1. 10/18/2012 Yine Sun, APC seminar Electron Beam Current-Profile Shaping Via Transverse-to-Longitudinal Phase-Space Exchange Yine Sun Fermi National Accelerator Laboratory Batavia, IL, USA Yine Sun, Tel-Aviv, Israel, LINAC20129/11/2012

2. 10/18/2012 Yine Sun, APC seminar Advanced Electron Beam Phase-Space Manipulations: Experiments and Simulations Yine Sun Fermi National Accelerator Laboratory

3. 10/18/2012 Yine Sun, APC seminar Motivation An electron beam straight from a photoinjector does not always have the phase-space properties required in its applications. Phase-space manipulation is necessary to achieve certain beam distribution, for example: – the beam can be compressed longitudinally  higher peak current  FELs; – round beam  flat beam (to drive planar dielectric wakefield or image charge undulator, and possible simplify the storage ring design for linear colliders…); – the longitudinal and transverse Emittances of the beam can be EXchanged (EEX); – when EEX is combined with the flat beam  repartition of emittances in 3D can be achieved; – the above manipulations are in the root-mean-square sense; the beam profiles can be more precisely tailored, such as bunch train or linearly ramped current etc. Aside from beam utilization, other possible applications include beam diagnostics etc. 3

4. 10/18/2012 Yine Sun, APC seminar Emittance EXchange (EEX) EEX therory: – 2002: Cornacchia and Emma, PRSTAB 5, 084001. Partial exchange : chicane – 2006: Kim, AIP Conf. Proc. No. 821. Complete exchange: double-dogleg 2010: Double-dogleg EEX experiment demonstration: – J. Ruan et al, PRL 106, 244801 (2011). 2010: Applications of EEX in beam current profile modulation: – Y. Sun et al, PRL 105, 234801 (2010). 4 deflecting cavity

5. 10/18/2012 Yine Sun, APC seminar Double-dogleg Emittance Exchanger 5 Four dipoles + one deflecting cavity: Under thin-lens approximation, with proper matching of the deflecting cavity strength (k) and the dogleg dispersion (D), i.e., 1+kD=0, the diagonal sub-block elements of the exchanger’s transfer matrix are zero ↔ the initial horizontal phase space is mapped into the longitudinal phase space, vice versa. LL2S α

6. 10/18/2012 Yine Sun, APC seminar EEX beamline at A0 photoinjector, Fermilab 6 1.3 GHz 9-cell SC cavity 1.3 GHz rf gun with CsTe cathode matching quadrupoles X3: multislits insertion XS4 Energy diangnostics Vertical spectrometer X5 X24: temporal diagnostics Autocorrelator Bolometer X23 3.9 GHz, 5-cell deflecting cavity Cs 2 Te cathode, Nd:YLF drive-laser Beam energy 14 - 15 MeV Charge 100pC - 1nC

7. 10/18/2012 Yine Sun, APC seminar A0 EEX experiment and simulation at low charge 7 J. Ruan, A. Johnson, Y.-E Sun, A Lumpkin, R. Thurman-Keup, FEL’2011, Shanghai, China Before EEXAfter EEX beam beamlets constructed horizontal Phase-space Norm. Emit. Before EEX (exp.)After EEX (exp.)After EEX (simu.)  x n (μm) 2.9 ±0.1 11.3 ±1.1 13.5  z n (μm) 13.1 ±1.3 3.1 ±0.3 2.9  y n (μm) 2.4 ±0.1 2.9 ±0.5 3.0

8. 10/18/2012 Yine Sun, APC seminar Thick-lens Effect 8 So far, only thin-lens approximation is considered for the transfer matrix of the deflecting cavity. LL2S α When taking into account of the finite cavity length l, the matrix is The R 43 leads to non-vanishing terms in the diagonal blocks of the EEX matrix. Adding accelerating cavity to the deflecting cavity solves the problem. (A. Zholents, ANL/APS/LS-327, May 2011)

9. 10/18/2012 Yine Sun, APC seminar Minimize Thick Lens Effect: Chirped Beam 9 30 DOC (min ΔE)35 DOC 45 DOC 40DOC Given the pure deflecting-mode cavity we have at A0, attempt to minimize thick lens effect is made by introducing an energy & z-position correlation (chirp) in the beam, i.e., δ  hz.

10. 10/18/2012 Yine Sun, APC seminar Chirped Beam EEX: measurements vs simulations The emittance ratio are compared at different energy chirp. Best EEX results are measured at 40 Degrees-Off-Crest (DOC). 10 GPT Simulated @ 450pC Measured @ 450 pC Ruan, Thangaraj, et al.

11. 10/18/2012 Yine Sun, APC seminar Sub-ps Bunch Train Generation via PEX 11 Inserting multi-slits mask into the beamline, the horizontal beam profile is sliced into several beamlets. After the Phase-space Exchange (PEX), horizontal and longitudinal phase-space are swapped. Therefore, beam is now a bunch-train. Quadrupole channels before and after EEX to minimize the contribution from x i ’. multi-slits mask deflecting cavity transversely shaped beam longitudinally shaped beam L L2S α

12. 10/18/2012 Yine Sun, APC seminar Final bunch length control via initial horizontal beam focusing 12 The final beam longitudinal properties can be controlled by initial horizontal beam parameters. While scanning the currents of two quadrupoles upstream of the double-dogleg beamline, the final bunch length after PEX can be monitored using an autocorrelator+bolometer system. Fig. (a) shows such a quadrupole scan without slits inserted, and (b) with the slits. The small island appeared with the slits inserted is related to the coherent radiation from the bunch-train structure.

