1 Introduction

In the past few decades, the study of nuclei far from the stability line has become a rapidly growing research area owing to continuously developing radioactive beam techniques. New decay modes were discovered in rare isotopes with extreme proton-to-neutron ratio such as β-delayed (multi-)particle emission and direct (multi-)particle emission processes [1-3]. Investigations on these decay modes can provide rich spectroscopic information, such as level energies, spins, parities, and level densities and information on their emission mechanism as well [4-6]. The comparison between experimental data and theoretical prediction for exotic nuclei can improve our understanding of the structure and behavior of nucleons inside the nucleus. It can also help us to learn various features of nuclei, such as neutron halos, or clustering phenomena [7]. Nuclei very far from the stability line are also particularly important in the nuclear-synthesis of elements; therefore, exploring exotic nuclei and their decay properties is significant for the thorough understanding of certain fundamental questions in astrophysics, such as the creation of isotopes, stellar evolution, star explosion, and other aspects of the universe [8-10].

Generally, there are two detection methods to study the β-delayed charged particle decay of proton-rich nuclei: the implantation-decay method, mainly using a silicon detector array, and the time-projection chamber (TPC) detection method. The basic principle of the implantation-decay method can be described as follows. The secondary unstable nuclei beams are implanted and stopped into a silicon detector, and the β-delayed charged particle can be detected by an implantation detector at the exact implantation position following the implantation. The energies of the β-delayed charged particles can be detected accurately, as the decay is detected inside the implantation detector and no dead-layers of the silicon detectors are passed through when the β-delayed charged particles are stopped in the silicon detector. Depending on the purity of the secondary radioactive beams, two implantation (continuous or beam on/off) modes can be chosen in different experiments. If the beam purity is not appropriate, the continuous implantation mode is more suitable. An analysis process, which can remove the contaminant backgrounds efficiently is described in this article later. The beam on/off implantation is adopted when the secondary beams are well purified and the contaminants can be suppressed properly in this mode; however, the utilization of the beam can be lower in beam on/off mode. Compared to the implantation-decay method, detection with the TPC is a newer technique to investigate the rare decay modes of proton-rich isotopes with extremely high isospin, such as the study of two-proton emission from ⁴⁵Fe [11] and ⁵⁴Zn [12] at the Large Heavy Ion National Accelerator (GANIL) in France. Measurements by TPC can provide a more direct and clear insight of the decay processes, especially when multi-particle emissions from proton-rich isotopes are studied. Angular and momentum correlations between the particles in in the multi-particle emission from the decay process can be established easier by using a TPC measurement.

²²Al was the first exotic odd-odd Tz=-2 β-delayed proton emission precursor, which was first observed by Cable et al. in 1982 [13]. In this experiment, only two high-energy peaks (8.212 (16) MeV, 8.537 (22) MeV) in their decay-energy spectrum were clearly confirmed. An approximate half-life value (${70}_{-35}^{+50}$ms) of ²²Al was determined. One year later, Cable et al. observed a β-delayed two-proton (2p) emission mode from ²²Al, which was the first discovery of this kind of decay mode from highly proton-rich nuclei [14]. In 1997, Blank et al. performed an experiment by implanting ²²Al into silicon detectors and into a micro-strip gas counter(MSGC) [15]. Using drift-time analysis in the MSGC, β-delayed α emission from ²²Al was observed for the first time, while a half-life value (T_1/2=59±3 ms) was also obtained in the experiment. In 2006, an experiment was performed using better purified ²²Al and γ detection by Achouri et al. [16], In this experiment an implantation-decay method and beam on/off mode were used, and more proton peaks, γ-particle coincidences and T_1/2=91.1±0.5 ms was obtained. Although several experiments had been performed focusing on ²²Al, more experimental data is required to understand the large dispersion of the previous half-life values and other details of the decay process.

