Questions in positron-emitter yield fractions from RESNUCLEi

Scoring
1

Versions

Please provide the used software versions.

FLUKA:4-5.1
Flair:3.4-5

Description

Dear FLUKA experts,

I would like to ask for your help regarding the calculation of positron-emitting isotope yields using RESNUCLEi, and a discrepancy I observe when comparing my results with published literature obtained using GATE.

My goal is to obtain the relative yields of positron-emitting nuclei produced by a 12C ion beam irradiating a PMMA target.

Using RESNUCLEi scored in the entire PMMA target region, I extract the isotope yields from the Residual nuclei distribution (A\Z table) and normalize each positron emitter to the sum of all β⁺ emitters. My results are:

Nucl Yield Percent (β+)

11C 5.94e-02 80.05 %

15O 6.32e-03 8.52 %
10C 3.24e-03 4.37 %
8B 2.55e-03 3.44 %
13N 1.35e-03 1.82 %
12N 5.04e-04 0.68 %
9C 4.05e-04 0.55 %
17F 2.69e-04 0.36 %
14O 1.70e-04 0.23 %

However, published results obtained with GATE (see attached figure) show that the yield fraction of 8B is very small.
From a physical point of view, 8B is not expected to contribute such a large fraction, therefore I believe my result is likely incorrect, but I am not able to identify the source of the discrepancy.

I have attached my FLUKA input file for reference. I would greatly appreciate your help in understanding:

  • whether my procedure for extracting positron-emitter yields from RESNUCLEi is correct, and

  • what could cause the discrepancy of nuclide yield, such as 8B and 14O compared to GATE and literature results as followed:
Positron emitter Half-life Yield (%) Fraction of total yield (%)
11C 20 min 13.437 71.14
10C 19 s 1.681 8.9
15O 173.6 s 2.375 12.57
14O 91 s 0.359 1.9
13N 10 min 0.806 4.27
12N 11 ms 0.213 1.13
8B 770 ms 0.017 0.09
Thank you very much for your time and help.

Best regards,
Yang

A002.inp (4.3 KB)

Input files

3

Hi @yang_xy

Thank you for your question! Sorry for the late answer, i was on vacations.

I will have a look into your problem in the coming week.

Cheers,

Jerzy

4

Hi again,

Could you please point me to the published results that you are mentioning? Is there an article that I could look into? Just to make sure what was their approach.

Cheers,

Jerzy

5

Thanks for your reply, The results I used for comparison come from this article: In-Beam PET Imaging in Carbon Therapy for Dose Verification (IEEE Transactions on Radiation and Plasma Medical Sciences, 2018). Here is the source website: In-Beam PET Imaging in Carbon Therapy for Dose Verification | IEEE Journals & Magazine | IEEE Xplore . Thank you very much for your help.

6

Hi @yang_xy

After looking into your input file and the paper I noticed a few issues with both. In my view, generally the paper does not describe the results clearly enough, but I will focus just on the Table II as this is what you were trying to reproduce with FLUKA. The Table II title says that it shows

Positron-Emitting Nuclei produced by a pulse of 259.5 AMeV 12C ions in the PMMA phantom

and to me that implies we need to take the number of nuclei of each isotope that we find at the end of irradiation profile and normalize it to 100 * 1E8 ions/second (as the paper defines 1E8 a single pulse). However, the numbers in Table II obtained with the simulations seem to be the same as the ones compared to the measurements quoted in Table I and these are defined as

the positron activity yields (...) when carbon beams impacted the PMMA phantom

and here I must admit I get a bit confused as I don’t know what the authors mean by positron activity yields.

Now coming to your input file:

First of all, you define the beam energy as 230 AMeV which does not correspond to any energy quoted in the paper. This is however a minor issue I think.

What is more important is your RESNUCLEI card will give the production rate Pas you do not assign any DCYSCORE card to it. Therefore, it will give you the P that you can find in the paper in equation (1) and (2) and in the end I don’t think this is what can be found in Table II in the Yield column. Initially I thought it is the case, but again, the paper is not very clear about it.