13. 10/18/2012 Yine Sun, APC seminar 13 Experimental demonstration of the sub-ps bunch train: energy domain Multislit mask (1)cavity off: horizontal modulation on X23 and XS4; but no energy modulation on XS4; (2)cavity on: NO horizontal modulation on X23 and XS4; but clear energy modulation appears on XS4 Transverse-longitudinal Phase-space Exchange

14. 10/18/2012 Yine Sun, APC seminar 14 The bunch train temporal structure is measured via the CTR signal measured downstream of EEX. A liquid-helium-cooled bolometer is used as the detector of the Michelson autocorrelator. Multipeaks of the autocorrelation function are measured when the slit mask is inserted, compared to a single peak when the mask is out. Experimental demonstration of the sub-ps bunch train: time domain

15. 10/18/2012 Yine Sun, APC seminar Experimental demonstration of the sub-ps bunch train 15 energytime: autocorrelation function Y.-E Sun et al, PRL 105, 234801 (2010).

16. 10/18/2012 Yine Sun, APC seminar 16 Experimental demonstration of the sub-ps bunch train: time domain The bunch separation is extracted from the autocorrelation function measured in the range of [350  760]  m. The separation can be easily tuned by changing the currents of one single quadrupole upstream of PEX. The corresponding CTR measured from these bunch trains is a narrow-band (δf/f≈20% at 0.5THz) with tunable frequency of [0.37 0.86] THz. * Assuming Gaussian distribution, the minimum individual bunch rms duration measured is less than 300 fs. *P. Piot et al., Applied Physics Letters 98, 261501 (2011)

17. 10/18/2012 Yine Sun, APC seminar Direct Beam Shaping at the Cathode 17 before EEX cathode 1 mm By inserting a 1mm diameter rod into the photo-cathode drive laser, the beam is divided into two beamlets on the cathode, then transported through the EEX beamline. after EEX, two beamlets in time was measured via electro-optical effects: bunch separation, single bunch rms length. *T. Maxwell et al., WEPP030, IPAC 2012, New Orleans, USA.

18. 10/18/2012 Yine Sun, APC seminar From A0 to High Brightness Electron Source Lab. 18

19. 10/18/2012 Yine Sun, APC seminar Optically-modulated bunches at HBESL* Method is similar to our demonstration of narrow-band THz radiation; The key element is the nano-structured cathode (field- emission array # ) triggered with <30 fs laser; Phase-space exchange to obtain an optically-modulated bunch train; Goal is to observe COTR 19 *Y.-E Sun, et al., WEPD19, FEL’2012 # W. Graves, et al., PRL 108, 263904 (2012).

20. 10/18/2012 Yine Sun, APC seminar 20 Transformation of a round angular-momentum dominated beam to a flat beam experiment ASTRA simulation 0.41±0.02 ε x (mm mrad)0.41±0.02 ε y (mm mrad) 41.0±0.5 ε y /ε x 100±5 Y.-E Sun, Ph. D. dissertation, University of Chicago(2005); available at http://lss.fnal.gov/archive/thesis/fermilab- thesis-2005-17.shtml

21. 10/18/2012 Yine Sun, APC seminar 6D Beam Manipulation at ASTA Flat beam transformation + transverse-to-longitudinal exchange 21 50 MeV 750 MeV1 GeV /PEX

22. 10/18/2012 Yine Sun, APC seminar Chicane EEX at ASTA Low Energy Beamline (50 MeV) 22 deflecting + accelerating cavity Q2Q3 Q4 Design stage: Adding accelerating cavity to the deflecting cavity solves the problem. (A. Zholents, ANL/APS/LS-327, May 2011) Q1-Q2: control the dispersion at cavity location; Q3-Q4: tune the beamline to obtain anti-block diagonal matrix; The other quads at the two ends are available for beam matching. Elegant Simulation Q1

23. 10/18/2012 Yine Sun, APC seminar Summary 23 A suite of beam phase-space manipulation experiments has been demonstrated: a partial transfer from one transverse phase-space to another (round-to-flat beam), and a complete phase-space exchange between transverse and longitudinal dimensions. Numerical simulations compared well with experimental measurements, and provided confidence in the design of next generation 6D beam-manipulation experiment, which involves the six coordinates (x,x’,y,y’, z,δ). Development in theory such as accelerating + deflecting hybrid-cavity etc. will be incorporated into the new EEX beamline design. Multi-stage of round-to-flat and EEX offers the possibility of arbitrary phase-space repartition, and innovative beam current profiles can be created to satisfy the demands of certain application. The future of “beamnastics” is bright.

24. 10/18/2012 Yine Sun, APC seminar Credits and Acknowledgements 24 Thanks to our Fermilab colleagues: P. Piot, J. Ruan, R. Thurman-Keup, A. Lumpkin, A. Johnson, H. Edwards, T. Maxwell and J. Thangaraj for their contributions to the EEX-related experiments. M. Church, S. Henderson and V. Shiltsev for their support. E. Harms, E. Lopez, R. Montiel, W. Muranyi, J. Santucci, C. Tan and B. Tennis for their technical support. Thank you for your attention!