Skip to main content Skip to secondary navigation

Accelerator-on-a-chip

Main content start

The ACHIP project aims to replace conventional radio-frequency based electron accelerators with accelerators operating with optical and near-infrared sources. Not only does this scale down the wavelength of the source, but also all other components of the accelerator. Therefore, an accelerator which spans kilometers, such as SLAC's linear electron accelerator, could one day fit inside a volume roughly the size of a shoebox. However, due to high losses in metals at optical frequencies, the elements which make up the accelerator can not simply be miniaturized, but must be re-invented in appropriate material systems, specifically dielectric materials such as silicon, silicon dioxide, and silicon nitride.

We use inverse design techniques to meet the specifications required of the components to accomplish the coupling, splitting, and phase-tuning of the light from a fiber-laser to the accelerating structure. In this approach, the inputs and outputs of an optical device are specified, along with the desired efficiencies. By starting with some initial condition for the device and simulating the electromagnetic fields, we then adjust the topology of the device such that the specified performance is better matched. We iterate on this process, guided by an efficient scheme to inform us on how to exactly adjust the device's permittivity, until no further optimization can occur or the desired performance is met.

The Vuckovic Lab's contribution to the ACHIP collaboration is working towards demonstrating the first waveguide-integrated accelerator. To accomplish this, we couple in light from a free-space beam or fiber on-chip through a grating coupler, which then couples into a high aspect ratio slab waveguide mode to produce a large spatial extent of the field. This large spatial mode then interfaces with the accelerator structure which results in phase-synchronous fields in the gap region to accelerate the electron. The grating couplers were designed through inverse design to be broadband to ensure coupling with the ultrafast laser pulses used in the experiment. In addition, the accelerator structure was also designed using this technique to ensure fabricability, maximum acceleration gradient, broadband performance, and minimization of transverse deflecting fields. SEM images of the structures used in this experiment are seen below:

Figure: (A) SEM image of waveguide-integrated accelerator. Light is coupled into the slab waveguide mode through the grating coupler. The propagating mode is then directly interfaced with the accelerating structure to produce fields in the gap region which accelerate electrons. Both the grating coupler and accelerating structure are designed using the inverse design methodology. (B) Zoom-in of the accelerating structure.