UNIVERSITY OF CALIFORNIA Los Angeles Direct Laser Acceleration in Laser Wakefield Accelerators A thesis submitted in partial satisfaction of the requirements for the degree of Doctor of Philosophy in Electrical Engineering by Jessica Leigh Shaw 2016 ABSTRACT OF THE DISSERTATION Direct Laser Acceleration in Laser Wakefield Accelerators by Jessica Leigh Shaw Doctor of Philosophy in Electrical Engineering University of California, Los Angeles, 2016 Professor Chandrashekhar Joshi, Chair In this dissertation, the direct laser acceleration (DLA) of ionization-injected electrons in a laser wakefield accelerator (LWFA) operating in the quasi-blowout regime has been investigated through experiment and simulation. In the blowout regime of LWFA, the radiation pressure of an intense laser pulse can push a majority of the plasma electrons out and around the main body of the pulse. The expelled plasma electrons feel the electrostatic field of the relatively-stationary ions and are thus attracted back towards the laser axis behind the laser pulse where they overshoot the axis and set up a wake oscillation. When ionization injection is used, the inner-shell electrons of higher-Z dopant atoms are tunnel ionized near the peak of the laser pulse. Those electrons slip back relative to the wake until they gain enough energy from the longitudinal wakefield to become trapped. Those electrons that are trapped off-axis will undergo betatron oscillations in response to the linear transverse focusing force of the ions. Through experiments and supporting simulations, this dissertation demonstrates that when there is a ii significant overlap between the drive laser and the trapped electrons in a LWFA cavity, the accelerating electrons can gain energy from the DLA mechanism in addition to LWFA. When laser pulse overlaps the trapped electrons, the betatron oscillations of the electrons in the plane of the laser polarization can lead to an energy transfer from the transverse electric field of the laser to the transverse momentum of the electrons. This enhanced transverse momentum can then be converted into increased longitudinal momentum via the v x B force of the laser. This process is known as DLA. In this experimental work, the properties of the electron beams produced in a LWFA where the electrons are injected by ionization injection and become trapped without escaping the laser field have been investigated. The maximum measured energy of the produced electron beams scales with the overlap between the electrons and the laser. Undispersed electrons beams are observed to be elliptical in the plane of the laser polarization, and the energy spectrum splits into a fork at higher energies when the electrons beams are dispersed orthogonal to the direction of the laser polarization. These characteristic features are reproduced in particle-in-cell (PIC) code simulations where particle tracking was used to demonstrate that such spectral features are signatures of the presence of DLA in LWFA. Further PIC simulations comparing LWFA with and without DLA show that the presence of DLA can lead to electron beams that have maximum energies that exceed the estimates given by the theory for the ideal blowout regime. The magnitude of the contribution of DLA to the energy gained by the electron was found to be on the order of the LWFA contribution. In the LWFAs studied here, both DLA and LWFA participate in accelerating the bulk of the electrons in the produced electron beam. The presence of DLA in a LWFA can also lead to enhanced betatron oscillation amplitudes and increased divergence in the direction of the laser polarization. iii The dissertation of Jessica Leigh Shaw is approved. Warren B. Mori Pietro Musumeci Oscar Stafsudd Chandrashekhar Joshi, Committee Chair University of California, Los Angeles 2016 iv Table of Contents Ch. 1: Introduction ........................................................................................................................................ 1 Ch. 1 References ......................................................................................................................................... 19 Ch. 2: Simulations ....................................................................................................................................... 23 Introduction ......................................................................................................................................... 23 Section 2.1: Satisfying the DLA Resonance Condition in a LWFA ................................................... 24 Section 2.2: Electron Beam Properties in the Presence of DLA ......................................................... 35 Section 2.: Resolution Effects on DLA in PIC Simulations of LWFA ............................................... 42 Conclusion .......................................................................................................................................... 54 Ch. 2 References ......................................................................................................................................... 55 Ch. 3: Experimental Methods ..................................................................................................................... 56 Introduction ......................................................................................................................................... 56 Section 3.1: Experimental Setup ......................................................................................................... 56 Section 3.2: Curve-Fitting Method for Finding Maximum Electron Energy ...................................... 60 Conclusion .......................................................................................................................................... 64 Ch. 3 References ......................................................................................................................................... 65 Ch. 4: Initial Results ................................................................................................................................... 66 Introduction ......................................................................................................................................... 66 Section 4.1: Initial Experimental Results ............................................................................................ 