²²Ne(α,n)²⁵Mg at the INFN Bellotti Ion Beam Facility

Abstract

Neutron capture reactions are the main contributors to the synthesis of heavy elements through the s-process. ²²Ne$(\alpha ,\mathrm{n})$²⁵Mg is the main neutron source in stars, together with ¹³C$(\alpha ,\mathrm{n})$¹⁶O. At energies ${\mathrm{E}}_{\mathrm{cm}}$ < 700 keV, limited data are available, i.e., reaction cross-section upper limits from direct experiments and highly uncertain estimates from indirect sources exist. The ERC project SHADES is currently performing direct cross-section measurements at these energies. We will present details on the ongoing experiment and discuss target characteristics, experimental backgrounds, and preliminary analyses on the detector efficiency and the 832 keV resonance.

1. Introduction

In stars, the ²²Ne$(\alpha ,\mathrm{n})$²⁵Mg reaction is of primary relevance in the synthesis of elements beyond iron. In particular, together with ¹³C$(\alpha ,\mathrm{n})$¹⁶O in AGB1 stars, it is responsible for the synthesis of 90 < A < 204 elements, in what is usually referred to as the s-process. Its negative Q-value of −478 keV requires a temperature of ${\mathrm{T}}_9\sim 0.3$ for ²²Ne$(\alpha ,\mathrm{n})$²⁵Mg to be activated, which is typical of the short helium-flash episodes in thermally pulsating AGB stars [1]. In massive stars, it is also the main source of neutrons and is responsible for the synthesis of 60 < A < 90 elements via the reaction chain ¹⁴N${(\alpha ,\gamma )}^{18}\mathrm{F}{\left({\beta }^{+}\nu \right)}^{18}\mathrm{O}{(\alpha ,\gamma )}^{22}\mathrm{Ne}{(\alpha ,\mathrm{n})}^{25}\mathrm{Mg}$, which is initiated with the ¹⁴N from the CNO cycle [2].

²²Ne$(\alpha ,\gamma )$²⁶Mg directly competes with ²²Ne$(\alpha ,\mathrm{n})$²⁵Mg in ²²Ne destruction. With its positive Q-value of 10.6 MeV, it activates at a lower temperature and can deplete the ²²Ne before the star temperature is high enough for the neutron channel to be open. An accurate knowledge of the low energy cross section of both reactions is then required for an accurate understanding of the s-process in several stellar scenarios.

Cross-section measurements of ²²Ne$(\alpha ,\mathrm{n})$²⁵Mg below 1 MeV are extremely challenging. As shown in Figure 1, except for a strong resonance located at $E_{\alpha }$ = 832 keV, only upper limits have been measured so far [3,4,5]. The background induced by cosmic-rays, the ion beam, and by the intrinsic radioactivity of detectors, limited the sensitivity of experiments and the possibility to investigate low cross-section regimes. Despite the contribution of indirect investigations towards improving our understanding of low energy resonant capture mechanisms [6,7], direct measurements are still required to solve present discrepancies [8]. To improve our understanding of the s-process, a new experimental campaign on the direct measurement of ²²Ne$(\alpha ,\mathrm{n})$²⁵Mg is ongoing at the IBF [9,10] of the LNGS-INFN [11] by the ERC funded project SHADES and the LUNA collaboration [12,13,14] (and references therein). Together with ASFIN [15,16], n-TOF [17,18], ERNA [19,20,21,22,23], and PANDORA [24], LUNA is one of the main contributors to nuclear astrophysics research in Italy. In this work, we give an overview of the experiment and some very preliminary results.

Figure 1. Extant literature on ²²Ne$(\alpha ,\mathrm{n})$²⁵Mg [3,4,25,26,27]. Cross section is shown as a function of $\alpha$-beam energies. Lower triangles identify upper limits. Harms et al. data [26] are not to scale.

2. Experimental Methods

The sensitivity of previous cross-section measurements has been limited by the high neutron flux from either the cosmic rays or induced by the ion beam impinging portions of the experimental setup [28]. The SHADES project designed different strategies to mitigate this issue and maximize the signal-to-noise ratio: performing measurements in an underground location with a very high intensity ion beam; using a high-purity, enriched ²²Ne windowless gas target; and employing a hybrid ³He proportional counter-EJ-309 liquid scintillator detection array (See Figure 2 in [29]). These are key ingredients described in the following section.

The natural neutron flux is reduced up to four orders of magnitude in LNGS [30,31], making it an ideal place to perform nuclear astrophysics experiments. At this level, the intrinsic natural radioactivity of a detector became a major source of background, e.g., [32] (and references therein). We performed an investigation to identify and quantify these contaminants and their effects [33], using a similar analysis that had already been performed in the past by the LUNA collaboration for similar ³He proportional counters [34]. Briefly, after an initial calibration with certified neutron sources, a PSD algorithm was set to separate neutron and $\gamma$-ray events. The scintillators were then moved underground in LNGS in the LUNA400 accelerator hall [35,36] (and references therein) and a 1-month-long background spectra was acquired. The neutron events exceeded the expected one and well defined spot were seen in the PSD matrix. A time-coincidence analysis between acquired events has shown that those spots were due to ff particles emitted by actinides in the detector case, which caused a background level of 64 events/h in the relevant region of interest. Despite this, the proportional counter-scintillator coincidence measurements could completely suppress this contribution.

