Ghost Beam Studies
- GBS - March 8, 2019 -- 3/8/19 List of questions to be answered through experimentation
- List of GTS Ghost Beam Experiments - 3/11/19 -- List of Experiments that will test the theory described below.
Current Explanation and Observations
The following is the currently accepted theory on the formation of the observed "Ghost Beam". To organize the explanation by what has been proven/shown to be true, the explanation text is color coded based on theoretical predictions, simulations, and experimental data/observations:
- Theoretically predicted, but not yet shown in simulations or in experiment
- Observed in simulation, but not yet explained in theory or shown in experiment
- Observed in experiments, but not yet explained in theory or in simulation
- Theoretically predicted and shown in simulations, but has not yet been observed experimentally
- Observed in simulation and experiment, but has not yet been explained theoretically
- Theoretically predicted and observed in experiment, but has not yet been shown in simulations
- Theoretically predicted, shown in simulations, and observed in experiment
- Speculation based on theory, simulation, and intuition, but has not yet been explicitly shown or proven to be true
At GTS, electrons in a real electron beam can ionize residual gas, resulting in ions and secondary electrons. After the real electron beam is turned off, ions and secondary electrons can be trapped in various places in the accelerator due to the magnetic mirror effect. The three main places the ions and secondary electrons can be trapped are between the anode and magnetizing solenoid and within the first two solenoid lenses. Eventually, the ions and secondary electrons recombine and emit light, some of which is incident on the photocathode, producing a "ghost beam" that we see on the viewers. (For a detailed explanation, see oral qualifier paper here.)
Although the ghost beam has been observed to last for many hours , it has not yet been theoretically predicted to be self-sustaining solely through recombination light from trapped ions/electrons. This suggests that there should be either a direct or an indirect source of ions or electrons to sustain the ghost beam for a long period of time. One possibility is that ions produced close to the photocathode during the real electron beam can "scrub" the photocathode due to back-bombardment, which can desorb surface oxygen molecules and lower the photocathode's work function. As a result, field emission from the photocathode is possible due to a high field gradient within the cathode-anode gap once the real electron beam is off. These electrons, along with electrons due to recombination light, may make up the ghost beam that we see on the viewers. While the real electron beam is off, residual oxygen molecules can adsorb to the photocathode increasing its work function. In order to keep the work function of the photocathode low and keep the photocathode emitting electrons through field emission, it may be that the field emitted electrons ionize more residual gas molecules, which then bombard the photocathode, releasing adsorbed oxygen. Another possibility is ions and electrons that are trapped between the anode and magnetizing solenoid can tunnel through the anode potential (due to the large difference between the absolute values of the cathode and anode potentials) and strike the photocathode.
- The ghost beam is made of electrons that originate from the photocathode. We know this due to how the ghost beam steers with corrector magnets upstream of the viewers.
- The higher the vacuum levels during the prior electron beam run, the more intense the ghost beam is on the viewers.
- The ghost beam only appears when BOTH the magnetizing solenoid and anode bias are above a certain current and voltage respectively. When either is decreased, the ghost beam intensity decreases and vice versa. When either drops below a certain threshold current/voltage, the ghost beam disappears and does not reappear.
- The ghost beam has been shown to be absent initially and only appear after adjusting the current through the focusing solenoids, in which case the ghost beam intensity will gradually increase to a certain point and then slowly decrease.
- The ghost beam current has been measured using a faraday cup to be on the order of a few nA.
GPT Simulation Observations
- Low energy electrons and ions (i.e. having kinetic energies on the order of 10s of eV) are able to be trapped within the first two solenoid lenses and between the anode and gun solenoid. More electrons and ions get trapped within the solenoid lenses than between the anode and gun solenoid.
Documents, Derivations, Tech Notes, and Plots
- Magnetic Mirror Derivation - 9/19/18 Derived formulas for the criteria for ions to become trapped between two solenoids.
- Anode-GS Systematic Study - 11/14/18 Anode Bias/Gun Solenoid current measurements and plots
- Test 1a Data - 11/09/18 Test 1a Data (Ghost Beam intensity vs prior electron beam current and duration)
- Ghost Current vs Gun Vacuum data - 11/20/18 Measurements & Plots of Ghost Current on the Faraday cup vs. Gun Vacuum Levels
- Oral Qualifier Paper -- 2/27/19 Oral Qualifier Paper on Ghost Beam - Contains theory on ion production rates and ionization cross section, the magnetic mirror effect, secondary electron production, and recombination.
- Recombination Rate Calculation - 3/11/19 Estimation of the recombination rate of ions and electrons assuming that the ghost beam is solely due to photoemission by light that is emitted during recombination.
Designing a steel shield that is predicted to create a magnetic mirror trap by absorbing the magnetic field of the gun solenoid. If electrons and ions are trapped together, then they may recombine and emit light that can pass through viewports connected to the beamline and be detected by photodiodes.
- FEMM Model -- 4/29/19 FEMM sketch of the steel shield surrounding the spool with viewports. The top and bottom horizontal lines denote the maximum height (radius) that the shield can have without being in contact with the gun solenoid.
- Optimizing Final Steel Shield Design -- 5/21/19 Plots of Bz vs z along the center line of the spool for various steel shield designs - trying to maximize the height of the spool while minimizing the amount of steel needed to create the spool. Design specs: 3/8" endplate thickness, 3/8" tube thickness, 3.5" between outer edge of spool and inner edge of tube.
- Steel Shield Initial Drawing -- 8/14/19 Initial design of steel shield around spool by Danny
JLab Logbook Entries
For a detailed list of JLab logbook entries that correspond to ghost beam measurements and observations, see List of Ghost Beam Measurements.