The phase-contrast trick that won a 1953 physics Nobel is being ported to electron microscopes

Almost a century ago, phase contrast — the trick that lets light microscopes reveal transparent living cells — was recognized with a Nobel Prize in physics, according to the UC Berkeley announcement carried by ScienceDaily. Seventy-three years after that 1953 prize, the same trick is being ported to electron beams. A team at UC Berkeley and the Chan Zuckerberg Biohub has built a laser add-on, called a crossed-laser phase plate, that installs onto a Thermo Fisher Krios G4 cryo-EM (cryogenic electron microscopy, which flash-freezes biological samples so their molecules can be imaged at near-atomic resolution) and turns the faint outlines of single proteins into sharper, more directly interpretable maps.

The result, described in a Nature Communications paper and a companion bioRxiv preprint, is a controlled phase shift applied to the electron beam as it crosses a sample. The phase plate holds a stable 89-degree shift, plus or minus 3 degrees, on amorphous carbon films while using roughly half the per-cavity laser power of earlier single-laser designs. On the standard apoferritin benchmark, the team reports a 1.79-angstrom reconstruction, and they measure an information limit of 0.9 angstroms with the plate installed.

On a real grid, run head-to-head with and without the new plate, the team got 1.87 angstroms with the device on versus 1.93 angstroms with it off, on roughly 27,000 particles each, according to the preprint. The mechanism's edge is not in headline-grabbing resolution benchmarks. It is in low-frequency contrast, the range that matters for small proteins and for thick cellular samples. The preprint reports enhanced low-frequency signal in Fourier ring correlation on roughly 350-nanometer-thick E. coli cells with the plate on, plus increased low-frequency template-matching signal used in cryo-ET (cryo-electron tomography, which reconstructs 3D views of molecules inside intact cells).

The two-laser geometry exists to solve a specific failure mode. A single-laser phase plate leaves a ghost image on the detector, an artifact of Kapitza-Dirac diffraction, the same quantum effect that scatters electrons off a standing wave of light. The Nature Communications paper predicts that stacking two cavities in a crossed configuration suppresses first-order ghost contrast by roughly 2 to 3 times relative to a single-laser plate, and by about 20-fold when two images are averaged — model predictions for the built device, not measurements of it. The first prototype offsets the two lasers 100 micrometers along the electron-beam axis to avoid interference, and the theory paper describes it as tested only up to about 14 kilowatts per cavity, below the circulating power at which mirror damage was seen in the earlier single-laser design.

The current prototype is early-stage. Its numerical aperture sits at 0.056 and 0.068 depending on configuration, and the authors say preliminary data suggests pushing to about 0.08 is feasible. Going higher, or shifting the laser wavelength below 532 nanometers, will require additional design changes. There is no public timeline, pricing, or instrument-availability claim for a commercial version, even though the work implies a path onto Thermo Fisher's installed Krios base.

The downstream payoff, if the engineering tracks, is biological. Cryo-EM today still struggles with the smallest human proteins, and imaging how those proteins arrange and interact inside their natural cellular context remains hard. A phase plate that lifts low-frequency contrast at the relevant scales is the kind of tool that could let researchers see those small proteins cleanly and, through cryo-ET, watch molecular machines in place. The historical parallel is the point: the trick that took light microscopy from blurry outlines to whole-cell biology is now being scaled to electrons. The proof-of-concept on a Krios G4 is in. The next engineering milestone the team flags is reaching a numerical aperture of about 0.08.