Tabletop accelerators are an exciting new field of research in which physicists use devices the size of a shoe box, or something just a bit bigger, to accelerate electrons to high energies. The ‘conventional way’ to do this has been to use machines that are as big as small buildings, but are often bigger as well. The world’s biggest machine, the Large Hadron Collider (LHC), uses thousands of magnets, copious amounts of electric current, sophisticated control systems and kilometres of beam pipes to accelerate protons from 0.09 TeV – their rest energy – to 7 TeV. Tabletop accelerators can’t push electrons to such high energies, required to probe exotic quantum phenomena, but they can attain energies that are useful in medical applications (including scanners and radiation therapy).

They do this by skipping the methods that ‘conventional’ accelerators use, and instead take advantage of decades of progress in theoretical physics, computer simulations and fabrication. For example, some years ago, there was a group at Stanford University that had developed an accelerator that could sit on your fingertip. It consisted of narrow channels etched on glass, and a tuned infrared laser shined over these ‘mountains’ and ‘valleys’. When an electron passed over a mountain, it would get pushed more than it would slow down over a valley. This way, the group reported an acceleration gradient – amount of acceleration per unit distance – of 300 MV/m. This means the electrons will gain 300 MeV of energy for every meter travelled. This was comparable to some of the best, but gigantic, electron accelerators.

Another type of tabletop accelerators uses a clump of electrons or a laser fired into a plasma, setting off a ripple of energy that the trailing electrons, from the plasma, can ‘ride’ and be accelerated on. (This is a grossly simplified version; a longer explanation is available here.) In 2016, physicists in California proved that it would be possible to join two such accelerators end to end and accelerate the electrons more – although not twice as more, since there is a cost associated with the plasma’s properties.

The biggest hurdle between tabletop accelerators and the market is also something that makes the label of ‘tabletop’ meaningless. Today, just the part of the device where electrons accelerate can fit on a tabletop. The rest of the machine is still too big. For example, the team behind the 2016 study realised that they’d need as many of their shoebox-sized devices as to span 100 m to accelerate electrons to 0.1 TeV. In early 2020, the Stanford group improved their fingertip-sized accelerator to make it more robust and scalable – but such that the device’s acceleration gradient dropped 10x and it required pre-accelerated electrons to work. The machines required for the latter are as big as rooms.

More recently, Physics World published an article on July 12 headlined ‘Table-top laser delivers intense extreme-ultraviolet light’. In the fifth paragraph, however, we find that this table needs to be around 2 m long. Is this an acceptable size for a table? I don’t want to discriminate against bigger tables but I thought ‘tabletop accelerator’ meant something like my study table (pictured above). This new device’s performance reportedly “exceeds the performance of existing, far bulkier XUV sources”, that “simulations done by the team suggest that further improvements could boost [its output] intensity by a factor of 1000,” and that it shrinks something that used to be 10 m wide to a fifth of its size. These are all good, but if by ‘tabletop’ we’re to include banquet-hall tables as well, the future is already here.