Team from the University of Tokyo demonstrates nanoscale double-slit interference

The double-slit experiment conducted by Thomas Young around 1805 showed that when light is passed through two narrow gaps (slits), interference fringes with alternating bright and dark portions are projected, demonstrating that light possesses wave properties. It is a famous experiment that even appears in high school textbooks.

An international joint research group including Director and Professor Naoya Shibata, Special Researcher Koudai Tabata, Associate Professor Takehito Seki, and Project Associate Professor Ryo Ishikawa at the Institute of Engineering Innovation, School of Engineering, the University of Tokyo, and Associate Professor Toma Susi at the University of Vienna, Austria, succeeded in an ultramicroscopic double-slit experiment likening adjacent atoms to two gaps. The experiment showed that it is possible to read atomic vibrations from stripe patterns made by electrons.

Shibata said: "In the future, applying this method to analyzing locations where heat flow tends to be hindered, such as interfaces and defects of semiconductor materials, can be considered. Through this, directly investigating where heat is likely to pass and stagnate inside materials at the atomic level becomes possible. In the future, we expect to contribute significantly to more advanced heat dissipation design of semiconductor devices and development of new materials utilizing heat efficiently." The findings were published in Nature.

Double-slit interference has been demonstrated with electrons, neutrons, atoms, and huge molecules, supporting the foundation of quantum mechanics that matter also behaves as waves. Particularly regarding electrons, single-electron interference experiments (1989) by Dr. Akira Tonomura in Japan and colleagues are famous. Modern interference experiments have developed into interferometric measurements that precisely measure physical quantities of targets from the clarity of brightness and darkness of interference fringes (visibility of interference fringes) and positions of fringes (phase).

If atoms in crystals could be used as interferometers (double slits), it is considered that properties of matter could be investigated at the single atomic bond level. Because it is necessary to confine interactions within micro-regions of atomic scales, realization was difficult.

The research group performed 4D-STEM measurement recording CBED patterns for each irradiation position while scanning an electron beam focused to atomic size over silicon crystals, using STEM equipped with a pixelated detector. From among the enormous recorded patterns, they selected only CBED patterns when the electron beam was incident precisely in the middle of this atom pair. When electrons proceed through a crystal, attractive force potential of atomic nuclei acts like a waveguide, and electron beams are confined along two atomic columns (channeling effect). As a result, on the exit surface of the sample, two small wave sources separated by 136 picometers (pm) are born, and these interfere with each other to create stripe patterns.

Averaging CBED patterns of 356 atom pairs, clear interference fringes appeared up to 3rd-order bright lines perpendicular to the direction connecting atomic columns. The period of fringes was the same as the value expected from the atomic column interval of 136 pm, showing that this atom pair worked precisely as a double slit. On the other hand, when placing the electron beam directly above one atomic column, interference fringes disappeared, confirming that double-slit interference is established only at the special position in the middle of adjacent atom pairs.

Furthermore, the researchers extracted information about correlated atomic thermal vibrations from the interference fringes. Atoms in a crystal are constantly vibrating due to thermal energy, even at room temperature. Through simulations and experiments, the team found that neighboring atoms vibrate in a correlated manner, and that this correlated motion plays a key role in preserving electron-wave interference.

Traces of this coordinated motion appear in different locations of interference patterns depending on the direction of atomic vibrations. If two atoms move in the same direction along the line connecting atomic columns, the slit interval of 136 pm is maintained, surviving up to high-order interference fringes. If they move in opposite directions, the interval fluctuates and high-order fringes are lost.

On the other hand, if they move in the same direction perpendicular to that, the direction of the double-slit axis is maintained, but if they move in opposite directions, the axis tilts and interference fringes blur in angular directions. Because regions where effects appear differ between two directions, degrees of coordinated vibration in two directions can be read independently from a single interference pattern.

Utilizing this property, the research group analyzed the appearance of interference fringes in detail, succeeding in experimentally determining correlation coefficients representing how aligned vibrations of adjacent atoms are. Furthermore, it was found that vibrations are more likely to align in the direction parallel to bonds than in the perpendicular direction, and that degree remains almost unchanged even when raising the temperature to 900 kelvins (K).

The way neighboring atoms' thermal vibrations are correlated reflects the nature of the chemical bonds between them and remains largely independent of temperature. This correlation coefficient is closely related to force constants representing stiffness of atomic bonds. Reading correlation coefficients from interference fringes enables evaluation of how stiff atomic bonds are. The researchers found that atomic vibrations tend to be correlated along the direction of chemical bonds. This reflects the fact that bonds are stiffer against stretching and compression than against lateral displacement.

Shibata commented, "We did not start out with this idea in mind. Rather, it emerged during our efforts to develop methods for observing atomic vibrations. We had already succeeded in measuring the magnitude and anisotropy of vibrations of individual atoms using STEM, and we began to wonder whether combining that approach with the latest detector technology might allow us to observe even more subtle vibrational phenomena. In conventional electron microscopy, simulations typically assume that individual atoms vibrate independently. However, in real materials, neighboring atoms are connected through chemical bonds and should vibrate collectively, much like waves propagating through a network of springs. We therefore conducted simulations under a variety of conditions to determine whether these coordinated vibrations could be detected."

He continued: "We found that an unusual contrast appeared only when the imaging conditions were carefully tuned and the electron beam was positioned precisely between neighboring atoms. Investigating the origin of this contrast led us to a deeper understanding of the physics underlying double-slit interference and the optical conditions required to observe it. There are still significant experimental challenges, however. These include handling the enormous volumes of data generated during measurements and precisely controlling factors such as sample tilt. To address these issues, we are developing systems that can analyze and visualize large-scale 4D-STEM datasets in real time, while also feeding the results back to the microscope to automatically optimize the optical system. We believe these technical hurdles can be overcome in the near future."

Journal Information
Publication: Nature
Title: Atomic-scale double-slit interferometry with a focused electron probe
DOI: 10.1038/s41586-026-10914-9