by LorenzoA (September 2024)

In this page I report the alpha analysis regarding the background run of LaBr3 from 26-08-2024.

The total measurement time is 30 minutes.

The total events are 207453 (115.25 Hz).

The total gamma events are 167404 (93.01 Hz).

The total alpha events are 40031 (22.24 Hz), corresponding to 1 event every 0.04496 seconds.

1) Pulse Shape Discrimination between γ/α

Here (left) I plot the PSD parametere (chi2amp/integral) in function of the integral. We can see the distinction of the gamma/beta band and the alpha band. To select the alphas I set the discrimination parameter PSD=0.052 and take the events that have integral>42000, in this way I am able to plot the alpha spectrum (right).

Figure 1
Pulse Shape Discrimination γ/α	Alpha spectrum

Looking at the alpha spectrum we can see that it is composed of a sort of 5 peaks at around 52000, 54000, 59000, 61000 and 67000 (integral value). We know that all the lanthanide based scintillators are subject to Ac contamination, in particular Ac227. So the responsible of these peaks have to be searched in the decay chain of Ac227:

Ac227 𝛽 (21.8y)⟶ Th227 𝛼 (18.7d)⟶ Ra223 𝛼 (11.4d)⟶ Rn219 𝛼 (3.96s)⟶ Po215 𝛼 (1.78ms)⟶Pb211 𝛽 (36.1m)⟶ Bi211 𝛼 (2.14m)⟶ Tl207 𝛽 (4.77m)⟶ Pb207

The alpha emitted from Po215 has the highest energy, and so it can be associated with the last peak visible in the spectrum. Starting from this I use delayed coincides to reconstruct the alpha triplet Ra223-Rn219-Po215-Pb211.

2)Search for the Ra223 → Rn219 → Po215 → Pb211 alpha triplet

The energies and half-life times of the 3 alpha decays are the following:

The first thing I do is select the alpha particle emitted from Po215. This is done by simply taking the events of the alpha spectrum that fall in the range of [65000-69000]. Having selected this events I search for delayed coincidences: for each event I look for the father in a time window of dt<10*t_1/2, and in the integal range of [54000, 65000]. In the figure below (left) I plot the time difference dT with the previous alpha event, versus the integral value of the previous alpha event. On the right instead I plot the histogram of the time difference distribution, expecting an exponential decay with t_1/2=1.78 ms and a constant background.

Figure 2
dT distribution	histogram t1/2=1.78ms

Instead what I found is an half life of 5.07±0.15 ms, which is not compatible with the correct value. This is the same problem that arises on the GAGG analysis (for reference and possible explanation see Silvia notes on Ra223 → Rn219 → Po215 → Pb211 alpha triplet for the GAGG).

To be sure that the signal is really due to the 215Po → 211Pb decay I search further back looking for the Rn219 decay, which has an half life of 3.96s. The best way to do this would be to select the events for which I already found the first coincidences, and then for each of these I would search the father in a time window of 10*t_1/2 and in the integral range [44000, 54000], where this time t_1/2=3.96 s. This approach is not feasible for the LaBr3 crystal because for each event I would have approximately 200 coincidences and it's impossible to know which one is really the father, and this happens because the crystal has such an high counting rate for alphas. Instead I selected each event for which I already found the coincidence 215Po → 211Pb, and for everyone of this events I computed the time difference with every event in the range integral [44000, 54000] up to dT=6*t_1/2. Since in the range [44000, 54000] there are approximately 15000 events (8.4 Hz) we have 200 possible fathers, and so we expect a total number of event of 200*number of coindicences 215Po=1023800, and in fact I found 1042420. In the figure below I plot the histogram of the time distribution of such events.

Figure 3

The red horizontal line represent the estimated background, with the dashed lines representing the error. I found an half life of 4.83±2.08 s, which is compatible with the expected value. As a cross check that everything is right we can see if by subtracting the estimated background from every bin we get the expected number of events, which is around 5100. By doing this, considering the error on the background, I find a range of possible events [4800,11000] consistent with the expected value.

3)Alpha calibration for LaBr3

From the delayed coincidence analysis we have identified 3 peaks of the alpha spectrum that we can use to calibrate in energy. Here I plot the distribution of the three consecutive decay signals: ²¹⁵Po (left), ²¹⁹Rn (center) and ²²³Ra (right).

Figure 4

The three histograms are fitted using a gaussian function plus a constant background and an error function, the results are shown in the statistic box. We can see that the last histogram shows a poor goodness of fit, this is kind of expected given the large amount of background that we have in the case of ²²³Ra. Using the mean values found, I now have 3 points to calibrate the ADC for alpha particles:

²¹⁵Po Q=7526 → ADC=67790

²¹⁹Rn Q=6946 → ADC=60790

²²³Ra Q=5979 → ADC=51770

There are now only two peaks left in the spectrum which were not idenfied. As I said before we know from the literature that for the major part the contamination in LaBr3 crystal is given by ²²⁷Ac decay chain. In this decay chain there are only 5 alpha decay, of which 3 have already been identified, so it is easy to associate the other two. ²²⁷Th decays alpha with a Q=6146 keV, while ²¹¹Bi alpha decay has a Q=6750 keV.

So the peak at around 54000 ADC (figure 1 right) is associated with the alpha from ²²⁷Th, while the peak at 59000 is from ²¹¹Bi. Indeed we can see that from the histograms that we get from the delayed coincidence analysis these two peak are not present because they are not correlated with the alpha triplet that I was studying. The decay scheme of ²²⁷Th, ²¹¹Bi permit only this empirical association.

To fit the peak from ²²⁷Th I select the alpha spectrum in the range ADC [50000,56000] and perform a double fit including the alpha from ²²³Ra (left figure). For ²¹¹Bi I do the same in the range ADC [56000, 64000] (figure on the right), where the second peak is given by ²¹⁹Rn.

Figure 5

Here for the fit I used a function of the type: write the function.., the parameters are reported for the statistic box. For ²¹⁹Rn, ²²³Ra I obtain the mean value compatible with the ones that I got with the delayed coincidence analysis. For ²²⁷Th and ²¹¹Bi I get:

²¹¹Bi Q=6750 keV → ADC=59490

²²⁷Th Q=6146 keV → ADC=53940

Finally I obtain a preliminary alpha calibration curve for the LaBr3 crystal:

Figure 6

Clearly the calibration curve is pretty bad, but we have to keep in mind that the alpha from ²²³Ra is off the line also for the GAGG, this because in the decay there is a high probability that also a gamma is emitted. The same thing happens also to ²²⁷Th. Also the statistics is very low because this analysis was done using a background run of only 30 minutes. All these reasons lead to the poor goodness of fit shown in the figure.

Clearly the objective of this note was to only recognize the peaks in the spectrum, not to do a complete alpha calibration.