rp Process

Explosive scenarios and the rp process

X-ray bursts are frequently observed thermonuclear flashes ignited on the surface of accreting neutron stars with periods of hours to days. Type I X-ray bursts are powered by the rapid proton-capture process (rp-process), a sequence of rapid proton captures and β+ decays near the proton dripline. Once the underlying nuclear physics is understood, comparisons of burst observations with models offer a unique pathway to constrain neutron-star properties such as accretion rate, accreted composition, or radii.

Direct proton capture rates have been difficult to measure because of the limited intensities of low-energy rare-isotope beams (see review in [Sch06]). Most rates in the rp-process are still based exclusively on theory, but those are rather unreliable. Shell-model calculations, which can be used up to A~60, can estimate the excitation energies of resonant states. However, reaction rates are so sensitive to resonance energies that the rather small uncertainties in the shell-model predictions of around 100-200 keV still can translate into reaction-rate uncertainties of many orders of magnitude.

The vast majority of the reaction rates can also be constrained with powerful indirect techniques. Among them are direct reactions with fast beams and γ-ray tagging that allow measuring the low-lying resonance structure of the unstable nuclei relevant for proton capture rates [Cle04, Che12, Lan14]. The important resonances most often decay via sizable γ-ray branches that can reveal resonance energies with a precision at the keV level. In addition, the use of direct reactions provides access to spectroscopic factors or asymptotic normalization coefficients which also enter the calculations of capture reactions rates to some extent.

The role of in-beam γ-ray spectroscopy with large arrays was first pioneered with Gammasphere using fusion-evaporation reactions with light ions, where as many levels as possible are populated, especially those of low spin near the particle evaporation threshold which play a role in capture reactions. For example, the complete level structure of 20Na below the proton threshold was determined [Sew04], a nucleus central to the reaction sequence 15O(α,γ)19Ne(p,γ)20Na dominating the early breakout of the hot CNO cycle into the rp-process in X-ray burster scenarios. Later, the rate of the 21Na(p,γ)22Mg reaction of importance for oxygen-neon nova outbursts was evaluated [Sew05] based on the determination of the full structure of 22Mg in the region relevant for astrophysical burning. In this instance, γ-ray spectroscopy proved essential to derive the precise energies of the most important astrophysical resonances. More recent measurements have focused, for example, on the 26gAl(p,γ) and 26mAl(p,γ) reactions [Lot09, Lota09], motivated in part by the observation of 26Al decay by the COMPTEL all-sky map project, which reported irregular emission along the plane of the Galaxy. These studies and several others of the same type contribute significantly to the understanding of the rp-process by determining, with high precision, the energies of the states of interest for proton capture and by establishing their spins and parity through angular-correlation measurements. In this context, measurements with GRETA provide powerful new capabilities: not only does GRETA provide higher efficiency at the γ-ray energies most commonly involved (Eγ> 1 MeV), but the ability to measure polarization enables spin and parity determination in instances where angular correlations are impossible (1-step decays) and angular distributions provide no useful information (lack of spin alignment of low-spin states).

Exploiting the high luminosity afforded by fast-beam, thick-target measurement, Figure 1 displays spectra taken with GRETINA during the pioneering γ-ray spectroscopy of the neutron-deficient nucleus 58Zn populated via a proton pickup reaction onto 57Cu projectiles at 30% of the speed of light [Lan14]. GRETINA’s γγ coincidence efficiency and peak-to-total together with the excellent resolution after Doppler reconstruction allowed a level scheme of 58Zn to be constructed for the first time, including the precise identification of 2+ states just above the proton separation energy that are critical for the 57Cu(p,γ)58Zn reaction rate. As a result, the uncertainty of this important reaction rate, that sets the effective lifetime of 56Ni in Type I X-ray bursts, could be reduced by several orders of magnitude [Lan14].

For slightly heavier nuclei away from magic numbers,in the important A~60 region, for example, many resonances of astrophysical importance [Par08a, Par08b, Cyb14] are not accessible to date as the level density of states in the critical region is too high and the γγ efficiency of present arrays is not sufficient to reconstruct the complicated decay schemes. Among the important reactions are, for example, 67As(p,γ)68Se, 63Ga(p,γ)64Ge and 59Cu(p,γ)60Zn, where precise excitation energies of important capture states just above the proton separation energy of the N=Z nuclei 68Se, 64Ge and 60Zn are needed. GRETA used in conjunction with direct fast-beam reactions will open up this mass region for the required, selective in-beam γ-ray spectroscopy studies for the first time. Also, firm spin assignments are needed. For lighter nuclei, spins are often assigned in comparison to theory, but in the regime of high level density, as present in this region of the nuclear chart, this approach fails as the level spacing in the region of interest is comparable or smaller than the uncertainty on the calculated excitation energies. Here, GRETA’s angular coverage and position sensitivity, together with the energy resolution, will allow firm spin determinations from γ-ray angular distributions.


[Che12] J. Chen et al., Phys. Rev. C 85, 045809 (2012).

[Cle04] R. R. C. Clement et al., Phys. Rev. Lett. 92, 172502 (2004).

[Cyb14] R. Cyburt et al., to be published.

[Lan14] C. Langer et al., Phys. Rev. Lett. 113, 032502 (2014).

[Lot09] G. Lotay et al., Phys. Rev. Lett. 102, 162502 (2009).

[Lota09] G. Lotay et al., Phys. Rev. C 80, 055802 (2009).

[Par08a] A. Parikh et al., New Astronomy Reviews 52, 409 (2008).

[Par08b] A. Parikh et al., The Astrophysical Journal Supplement Series 178, 110 (2008).

[Sch06] H. Schatz and K. E. Rehm, Nucl. Phys. A777, 601 (2006).

[Sew05] D. Seweryniak et al., Phys. Rev. Lett. 94, 032501 (2005).

Figure 1: Event-by-event Doppler reconstructed spectrum of 58Zn, using the interaction points from GRETINA's online signal decomposition. This first spectroscopic data of 58Zn was taken with GRETINA at NSCL using the d(57Cu, 58Zn + γ)X reaction at 70MeV/u [Lan14]. Upper panel: Total γ-ray spectrum with the higher-energy region expanded. The inset shows the resolution of a double structure when restricting the polar detection angle theta to ~70o. Lower panel: γγ coincidence spectrum. Figure adapted from [Lan14].

Figure 2: Opportunities to identify and characterize resonances important for the rp-process with GRETA at the S800 at FRIB, using (d,n) reactions performed in the scheme pioneered in [Lan14]. Isotopes marked in red will be accessible at FRIB at rates exceeding 105 pps and will be available for the very detailed γ-ray spectroscopy needed to conquer the high level density in the energy region of interest. 59Cu and 63Ga projectiles, needed for the critical 59Cu(p,γ) and 63Ga(p,γ) reaction rates, are circled. Almost all nuclei along the rp-process path will be available at sufficient intensity for the needed measurements.