New High Mass Records for DRAGON

DRAGON achieved two records earlier this year with the highest mass proton capture reaction measured at the facility immediately followed by the highest mass beam delivered to the facility. These measurements were part of tests designed to demonstrate the feasibility of a measurement of the 76Se(a,g)80Kr reaction, which would be the highest mass capture reaction ever measured at DRAGON and a very technically challenging experiment. 

A technical description of the tests follows, taken from a report issued to TRIUMF's Experimental Evaluation Committee in summer 2011: 

58Ni run

In early April of 2011, DRAGON took 12 shifts of beamtime in order to study the feasibility of measurements at high mass and high charge state. 8 shifts of 58Ni beam at 1.42 MeV/u were taken in order to (a) determine the magnitude of the beam suppression of DRAGON for a proton capture reaction that has roughly the same Dp/p (~1%) for neighboring charge states as the proposed 76Se(a,g)80Kr reaction (as it is expected that the dominant source of background will be neighboring charge states scattered into the energy acceptance of the separator), (b) detect 59Cu recoils at the end of the separator using local time-of-flight and DE-E ionization chamber methodology as would be used in the final experiment, and (c) re-familiarize the group with the use of post-target silicon-nitride charge state booster foils required to achieve charge-states high enough to overcome the rigidity limitations of the separator. 

The 58Ni beam was delivered at q=10+ with an intensity of 700 epA (4.4E8 pps). Transmission through the DRAGON gas target was 100%, and an operating pressure of 6 Torr H2 was used. A 50 nm thick SiN charge state booster foil was used on the downstream side of the gas target enabling q=20+ beam to be bent round the first dipole magnet to the energy dispersed focus. The separator was then scaled for 59Cu recoils and the leaky beam rate at different separator slit settings was investigated.

Initially, with no local time-of-flight system in place, and regular DRAGON (large acceptance) slit settings, it was found that the rate of leaky beam particles reaching the focal plane detector was 3 kHz for this selected recoil charge state and beam intensity. With a modest reduction in mass slit settings to 10mm, far above what would be close to cropping recoil transmission efficiency, this was reduced to 60% of that rate. With the MCP local-TOF system in place the rate was reduced to 300 Hz, implying that the leaky beam is dispersed in the horizontal plane and more intense on the low energy side of that plane, thus being mostly blocked by the 1” aperture of the initial MCP foil (note that recoils are focused to a small cone angle at this point so are not in danger of being cropped by this aperture), equivalent to using the final slits for suppression purposes.

Further tests with charge slits and mass slits were performed to optimize the reduction in leaky beam particles without a subsequent reduction in recoil transmission. In this way optimum settings of 20 mm and 8 mm were found for the horizontal charge slits and mass slits respectively. It was found that at 6 mm the mass slits began to cut into recoil transmission at the level of about 33%, close to what is predicted by simulation. The maximum recoil cone angle of the 58Ni(p,g)59Cu reaction is around 1.6 mrad, well below the 20 mrad acceptance of DRAGON.

Analysis of focal plane energy spectra showed qualitatively that indeed there were several different leaky beam peaks, expected to be the peaks from neighboring charge states to the selected one, and that these were dominating the rate.

Eventually, leaky beam rates were found to be on the order of 5 Hz. This means that if the beam intensity were scaled up by around 500 (as required for the proposal), the rates for this reaction would still be within the envelope that the DRAGON DAQ system can handle.

Detection of Recoils

Fig. 1 shows the separator time-of-flight between a detected particle at the focal plane in coincidence with a gamma ray at the target position, with no other cuts applied, for a selection of the 58Ni data. It clearly shows a sharp peak corresponding to 59Cu recoils (the smaller, broader peak is caused by events which are detected by the Ionization Chamber but not the MCP, thus having worse timing resolution and an offset, reflecting the efficiency of the MCP). The time-uncorrelated leaky beam events can be seen as a flat background, showing the true nature of the excellent beam suppression in coincidence mode of DRAGON.

Figure 1: time-of-flight between detected particle at DRAGON focal plane and coincident gamma ray at target position. X-axis range is 10 microseconds.

Rough numbers have been extracted for the overall beam suppression performance of DRAGON during this run. They are:

Average beam suppression in ‘singles’ mode ~ 1.2E8

Average beam suppression in ‘coincidence’ mode ~ 3.0E10.

The conclusions from this run are:

(a) the leaky beam rate per incident beam particle for this reaction, with similar Dp/p to the proposed 76Se reaction is sufficiently small that a scale up to the 76Se experiment at higher intensity is not deemed problematic;

(b) the coincidence mode beam suppression is extremely good, better than anticipated, at this high mass (it should be noted that at this point, 58Ni was the highest mass beam ever delivered to DRAGON).

84Kr charge-state distributions and suppression tests

In addition to the 58Ni beam, a 1.2 enA 84Kr 15+ (5E8 pps) beam of 1.25 MeV/u was delivered to DRAGON and sent through the gas target with the SiN booster foil inserted and helium gas used. Charge state fractions for q=23,24,25,26 were measured. The results are shown in table 1. The measurement was repeated with a beam energy of 1.46 MeV/u. The measured charge state fractions are lower than previously calculated. However, the conditioning tests indicate that higher than anticipated fields are feasible, thus allowing us to use a lower charge state than expected (23+ or 24+). As the charge state fraction is higher for lower charge states, there is an effective increase in yield.


Selected Charge state

Current on faraday Cup at energy-dispersed focus (enA)

Percentage of incident current (normalized to particle-nA)

Dipole Field (Gauss)




Above limit of NMR probe head













Table 1: Measured Kr charge state fractions at 1.25 MeV/u after SiN booster foil.

The Electric Dipole maximum voltage achieved during this run was 210 kV (although it had previously reached 220 kV in a conditioning test). This enabled us to then scale the separator for 88Sr recoils that would be present if we were trying to measure the 84Kr(a,g)88Sr reaction. We used a beam energy to 1.25 MeV/u for this test and could select q=25 in the separator settings. The purpose of this test was to determine what the beam suppression was at the settings for these even higher mass recoils. The result was a measured suppression of ~1.1E9 in singles mode, and 1.4E10 in coincidence mode. Again, these numbers show sufficient ‘raw’ (meaning no further cuts applied) suppression in order to perform a measurement of the 76Se reaction at similar energies. The leaky rate here was much lower than for the nickel runs (~0.5 Hz) and there was little indication of background from lower charge states (presumably due to these charge states being outside the envelope of DRAGON). These conditions are similar to those of the 84Kr(a,g)88Sr experiment (i.e. close to the limits of DRAGON) and the leaky rate here is thus more representative of expected running conditions. Scaling up the beam intensity to 2.5E11 pps would thus suggest a singles leaky rate of around 250 Hz, well within the DAQ capabilities.


Both tests demonstrate that the DRAGON separator has sufficient suppression to perform the 76Se(a,g)80Kr experiment as proposed and so stage 2 approval is requested at this time. Development of a 76Se beam of the necessary ( 2.5E11 pps) intensity is required and it is requested that resources are allocated to allow this be done during schedule 120. The first measurements of the 76Se(a,g)80Kr experiments could then be undertaken in the spring of 2012 before RIB is available.