Electron bunch structure in energy recovery linac with high-voltage dc photoelectron gun

The internal structure of electron bunches generated in an injector line with a dc photoelectron gun is investigated. Experiments were conducted on the ALICE (accelerators and lasers in combined experiments) energy recovery linac at Daresbury Laboratory. At a relatively low dc gun voltage of 230 kV, the bunch normally consisted of two beamlets with different electron energies, as well as transverse and longitudinal characteristics. The beamlets are formed at the head and the tail of the bunch. At a higher gun voltage of 325 kV, the beam substructure is much less pronounced and could be observed only at nonoptimal injector settings. Experiments and computer simulations demonstrated that the bunch structure develops during the initial beam acceleration in the superconducting rf booster cavity and can be alleviated either by increasing the gun voltage to the highest possible level or by controlling the beam acceleration from the gun voltage in the first accelerating structure.

Figure 1

Layout of the ALICE injector beam line. Diagnostic units with YAG-02 and YAG-03 contain also vertical slits on motorized stages. DIP-01 and screen YAG-05 constitute an energy spectrometer.

Figure 2

(a) Beam image at the exit from the booster on the YAG-02 screen; (b) single-pixel horizontal profiles acquired along two horizontal lines shown in (a)—one through the center of the beam (black line) and with an offset from the center (red line); (c) beam image on one of the postlinac OTR screens; (d) beam image on the YAG-05 screen in a dispersive section of the ALICE injector.

Figure 3

Simulated (ASTRA) development of the mean beam energy (top, black line), the rms bunch length (top, red line), and the bunch energy spread (bottom) from the photoinjector cathode to the YAG-05 screen of the energy spectrometer. The first cavity (BC1) is set to $-10°$ off-crest phase. The BC2 cavity is set to $+90°$ zero-cross phase. The BC2 gradient is chosen such that the longitudinal phase space is rotated to the negative energy chirp at the exit from BC2. The gun voltage is 230 kV, and the final beam energy is 4.0 MeV. The buncher voltage is 31.7 kV.

Figure 4

Beam images on energy spectrometer screen YAG-05 and corresponding energy spectra at four different buncher voltages. dc gun voltage is 230 kV, and booster cavities BC1 and BC2 phases are set to $-10°$ and $+40°$, respectively. The second solenoid SOL-02 is set to the current of 2.20 A.

Figure 5

The energy difference between two peaks in the beam energy spectra as a function of the buncher peak voltage and SOL-02 current. The BC2 phase is set to $+40°$ here.

Figure 6

The energy separation between beamlets as a function of the booster cavity BC2 peak voltage at two buncher voltages of 44.9 and 50.2 kV. Cavity BC2 is set to the streaking mode (zero-cross phase).

Figure 7

The dependence of energy difference between beamlets on bunch charge. Corresponding beam images on energy spectrometer screen YAG-05 are shown for high and low bunch charges.

Figure 8

The effect of the first booster cavity BC1 gradient on the beam energy spectrum. (a) Representative energy spectra at $V_b=44.9\mathrm{kV}$. (b) Energy difference between beamlets as a function of the booster cavity BC1 voltage at three different buncher voltages. Data were collected at a constant beam energy of 4.0 MeV at the exit from the booster second cavity BC2 and constant energy chirp introduced by BC2.

Figure 9

(a) Beam image at the exit from the booster on the YAG-02 screen with nominal booster cavity phases of $-10°/+25°$, a beam energy of 6.5 MeV, but with a higher than nominal buncher voltage. The image is gamma-corrected to accentuate the core and the halo of the beam. (b)–(e) Four beam images on the YAG-04 screen in the middle of the injector dogleg while the beam at the YAG-02 position is progressively blocked from right to left with the slit assembly.

Figure 10

Variation of the intensity of energy spectrum peaks while the beam is progressively blocked by a slit assembly located at the YAG-03 position.

Figure 11

Two-beamlets structure in an energy spectrum at 325 kV gun voltage and a significantly reduced buncher voltage of 31.7 kV.

Figure 12

FWHM energy spread as a function of the booster cavity BC2 peak voltage at three buncher voltages of 31.7, 54.9, and 63.4 kV. Cavity BC2 is set to the streaking mode (zero-cross phase). The gun voltage is 325 kV.

Figure 13

Horizontal normalized emittance at the exit from the booster as a function of the buncher peak voltage at gun voltages of 230 (black) and 325 kV (blue) using a slit-screen technique. The red dots represent the emittance measurement at 325 kV by a quadrupole scan technique.

Figure 14

Simulated (ASTRA) development of the mean beam energy, the rms bunch length, and the bunch energy spread from the photoinjector cathode to the exit from the booster. The buncher voltage is 50.2 kV, thus causing the longitudinal crossover before the booster. The first cavity (BC1) is set to $-10°$ off-crest phase. The BC2 cavity is set to $+90°$ zero-cross phase. The BC2 gradient is chosen such that the longitudinal phase space is rotated to the negative energy chirp at the exit from BC2. The gun voltage is 230 kV, and the final beam energy is 4.0 MeV.

Figure 15

Electron bunch distributions at the dc gun voltage of 230 kV and two buncher voltages of 31.7 (red) and 50.2 kV (black). The distributions are shown at the exit from the booster. From top to bottom: rms transverse beam size and current distributions along the bunch, longitudinal phase space, and energy spectra. The head of the bunch is at positive values of the longitudinal position.

Figure 16

Evolution of the longitudinal phase space at 230 kV dc gun voltage and 31.7 kV peak buncher voltage (parameters correspond to Fig. 3). Distributions are shown at the exits from the buncher ($s=1.60\mathrm{m}$), the first four cells of BC1 ($s=2.50$, 2.62, 2.72, and 2.84 m), and the last cell of BC1 ($s=3.60\mathrm{m}$). The phase space at the exit of BC2 is shown in Fig. 15. The head of the bunch is at positive values of the longitudinal position.