E-324: Optical visualization of beam-driven plasma wakefield - - PowerPoint PPT Presentation

e 324 optical visualization of beam driven plasma
SMART_READER_LITE
LIVE PREVIEW

E-324: Optical visualization of beam-driven plasma wakefield - - PowerPoint PPT Presentation

FACET-II Science Workshop 30 October 2019 E-324: Optical visualization of beam-driven plasma wakefield accelerators Mike Downer, R. Zgadzaj + J. Brooks, I. Pagano (PhD students) University of Texas at Austin + SLAC staff: M. Hogan, V.


slide-1
SLIDE 1

FACET-II Science Workshop 30 October 2019

E-324: Optical visualization of beam-driven plasma wakefield accelerators

Mike Downer, R. Zgadzaj + J. Brooks, I. Pagano (PhD students) University of Texas at Austin + SLAC staff: M. Hogan, V. Yakimenko & others + FACET-II collaborators: M. Litos (UC-Boulder) & students, K. Marsh (UCLA) & others + computational collaborators: T. Silva, J. Vieira (IST); K. Lotov & students (Budker Inst.) E-324 Scientific Goal: Observe, analyze, understand on-axis PWFA structures* that were invisible to us in E-224 (FACET-I)

  • I. Summary of E-224 Results from FACET-I

* (1) ion-density peak, hollow channel (∆t ~ 10-20 ps); (2) blowout-regime electron-wake (∆t ~ 0+) (1) lower ne; (2) shorter interaction region; (3) larger probe angle; (4) better optical resolution

  • II. E-324 FACET-II experimental plans
  • R. Zgadzaj et al., “Dissipation of electron-beam-driven plasma wakes,” submitted for publication.
slide-2
SLIDE 2

normalized intensity [arb. units]

2 1

CCD e- bunch probe f/40 lens heat-pipe oven

Li vapor

na = 0.8e17 cm-3

0 +z1

  • z1

In FACET-I, we imaged near-field diffraction patterns of an ion wake in a single shot ∆t

Ee = 20 GeV Q = 2.4 nC σr = 30 µm σz = 55 µm Epr = 1 mJ λpr = 0.8 µm σr = 0.5 cm τpr = 0.1 ps jitter ~ 0.1 ps

1.5 m

θ = 8 mrad

0.03 ns 0.3 ns 0.6 ns 0.9 ns 10 µs ∆t < 0

image plane

radial position [mm]

2 2 2 2 longitudinal position in heat-pipe oven [cm] 30 30 30

  • 30
  • 30
  • 30

ptychography?

slide-3
SLIDE 3

LCODE-simulated ne(r) profiles reconstruct

  • bserved fringe patterns at ∆t ≥ 0.1 ns

600 ps 300 ps 100 ps 900 ps

  • riginal ions only

including impact ionization

CALCULATED MEASURED

[arb. units]

2 1

normalized intensity

2 2 2 2 2 2 2 2

equivalent radius in interaction region [mm]

calculations by K. V. Lotov,

  • V. K. Khudyakov, A. Sosedkin
slide-4
SLIDE 4

Our FACET-II scientific goal: observe, analyze & understand on-axis PWFA structures*

FACET-I geometry

planned for FACET-II

* 1) At ∆t ~ 10-20 ps: on-axis ion density peak, quasi-hollow channel favorable for focused positron accleration * 2) At ∆t < 1 ps: strongly nonlinear electron wake

a) pre-ionized vs. self-ionized plasma b) e- vs. e+ driven wakes c) transverse wake instabilities: dependence on drive bunch shape, pre-formed plasma

  • A. A. Sahai, Phys. Rev. Accel. Beams 20, 081004 (2017)
  • T. Silva, J. Vieira et al., forthcoming presentation at this workshop.
slide-5
SLIDE 5

FACET-II’s shorter Li oven & lower ne enable steeper θprobe , optical access to interior plasma structures

Enlarged output cube (6” à 8”) & output mirror (3” à 4” diam.) improve probe collection efficiency. Lower f # lens (60 à 25) improves imaging resolution.