66 Section 4.2: Limitations on Initial Experimental Results ................................................................... 72 Conclusion .......................................................................................................................................... 74 Ch. 4 References ......................................................................................................................................... 75 Ch. 5: Experimental Results ....................................................................................................................... 76 Introduction ......................................................................................................................................... 76 Section 5.1: Divergence of Produced Electron Beams ....................................................................... 77 Section 5.2: Characteristic Spectral Features of DLA ........................................................................ 79 Section 5.3: Simulations of Characteristic Spectral Features ............................................................. 80 Conclusion .......................................................................................................................................... 91 Ch. 5 References ......................................................................................................................................... 92 Ch. 6: Conclusion........................................................................................................................................ 93 Ch. 6 References ......................................................................................................................................... 96 v List of Figures Chapter 1 Figure 1. 1:Concept of two-stage LWFA ...................................................................................... 1 Chapter 2 Figure 2. 1: Energy spectrum of the electron beam produced in the simulation. ......................... 26 Figure 2. 2: Plot of a portion of the test electron’s trajectory in the lab frame (blue curve). The red dot marks a zero of the betatron oscillation, and the instantaneous betatron wavelength λ /2 can be β determined by finding the next-closest minima and maximum as marked by the red lines ......... 27 Figure 2. 3: Theoretical (dotted blue line) and measured-in-simulation (solid blue line) ω as a β function of distance. ...................................................................................................................... 28 Figure 2. 4: Illustration of method used to calculate v . The solid blue curve shows the advance ϕ Δz of the transverse electric field relative to the dashed blue curve 500 time steps (i.e. Δt) earlier. This method yields an average value of v as long as the electron samples a particular laser cycle. ϕ When the electron slips into the next laser cycle, there is a discontinuity in the measurement of v ϕ due to this method. ........................................................................................................................ 29 Figure 2. 5: Plot of the scaled quantities v /c (red curve), ω (magenta curve), and v /c -1 (black || 0 ϕ curve) as a function of propagation distance in the simulation. ω _and v /c -1 are scaled to their 0 ϕ maximum values of 3.3 x 1015 Hz and 1.0017, respectively. The plot starts at 160 μm into the simulation, which is where the test electron was born and which is located at the start of the constant-density region of the plasma. Where the electron is born, v /c = 1.0017 and ω = 2.3 x ϕ 0 1015 Hz. By the end, v /c = 1.0004 and ω = 3.3 x 1015 Hz. ........................................................ 30 ϕ 0 vi Figure 2. 6: Plot of the scaled quantities ω (solid blue curve), v ⁄c (red curve), v ⁄c -1 (black β ∥ ϕ curve), and ω (magenta curve) as a function of propagation distance in the simulation. ω , ω , 0 β 0 and v /c-1 are scaled to their maximum values 3.1 x 1013 Hz, 3.3 x 1015 Hz, and 1.0017, ϕ respectively. The plot starts at 160 μm into the simulation, which is where the test electron was born and which is located at the start of the constant-density region of the plasma. Also plotted for reference is the DLA contribution calculated using Equation 2.2 (green curve) scaled to its maximum value of 93 MeV. For comparison, the expected ω (blue dotted curve) is also plotted β after being scaled to 3.1 x 1013 Hz, which was the same value used to normalize the measured ω β . The region over which the test electron first loses net energy to the transverse laser field is shaded gray. .............................................................................................................................................. 31 Figure 2. 7: Nω (blue curve) and (cid:3435)1-v ⁄v (cid:3439)ω (red curve) calculated for the test electron as a β ∥ ϕ 0 function of the distance into the simulation. Both curves were calculated using the values measured in Figure 2. 6. Also shown (black curve) is the laser frequency ω witnessed by the 0,frame electron in its frame. The green curve is the DLA contribution to the total energy gain scaled to its maximum value, and the shaded region marks where the electron first loses net energy to the transverse laser field .................................................................................................................... 33 Figure 2. 8: Example measurement of the laser phase sampled by the test electron. The blue curve shows one complete laser cycle. The red line marks the location of the test electron, and the phase ΔΦ that it samples is measured relative to the point Φ = 0 assigned to the laser cycle. ............... 33 Figure 2.9: Plot of the phase ϕ/π (red curve) of the test electron relative to the sampled laser cycle and of the transverse momentum (dashed blue curve) of the test electron scaled to its maximum value p ⁄mc=36.4. As seen in Figure 2. 8, where the normalized Φ/π is between -0.5 and 0.5, the ⊥ transverse laser field is positive. The solid blue regions along the transverse momentum curve indicate where the test electron is gaining energy from DLA. The discrete jumps in the phase curve mark where the test electron slipped back one laser cycle. The green curve is the DLA contribution to the energy gained by the test electron scaled to its maximum value, and the shaded area indicates where the test electron first experiences net energy loss to the transverse laser field. ....................................................................................................................................................... 