1. TwinEBIS: Ion Source Modelling
Good morning! My name is Rebecca Taylor, I am from the University of Birmingham in the UK, studying Particle Physics in an integrated Masters Course.
My department is Hadron Source and Linacs within Accelerator and Beam Physics. After this summer school I plan to do accelerator design in either hadron therapy or muon colliders.
At CERN I have most enjoyed networking, meeting students, co-workers, senior physicists, and learning about what people are passionate about.
Today I am going to be telling you about TwinEBIS, a very small and stand-alone experiment, which offers some interesting physics. Here you can see its simulated model, and how it stands in real life (with considerably more wires).
My part of the summer student project has been split into three tasks, which I will explain to you later.
Rebecca Taylor
BE-ABP-HSL
MSci Physics with Particle Physics
University of Birmingham, UK

2. Electron Beam Ion Source
What is an EBIS:
Electron Beam Ion Source
Electron Gun:
High current density
Low energy (5-15keV)
Higher ionisation cross-section
Solenoid:
For e- beam compression
Produces highly charged ions in short pulses
Transforms radioactive 1+ ions into q+ charge state
Good for storing up particles then releasing them
TwinEBIS:
Replica of REX-EBIS at ISOLDE
Test bench, easily accessible
Gain high charge state for good acceleration
Well if you learn anything from this lecture, I hope you learning that an EBIS stands for “Electron Beam Ion Source”.
An electron beam is produced from an electron gun, with a high current density and low energy (as ionization cross section decreases with energy). The solenoid keeps the electron beam compressed throughout the drift tube. The electron beam ionizes either a gas, or charge breeds low-charge ions, and it can produce these ions in very short pulses.
TwinEBIS is a streamlined replica of the REX-EBIS charge breeder at ISOLDE, and it can be used as a training and test bench.

3. TwinEBIS for Hadron Therapy
Hadron Therapy: Radiotherapy treatment using beams of heavy positive ion nuclei.
Can be a compact source of carbon ions:
Stripped of electrons to avoid their interaction with patient
Compact to fit into a hospital
Utilizes the very short ion pulses (<5μs)
To be attached to a 750 MHz RFQ linear accelerator and brought up to therapy energies.
Its purpose has now changed as we realized its potential for being a source of carbons for a compact hadron therapy accelerator. Within our group, some members are looking into connecting our ion source to an RFQ accelerator, bending it into an appropriate shape and modulating its energy using a chain of klystrons.
EBIS are very suited for hadron therapy when coupled with a linac as they can store a lot of charge, and release it in short pulses, which is required to change the energy between carbon doses. It is important to ensure that the carbon nucleus is completely stripped of electrons, to avoid them interacting and causing unwanted side-effects within the patient.
In the image you can see to scale the EBIS compared to the rest of the accelerating structure.
Vittorio Bencini : A novel linear accelerator for carbon ion therapy

4. Layout of TwinEBIS
So what is the geometry of TwinEBIS? I spent a few weeks modelling its internal vacuum chamber using a CAD programme (for reasons you will see later).
The MEDeGUN is the electron beam source, using an anode and a cathode, and brings the electron beam up to anywhere between 5 to 15keV in energy (so quite low). You can see the electron gun in green in the picture.
To produce carbon ions we inject methane through the gas pipe, in the center of this vacuum chamber, and these molecules will fly around the entire vacuum system. The penning gauges will measure pressure within this vacuum. As our drift tube has holes in it, the methane gas will fly through these and then get ionized by the electron gun.
At the end of the EBIS, the electrons are collected by the collector, and the ionized carbon will be picked up and measured by our time of flight mass spectrometer.
I say carbon 6+ here, but in reality there will be all sorts of other things, such as C4+, C5+ and hydrogen ions.
I have been tasked with the problem of simulating the pressure within the vacuum chamber, which means I can predict the ionization rate and therefore the number of ions we are getting out.