A newly designed silicon box detection system was used to measure β-delayed particle emission from ²²Al. ²²Al is also a β-delayed two-proton decay precursor and the surrounding silicon detectors in the silicon box detector are designed to detect the relative angular and momentum correlations between the two β-delayed protons from ²²Al, when the two protons escape the centered implantation detector and detected by the surrounding detectors. In our experiment, the half-life value of 90.8±1.3 ms was given, which is consistent with the previous result by Achouri et al. [16]. The energy spectrum of β-delayed protons from ²²Al obtained in our experiment shares the main features of all previous experiments for ²²Al. In addition, in our experiment γ rays were also detected by five Clover-type detectors and the proton γ-ray coincidence was also obtained. The results of our experiment indicate that the newly designed silicon detector array achieves a good performance in continuous-implantation mode, when the secondary beam is severely contaminated by other unstable β-delayed charged particle decay precursors. The experimental setup, data analysis, and discussion is discussed later in this article.

2 Description of the experiment

The experiment was performed at the National Laboratory of Heavy Ion Research (HIRFL) of the Institute of Modern Physics, Lanzhou, China. ²²Al was produced by the projectile fragmentation of a primary beam of ²⁸Si at 76 MeV/nucleon, with an average intensity of 50 enA impinged on a 384.8 mg/cm² Be target. The setting of the dipole magnet was optimized for ²²Al. A series of aluminum foils were installed upstream as an energy degrader to ensure that most of the ²²Al ions were implanted and stopped into a double-sided silicon detector (DSSD0). A total number of 1.3×10^{5 22}Al was recorded in the experiment. The average implantation rate of ²²Al was 18/min, and the average purity of ²²Al was 0.26%. The main contamination in the secondary beam were ²¹Mg(∼7%), ²⁰Na(∼34%), ¹⁹Ne(∼56%), and other stable isotopes. Particle identification were performed by using energy-loss and time-of-flight (TOF) correlations.

A schematic drawing of the experimental setup is shown in Fig. 1. A 4π silicon box detector was designed for implantation and decay detection. In front of the box, three silicon detectors (Δ E1-Δ E3) and two scintillation detectors (T1 and T2) were installed for particle identification. The silicon box detector was composed of one 69 μM thick double-sided silicon detector (DSSD0) in the center of the box, surrounded by further silicon detectors. The structure of the silicon box was designed to detect both the β-delayed one-proton and β-delayed two-proton emission processes from the ²²Al. A 42.5 (1)° angle between the centered and the bottomed DSSD planes was set to achieve a better acceptance of the emitted charged particles. Four double-sided silicon detectors (DSSD1–DSSD4) followed by four quadrant silicon detectors (QSD1–QSD4), and two silicon circular dichroism (CD) detectors (DSSD5 and DSSD6) were used to cover the 4π solid angle to achieve the highest detection efficiency. The thicknesses of DSSD1–DSSD6 were 64 μM, 61 μM, 304 μM, 525 μM, 317 μM, and 315 μM, respectively. The thicknesses of QSD1-QSD4 were 1533 μM, 1546 μM, 314 μM, and 309 μM, respectively. The energy-resolution of each DSSD was 50 KeV. Five high-purity (HP) germanium clover detectors were installed outside the silicon box to detect γ-rays emitted from the decay of the implanted nuclei during the experiment. The total efficiency for γ-rays at 1.528 MeV was estimated to be 1.1(0.2)%.

Figure 1:Full-screen View

(Color online) Experimental setup on RIBLL1. A detailed description can be found in the text.

All silicon detectors were installed on printed circuit boards. The pre-amplifiers for the silicon detectors are SPA02-type, designed by the China Institute of Atomic Energy. A circulating alcohol cooling system was used to cool the silicon detectors and the pre-amplifiers during the experiment, to maintain their temperature at ∼0 °C and ∼30 °C, respectively. Two different electronic signal gain factors were set for the DSSD0 to detect high-energy (greater than 100 MeV) implantation ions and low-energy (several MeV) β-delayed particles. Both the implantation signals and the decay signals were used to trigger the Versa Module Europa (VME) data acquisition (DAQ) system.

A Monte Carlo simulation of proton detection efficiency was performed with the Geant4 software toolkit for our experimental setup. In the simulation, a Gaussian distribution was assumed for the implantation depth alongside the beam direction in the DSSD0. The simulation results are shown in Fig. 2, where the full square line represents the proton detection efficiency for the DSSD0 and the full triangle line represents the proton detection efficiency for the silicon box. As can be seen in Fig. 2, most protons, which escaped the implantation silicon detector can be detected by the surrounding silicon detectors.