Bypass pipe

2.34 m 1.22 m

3.81cm

1.25” (3.175cm) 1.125”

Pipe wall wick

Bypass pipe

Oven pipe transverse dimensions the same for FACET I and II

Oven pipe

Cooling jacket Gate valve

FACET I

FACET II

Bellows angle when bypass pipe is used during alignment Bellows angle when oven pipe is used during experiments

Oven pipe

8” cube

6” cube

f/# 60

f/# 25

4”

3”

beam-ionized Li (ne = 8e17 cm-3) laser pre-ionized Li (ne = 5e17 cm-3, rp ~ 200 µm) limiting optical aperture

slide-6
SLIDE 6

Simulated probe images show signatures of evolving on-axis ion density maxima

Li refractive index

t1 ≈ 7 ps

  • θpr = 20 mrad
  • ne = 3.5 x 1016 cm-3
  • laser pre-ionized Li plasma
  • rp ≈ 200 µm

OSIRIS plasma simulations by T. Silva, IST probe simulations by R. Zgadzaj, UT-Austin

t2 ≈ 16 ps t3 ≈ 22 ps

200 200 µm

  • λpr = 800 nm

light interior

  • n-

axis peak hollow channel

slide-7
SLIDE 7

Probe at θpr = 8 mrad less sensitive to

  • n-axis plasma structures
  • θpr = 8 mrad
  • ne = 3.5 x 1016 cm-3
  • laser pre-ionized Li plasma
  • rp ≈ 200 µm
  • λpr = 800 nm

Li refractive index

t1 ≈ 7 ps

OSIRIS plasma simulations by T. Silva, IST

t2 ≈ 16 ps t3 ≈ 22 ps

200 200 µm probe simulations by R. Zgadzaj, UT-Austin

dark interior

slide-8
SLIDE 8

Shape of plasma column edge influences how

  • n-axis structures modify probe image

Li refractive index

200

t1 ≈ 7 ps

OSIRIS plasma simulations by T. Silva, IST

  • θpr = 20 mrad
  • ne = 3.5 x 1016 cm-3
  • laser pre-ionized Li plasma
  • rp ≈ 200 µm
  • λpr = 800 nm

probe simulations by R. Zgadzaj, UT-Austin

plasma edge: super- Gaussian order 20 plasma edge: super- Gaussian order 8

200 200 µm

Image of probe (∆t ≈ +20 ps) can be normalized to image of reference pulse (∆t ≈ -1 ps) to correct for this.

slide-9
SLIDE 9

Li refractive index

t1 ≈ 7 ps

OSIRIS plasma simulations by T. Silva, IST

  • θpr = 20 mrad
  • ne = 3.5 x 1016 cm-3
  • laser pre-ionized Li plasma
  • rp ≈ 200 µm
  • λpr = 800 nm

probe simulations by R. Zgadzaj, UT-Austin

plasma edge: super- Gaussian order 8

200 µm 200

simulated reference pulse image

slide-10
SLIDE 10

* for 4 mm wide probe (4σ diameter) and wake axis @ θpr = 20 mrad

Overlap region ~ 20cm* 20 mrad probe probe e-bunch e-bunch Walk-off ~38μm

At 0.5 > ∆t > 0.2 ps, probe remains within 1st bucket of electron wake for 20 cm

Geometric walkoff @ θpr = 20 mrad: +40 µm Group-velocity walkoff: -2 µm

slide-11
SLIDE 11
  • bkgd. plasma
  • nly; no wake
  • bkgd. plasma

+ electron wake fully-blown-out bubble Li refractive index

200 200 µm probe simulations by R. Zgadzaj, UT-Austin

  • θpr = 20 mrad
  • ne = 3.5 x 1016 cm-3
  • laser pre-ionized Li plasma
  • rp ≈ 200 µm
  • λpr = 800 nm

Under FACET-II conditions, a fully-blown-out e-beam-driven wake will be visible ∆t ≈ 0.3 ps

slide-12
SLIDE 12

The simulations motivate additional experimental upgrades for PWFA imaging

PROBE

Lp = Plasma oven length θpr = 20 mrad

(1) double θpr à probe interior plasma structures (2) reduce f # à higher image resolution PROBE independent of ionizing laser L1

ionizing laser beam

a(3) install NL phase con-

trast optics à improve sensitivity to ne < 1016 cm-3 (4) install multiple image-

  • bject planes + reference

à distinguish on-axis from edge structures. UPGRADES:

CCD 1 CCD 2 CCD 3 multiple image planes

apre-tested in our U. Texas lab Li, Opt. Lett. 38, 5157 (2013)

FS

phase- contrast

slide-13
SLIDE 13

Later FACET-II Experiments

He/H2 chamber with transverse optical access will enable Faraday profiling* of magnetized PWFAs

(1) double θpr à probe interior plasma structures (2) reduce f # à higher image resolution (3) install NL phase con- trast optics à improve sensitivity to ne < 1016 cm-3 (4) install multiple image-

  • bject planes + reference

à distinguish on-axis from edge structures. UPGRADES: transverse probe ~ longitudinal probe Cotton- Mouton

  • ptics

! kpr ⊥ ! Bφ ! kpr || ! Bφ CCD CCD Faraday ro- tation optics

b(5) install Faraday/C-M

probes à sensitive profil- ing of low-ne structures

bpre-tested in Texas PW expts. Chang, PhD dissertation (2018) LWFA expts.: Kaluza, PRL 105, 11502 (2010); Buck, Nat. Phys. 7, 453 (2011) [ne > 1019 cm-3] Chang, PhD (2018) [ne ~ 1017 cm-3]

* probe of: (1) bubble sheath structure & dynamics (2) external or Trojan horse e--injection

slide-14
SLIDE 14

probe pump ~10cm Gas cell Anamorphic Imaging System Cubic Polarizer Cubic Polarizer CCD CCD

anamorphic imaging

probe CCD CCD polarizer polarizer Texas PW pump

100 J, 150 fs

10 cm gas cell

(ne0 = 2•1017 cm-3)

2 GeV e-

Faraday rotation picks out dense bubble wall in tenuous plasma

Faraday probe setup

  • separation of ± lobes: bubble size
  • |∆ϕFaraday|: bubble wall density
  • width of each lobe: bubble wall thickness
  • longitudinal variations: bubble evolution

Based on technique developed by: Kaluza, PRL 105, 115002 (2010); Buck, Nat. Phys. 7, 453 (2011) in ne > 1019 cm-3 plasma; results shown from Chang, PhD dissertation, UT-Austin (2018)

Faraday rotation results

4 measure- ments e-bunch

slide-15
SLIDE 15

SUMMARY

(1) on-axis ion-density peak (2) hollow-channels (3) e-wake structure & propagation dynamics (4) transverse & longitudinal e-wake instabilities

  • In FACET-II, we will visualize on-axis e- and ion-wake structures,

taking advantage of larger θpr, higher-resolution imaging, lower ne than in FACET-I.

∆t ≥ 30 ps ∆t < 1 ps

  • Our simulations show that our probe will illuminate these structures

well, and provide clear optical signatures of them...

  • ... but there are challenges. Probe images are...

... low contrast à use phase-contrast imaging ... convolved with plasma-edge signatures à use probe-reference pulse pair.

slide-16
SLIDE 16

OSIRIS1 simulations show nonlinear e-wake excitation & early (∆t < 40 ps) ion response

simulations by T. Silva, J. Vieira (IST)

1Fonseca et al., PPCF 55, 124011 (2013)

1015 1016 1017 1018

e-bunch density [e cm-3] incident e-bunch ionization & NL e-wake excitation Er(z) (∆t = 1 ps) ρLi+(z) (∆t = 40 ps) <Er>

plasma density ne [1017 cm-3]

1 2 <ρLi+>

OSIRIS1 framework

· Massivelly Parallel, Fully Relativistic PIC Code · Visualization & Data Analysis Infrastructure · Developed by Osiris Con- sortium, UCLA + IST

experiment reconstruction 20

  • 20

longitudinal distance [cm] radial distance [mm]

2 2 2 2 ∆t = 40ps

moving window fixed window fixed window

  • II. Overview of FACET-I results (E-224)