Dielectric Accelerators at CLARA G. Burt, Lancaster University On behalf of ASTeC, Lancaster U., Liverpool U., U. Manchester, and Oxford U.
Dielectric Accelerators Types Photonic structures Dielectric Wakefield Accelerators (like CLIC but with dielectric) Dielectric RF Linacs (replace RF structure with dielectric) Laser driven dielectric accelerators (very high frequency, optical or THz) Also Dielectric Wall Accelerators (high voltage switches) but will not be discussed here Why? Dielectrics can have very high gradients if the right material is used (5.5 GV/m shown in experiments in single shot) They can operate at high frequencies, THz higher (smaller) Can potentially have lower wakefields (for photonic structures) Simpler to manufacture (in some cases)
Dielectric resonator example Aperture 10mm gives a 7GHz resonance, 5mm gives 15 GHz in diamond Simple cylinder allows coupling to the beam even at low(ish) beta Q=3000-5000 R/Q=100, Rsh~0.5 MOhm Peak field on axis Fibres can be used to increase the Q
Undulators and deflectors Transverse deflecting cavity Microwave Undulator cavity
So what are the problems? Multipactor Dielectrics normally have a high SEY, and can charge up. The simple shape is also prone to resonant trajectories. Breakdown at Interfaces Fields are enhanced at vacuum/dielectric gaps causing breakdown, structures need to be gap free including couplers Losses Cooling can be an issue so low losses are required. High gradient also means high power. Omniguide a potential solution or artificial dielectrics (cast dielectric) BBU Unless using a PBG high impedances for HOMs are a problem. Omniguide also a solution. Needs exact calculation of the wakefield (DWA) So that the 2 nd bunch can be Transformer ratio (DWA)
Material Choice Needs high thermal conductivity, low loss tangent, high damage threshold, high dielectric breakdown and no DC charging from beam. Common choices Silicon, Quartz, CVD Diamond and SiO2. Quartz 1.3 W/(m K) charging? Cordierite 3 W/(m K) DC conductivity Alumina 18-30 W/(m K) DC conductivity Sapphire 23-25 W/(m K) charging? Copper 401 W/(m K) metal Diamond 2200 W/(m K). no charging
Stanford, SLAC, UCLA results
Structures Woodpile Grating Fibre Silicon buried GALAXIE
Reverse Dual Grating Silicon (n=1.527) which is the standard material for photolithography fabrication process was chosen for this structure
Required laser parameters & initial particle energy Laser characteristics Pulse energy Average power Pulse width Repetition rate Required parameters 1.7 μj 1.7 W 100 fs 1 MHz The required laser parameters to pump a 1 cm long and one laser wavelength high structure from one side Electron bunches at energies higher than 1MeV are perfectly synchronized with the acceleration field in this structure
THz vs optical Optical THz RF As a bunch is normally a few hundred microns long the energy spread introduced by an optical acceleration is large. Some electrons are decelerated. At THz the spread is smaller and all are accelerated
Photonic band gap structures A photonic bandgap (PBG) structure is a one-, two- or three-dimensional periodic metallic and/or dielectric system (for example, of rods). 1D example: Bragg reflector
Periodic surface Lattice free space Waveguide incident wave reflected wave wall free space A D j % B + p defect n 1 n 2 n 1 n 2 n 2 n 1 n 2 dielectric Acceleratin g or Deceleratin g Electron forward propagating wave I. Konoplev at Oxford is proposing using a defect mode in a periodic surface lattice to confine the fields. Feedback and output lines Electron beam drift channel z y x h = 1 mm d z = 0.75 mm E y d z h Feedback and output lines
PBG resonators and waveguides 2D PBG structures (arrays of rods) are of main interest for accelerator applications. If a wave of certain frequency cannot propagate through a photonic crystal wall, then a mode can form in a crystal defect. This way we can construct a PBG resonator or PBG waveguide. PBG resonator Higher order mode PBG resonator PBG waveguide
PBG crab cavities A PBG dipole cavity would allow the construction of a crab cavity with no trapped higher order modes. However, one must be careful not to trap other modes in the band-gap as well. Lancaster and Huddersfield
PBG Crab Cavities A solution was found, where the rods around the defect (two missing rods) where enlarged. This pushes the modal frequencies down allowing the monopole to be pushed out of the bandgap. Lancaster and Huddersfield
Dielectric Wakefield Accelerators (DWA) Principle Drive bunch (short and high bunch charge) excites wakefield in simple dielectric structure (via Cherenkov mechanism) Trailing bunch (short and low bunch charge) is accelerated in the wakefield a bit similar to LWFA Accelerating gradient scales with high charge, short beams Natural frequency choice : ~ THz range hence : structures (and beams) ~1mm or below (transversely) > GV/m field breakdown thresholds demonstrated
Wakefield (MV/m/nC) The Wakefield Theorem and the Transformer Ratio metal jacket e a b DWFA W + W - The R< 2 limit has kept interest in collinear wakefield accelerators to a minimum. R = W+ W - = (Maximum energy gain behind the drive bunch) (Maximum energy loss inside the drive bunch) < 2
Proposed DWA R&D on VELA/CLARA DWA is an advanced compact accelerator concept and continuously gains momentum worldwide Installation of the linac in CLARA front end (beginning 2015) will make VELA/CLARA beam suitable for a variety of DWA studies Beam energy ~40MeV High bunch charge >100pC Low emittance and short bunch lengths down to sub-ps levels Collaborative effort within CI (and potentially with others) - Lancaster Uni., Manchester Uni., Liverpool Uni., Tech-X Corp., and JAI have already expressed interest in collaborating within this project New avenues for VELA exploitation (e.g. high power narrow band THz radiation) DWA test-bed as a platform for development of advanced diagnostic techniques - electron bunch and THz DWA studies will enhance overall scientific output from VELA (and CLARA) L b a e-beam