Figure 2:Full-screen View

Monte Carlo simulation results of proton detection efficiency for the DSSD0 (full square line) and the silicon box (full triangle line).

²⁰Na is one of the main contaminations in the experiment, and it is also a β-delayed α emission nuclei widely studied by several different experiments [18-21]. Several strong β-delayed α branches are available and their energies are exactly known from various studies, which can provide a perfect standard source to calibrate the in-beam of the DSSD0. The surrounding silicon detectors (DSSD1–DSSD6, QSD1–QSD4) were calibrated by a standard composite α source (Pu-239, Am-241, and Cm-244). The calibration of the ToF-Δ E spectrum was done by an LISE++ simulation [22]. The energy of the clover detectors were calibrated by a ¹⁵²Eu standard source.

3 Results and discussion

In the experiment, there were strong contaminations in the secondary radioactive beams where ²¹Mg and ²⁰Na are also β-delayed proton and β-delayed α precursors; therefore, a special correlation method was used in the data processing to subtract the backgrounds. The correlation method was implemented as follows. The ions of interest were implanted and stopped into one pixel of the centered silicon detector (DSSD0). The implanted ion was identified and flagged by the ToF-Δ E correlation method. Following the implantation event, in a time interval of 4000 ms, all the decay events in the same pixel of the implantation event were correlated with the implantation event.

The time interval was set to 4000 ms, which is sufficiently long compared with the half-life value of the ²²Al. In the analysis procedure, the correlation between an implantation event and its corresponding decay event can not be established directly in an experiment with continuous-implantation. The decay event in the same pixel could belong to any other implantation events, which implanted before the decay event, if the time interval between the decay event and the implantation event is suitable for the half-life of the implantation ion. When a decay event was correlated with all possible implantation events, several uncorrelated events exist, which contribute the constant background in the decay time spectrum. In the decay time spectrum, all correlated events contribute to an exponential decay trend, and the uncorrelated events contribute to a constant background. The events in the range of 2000–4000 ms in the decay time spectrum in Fig. 3 were considered to be the constant background. By subtracting the decay energy spectrum of the selected background from the decay energy spectrum of the correlated and uncorrelated events, the contaminations can be efficiently removed in the energy spectrum of the β-delayed charged particles for ²²Al. A similar data analysis procedure was used in the experiment by Dossat et al. in 2007 [23].

Figure 3:Full-screen View

Decay time spectrum of ²²Al. A fit with a formula composed of an exponential decay and a constant background gives a half-life value of 90.8±1.3 ms for ²²Al. The lower figure shows the residuals between the data and the fit function.

The decay time distribution in Fig. 3 includes the correlated (exponential curve) and uncorrelated (constant background) events. The function

describing the standard radioactive decay law and a constant background was used to fit the decay time spectrum. A half-life value of 90.8±1.3 ms was obtained from the fit for ²²Al. The lower figure in Fig. 3 shows the residuals between the data and the fit function.

All previous experimental data of half-life values for ²²Al, together with our result are listed in Table 1, where there a large dispersion in the three previous half-life values for ²²Al can be seen. The half-life value obtained by us is consistent with the previous value by Achouri et al.; however, the error bars are larger. The experimental half-life value for ²²Al by Achouri et al. (2006) is confirmed by this experiment.

As mentioned above, the decay time spectrum in Fig. 3 includes both correlated and uncorrelated events. Following the background subtraction from the energy spectrum of β-delayed charged particles from correlated and uncorrelated events, the energy spectrum of β-delayed charged particles from ²²Al was obtained as shown in Fig. 4, where up to 15 peaks can be seen. It exhibits the same main features as those observed in a previous measurement [16]. We denote the various peaks with peaks numbers i (where i=1,2,...,15). A Gaussian distribution was used to extract the peak energy for each peak. The branching ratios were obtained by dividing the areas of each peaks by the numbers of ²²Al implanted, with considering the detection efficiency. Peak 12 at 5.541±0.025 MeV is due to the β-delayed α particle emission from the contaminant ²⁰Na, and it is verified by the fitness of the decay time. As peak 12 is considered as one of the peaks from the decay of ²⁰Na, peak 6 is expected to be contaminated by ²⁰Na due to a significant peak energy at 2.685 MeV in the energy spectrum of the charged particles from the decay of ²⁰Na, which is very close to peak 6. This contamination can be subtracted by the ratio of the relative intensities from ²⁰Na. The branching ratios of other peaks in the energy spectrum of the charged particles from the decay of ²⁰Na are very small compared with the peak energies at 2.685 MeV and 5.547 MeV [18]; thus, the other peaks in the contamination of ²⁰Na are only considered in the error bars in our results. All peaks and branching ratios were listed and compared with previous measurements in Table 2, where the energies are the total decay energies in the center-of-mass frame.

