Michigan Electron Long Beam Accelerator Research Articles

Metal-oxide-junction, triple point cathodes in a relativistic magnetron

Triple point, defined as the junction of metal, dielectric, and vacuum, is the location where electron emission is favored in the presence of a sufficiently strong electric field. To exploit triple point emission, metal-oxide-junction (MOJ) cathodes consisting of dielectric "islands" over stainless steel substrates have been fabricated. The two dielectrics used are hafnium oxide (HfO(x)) for its high dielectric constant and magnesium oxide (MgO) for its high secondary electron emission coefficient. The coatings are deposited by ablation-plasma-ion lithography using a KrF laser (0-600 mJ at 248 nm) and fluence ranging from 3 to 40 J/cm(2). Composition and morphology of deposited films are analyzed by scanning electron microscopy coupled with x-ray energy dispersive spectroscopy, as well as x-ray diffraction. Cathodes are tested on the Michigan Electron Long-Beam Accelerator with a relativistic magnetron, at parameters V=-300 kV, I=1-15 kA, and pulse lengths of 0.3-0.5 micros. Six variations of the MOJ cathode are tested, and are compared against five baseline cases. It is found that particulate formed during the ablation process improves the electron emission properties of the cathodes by forming additional triple points. Due to extensive electron back bombardment during magnetron operation, secondary electron emission also may play a significant role. Cathodes exhibit increases in current densities of up to 80 A/cm(2), and up to 15% improvement in current start up time, as compared to polished stainless steel cathodes.

Long-pulse, high-power, large-orbit, coaxial gyrotron oscillator experiments

Long-pulse, large-orbit, coaxial gyrotrons are currently under investigation. The electron beam is generated by the Michigan Electron Long Beam Accelerator (MELBA) with parameters: V=-0.8 MV, I/sub anode//spl les//spl sim/4 kA, I/sub tube/=0.2-2 kA, and pulse length=0.5-1 /spl mu/s. Large-orbit, axis-encircling electron beams are generated by a magnetic cusp. Experimental gyrotron performance with coaxial cavities (unslotted and slotted) is compared to a noncoaxial cavity. The coaxial gyrotron demonstrated superior current transport and microwave production over the noncoaxial gyrotron. The coaxial rod apparently raises the limiting electron beam current in the diode, allowing higher currents to be extracted. The unslotted, coaxial gyrotron showed microwave power levels of 20-40 MW with pulse lengths of 10-40 ns, This coaxial gyrotron operated in two main modes: TE/sub 111/ and TE/sub 112/ with frequencies of 2.34 and 2.5 GHz, respectively. The gyrotron frequency is tunable between the respective modes by changing the magnetic field. The slotted, coaxial gyrotron showed the highest power of 60-90 MW and extremely short pulse lengths of 10-15 ns. For all three gyrotrons, the microwave pulse-shortening mechanisms of mode hopping and mode competition are definitively identified by time-frequency analysis of heterodyned microwave data.

Polarization control of microwave emission from high power rectangular cross-section gyrotron devices

Results are summarized of experiments on a gyrotron utilizing a rectangular-cross-section (RCS) cavity region. The major issue under investigation is polarization control of microwave emission as a function of magnetic field. The electron beam driver is the Michigan Electron Long Beam Accelerator (MELBA) at parameters: V=0.8 MV, I/sub diode/=1-10 kA, I/sub tube/=0.1=0.5 kA, and t/sub e/-beam=0.4-1.0 /spl mu/s. The annular e-beam is spun up into an axis-encircling beam by passing it through a magnetic cusp prior to entering the RCS interaction cavity. Experimental results show a high degree of polarization in either of two orthogonal modes as a function of cavity fields. The RCS gyrotron produced peak powers of 14 MW in one polarization (TE/sub 10/) and 6 MW in the cross-polarized mode (TE/sub 01/). Electronic efficiencies for this device reached as high as 8% with transverse efficiency of 16%. Experimental results on the beam alpha (/spl alpha/=V/sub /spl perp///V/sub /spl par//) diagnostics, where alpha is the ratio of the e-beam's transverse velocity to its parallel velocity, agree well with the single electron trajectory code. MAGIC code results are in qualitative agreement with microwave measurements. Microwave emission shifts from the dominant fundamental mode polarization (TE/sub 10//spl square/ ), to the next higher order mode polarization (TE/sub 01//spl square/) as the solenoid magnetic field is raised from 1.4-1.9 kGauss. Frequency measurements using heterodyne mixers support mode identification as well as MAGIC code simulations.

Optical spectroscopy of plasma in high power microwave pulse shortening experiments driven by a μs e-beam

Microwave pulse shortening experiments have been performed on a rectangular-cross-section (RCS) gyrotron driven by the Michigan Electron Long Beam Accelerator (MELBA) at parameters V=-800 kV, I/sub tube/=0.3 kA and pulselengths of 0.5-1 /spl mu/s. Pulse shortening typically limits the highest (10 MW level) microwave power pulselength to 100-200 ns. Potential explanations of pulse shortening are being investigated, particularly plasma production inside the cavity and at the e-beam-collector. We report the first optical spectroscopy diagnostic measurements inside an operating gyrotron as a means of exploring plasma effects on pulse shortening. Plasma hydrogen H-/spl alpha/ line radiation has been characterized in both time-integrated and temporally-resolved measurements and correlated with microwave power/cutoff. Hydrogen is believed to originate from water absorbed on internal tube surfaces in the gyrotron.

Propagation of microsecond electron beams in gases and excimer laser-ionized channels in the ion-focused regime

Experiments have been performed to investigate the transport and stability of microsecond-pulselength intense relativistic electron beams in the ‘ion-focused regime’ (IFR) in low-pressure neutral gas and laser-preionized channels. Beams have been generated by the Michigan Electron Long Beam Accelerator (MELBA) which operates with parameters: voltage = −0·7 to −1 MV electrically compensated for collapsing diode impedance over the first 1·5 μs, pulselength = 0·1 to 4μs limited by diode closure, and current = 0·2 to 30 kA.Long pulse beam propagation experiments have demonstrated the following effects:1) At pressures of 1–230 m Torr in air and helium with no laser-preionized channel:a) the electron-beam transport efficiency peaks after times longer than the time calculated for beam-induced ionization to reach fe. = 1. Up to 100% efficient transport occurs during short spikes over a 1 m propagation distance.b) Within 100 ns of efficient transport the propagated beam exhibits rapid beam cutoff in air and helium. This initial pulse of propagated current is followed by a series of 10 ns to 60 ns duration spikes for the remainder of the injected beam pulse. At higher gas pressures usually less current is propagated in the later peaks. At the lowest pressure for which significant current is propagated (7·6 m Torr air, 14 m Torr helium) the maximum transported current occurs in one of the later peaks.2) In initial experiments with excimer-laser-ionized channels in diethylaniline at pressures of 3–8 × 10−4 Torr the electron-beam transport over 1·2 m is efficient (80% to 100%) for at least the first 0·5 μs and does not exhibit the erratic transport observed in the electron-beam induced channels. For fixed laser energy, a narrow pressure window exists for efficient and stable long pulse electron-beam transport. Pressures below optimal give inefficient transport while higher pressures cause ejection of the beam from the ionized channel at later times.