Science1 publisher2 min readPublished
Electron diffraction placed a cobalt hydride's hydrogen nearer the neutron reference than X-rays did
Yale chemists checked hydride placements in one open-shell cobalt crystal against a neutron structure, and the microcrystal electron diffractometer came closest. The comparison rests on that single compound.
The Scientist · Science desk

What happened
- Yale chemists compared several ways of locating the hydrogen bound to cobalt in a single crystal, including a microcrystal electron diffractometer the university had recently acquired.
- The reliable detection came with two conditions: the data had to be high quality and the refinement had to use a dynamical scattering model.
- MicroED also works with crystals much smaller than traditional X-ray crystallography requires.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- capability A catalysis group that cannot get time at a neutron source has an in-house route to a hydride position, if it can run dynamical refinement on its own electron data.
- constraint The advantage is tied to the analysis as much as the hardware, so buying the diffractometer without the dynamical refinement does not buy the accuracy reported here.
- decision For a disputed hydride, the next move is no longer a beam-time application, and groups now have to judge whether an electron structure will persuade a referee.
The reason to try electrons on this problem is physical. "The advantage of using electrons is that they interact more strongly with samples and diffract through a different mechanism than X-rays. This makes them more effective at determining the accurate location of the H atom," said Ryan Donnelly, the first author and a graduate student in Patrick Holland's lab at Yale [6]. In a standard X-ray map, the large metal atom obscures the view of the tiny hydrogen [3].
The comparison worked because it had an outside answer to check against. Neutron crystallography, the gold standard for precise measurements of this kind, needs a specialized facility, and there are only a handful in the world [4]. Oak Ridge National Laboratory in Tennessee, one of them, helped validate the study [5]. Donnelly then set his electron diffraction structure against several ways of analyzing X-ray data, and the hydride came out closer to the neutron position [9]. "We discovered that a microcrystal electron diffractometer (MicroED) can accurately locate the hydride better than X-ray methods," said Holland, the senior co-author and Whitehead Professor of Chemistry [2].
Two conditions have to hold for that. MicroED can reliably detect the hydride when the data are high quality and the refinement uses a dynamical scattering model [10].
Donnelly used one cobalt hydride crystal, and the paper's title names an open-shell cobalt complex [7][12][1]. The test case is a single compound, and the metal is cobalt. The phys.org account reports the electron placement as closer to the neutron result without stating the distances or the uncertainties [2]. Those figures are what another group would need before treating an electron structure as decisive in a dispute over where a hydride sits.
Access is what the authors press. The study describes MicroED with dynamical refinement as a practical, broadly deployable approach for detecting metal hydrides, potentially expanding access beyond specialized neutron facilities [11]. The instrument also works with crystals much smaller than traditional X-ray crystallography requires [8]. Yale's group has used it on a microcrystalline metal-organic framework, a rhenium catalyst and a natural product from total synthesis, three structures before this one [13][14]. "The MicroED system has been a boon to our research," said Holland [15]. The paper is in ACS Central Science [12].
What to watch
- Whether the same MicroED plus dynamical refinement pipeline reproduces neutron hydride positions on metals other than cobalt.
- Whether the published distances and error estimates let other groups set a tolerance for hydride placement.
- Whether labs without a dynamical refinement workflow reproduce the placement on their own crystals.