Research

The reactions that burn hydrogen in the CNO cycle have cross sections that fall exponentially towards stellar energies — at the Gamow window they can reach the picobarn scale, which at the surface of the Earth is indistinguishable from cosmic-ray background. LUNA's answer is 1.4 km of dolomite overburden, which suppresses the muon flux by six orders of magnitude and makes the measurement possible.

For ¹²C + p and ¹³C + p radiative capture, at LUNA and at Felsenkeller, I led the analysis and half of the measurement campaign. The results substantially reduced the uncertainty of the proton-capture rates on both carbon isotopes and of the astrophysical ¹²C/¹³C ratio that depends on them. I now lead the ¹⁹F + p working group at LUNA's 400 kV accelerator, underground at Gran Sasso.

I am also the collaboration's DAQ Coordinator: I designed and run the data-acquisition and slow-control infrastructure of LUNA's current campaigns and of the 3.5 MV machine at the Bellotti Ion Beam Facility (BIBF). It lets shifts be taken remotely, which is no small thing for an experiment under 1.4 km of rock.

AZURE2 is where a set of measured points becomes a cross section: a multi-level, multi-channel R-matrix model fitted to data across every channel that shares a compound nucleus, then extrapolated to the energies at which stars burn. It is used across academia, national laboratories and the IAEA.

Since taking over as maintainer I have worked on the parts that decide whether a modern analysis is feasible at all: analytic gradients and the residual Jacobian computed inside the forward model, so a fit or an MCMC no longer pays a factor of a few hundred for finite differences; a tabulate-once reaction-rate integrator; polarization observables; identical-particle scattering. I also look after the graphical interface and the documentation, and support and train the user community.

The evaluations on this site run the same code on real analyses, published and unpublished.

The 6.79 MeV state in ¹⁵O sits about 0.5 MeV below the proton threshold. Capture into this state carries most of the low-energy S-factor of ¹⁴N(p,γ)¹⁵O, the slowest step of the CNO cycle and therefore the reaction that sets the cycle's rate and the CNO neutrino flux. As a subthreshold resonance, the state also shapes the ground-state transition, and that contribution scales with its radiative width, i.e. with the inverse of its lifetime. The lifetime is shorter than a femtosecond and the published values disagree; a γ-ray tracking array has the position resolution to measure it precisely from the Doppler shift.

I am the spokesperson for the measurement made with AGATA at Legnaro and led the full analysis: Doppler correction, kinematic reconstruction, Geant4 simulation and the Bayesian extraction of the lifetime itself. The analysis is complete and the manuscript is under review at Nature Physics.

Every laboratory measurement of a low-energy cross section is made with atoms, not bare nuclei, and the electron cloud screens the Coulomb barrier. The enhancement it produces grows as the energy falls, which is where astrophysics needs the number, and measured screening potentials have persistently come out larger than theory allows.

At CRYRING the ¹⁵N + p reaction is studied in inverse kinematics with a stored, cooled beam, which changes the atomic configuration of the collision and so tests the discrepancy in a way a conventional target experiment cannot.