34 vii Figure 2. 10: (a) Plot of the maximum energy gain (black curve), LWFA contribution (red curve calculated using Equation 2.1), and DLA contribution (blue curve calculated using Equation 2.2) for the tracked electrons in the 2D OSIRIS simulation of a 30 fs laser pulse with an a of 2.1 0 propagating through 1800 μm of 8 x 1018 cm-3 plasma. The dotted black curve marks the plasma density profile. (b) Plot of the maximum energy gain, LWFA contribution, and DLA contribution as a function of the distance for a 45 fs laser pulse with an a of 2.1 propagating in the same plasma 0 as (a). In Regions 1 and 3, the electrons are gaining energy from DLA. In Region 2, the electrons are slowly losing energy to the transverse laser field. Note that all the electrons in (b) oscillate nearly in phase with one another. ................................................................................................. 38 Figure 2. 11: Plot of the betatron wavelength measured in the T = 0.8 simulation (black dots) p versus the average electron energy over that oscillation. The blue curve shows the theoretical betatron wavelength calculated using Equation 2.4 assuming complete blowout. The red curve shows the theoretical betatron wavelength accounting for the partial (60%) blowout seen in the simulation. ..................................................................................................................................... 40 Figure 2. 12: Plot of the maximum possible DLA and LWFA contributions versus T . The blue p dots mark the maximum possible DLA contribution in each simulation, which occurs at 1163, 832, 611, and 513 μm into the constant density regions of the 1.2 x 1019 (T = 0.6), 1.6 x 1019 (T = 0.7), p p 2.0 x 1019 (T = 0.8), and 2.4 x 1019 (T = 0.9) cm-3 simulations, respectively. The red triangles p p mark the maximum LWFA contribution, which occurs at 1457, 825, 561, 303, and 269 μm into the constant density regions of the 8 x 1018 (T = 0.5), 1.2 x 1019 (T = 0.6), 1.6 x 1019 (T = 0.7), p p p 2.0 x 1019 (T = 0.8), and 2.4 x 1019 (T = 0.9) cm-3 simulations, respectively. ........................... 41 p p Figure 2. 13: (a) Plot of the electron trajectories for the 5 highest-energy electrons in the τ = 35 laser fs (green curves) and τ = 60 fs (black curves) simulations. (b) Normalized transverse laser momentum as a function of the propagation distance for the τ = 35 fs (green curves) and τ = laser laser 60 fs (black curves) simulations. Note that the highest-energy electrons are trapped much earlier in the τ = 35 fs case. ................................................................................................................. 46 laser Figure 2. 14: Plots of the trapped charge after it exits the plasma for the (a) standard-, (b) double- , and (c) quadrupole-resolution simulations. τ = 60 fs in all cases........................................... 47 laser Figure 2. 15: Final energy spectra of the electrons accelerated in the first bucket of the wake for the standard 30 pts/λ (black), double 60 pts/λ (magenta), and quadrupole 120 pts/λ resolution (cyan) simulations. The two higher-resolution spectra have converged. The small random variation between the three spectra arises because only 20% of the macroparticles are extracted and saved. ...................................................................................................................................... 48 viii Figure 2. 16: First Row: Plots of total electron energy (black curves), LWFA contribution to the electron energy (red curves), and DLA contribution to the electron energy (blue curves) as a function of the propagation distance for the (a) standard-, (b) double-, and (c) quadruple-resolution cases. Curves are shown for the 5 highest-energy electrons in each simulation. Black dotted line shows the plasma density profile used in the simulation. For all three cases, there was ~ 10% residual electron density on axis inside the first period of the wake. Second Row: Plots of the trajectories of one of the 5 highest-energy electrons for the (d) standard-, (e) double-, and (f) quadruple-resolution cases. Third Row: Plots of the normalized transverse momentum of the same electrons as in row 2 as a function of propagation distance for the (g) standard-, (h) double-, and (i) quadruple-resolution simulations. ............................................................................................ 52 Chapter 3 Figure 3. 1: Typical experimental setup. The thick red line shows the main laser pulse being focused by the f/6 OAP system at the entrance of the gas cell. The thin red line shows how a portion of the main pulse transmitted through a turning mirror is used for the Michelson interferometer. A typical interferogram is shown. The electrons are dispersed by the dipole magnet shown in the figure onto a scintillator or a lanex and imaged by a PI-MAX 3 camera. The dipole magnet typically was located 3.2 cm downstream from the gas cell, and the distance from the end of the magnet to the screen was 7.0 cm. A typical measured electron spectrum is also shown. ........................................................................................................................................... 57 Figure 3. 2: (a) Photograph of gas cell developed at UCLA [3]. (b) Top view of gas cell showing interaction region. (c) Schematic of cross section of interior of gas cell. .................................... 58 Figure 3. 3: Example density profiles from an 1140 μm gas cell. The plasma density profiles are uniform along the length of the cell. The white lines in the two center images indicate the region over which the average density is calculated. An average is used to compensate for the slight density depression on axis that is caused by the Abel inversion technique [5-7] used to determine the plasma density from the interferogram. .................................................................................. 59 Figure 3. 4: Plot of the plasma density in the gas cell versus a 100% He fill pressure, which shows that the plasma density in the gas cell scales linearly with the backing pressure. ........................ 60 ix