Research

Explosive Nucleosynthesis

The study of nuclear reaction rates in explosive astrophysical environments is one of the important frontiers of nuclear astrophysics [e.g., M. Wiescher, H. Schatz, and A.E. Champagne, Phil. Trans. Roy. Soc. London A 356 (1998) 2105]. Observations suggest that nova explosions and x-ray bursts occur in close binaries in which the hydrogen-rich material from the outer layers of an extended star is overflowing its Roche lobe and accreting onto the surface of its companion, a white dwarf or a neutron star, respectively. In such explosive events, the temperatures and densities involved are sufficiently high so that proton and alpha-particle induced nuclear reactions can be fast enough to bypass beta-decay processes. These nuclear reactions involving radioactive nuclei can greatly increase both the rate of energy generation and the total amount of energy produced and can have a dramatic impact on both the isotopic and elemental abundances produced.

During the past three years, our group is focusing its efforts primarily on this area of explosive nucleosynthesis, determining the rates of reactions involving radioactive nuclei in events such as nova explosions and x-ray bursts. Our work includes both stable-beam experiments (locating important resonances and measuring their spectroscopic properties) and direct radioactive-beam measurements of those rates.

Recently, our experiments have concentrated on the properties of the key ¹⁸F(p,a) resonance (E_cm=664 keV) and on the reaction sequence, ¹⁴O(a,p)¹⁷F(p,g)¹⁸Ne(a,p)²¹Na(p,g)²²Mg which can be important both as a breakout path from the Hot CNO cycle to the rp-process and as a possible path for the production of ²²Na in x-ray bursts. The 1274-keV gamma ray from the decay of ²²Na is almost certainly the next gamma-ray line which will be discovered with the new generation of Gamma-Ray satellites which will be launched in the next 2 years. During the next three years we expect to continue to pursue these areas, using both coincidence spectroscopy measurements with our new silicon-strip array and direct radioactive beam studies.

a. The ¹⁸F(p,a) Reaction

The observation of gamma rays associated with nova explosions would provide a very important constraint on nova models. The most intense gamma rays immediately after a nova explosion are expected to be the 511-keV annihilation radiation associated with the positron decay of ¹³N and ¹⁸F. However, the short ¹³N half-life means that the nova envelope will be too opaque during the time scale of the ¹³N decay (t_1/2 » 10 minutes) to let those gamma rays escape, and therefore the most significant source of 511-keV gamma rays during the first several hours after the explosion is expected to be the decay of ¹⁸F (t_1/2 » 110 minutes). The amount of ¹⁸F that will be available as a source for such decays in the outer envelope is limited by the destruction of ¹⁸F during the explosion via the ¹⁸F(p,a)¹⁵O reaction. Without a better determination of the rate of this reaction, it is impossible to understand why recent attempts to detect this radiation have failed.

Our earlier spectroscopy studies and the initial ¹⁸F-beam experiments at both Argonne and Louvain have suggested that an s-wave resonance at E_cm=660 keV plays the dominant role in determining the rate of the ¹⁸F(p,a) reaction in its competition with the ¹⁸F(p,g) reaction in nova explosions and x-ray bursts; the competition between thesereactions determines whether these seed nuclei are recycled back to the Hot CNO cycle or are passed on to the rp-process via ¹⁹Ne. In our more recent studies of this resonance, we have utilized both stable-beam measurements [transfer reactions such as ¹²C(¹⁰B,t)¹⁹Ne] and direct measurements at this resonance using radioactive beams to study both ¹⁸F+p elastic scattering and the ¹⁸F(p,a) reaction.

The results of our ¹⁸F beam measurements at HRIBF (E_cm = 664.7 ± 1.6 keV, G = 39.0 ± 1.4 keV, G_p/G_tot = 0.39 ± 0.02) are sufficiently more precise than the previous measurements that it can be argued that they can be considered to have superceded the results of the earlier experiments.

b. The ¹⁴O ® ²²Mg Reaction Sequence

In current models of nova explosions, the energy generation and nucleosynthesis at temperatures of T£ 0.4 GK are determined by the Hot CNO cycle,

¹²C(p,g)¹³N(p,g)¹⁴O(b⁺n)¹⁴N(p,g)¹⁵O(b⁺n)¹⁵N(p,a)¹²C.