Figure 4:Full-screen View

Energy spectrum of the β-delayed charged particles from ²²Al detected in the central and surrounding silicon detectors. Peak 12 is from the contaminant ²⁰Na.

Figure 5 shows four γ rays detected by the five germanium clover detectors triggered by β-delayed particles from ²²Al. γ ray 1 at 0.332 MeV originates from the γ transitions from the first excited state to the ground state in ²¹Na, and γ ray 3 at 1.384 MeV originates from the γ transitions from the second excited state to the first excited state in ²¹Na. γ ray 2 at 0.511 MeV originates from positron–electron annihilation, while γ ray 4 at 1.633 MeV originates from the γ transitions from the first excited state to the ground state in ²⁰Ne.

Figure 5:Full-screen View

γ-ray spectrum detected by the germanium clover detectors triggered by the β-delayed charged particles from ²²Al. Details of γ transitions are discussed in the text.

Figure 6 shows the charged particle spectrum of ²²Al in coincidence with the γ-ray energy at 0.332 MeV and it also shows the decay branches from excited states from ²²Mg to the first excited state in ²¹Na. Considering the difficulties of removing the background of the gamma and charged particle spectrum completely, the charged particle spectrum of ²²Al in coincidence with the γ-ray energy at 0.332 MeV also contain charged particles, which do not correlate with the 0.332-MeV γ-rays. We can distinguish between correlated charged particle peaks and uncorrelated charged particle peaks in Fig. 6 by comparing the relative branching ratios of each proton peaks before and after the γ coincidence in Figs. 4 and 6. In Fig. 6, charged particle peak 1 is significantly more suppressed than charged particle peaks 2, 3, and 6, after the gate of 0.332 MeV γ-rays on the energy spectrum of β-delayed charged particles from ²²Al. Thus, the 0.332-MeV γ rays are correlated with proton peaks 2, 3, and 6. This result is consistent with the information of coincidences in a previous experiment performed by Achouri et al. [16].

Figure 6:Full-screen View

Energy spectrum of the β-delayed charged particle emission from ²²Al, in coincidence with the γ-ray energy at 0.332 MeV.

In this experiment, the coincidence of the charged particle peak 10^* at 4.461 MeV with the 1.633-MeV γ-rays from ²⁰Ne was observed, which indicates a β-delayed two-proton decay branch of ²²Al to the first excited state in ²⁰Ne, as observed in a previous experiment [16]. The energy of the charged particle peak 9 (E = 4.009±0.025 MeV) is in agreement with the β–α transition measured by two previous experiments [15,16]. However, in the present experiment, no clear γ rays at 1.887 MeV were observed due to the low statistics. Thus, no β-.α transition was observed in this experiment. In future experiments, more statistics are required to observe these decay branches.

4 Conclusion

β-delayed charged particle emission from the proton-rich nucleus of ²²Al has been investigated experimentally on RIBLL1. ²²Al was produced by projectile fragmentation of a ²⁸Si primary beam on a Be target. In the experiment, a continuous-implantation method was used. The energy spectra of the β-delayed charged particles from ²²Al, the γ-rays correlated with β-delayed charged particles from ²²Al, and the β-delayed charged particles from ²²Al correlated with γ-rays at 0.332 MeV are presented in this work. The half-life value for ²²Al was measured as 90.8±1.3 ms, which is in good agreement with the previous experimental result of 91.1±0.5 ms by Achouri et al. [16]. The results of our study on the decay of ²²Al in the presented experiment show that the silicon detector array performs effectively in the measurement of β-delayed charged particle emission in the continuous implantation-decay method.