MOGA Optimization of UED Beamlines Colwyn Gulliford 2242021

MOGA Optimization of UED Beamlines Colwyn Gulliford 2/24/2021 Accelerator Seminar 1

Interest in UED Ultrafast electron diffraction (UED) promises new insight in many research topics: • Ultrafast molecular dynamics • Transient structural analysis • Study of excited states Competitive/Complementary to X-ray diffraction but with “table top” sized experiments (beamline only a few meters long): COOL! • High bunch charge (105 – 106 electrons / bunch) What are the limits to beam quality in UED beam lines? 2/24/2021 Accelerator Seminar 2

Brightness Limit Normalized RMS transverse emittance Average Brightness Coherence Length (nm) In charge saturation limit: Photocathode Laser Accelerating Field 2/24/2021 Space Charge Repulsion + Image Accelerator Seminar 3

Why MOGA Multiobjective Genetic Algorithms (MOGAs) Allows the discovery of solutions optimizing multiple objectives without assigning weights Set of solutions s. t. for each objective, no other solution is better forms the Pareto Frontier GA: global minimization 2/24/2021 Accelerator Seminar 4

Previous Success: MOGA + GPT • Agreement between simulation and measurement • Measured low emittances • Apply same machinery to UED set-ups 2/24/2021 Accelerator Seminar 5

Beamline Layouts Solenoid 1 Photogun Buncher Solenoid 2 Screen (“Sample”) z Vary: Element strengths/positions, Laser Size/Shape NCRF Gun: 1. 6 Cell, 100 MV/m, 2. 856 GHz Cryogun: 10 MV/m, MTE = 5 me. V MTE = 35 me. V Uses a 3. 0 GHz NCRF bunching cavity 2/24/2021 Accelerator Seminar 6

CU Cryogun Test Beamline • Gun moved to new Bright Beams Lab • Modifications/Improvements 2/24/2021 Accelerator Seminar 7

Optimized Emittance *For long final bunch length (<= 500 fs) Cryogun NCRF gun KE = 225 ke. V KE = 4. 5 Me. V • Optimizer selects long initial laser pulse length (ps) • q 2/3 scaling in both cases (cryogun with larger final spot size) • RF gun out performs cryogun: 5 x larger MTE, roughly 6 -7 x cathode field -> scaling law predicts Cryogun emittance to be 2. 6 X, ~3 in plot • For cryogun, emittance suffers above 150 f. C when focusing down to 25 µm 2/24/2021 Accelerator Seminar 8

Cryogun Coherence Results Optimized Coherence Length as a function of final bunch length for 105 and 106 electrons and various final spot sizes Look at Examples • For 105 electrons, see linear scaling with σx , solutions give ~10 nm, σx ≤ 50 fs, σx ≤ 25 µm • For 106 electrons, coherence performance is varied – anticipate this from optimal emittance data 2/24/2021 Accelerator Seminar 9

CU Cryogun Examples Final Phase Space Initial Phase Space RMS emittance Avg Slice Emittance • Initial Emittance: 0. 53/0. 58 nm • Final Emittance: 1. 05/5. 27 nm • Emittance Growth: 200/914% • Continuous emittance growth • See solenoid aberration 2/24/2021 Accelerator Seminar 10

NCRF Gun Results Optimized Emittance vs. Final Bunch Length Optimized Coherence vs. Final Bunch Length Example • Beam here is more easily compressed at 106 electrons than in cryogun case: Lc, x = ~2 nm σx ≤ 6 fs, σx ≤ 25 µm • Estimate Lc, x = 10 nm for 105 electrons based on q 2/3 scaling 2/24/2021 Accelerator Seminar 11

NCRF Gun Example • Initial Emittance: 1. 2 nm • Final Emittance: 5. 2 nm • Emittance Degradation of 430% • See emittance growth along beamline during comp. 2/24/2021 Accelerator Seminar 12

Summary 225 ke. V, Simulated: Beamline Charge en, x, i en, x, f Growth CU Cryogun 16 f. C (105 e) 0. 53 nm 1. 1 nm 210% 160 f. C (106 e) 0. 58 nm 5. 3 nm 914% 9. 5 Me. V, Measured: Beamline Charge (f. C) en, x, i en, x, f Growth CU Injector 100 p. C 0. 24 um 0. 4 um 160% 300 p. C 0. 42 um 0. 78 um 180% 4. 5 Me. V, Simulated: Beamline Charge (f. C) en, x, i en, x, f Growth NCRF Gun 160 f. C 1. 22 nm 5. 2 nm 426% 2/24/2021 Accelerator Seminar 13

Conclusion • • “Standard” 3 D space charge sims show promising results for SS-UED Use of MOGA allows exploring scaling laws -> q 2/3 Assumed aggressive values for MTE’s and E-fields - still work to be done As understanding of photocathode emission grows, include other effects There are schemes to push for even smaller emittances: field enhancing structures To make best use of these + beams shown here, need better emittance preservation during transport, not just more brightness at cathode! • Thanks for your attention • Thanks to the Cornell High-brightness Beam Group and NSF 2/24/2021 Accelerator Seminar 14