As the temperature and density on the surface of the accreting star increase, however, a-particle induced reactions on the ¹⁴O and ¹⁵O nuclei become faster than the corresponding b⁺ decays, and in X-ray bursters the star may then breakout from the Hot CNO cycle to the rp-process, providing a way to further enhance the rate of energy generation. Based on present nuclear physics information, the initial breakout path is thought to occur through the ¹⁵O(a,g)¹⁹Ne reaction. At higher temperatures (T³ 0.8 GK) the

¹⁴O(a,p)¹⁷F(p,g)¹⁸Ne(a,p)²¹Na(p,g)²²Mg

sequence provides an alternate bridge to the rp-process. The relative roles of the ¹⁵O- and ¹⁴O-bridges to the rp-process depend on the cross-sections of these reactions under extreme conditions of temperature and density.

The ¹⁴O ® ²²Mg Reaction Sequence Þ The ¹⁴O(a,p)¹⁷F Reaction

At high temperatures and densities, ¹⁴O may be burned by the ¹⁴O(a,p)¹⁷F point and increasing the energy generation in X-ray bursts, altering the isotopic abundances produced, and possibly leading to the synthesis of heavier elements via the rp-process. The rate of the ¹⁴O(a,p)¹⁷F reaction is expected to depend critically on the properties of certain resonances in the compound nucleus, ¹⁸Ne. In particular, a J^p=1^- state near E_x=6 MeV (together with its interference with the non-resonant direct reaction cross section) is expected [8] to dominate the rate of this reaction at temperatures in the range T<1.0 GK.

We have therefore studied the ¹H(¹⁷F,a)¹⁴O reaction (the time-inverse of the ¹⁴O(a,p)¹⁷F_gs reaction), using a radioactive ¹⁷F beam at HRIBF. Recoiling ¹⁴O ions were detected in coincidence with a-particles in an array of silicon strip detectors. An excitation function of the (p,a) reaction cross section was measured over the entire energy range of interest for the ¹⁴O(a,p)¹⁷F reaction, 1.0 £ E^a_{cm £} 2.6 MeV. Properties of ¹⁸Ne states with 6.0 £ E_{x £} 7.6 MeV were determined, including the important J^p=1^- state. The direct reaction cross section and its interference with the 1^- state were also measured, allowing - for the first time - the ¹⁴O(a,p)¹⁷F_gs reaction rate to be determined with reasonable certainty for temperatures important for novae and X-ray bursts. The design of this experiment was similar to the ATLAS ¹H(¹⁷F,a)¹⁴O experiment [9], but the combination of the better beam emittance at HRIBF tandem and the 21 days of beam time that were made available there allowed the HRIBF experiment to produce a more definitive data set.

One aspect of the determination of this reaction rate that still needs to be addressed is the role of the ¹⁴O(a,p₁)¹⁷F^*(g)¹⁷F_gs reaction. This will be addressed in a study of ¹⁷F(p,p₁)¹⁷F* inelastic scattering

The ¹⁴O ® ²²Mg Reaction Sequence Þ The ¹⁷F(p,g)¹⁸Ne Reaction

For this sequence to proceed further, the ¹⁷F(p,g)¹⁸Ne reaction must be faster than ¹⁷F beta decay. The rate for this reaction is determined by the s-wave (J^p=3⁺) resonance which Bardayan found at E_c.m.=599.8±1.5 keV (E_x=4.524 MeV ) with G=18±2 keV, as part of his PhD thesis work at HRIBF. While this resonance is now precisely located, there still remains a factor of ~ 2 uncertainty in the strength of this resonance, associated with the unmeasured G_g for this state. Our proposal to measure the strength of this resonance by a direct ¹H(¹⁷F, ¹⁸Ne) g experiment was approved by the HRIBF PAC-4 (Feb. 1999) and identified as the #2 priority for that facility (second only to the ¹H(⁷Be,⁸B)g proposal). The ¹H(¹⁷F, ¹⁸Ne) g experiment is currently waiting for the implementation of the windowless gas target system (Fall-2000) and the development of adequate ¹⁷F beam intensity (Summer-2001 ?).

The ¹⁴O ® ²²Mg Reaction Sequence Þ ¹⁸Ne(a,p) and ²¹Na(p,g) Resonances

The structure of ²²Mg has previously been studied primarily through the two-nucleon transfer reactions (³He,n) and (p,t) which proceed primarily through the direct reaction mechanism, selectively populating 2-particle and 2-hole states in ²²Mg. Although much progress has been made in determining the structure of ²²Mg through these experiments, including a recent TRIUMF/CNS-Tokyo remeasurement of the ²⁴Mg(p,t)²²Mg reaction, the spectroscopic information needed for astrophysics studies is still incomplete. Additional studies using nuclear reactions that proceed through different mechanisms (and therefore have different selectivities for populating ²²Mg excited states) could shed additional light on the location of key ²¹Na+p and ¹⁸Ne+a resonances which may be relevant to explosive nucleosynthesis.