Then you define the irradiation profile with the IRRPROFI card:

Screenshot From 2026-01-15 15-52-26

that corresponds to 5 seconds beam time with 1e6 ions per second. According to the paper, the beam profile that you need is 100 pulses of 1 seconds each 1e8 ions/sec with 2 seconds break between them. So the profile would have to be defined with blocks like

Screenshot From 2026-01-15 15-54-54
Screenshot From 2026-01-15 15-54-54489×116 4.79 KB

repeated over and over until you reach 100 pulses.

Then there is your DCYTIMES card where you define t1: 0 and t2: 600 and I am not sure why would you need the t2as this is basically the time of 10 mins after irradiation and it does not appear anywhere in the analysis of the paper. You don’t use the t2 as I can see in your file so it is not a problem, but I would like to make sure that you are aware what you are doing with this card.

Then you have a bunch of USRBINwhen you score a single isotope activity at the t1:0which means exactly after the irradiation profile that you defie with IRRPROFIand I think this is correct. If you want the integrated activities for all the isotopes produced within the target region, you can also use RESNUCLEI and assign a DCYSCORE to it with time 0.

I tried to look at the activity at the end of the aforementioned irradiation profile of 100 pulses separated by 2 s pauses and then extract the number of generated nuclides N with the equation

N = A/lambda

where A is the activity at the end of irradiation and lambda is the decay constant of each nuclide, but still I was not able to reproduce the numbers from Table II.

If you can please help me understand better what the numbers in Table II actually mean, we can try together to reproduce the values with FLUKA.

Best,

Jerzy

7

Hi Jerzy,

Thank you very much for your detailed reply — it is really helpful.

Based on my understanding, the simulated “Yield” values in Table II seem to be consistent with the simulated values reported in Table I, so I assume both tables refer to the same physical quantity (i.e. isotope yield from FLUKA simulations).
Therefore, I would like to clarify the definition of yield used in this paper.

(1) Is the “yield” in Table I and Table II simply the residual nuclei distribution, i.e. time-independent? More specifically, does the paper define “yield” as the produced residual nuclei inventory (nuclide distribution), rather than an activity-related quantity?

(2) About RESNUCLEI scoring without DCYSCORE

In my current FLUKA input, I use the RESNUCLEI card only (without DCYSCORE), and I extract the isotope information from the sum.lis file section:

**** Residual nuclei distribution **** (nuclei / cmc / pr) ****

where I identify the isotope yield and statistical error based on Z and A.

Could you please confirm whether the unit: (nuclei / cmc / pr) refers to nuclides / cm³ / primary?

(3) RESNUCLEI together with DCYSCORE: activity becomes time-dependent

When I assign DCYSCORE to RESNUCLEI, the same sum.lis section changes its unit to:

(Bq/cmc). This suggests that once decay scoring is activated, the output becomes an activity (time-dependent) quantity.

So my last question is: Is it correct to say that only activity is time-dependent, while the “yield” (as nuclide inventory / production) is time-independent?

At the moment, my personal interpretation is that the “yield” in Table I and Table II should be time-independent (not an activity), and therefore it should correspond to the residual nuclei distribution rather than Bq.

I would really appreciate your comments on this interpretation.

Best regards,
Yang

8

Hi @yang_xy

Below I am trying to answer your questions:

(1) I am still not sure what the yield means in both Tables, this is the problem. I think it has something to do with activity, because, as I mentioned above, the paper states that Table II shows

the positron activity yields (...) when carbon beams impacted the PMMA phantom

and the other clue is that the production rates extracted from FLUKA do not really agree with the values in the Yield column (you tested it yourself).

(2) Yes, RESNUCLEwithout an assigned DCYSCORE will give you the output as nuclides / cm³ / primary and this is time independent. You have to be careful though, because the volume of your region has to be given by the user, the default is 1 and in this case the output will give you the integrated number of nuclide / primary in a given region (not relevant for ratios).

(3) After you assign a DCYSCOREto RESNUCLE, your output will be given in Bq / cmc (here again if you leave the default volume 1 you will get the activity in the whole region) and it will give you a snapshot of your activity at a given cooling time AFTER your irradiation scheme defined by IRRPROFI cards. Look into note 7 in the manual of the RESNUCLE card:

Once again, about the interpretation of the yield column in Table II:

I was reading a bit through the thesis that the authors are referring to when quoting the measured values in Table I and I am not able to reproduce the numbers that appear in the paper. The thesis is available here:

Look into page 73 Table 5-5. If you manage to reproduce the numbers that the paper authors calculated from the results available in this thesis, let me know. After we understand what we are supposed to calculate with FLUKA, we can try to reproduce the values in Table II.

