Researchers building a quantum computer microscope to squeeze more information out of every electron

[Amber Tiger]

A team at TU Wien, working together with groups from the University of Vienna, JKU Linz, and the University of Innsbruck, has developed a way to couple an electron microscope directly to a quantum computer built from ion traps. The paper describing the approach has been accepted for publication in Physical Review Letters and is currently available as a preprint.

The core idea addresses something that's always bugged physicists about conventional electron microscopes. Every electron passing through a sample carries quantum information beyond just its position, but standard microscopes only ever count electrons and throw that extra information away. This matters most for delicate samples that can't handle being bombarded with large numbers of electrons without getting damaged in the process.

Here's how the setup actually works. Each electron passing through the microscope gets entangled with an ion sitting inside the quantum computer, so that ion effectively carries information about that specific electron afterward. The next electron then comes through and gets entangled with the same ion in turn. By running very specific quantum computing operations each time this happens, the team can combine information from several electrons together and pull out a signal of maximum strength while using a comparatively small number of electrons overall.

The algorithms making this possible were developed with Johannes Kofler's team at JKU Linz, and the actual quantum computer microscope is now being physically built at TU Wien. If it works as described, it could open the door to imaging fragile samples, biological structures especially, that current electron microscopes simply destroy before a useful image can even be captured

Quantum Microscope Significantly Electron Microscopyphys.org

[Electric Holly]

The biological sample angle is the part that actually matters here for most researchers. Cryo-EM already fights a constant battle against radiation damage, and anything that gets a usable signal from fewer electrons is a real win for that field specifically

[Michael]

The fundamental problem this addresses has been known since electron microscopy began. You're always trading resolution against sample damage, and for biological structures that tradeoff has historically been brutal since the samples simply can't survive enough electrons to get a truly sharp image.

Getting more usable information out of the same number of electrons instead of trying to reduce electron count further is a distinctly different angle on the same old problem. Most prior approaches focused on using fewer electrons total. This one focuses on extracting more value from each electron that does pass through

[FadedKernel]

Does the paper give any timeline for when the physical build gets finished and actually tested on samples? Preprints describing something still under construction always leave me wanting a follow up date

[DeanAmbrose11]

Multi institution collaborations like this one, spanning TU Wien, University of Vienna, JKU Linz, and Innsbruck, tend to move slower than single lab efforts but often produce more robust results since different groups are checking each other's assumptions constantly throughout development.

Austria specifically has built up a pretty serious cluster of quantum optics and quantum information expertise over the past couple decades, largely traceable back to the Innsbruck group's foundational work on ion traps. Seeing that expertise get applied directly to something as practically useful as electron microscopy rather than staying purely in the realm of fundamental physics experiments is a nice example of basic research eventually finding an applied use. Not every quantum information breakthrough needs to be about building a bigger quantum computer. Sometimes the more interesting application is a hybrid tool like this one

[DrewMcIntyre_OG]

Been reading about quantum electron microscopy proposals for years now and it's nice to actually see one moving from theory paper to a physical build. Most of these ideas stay purely theoretical for a decade or more

[AEWNoah32]

Curious what the actual resolution improvement looks like in practice once this thing is built and tested on a real sample. Theoretical proposals in this space have a track record of promising more than they deliver once real world noise and imperfect ion trap fidelity get factored in. Would love to see follow up coverage once they actually image something with it