We are therefore currently members of several collaborations carrying out measurements of a variety of reactions using complementary facilities at Yale, Argonne, and TRIUMF, as outlined below. At Yale we have been using the ¹²C(¹⁶O,⁶He)²²Mg. reaction to locate new resonances in the energy regions just above the ²¹Na+p and ¹⁸Ne+a threshold, and we are now initiating a program to measure the proton and a-particle decay of these ²²Mg states. At Yale we are also collaborating with Caggiano and Rehm from Argonne in a study of ²²Mg states using the ²⁵Mg(³He,⁶He)²²Mg reaction, while at Argonne we are part of a collaboration led by Rehm using the time-reversed ²¹Na(p,a)¹⁸Ne reaction to study the ¹⁸Ne(a,p)²¹Na resonances. At TRIUMF we are part of collaborations led by Shotter and by D'Auria that will use a ²¹Na beam to study ²¹Na(p,p) elastic scattering and to directly measure the ²¹Na(p,g) reaction rate, respectively.

¹²C(¹⁶O,⁶He)²²Mg: By making use of the Yale Split Pole with its focal-plane detector (calibrated by utilizing ¹⁶O beams with energies of 72, 84, and 90 MeV to generate a set of 9 peaks corresponding to the ground state and first and second excited states of ²²Mg), we were able to find a total of 19 new levels in ²²Mg and to measure their centroid energies with a precision of 10-20 keV [Alan Chen, PhD Thesis, 1999]. One of these new levels is just above the ²¹Na+p threshold (E_x = 6.041 MeV; E_cm = 540 keV) and could be an important ²¹Na(p,g) resonance; this level is consistent with the level found at 6.046 MeV in the recent TRIUMF/CNS-Tokyo ²⁴Mg(p,t)²²Mg measurement and with the level found at 6.051 MeV in the parallel set of measurements which we have carried out in a study of the ²⁵Mg(³He,⁶He)²²Mg reaction at Yale, using our split-pole spectrometer in collaboration with Caggiano and Rehm from Argonne.

The figure to the right displays the 18 other new levels which we located in our measurements of the ¹²C(¹⁶O,⁶He)²²Mg reaction in the region just above the ¹⁸Ne+a threshold (0<E_cm<3 MeV) which was almost completely unexplored before our study. On the basis of these results the calculated rate N_A<sv> for this reaction is increased by roughly a factor of 10 compared to the earlier model calculations.

In order to better determine the rate of the ¹⁸Ne(a,p)²¹Na reaction and to thus gain a clearer picture of its role in explosive nucleosynthesis, further studies of the structure of ²²Mg are needed - in particular, experiments to measure the partial widths and spins and parities of these resonances, as well as their strengths. Progress in measuring the partial widths of these resonances can be made by using stable beam reactions to populate these states and to then detect their decay products. We are currently starting a series of such measurements at Yale using our split-pole spectrometer coupled to a large-solid-angle silicon array, as described below.

Complementary techniques for determining the strengths of these ¹⁸Ne(a,p)²¹Na resonances involve the use of radioactive beams of either ¹⁸Ne (e.g., the Ph.D. thesis work of Bradfield-Smith who studied the ¹⁸Ne(a,p)²¹Na reaction directly, bombarding a thick ⁴He target with a ¹⁸Ne beam at Louvain-la-Neuve) or ²¹Na (e.g., via the time-reversed reaction ²¹Na(p,a)¹⁸Ne - ATLAS Proposal #691, Rehm et al.). As an example of the complementary nature of the radioactive-beam measurements and stable-beam spectroscopy studies, after the initial ¹⁸Ne-beam measurements at Louvain found ¹⁸Ne(a,p) resonances at E_x=10.990, 11.050, and 11.130 MeV, our more extensive results were then used by the Louvain experiment to plan a second set of lower energy studies which are still under analysis. Our ¹²C(¹⁶O,⁶He)²²Mg results are also being using in planning the bombarding energies for the ANL ²¹Na(p,a)¹⁸Ne experiment.