Cheers,

Jerzy

9

Thank you very much for your detailed explanation and for sharing the thesis link — it is really helpful. I will carefully read the thesis and try to understand how the authors defined and derived the “yield” values reported in Table I/II.

Before that, I would like to confirm one additional point regarding the RESNUCLEI output.
When I use RESNUCLEI (without DCYSCORE) and extract the isotope yields from the sum.lis section:

**** Residual nuclei distribution **** (nuclides / cmc / pr)

I observe that the yield of B-8 is relatively high (second highest after C-10 in my case).
Could you please confirm if such a high B-8 yield is physically reasonable in this type of 12C + PMMA irradiation scenario, or could it be related to the way FLUKA reports very short-lived/unstable residual nuclei (e.g. prompt breakup products)?

Thank you again for your help.

Best regards,

Yang

10

In addition, I would like to clarify one point about the wording in the paper.
The authors also wrote a Chinese Master’s thesis covering the same study (unfortunately this Master’s thesis was wrote in Chinese). In that thesis, the descriptions of both Table I and Table II are simply referred to as “yield” (in Chinese), and the wording “positron activity yields” does not appear.
Therefore, I am wondering whether the term “positron activity yields” in the English paper might be a wording ambiguity, rather than indicating that the yield is strictly an activity-related quantity.

I hope this clarification is helpful for interpreting my question.

11

Hi @yang_xy

I did not forget about your question, I am trying to validate 8B production in FLUKA with available experimental data (which turns out to be rather sparse in this topic) to resolve your doubt about the 8B yield. I will post a conclusion soon.

However, still I would like to understand how we can arrive to the experimental values presented in the paper Table II from the numbers available in the aforementioned PhD thesis.

Cheers,

Jerzy

12

Hi Jerzy,

Thank you very much for the update — I really appreciate that you are taking the time to validate the ⁸B production against experimental data. That is extremely helpful.

Regarding Table II, I fully agree with you: the key point now is to understand how the authors derived the “experimental” values reported there from the results available in the PhD thesis. I am currently going through the relevant sections around Table 5-5 (page 73) .

Once I find any explicit definition or intermediate calculation steps in the thesis, I will summarize them and share the details here.

Thanks again for your support, and I’m looking forward to your conclusion on the ⁸B yield.

Best regards,

Yang

13

Hi again @yang_xy

I performed a cross section calculation with the interaction model of FLUKA with an intention to reproduce the results presented in this article:

It uses lower energy of the 12C ion beam than what we were discussing before, but has the advantange of describing the measurements and the setup very clearly.

From what I found, trying to reproduce the 8B production differential cross sections of 12C ions on carbon target and on oxygen target (Tables VII and VIII in the paper) and also comparing the integrated cross sections (Table V), I conclude that it is very unlikely that FLUKA overestimates the 8B yield in the setup of your input file (12C on PMMA target).

In the article, the data points for oxygen target have very high uncertainty, but my FLUKA results are consistently below these measured values. The integrated cross section obtained with FLUKA, 3.32 mbarn is about 2 times lower than the one presented in the Table V, 6.9 +/- 2.7 mbarn, but the quoted uncertainty is large.

In the case of the carbon taget (where the uncertainties are generally more under control), FLUKA underestimates the forward direction (first 2 bins in the article) with respect to the measured values and then for the remaining angles produces slightly higher values. The integrated cross section from FLUKA, 4.51 mbarn, is in agreement with the value quoted in the article 6.1 +/- 1.8 mbarn within the uncertainty.

Now to be precise, the article and my simulations to reproduce its findings use thin targets so the cross sections are proportional to the yields. In the case of your simulation, we are dealing with a thick target and the relation between the yield and the cross section is a bit more complex. However, I don’t think we have a reason to believe that these additional effects could lead to a factor of 10 yield overestimation. Once again, we must understand the values in the original article that you were trying to reproduce and then we can proceed with the FLUKA cross checks.

Best,

Jerzy