5. Task 1: Methane Ionisation
Initial:
CH4-electron interaction
Secondary:
Dissociative Ionisation
Dissociative Excitation
Plotted production/decay of fragments
Results depend on species branching ratios
Ions trapped in beam, neutrals fly out & lost
Conversion eff ≈ 25% C+ overall
Further charge breeding occurs after this stage
So the ionization of methane is not a trivial problem to solve. It turns out that there are multiple ways that ionization of methane can lead, and not all of them lead to the positive carbon ions that we are seeking.
Initially, these black lines, is when the methane is ionized, and it forms all of these different fragments, including neutral CH2 and CH3. These BR are for 30keV.
But it doesn’t stop there, the ions are trapped in the electron beam, so we get secondary ionization of these fragments. This can occur via dissociative ionization, which are these thick lines, or dissociative excitation, which forms neutral fragments as well. The end result is a messy chain of production and decay. Knowing the cross-sections of these events, I have plotted a graph of the production of fragments with time, and as neutral radical fragments escape the beam, they are considered lost. As a result, I have estimated we get about 25% C+ output (which depends directly on secondary branching ratios which have a very large uncertainty)

6. Pressure simulation: MolFlow
Using beam volume, temperature, efficiency, cross-section, breeding time, current density:
Minimum pressure for 109 carbon ion output:
7.7x10-9 mbar
MolFlow: A Monte-Carlo pressure simulator for ultra-high vacuums.
Pressure across drift tube:
10-8 – 10-7 mbar
Not finalized results
Now that I know the conversion efficiency, I can combine all these attributes which I know (beam volume, temperature, ionization rate, desired number density), I can calculate the minimum pressure that is needed for 10^9 carbon ion output. I have predicted this to be about 7.7 to the minus 9 milibar.
Now that I know this number, I can put a geometry model of my vacuum system in a simulation code called MolFlow. By entering a minimum methane gas flow rate of 10-10, and I know the pumping of the turbo pumps (mark on the screen). According to my model, I have measured the pressure across the drift tube, and found it ranges from 10-8 to This is larger than 7.7x10-9, so everything is good. I will compare the penning gauge pressure and the ion output to the actual experimental values in a few weeks.

7. Task 2: Low Energy injection/extraction line simulation
Electric quadrupoles focus beam to mass spectrometer focal point
Finding solutions for multiple ions at multiple energies for both injection and extraction
Electric fields easily adjustable
Conventional ion source- single charges
My second task is quite different, and involves the calculation of beam optics throughout an ion extraction beamline. This has applications beyond Hadron Therapy and will allow for more variety of ions to be included in the future.
We can see the EBIS here in blue, with the time of flight in one direction and the RFQ linac in the other. First the ion beam will bend through this electrostatic switch, 20 degrees. Electric quadrupoles will focus this ion beam to the magnetic dipole’s focal point, and this dipole will be used to measure the mass/charge ratio of the ions. The electrostatic bender will determine the energy resolution of the ion beam. The conventional single-charged ion source on the other side will allow for ions to be injected into the EBIS.
All components are electric to ensure flexibility, as there should be optical solutions for multiple ions at multiple energy spreads to go in both directions.
I drafted up this design, am modelling the switch field and the dipole, and I will choose the quadrupole focusing to ensure the beams travel through the beamline efficiently.

8. Task 3: ToF assembly & data analysis
Completed work:
Tested beam chopper response
Tightened vacuum flanges
Helped assemble gridded lens & other components
Future work:
Will be ready to produce e- beam in the next few weeks
Will take data from the ToF sensors
Measures particle velocity and energy
Expected to take first mass spectrum results by the end of the month
My final task involves a lot more hardware and analysis. Alongside my team, I have tested the response time for the beam chopper and assembled some of the components within the time of flight. We will be producing the electron beam in the next few weeks, and expect to have the first mass spectrum results by the end of the month!

9. Thank you and I welcome any questions!
Summary
Estimated pressure required to produce desired number of carbon ions.
Calculated percentage of methane molecules which become positive carbon.
Currently calculating beam optics to design ion injector/extractor.
Will analyse the mass spectrum from the ToF to know properties of the carbon beam
Thanks for the Summer Student Team for organizing this opportunity, and extra thanks to H. Pahl, G. Khatri & F. Wenander for advice and help.
Thank you and I welcome any questions!

10. CH4 ionisation cross-section
TwinEBIS Parameters
EBIS design values:
Pressure-dependent variables:
Parameter
Design Value
Main magnet
2 T
Trap length
0.25 m
Electron current
1 A
Electron energy
5-15 keV
Capacity C6+
1·109 ions/pulse
Repetition rate C6+
180 Hz
Parameter
Symbol
Value
Beam radius r
100 µm
Beam volume V
2.6 x10-8 m3
Breeding time
tbreed
5.5 ms
Ions output
Nion
109
Electron density Je
3 kA/cm2
CH4 ionisation cross-section σ
2.7 x cm2
Efficiencies η
50%, 50%, 25%
Temperature T
290 K