Laser plasma acceleration could be a disruptive technology, providing a route to build far more compact accelerators and open new avenues in fundamental research, industry, and health. Before it reaches that potential, however, certain properties of the plasma-driven electron beam, as it is delivered by current prototype accelerators, still need to be ironed out. To this end, DESY’s LUX experiment has made significant progress. Using a novel correction system, a research team was able to significantly improve the quality of electron bunches accelerated by a laser plasma accelerator, thereby bringing the technology a step closer to concrete applications, such as a plasma-based injector for a synchrotron storage ring. Conventional electron accelerators use radio waves that are directed into so-called resonator cavities. The radio waves transfer energy to the electrons as they fly past, increasing their velocity. To achieve high energies, many resonators have to be connected in series, making the machines large and costly. Energy compression of a laser-plasma accelerated electron bunch in an active radiofrequency cavity: High-energy electrons at the beginning of the pulse are decelerated while low-energy electrons at the end of the bunch are accelerated. Courtesy of Science Communication Lab for DESY. Laser-plasma acceleration promises a novel compact alternative. Short, intense laser pulses are shot into a small hydrogen-filled capillary generating a plasma — an ionized gas. When the laser pulse passes through the plasma, it creates a wake similar to the wake of a high-speed boat travelling though water. This wake can accelerate a bunch of electrons to enormous energies within a few millimeters. To date, the technology has had some drawbacks. “The electron bunches produced are not yet uniform enough,” said Andreas Maier, lead scientist for plasma acceleration at DESY. “We would like each bunch to look precisely like the next one.” Another challenge concerns the energy distribution within a bunch. Figuratively speaking, some electrons fly faster than others, which is unsuitable for practical applications. In modern accelerators, these problems have long been solved by using clever machine control systems. Using a two-stage correction, the DESY team has now succeeded in significantly improving the properties of the electron bunches produced by their laser-plasma accelerator. To achieve this, electrons accelerated by the LUX plasma accelerator are sent through a chicane consisting of four deflecting magnets. By forcing the particles to take a detour, the pulses are stretched in time and sorted according to their energy. “After the particles have passed the magnetic chicane, the faster, higher-energy electrons are at the front of the pulse,” said Paul Winkler, first author of the study. “The slower, relatively low-energy particles are at the back.” The stretched and energy-sorted bunch is then sent into a single accelerator module similar to those used in modern radiofrequency-based facilities. In this resonator, the electron bunches are slightly decelerated or further accelerated. “If you time the beam arrival carefully to the radio frequency, the low-energy electrons at the back of the bunch can be accelerated and the high-energy electrons at the front can be decelerated,” Winkler said. “This compresses the energy distribution.” The team was able to reduce the energy spread by a factor of 18 and the fluctuation in the central energy by a factor of 72. Both values are smaller than one permille making them comparable to those of conventional accelerators. “This project is a fantastic example of the collaboration between theory and experiment,” sad Wim Leemans, director of the Accelerator Division at DESY. “The theoretical concept was recently proposed and has now been implemented for the first time.” Most of the components used were from existing DESY stocks. The project team had to invest a great effort in setting up the correction stage and synchronizing the extremely rapid processes. “But once that was done things went surprisingly well,” said Winkler. “On the very first day when everything was set up, we switched on the system and immediately observed an effect.” After a few days of fine-tuning, it was clear that the correction system was working as intended. “This is also a result of the successful synergy between plasma acceleration and modern accelerator technology, as well as the collaboration between a large number of technical teams at DESY, who have extensive experience in building accelerators,” said Reinhard Brinkmann, former director of the accelerator division. “The results will help to further strengthen confidence in the young technology of laser-plasma acceleration,” Maier said. The research team already has concrete ideas for a potential application: The new technique could be used to generate and accelerate electron bunches to be injected into x-ray sources such as PETRA III or its planned successor, PETRA IV. To date, such particle injection has required relatively large and energy-intensive conventional accelerators. Laser-plasma technology now appears to offer a more compact and economical alternative. “What we have achieved is a big step forward for plasma accelerators. We still have a lot of development work to do, such as improving the lasers and achieving continuous operation,” said Leemans. “But in principle, we have shown that a plasma accelerator is suitable for this type of application.” The research was published in Nature (www.doi.org/10.1038/s41586-025-08772-y).