. SY120beam dump [9] will be used for the beam test. Meantime, aprototype RF monitor will be built and installed in the MIbeam dump for the radiation endurance test. Ultimately, ifall tests will be passed a prototype monitor will be built andapplied in the NuMI beam line.

REFERENCES[1] M.E. Convery, Proc. of Sci., FERMILAB-CONF-16-506-AD,

2016.

[2] https://pip2.fnal.gov/

[3] https://conference.sns.gov/event/20/picture/34.pdf

[4] M. Chung et al., Phy. Rev. Lett., vol. 111, p. 184802, Oct. 2013.

[5] B. Freemire et al., Phys. Rev. AB, 19, 062004, 2016.

[6] K. Yonehara et al., in Proc. IPAC15, pp. 1189–1191.

[7] A. Tollestrup, "Cavity Selection for Hadron Calorimeter",https://indico.fnal.gov/getFile.py/access?contribId=0&resId=0&materialId=slides&confId=14043

[8] S. Kopp et al., in Proc. EPAC04, pp. 2768–2770.

[9] A. Watts, Fermilab Beams Document 5040-v2.