SAMURAI – How do you make the first antimatter molecule?

M. Beyer
Oct 28, 2025

Synthesizing Anti-Matter Using Radiative AssociatIon – an NWO Vidi project

A reaction nobody has watched

Take a proton and a hydrogen atom and let them meet. Every so often the pair emits a photon, loses just enough energy, and stays together as H$_2^+$, the simplest molecule there is. This process, radiative association, is how the first molecules formed in the early universe, long before there was dust on which chemistry could happen more easily. Astrophysical models depend on how often it happens.

As far as we know, nobody has ever measured it.

The same question has a second, more exotic side. Antihydrogen atoms are now made and trapped routinely at CERN, and the next step is to bind them into molecules: $\bar{\mathrm{H}}_2^-$ would allow more sensitive tests of the symmetry between matter and antimatter (CPT) than single atoms do. Three routes have been proposed, and all three are reactions between hydrogen atoms and ions. Antiparticles are scarce, so it matters a great deal which route is the most efficient. Because the reactions should run in exactly the same way for matter and antimatter, we can find out with ordinary hydrogen.

Collisions involving hydrogen species. H₂* and H₂⁺* are the short-lived complexes through which the reactions pass. Reaction 8 → 7 is radiative association.
Collisions involving hydrogen species. H₂* and H₂⁺* are the short-lived complexes through which the reactions pass. Reaction 8 → 7 is radiative association.

Half a collision

The usual way to study a reaction is to aim two beams at each other and count what comes out. For hydrogen this has been done down to collision energies of about a millielectronvolt, but the energy resolution is not good enough to see the narrow resonances, where the colliding pair is briefly caught in a quasi-bound state and the reaction probability can rise sharply.

We approach it from the other side. A collision passes through a transition complex: the moment when the two atoms are close together and form, for a short time, a molecule. Instead of colliding atoms to reach that state, we start from a hydrogen molecule and put it there with lasers, choosing its energy and angular momentum, and then watch it fall apart. This is the second half of a collision. Because the quantum mechanics is the same forwards and backwards in time, two such half collisions tell us about the full one.

Full and half collision. In a half collision, laser excitation prepares the transition complex in a chosen partial wave and at a chosen energy, and only the separation of the fragments is observed.
Full and half collision. In a half collision, laser excitation prepares the transition complex in a chosen partial wave and at a chosen energy, and only the separation of the fragments is observed.

Resonances that beam experiments can easily miss, or that stay hidden in the measured cross sections, show up in this picture as sharp spectral lines. Their positions, widths and branching ratios can be measured with spectroscopic precision, and they pin down the collision far more tightly than a cross-section measurement alone.

Three steps

The three objectives of SAMURAI.
The three objectives of SAMURAI.
  1. H$_2$. Using two tunable extreme-ultraviolet lasers built in our lab, we excite H$_2$ to highly excited states above the H$^+$+H$^-$ and H(2$\ell$)+H(2$\ell$) thresholds and image the fragments – protons, H$^-$ ions, excited atoms and electrons – with velocity-map imaging. This also gives us access to excited electronic states of H$_2^+$ that have never been observed. Edward Teller calculated the first of them in his PhD thesis in 1930.
  2. H$_2^+$. We trap H$_2^+$ ions prepared in vibrational levels just below the H$^+$+H(1s) threshold and use XUV and millimetre-wave radiation to drive them across it. The last bound level of the excited A$^+$ state is predicted to be bound by only a few MHz. Once the hyperfine structure is included, one component stays bound and the other becomes a Feshbach resonance, which we can tune with a magnetic field.
  3. H$^+$ + H. Finally, we cool H$_2^+$ sympathetically with laser-cooled Be$^+$ ions in a Coulomb crystal, break it apart gently into a cold proton and a cold hydrogen atom, and try to put the two back together with light. That would be the first observation of radiative association, and it brings the toolbox of ultracold chemistry to hydrogen, which has so far resisted it.

Alongside the experiments, we develop a multichannel quantum-defect and scattering model that treats bound states, resonances, ionization and dissociation on the same footing, and compare it with ab initio calculations.

What you would work with

Most of the hardware exists: two narrow-band XUV sources (60–205 nm), dye lasers and an OPO, velocity-map imaging spectrometers, and a home-built wire trap for H$_2^+$ and Be$^+$. The work combines lasers, vacuum, ion traps, electronics and quantum theory, and you will learn all of them.

We are looking for a PhD student for this project. See the vacancy or contact m.beyer@vu.nl

Maximilian Beyer
Maximilian Beyer
Assistant Professor

My research interests focus on precision measurements of simple molecular ions to test quantum electrodynamics (QED) and to search for physics beyond the Standard Model.