Microbunching Instability in Relativistic Electron Bunches: Direct Observations of the Microstructures Using Ultrafast YBCO Detectors

Relativistic electron bunches circulating in accelerators are subjected to a dynamical instability leading to microstructures at millimeter to centimeter scale. Although this is a well-known fact, direct experimental observations of the structures, or the field that they emit, remained up to now an open problem. Here, we report the direct, shot-by-shot, time-resolved recording of the shapes (including envelope and carrier) of the pulses of coherent synchrotron radiation that are emitted, and that are a “signature” of the electron bunch microstructure. The experiments are performed on the UVSOR-III storage ring, using electrical field sensitive ${\mathrm{YBa}}_2{\mathrm{Cu}}_3{\mathrm{O}}_{7-x}$ thin-film ultrafast detectors. The observed patterns are subjected to permanent drifts, that can be explained from a reasoning in phase space, using macroparticle simulations.

Figure 1

Scheme of the experiment. A relativistic electron bunch is stored in the UVSOR-III storage ring. When the electron bunch density exceeds a given threshold, a dynamical instability leads to the spontaneous formation of a pattern in the longitudinal direction. In this Letter, the coherent emission (with $\approx 9\mathrm{mm}$ wavelength) due to the presence of the structure is directly monitored, using a new type of thin-film superconducting detector based on YBCO (see text).

Figure 2

Typical shape of a single pulse that is detected by the YBCO ultrafast detector (in red). (a), (b), and (c) represent the same signal at different scales. The color scale diagram (d) represents the evolution of the (c) signal over many successive turns in the storage ring. Note that the voltage is directly proportional to the electric field of the incoming radiation, and that both the envelope and the carrier are thus recorded. The dashed line in (d) corresponds to the signal displayed in (c).

Figure 3

Interpretation of the drifting motions observed in Fig. 2, from the model (2). (a) Distribution of the electrons in longitudinal phase space (position $z$ energy $\delta$) at a given time. The CSR instability leads to a microstructure in this phase space. (b) Longitudinal electron density distribution, i.e., obtained by the vertical projection of the distribution (a). (c) CSR electric field deduced from (b). Note that the structure is more clearly observable in (c) than in (b), because slow variations are cut off in the CSR process. The electric field versus time (d) reveals that the structures are permanently drifting, as in the experiment. This is essentially due to the permanent rotation of the electron distribution (a) at the synchrotron frequency.

Figure 4

Dynamics following a localized laser perturbation (slicing) at $t=0$, when the system is below the instability threshold. Note that no CSR emission is detected before $t<0$. Interpretations are given in Fig. 5.

Figure 5

Numerical simulation of transients after slicing the electron bunch with a laser, and interpretation in phase space. (a) Simulation of the detection signals. (b)–(i) Phase-space distribution (grayscale image) and associated CSR wakefield (red curve) at different times. For clarity, the reference distribution before laser perturbation (b) is subtracted from the distributions. (c),(d),(e), snapshots near $t=0$. (f),(g),(h), and (i) are successive snapshots corresponding to different instants of the “cusp” feature visible in (a) at $t=40–47\mu \mathrm{s}$.