Particle accelerators as photon sources are advanced tools in investigating the structure and dynamical properties of matter, and have enabled advances in science and technology for more than half a century. The present workhorses of these sources are storage ring-based synchrotron radiation facilities [1,2,3] and linear accelerator-based free-electron lasers (FELs) [4,5,6,7]. These two kinds of sources deliver light with high repetition rate and high peak brilliance and power, respectively. Some applications, however, do need high average power and high photon flux. Kilowatt extreme ultraviolet (EUV) light sources, for example, are urgently needed by the semiconductor industry for EUV lithography [8]. Another example is that to realize high energy resolution in synchrotron-based angle-resolved photoemission spectroscopy (ARPES), which is highly desired by fundamental condensed matter physics research, we need the initial radiation photon flux before monochromator is high enough. To obtain high average power and high photon flux, a high peak power or a high repetition rate alone is not sufficient. We need both of them simultaneously.

The key of the high peak power of FELs lies in microbunching, which means the electrons are bunched or sub-bunched to a longitudinal dimension smaller than the radiation wavelength so that the electrons radiate in phase and thus cohere [9,10,11]. The power of coherent radiation is proportional to the number of the radiating electrons squared, therefore can be orders of magnitude stronger than the equivalent incoherent radiation in which the power dependence on the electron number is linear. The Self-Amplified Spontaneous Emission (SASE) scheme [6, 7] of microbunching making the high-gain FELs so powerful, however, is actually a collective beam instability which degrades the electron beam parameters and the microbunching can only be exploited once. The repetition rate of the radiation is thus limited by the repetition rate of the driving source, i.e., the linear accelerator. There are now active efforts devoted to improve the repetition rate of FEL radiation, for example by implementing the superconducting technology. However, the realization of a high-average-power, continuous-wave (CW), narrowband, short-wavelength light source remains a challenge.

Fig. 1.1
figure 1

A schematic layout of a conventional storage ring (left) and an SSMB storage ring (right)

Full size image

A mechanism called steady-state microbunching (SSMB) has been proposed [12, 13] to resolve this issue. The idea of SSMB is that by a phase-space manipulation of an electron beam, microbunching forms and stays in a steady state each time going through a radiator in a storage ring. The steady state here means a balance of excitation and damping, a true equilibrium in the context of electron storage ring beam dynamics. The schematic layout of an SSMB storage ring and its operating principle in comparison to a conventional storage ring is shown in Fig. 1.1. SSMB replaces the conventional bunching system in a storage ring, namely the radiofrequency (RF) cavity, with a laser modulation system. As the wavelength of laser (\(\sim \mu \)m) is typically six orders of magnitude smaller than that of an RF wave (\(\sim \) m), a much shorter bunch, i.e., microbunch, can thus be anticipated by invoking this replacement together with a dedicated storage ring magnetic lattice.

The microbunching in SSMB is from the active longitudinal focusing provided by the laser modulator, just similar to the conventional RF bunching through phase stability principle [14, 15]. The radiation in SSMB, unlike that in an FEL, is a passive process and the radiator can be rather short, for example it can be a simple dipole magnet or a short undulator. The SSMB modulator is also much shorter than the radiator undulator in a high-gain FEL. Therefore, there is no FEL mechanism invoked in the bunching or radiation process in SSMB. If there is some unavoidable FEL effects, it needs to be controlled within a safe region to not destroy the steady state micobunches.

To provide adequate and stable longitudinal focusing such that microbunches can be formed and sustained, SSMB requires a powerful phase-locked laser to interact with electrons on a turn-by-turn basis. The realization of such a laser system usually demands an optical enhancement cavity. A laser cannot effectively interact with the co-propagating electrons if the electrons go through a straight line, as the electric field of a laser is perpendicular to the laser propagation direction. A modulator which bends the electron trajectory transversely is thus needed. The modulator is usually an undulator, which is a periodic structure of dipole magnets with oscillating polarity. Note that to avoid the head-on collisions, i.e., the Compton back-scattering, between the reflected laser and the electrons, a four-mirror optical cavity, instead of a two-mirror one, is chosen for the illustration in Fig. 1.1.

Note that we have not presented explicitly the energy replenish system for SSMB in the illustration. The modulation laser in principle can be used to compensate the radiation energy loss of the electrons, just like the traditional RF, but this may not be a cost-effective choice. Besides, the electron beam current and output radiation power will also be limited by the incident laser power. Instead, one may just use a traditional RF cavity for the energy compensation. If a larger filling factor of the electron beam is desired, the energy supply system could also be one or several induction acceleration cavities. In the present envisioned high-average-power SSMB photon source, induction linac is tentatively used as the energy compensation system and the filling factor of the electron beam in the storage ring can be rather large, for example larger than \(50\%\).

Once realized, SSMB can combine the strong coherent radiation from microbunching and the high repetition rate of beam circulating in a storage ring to provide high-average-power, high-repetition (MHz to CW) narrowband radiation, with the wavelength ranging from THz to soft X-ray. Such a novel photon source could provide unprecedented opportunities for accelerator photon science and technological applications. For example, SSMB is promising for generating kW-level EUV radiation for EUV lithography [16]. Energy-tunable high-flux narrowband EUV photons are also highly desirable in condensed matter physics study, such as used in high-resolution ARPES to probe the energy gap distribution and electronic states of superconducting materials. Ultrahigh-power deep ultraviolet and infrared sources are potential research tools in atomic and molecular physics. Moreover, new nonlinear phenomena and dynamical properties of materials can be driven and studied by high-peak and average-power THz sources. Besides high power, SSMB can also produce ultrashort (sub-femtosecond to attosecond) photon pulse trains with definite phase relations, which could be useful in attosecond physics investigations.

This dissertation is devoted to the theoretical and experimental studies of SSMB, with important results achieved. The contribution of this dissertation can be summarized as: first, answer the question of how to realize SSMB; second, reveal what radiation characteristics can we obtain from the formed SSMB; and third, experimentally demonstrate the working mechanism of SSMB in a real machine for the first time. More specifically, in Chaps. 2 and 3, we have conducted in-depth theoretical and experimental studies on single-particle effects vital for the formation and transportation of microbunching in a storage ring. Chapter 2 is on longitudinal dynamics, while Chap. 3 is devoted to transverse-longitudinal coupling dynamics. Chapter 4 is the theoretical and numerical investigation on the average and statistical characteristics of the radiation generated from the formed microbunching. In Chap. 5, we report our work on the first successful demonstration of the mechanism of SSMB, performed at the Metrology Light Source in Berlin. Finally, in Chap. 6 we present a short summary of the dissertation, together with some useful formulas and example parameters of SSMB storage rings aimed for kW-level infrared, EUV and soft X-ray radiation, respectively. Summarizing, the highlights of this dissertation are:

  • Presents the first proof-of-principle experiment of a promising accelerator light source mechanism.

  • Covers precision longitudinal and transverse-longitudinal coupling dynamics in a storage ring.

  • Provides useful formulas and example parameters for high-power infrared, EUV and soft X-ray light source design.

The work presented in this dissertation is of fundamental importance for the development of an SSMB-based high-power photon source.