The atomic nucleus and its electrons are often thought of as independent systems that are held together in the atom by their mutual attraction. Their interaction, however, leads to other important effects, such as providing an additional decay mode for excited nuclear states, whereby the nucleus releases energy by ejecting an atomic electron instead of by emitting a γ-ray. This ‘internal conversion’ has been known for about a hundred years and can be used to study nuclei and their interaction with their electrons1,2,3. In the inverse process—nuclear excitation by electron capture (NEEC)—a free electron is captured into an atomic vacancy and can excite the nucleus to a higher-energy state, provided that the kinetic energy of the free electron plus the magnitude of its binding energy once captured matches the nuclear energy difference between the two states. NEEC was predicted4 in 1976 and has not hitherto been observed5,6. Here we report evidence of NEEC in molybdenum-93 and determine the probability and cross-section for the process in a beam-based experimental scenario. Our results provide a standard for the assessment of theoretical models relevant to NEEC, which predict cross-sections that span many orders of magnitude. The greatest practical effect of the NEEC process may be on the survival of nuclei in stellar environments7, in which it could excite isomers (that is, long-lived nuclear states) to shorter-lived states. Such excitations may reduce the abundance of the isotope after its production. This is an example of ‘isomer depletion’, which has been investigated previously through other reactions8,9,10,11,12, but is used here to obtain evidence for NEEC.
C.J.C. and J.J.C. thank A. D. Ayangeakaa for input on the potential contributions of Coulomb excitations to the background and M. S. Litz and N. R. Pereira for discussions. We also thank J. Rohrer for assistance in setting up the Gammasphere experiment and the ATLAS operations staff for their efforts. This work was initiated under the US Army Research Laboratory (ARL) Director’s Research Initiative, award number DRI-FY14-SE-022. Further support was provided by ARL Cooperative Agreements W911NF-12-2-0019 and W911NF-16-2-0034, the US Department of Energy (DOE), Office of Science, Office of Nuclear Physics under contract number DE-AC02-06CH11357, the National Science Foundation under grant number PHY-1203100, the Australian Research Council under grant number FT100100991, and the Polish National Science Centre under grants 2011/01/D/ST2/01286 and 2017/25/B/ST2/00901. M.P., J.R. and A.B.H. received support through Ecopulse, Inc. under ARL contract number W911QX09D0016-0004. This research used resources of Argonne National Laboratory’s ATLAS facility, which is a DOE Office of Science User Facility, and of the HIAF at ANU.