The IBF can deliver an up to 500 $\mathsf{\mu }\mathrm{A}$ ⁴He⁺ beam [9]. This can increase the signal-to-noise ratio in the case of time-dependent background sources, such as cosmic-ray induced background and the intrinsic background of the detectors. At the same time, if the beam impinges on parts of the setup, it can cause a BIB that can drastically reduce the detection sensitivity [28]. Among the others, ¹³C and ¹⁰B are the most abundant species, which can cause an intense BIB for ²²Ne$(\alpha ,\mathrm{n})$²⁵Mg measurements.

During the measurement, we could exploit the whole beam current below the 830 keV resonance. At higher energies, the high reaction rate made the neutron flux from the reaction too high, increasing the detection dead time. For this, we limited the beam current to 150 μA. A total of almost 1000 C was accumulated, especially at the lower energies.

¹³C and ¹⁰B contaminants are commonly found in ion beam setup materials, and the associated $(\alpha ,\mathrm{n})$ cross-section is an order of magnitude higher than the one of interest. To reduce their presence, the beam line parts exposed to the ion beam and its halo were coated with oxygen-free high-conductivity copper cleaned with a nitric acid solution in an ultrasound bath. In addition, only aluminum was used for the vacuum sealings: the standard viton material is well known to release carbon compounds under a vacuum, which can accumulate on the walls of the ion beam line.

The 99.995% enriched ²²Ne gas is extremely expensive and makes a gas recirculation system necessary. Over time, this can lead to a build up of carbon chemical compounds from vacuum sealing, from the oil of the vacuum pumps, and the small yet unavoidable leaks of the vacuum system. In addition to the aluminum vacuum sealings, we employed silicon-industry-grade pumps, which are very demanding in term of cleanliness, and a noble gas purifier in order to minimize carbon content in the gas target. Gas purity was monitored during beam time by measuring the yield of the 830 keV resonance.

The detection array has the capability to remove BIB events and increase the measurement sensitivity. To put it into operation, we performed a preliminary performance measurement campaign at the Goethe University Frankfurt’s Van de Graaff accelerator facility FRANZ [37]. Here, neutrons in the energy range 50–700 keV were measured with a reduced version of the array. Energy calibrations were performed and the pulse shape analysis techniques were improved using two approaches: the standard PSD and a more sophisticated neural network analysis, which proved to be performing more at the lowest energies [38].

A 1-year-long ²²Ne$(\alpha ,\mathrm{n})$²⁵Mg measurement campaign was started at IBF in November 2024. It included a precise determination of the ${\mathrm{E}}_{\mathrm{cm}}$ = 706 keV resonance energy and its uncertainty. In addition, we performed long runs below 800 keV when searching for new resonances or, in case they are not accessible, in the search of lower upper limits for the cross section.

To determine the detection array efficiency, devoted measurements were performed at the HUN-REN Institute for Nuclear Research (ATOMKI) in Hungary, exploiting the ⁵¹Vn${(\mathrm{p},\mathrm{n})}^{51}\mathrm{Cr}$ reaction down to threshold energies. A solid target was employed and the detection array moved along the beam direction to evaluate the efficiency for neutrons generated in a region corresponding to the gas target section. Analysis is undergoing. Details and results will be reported in a devoted publication.

3. Results

Also, in LNGS, ²²Ne$(\alpha ,\mathrm{n})$²⁵Mg has been challenging. BIB proved to be the main background source, against which we iteratively applied the solution described in the previous section.

As preliminary results, Figure 2 reports an energy scan of the ${\mathrm{E}}_{\mathrm{cm}}$ = 706 keV resonance. Uncertainty bars represent the statistical uncertainty. We repeated the same investigation for several target thicknesses ranging from about 500 eV up to 40 keV. The analysis is still ongoing, but a reduction in the overall resonance energy uncertainty is expected.

Figure 2. Experimental yield around the ${\mathrm{E}}_{\mathrm{cm}}$ = 706 keV resonance energy.

BIB was particularly problematic at the lowest energies. The straggling of the ion beam in the gas target causes losses of the beam into the several aperture along the beam line, in particular on the one downstream of the reaction chamber. Being close to the detector array, BIB events had a higher detection efficiency compared to other portions of the beam line. To compensate for this, in addition to the anti-coincidence capability of the detector array, we performed measurements at the same energy and ion beam tuning parameters with both ²²Ne gas and ²⁰Ne gas. This allows for a better background subtraction, pushing the measurement sensitivity close to the ideal one. The measurement campaign ended on December 2025 and the analysis is currently ongoing.

4. Discussion

Low energy measurements of ²²Ne$(\alpha ,\mathrm{n})$²⁵Mg are required to properly constrain the stellar models of massive AGB stars, and to better understand the nucleosynthesis of trans-ferric isotopes. In IBF-LNGS, we performed a long measurement campaign to improve the state of the art, which is currently limited by the energy definition of the ${\mathrm{E}}_{\mathrm{cm}}$ = 706 keV resonance and the lack of cross section data at lower energies, where only upper limits are available. We preliminarily conclude that the energy precision improves to on the order of 100 eV and that the upper limit for the low-energy cross sections is reduced by a factor of 2. However, additional analyses are required to confirm these values and to further improve the sensitivity of the measurement. For this, future experiments may also consider the use of a supersonic gas jet target [39,40], which requires no apertures for the confinement of the gas close to the reaction region.

The s-process is still a hot topic in nuclear astrophysics [41], together with reactions such as 12C+12C [42,43] and ¹²C${(\alpha ,\gamma )}^{16}O$ [44,45]. With this work, we are pushing the accuracy and the precision of its determination to levels required for astrophysical models. For this, a comprehensive R-Matrix analysis (e.g., [46,47,48]) will be performed as a final step, with the goal to provide reliable extrapolations of the reaction rates at the stellar energies.