The ¹⁸Ne(a,p)²¹Na and ²¹Na(p,a)¹⁸Ne experiments are naturally limited to studying ²²Mg states above the ¹⁸Ne+a threshold. Additional studies of ²¹Na+p resonances can be carried out by measuring G_p/G_tot - using the stable-beam coincidence techniques described in section I.c. below. In fact this is currently planned as the first new-physics experiment which we will carry out after our array, after we finish commissioning tests. ²¹Na+p resonance studies will also be carried out in 2001, utilizing the ²¹Na beam at TRIUMF/ISAC first to study ²¹Na(p,p) elastic scattering using the TUDA silicon array and then to directly measure the ²¹Na(p,g)²²Mg reaction using the DRAGON spectrometer. We are collaborators on both of those experiments and recently carried out tests of the DRAGON focal-plane detector at Yale using our spectrometer.

c. Future Explosive Nucleosynthesis Projects

The focal-plane detector [see Section II.a. below] which we developed for use with the Yale Enge split pole provides a highly reliable and well understood instrument for carrying out spectroscopic measurements of many nuclei of importance to Nuclear Astrophysics. For example, various nuclei, such as ²²Mg, where the energy of the excited states is tagged using the ¹²C(¹⁶O,⁶He)²²Mg^* reaction or the nucleus ¹⁹Ne where states are tagged using reactions such as ¹⁹F(³He,t)¹⁹Ne^* or ¹²C(¹⁰B,t)¹⁹Ne^*, have been extensively studied using this instrument. Beginning in September 2001, John D'Auria will be spending a sabbatical year here at Yale during which time we will work on the analysis of the ²¹Na(p,g) data, as well as studies of ²⁰Na+p resonances in ²¹Mg using stable-beam reactions such as ²⁴Mg(³He,⁶He) together with our split-pole spectrometer in order to design of an experiment to subsequently measure the ²⁰Na(p,g) reaction rate at TRIUMF.

With the addition of a large-solid-angle array of silicon detectors in the scattering chamber, it will now be possible to observe the coincidence between events on the focal plane and charged-particle decays from the excited state of the residual nucleus, by gating on a peak in the focal plane spectrum and searching for particle decays from the corresponding excited state. (Initial tests using a single surface-barrier detector to perform coincidence measurements have indicated that a sensitivity to branching ratios of » 3% can be achieved.) Such coincidence experiments will measure the relevant partial widths (G_a and G_p) and branching ratios for that state. With such information, the calculation of the rates of reactions involving radioactive nuclei, such as ²¹Na(p,g)²²Mg, can become possible, complementing the use of a radioactive beam or target. The first coincidence experiment performed with our new array will be a measurement of the proton decays of ²²Mg states of importance as resonances for the ²¹Na(p,g)²²Mg^* reaction. This same reaction will then be used to look at the higher energy states in ²²Mg (above the a-particle threshold at 8.14 MeV) to measure the a-particle and proton branching ratios of states important as resonances for the HotCNO-cycle break-out reaction, ¹⁸Ne(a,p)²¹Na (see above figure).

Additional measurements of interest include experiments such as ¹⁹F(³He,t)¹⁹Ne^*(a)¹⁵O [looking at alpha decays from states important to the Hot CNO cycle break out reaction ¹⁵O(a,g)¹⁹Ne], or ¹⁶O(³He,³He)¹⁶O^*(a)¹²C, [to look at alpha decays from excited ¹⁶O states with the intention of determining the strength of the ¹²C(a,g)¹⁶O reaction], or ¹²C(¹²C,⁶He)¹⁸Ne ^*(a/p)¹⁴O/¹⁷F [to look at a-particle and proton decays from the excited states of ¹⁸Ne to determine the reaction rate for the ¹⁴O(a,p)¹⁷F and ¹⁷F(p,g)¹⁸Ne reactions], and ²⁵Mg(³He,⁶He)²²Mg ^*(a/p)¹⁸Ne/²¹Na [to study ²²Mg through an alternate entrance channel].

At the same time, we will continue to be involved as active members of RIB collaborations at ANL [working on the ²¹Na(p,a )¹⁸Ne reaction], at TRIUMF [working on the ²¹Na(p,p) and ²¹Na(p,g) reactions], and at ORNL/HRIBF [continuing our work on the ¹⁷F+p system, in particular the ¹⁷F(p,p') and ¹⁷F(p,g) reactions, and working on the ⁷Be(p,g) reaction, as well as working towards the development and use of beams of ²⁵Al and ³³Cl]. In anticipation of the possibility of future ²⁵Al and ³³Cl beams, we have recently begun collaborating in studies of the locations of ²⁵Al+p and ³³Cl+p resonances using the ²⁹Si(³He,⁶He)²⁶Si and ¹²C(²⁸Si,⁶He)³⁴Ar reactions, respectively. These measurements are being carried out at Yale, using our split pole spectrometer, collaborating with Caggiano et al. (ANL) (the ²⁶Si study) and with Bardayan et al. (UNC/TUNL) (the ³⁴Ar study).

Instrumentation

a. The Yale Lamp Shade Array (YLSA)

The initial array will be made up of five LEDA-type silicon strip detectors, mounted in a lampshade configuration in the backward hemisphere of the scattering chamber. The solid angle for looking at proton decays from ²²Mg excited states will be approximately 25% of the full solid angle in the Lab, giving a high detection efficiency. The high segmentation (16 azimuthal strips per detector) gives angular information for every event and reduces the effects of pileup in both the ADC's and the TDC's. The first three detectors have been purchased and will be in place by the end of this summer. The electronics to implement these detectors will consist of preamplifier circuits constructed in Brookhaven National Laboratory, mounted on a motherboard which allows 32 channels of electronics to be contained in a unit of dimension 6"x2"x6". The preamplified signals will go through a further stage of amplification using RAL109 amplifiers made by the Rutherford Appleton Laboratory (UK). The preamplifiers are at present undergoing final tests and should be available for use at the end of this summer. The amplifier electronics is currently being tested.

Due to the flight time (~100 ns) of the ⁶He particle from the target to the focal plane, it is necessary to delay the signals from the Si-array. The event rate in the focal plane detector is of order 100Hz, while that in the silicon array (with a solid angle of p steradians) is of order 100 kHz. Therefore, the focal plane, with its lower rate, will be used to define the trigger, while the time signal for each channel in the silicon array will be delayed by 300 ns to allow a time coincidence to be made. The timing delay will be implemented using active ECL delay chips, which take the ECL output of the RAL amplifier as their input and delay the signal by 300 ns. The rise time, and hence the resolution in time, is smaller using an active circuit to produce the delay than in the case where an inductance-resistor circuit is used, which typically degrades the voltage level in addition to increasing the rise time of the logic pulse.

b. YLSA Data Acquisition

To reduce hardware costs, a new front-end is being built for our acquisition system; this will be VME, rather than CAMAC, based. New ADCs and TDC's for the VME system (made by CAEN) allow 32 channels of electronics to be contained in a single width VME module at a fraction of the cost that would be possible using FERAbus electronics. Two further advantages area that the acquisition system is now contained in one crate rather than two and that the maximum acceptable event rate will be higher. The software for this new system is currently being written and is expected to be fully operational by the end of this summer.

c. Jam

We have written a computer based data acquisition (DAQ) system, Jam, using the Java programming language. Java has many high level features which allow rapid development of complex programs, allowing us to rapidly develop a user-friendly DAQ system with a familiar graphical user interface which we believe makes it easier to learn and use than other currently available DAQ systems. Our philosophy in developing the system is that simplicity of use is a higher priority than generality, at least as the program is initially presented to the user. Setting up the acquisition system is straightforward for the majority of cases. A paper describing the system has been accepted for publication in Nuclear Instruments and Methods in Research A. In the past year, we have been able to make a number of improvements to Jam. The Java graphical user interface has gone through a significant upgrade with the addition of "swing" classes, and we have modified Jam to take advantage of this upgrade. At present a stable beta version 1.2 of Jam is in use for data acquisition at WNSL. At present, Jam is used at WNSL (by the Nuclear Astrophysics group and by a number of visiting research groups) as well as at TUNL (by their Nuclear Astrophysics group).

Documentation of the system is available at http://jam-daq.sourceforge.net/

The new silicon array described in Section-b. above will have a significant number of new electronics channels (90 energy + 90 time). For both simplicity and economy, we have chosen to convert our data acquisition hardware to VME-based ADC and TDC modules. Jam's front end is already programmed on a VME microprocessor module, but at present it relies on a specific VME-to-CAMAC bus module, with all digitization taking place in a CAMAC crate. Due to the new electronics configuration, it will be necessary to modify Jam to allow it to accept data from VME modules in an event-by-event mode (as contrasted with a high-speed mode in which every signal gets a time stamp associated with it). In addition, a hierarchy will have to be introduced to the list of histograms viewable in the Jam application, in order for the greatly increased number of channels of information to be easily accessible. As noted above, this new software is currently under development and is expected to be operational by the end of the summer.