# Half life determination

1. Jul 4, 2014

### sakkoyun

Hello,
I have a gamma spectrum of one element (X). There are several gamma peak in the spectrum. For example,

800 keV peak, 1000 keV peak and 1200 keV.

I have an information about the peak area changing with time. Namely, I know the peak area in different times. So I can calculate the half life of the element by using the area and time information.

But I have a problem.

I got 15 hours from 800 keV peak.
14 hours from 1000 keV.
16 hours from 1200 keV peak.

Which half-life is correct. What should I do? Which is the most reliable? Should I get average value?

Best wishes

2. Jul 4, 2014

### Meir Achuz

The widths of different peaks will be different.
Calculate the half-life, by measuring only the height of each peak.
These should all decrease at the same rate.

3. Jul 4, 2014

### Staff: Mentor

Sure, but the widths should not vary with time. And if they did, you could use the height of the peak. There is no absolute value mentioned here, all values are relative.

You should find some estimate how precise your values are. For example, is your first value "15 hours plus/minus 2 minutes" or "in the range of 13 to 17 hours"? If those three values are compatible within the uncertainties, calculate the average. If the uncertainties differ significantly, a weighted average is better.
If the values are incompatible, it would be interesting to know more about the experiment. Do some lines come from decay products? That can give an activity that deviates from an exponential decay.

4. Jul 6, 2014

### sakkoyun

Thank you very much.
With the errors, 15 +-0.3
14+-0.1
16+- 0.5

The 1200 keV peak is more intense. The others are small peaks. Will I get an average of all them? Namely, (15+14+16)/3=15. Or, will I get the half-life as 16 due to the fact that 1200 keV is more intense. Or, will I get 14 for half-life due to the fact that this value is smallest error?

Best wishes

5. Jul 6, 2014

Staff Emeritus
--->

6. Jul 8, 2014

### Staff: Mentor

14+-0.1 and 16+- 0.5 are a bit far apart. It would be interesting to see the decay scheme. If there is a physical reason for them to have the same half-life, something went wrong, you were really unlucky (>3 standard deviations) or the uncertainties are non-gaussian.

7. Jul 9, 2014

### sakkoyun

Zn68 was created after reaction. 3 gamma peak are clear in gamma spectra. 805, 1077 and 1260 keV. After analyzing each three peaks individually, I got three different half-life values. (Analyzing is performed by investigating the peak area with time. And obtaining decay (growth) curve). Which one is more reliable? Or ?

8. Jul 9, 2014

### Staff: Mentor

And what did you have before?

Did you identify the gamma energies in a decay scheme?

How exactly did you do this?

9. Jul 10, 2014

### sakkoyun

Ga69 was bombarded by 10 MeV photon. A neutron was emiited from Ga69. The unstable Ga68 decayed to Zn68.

10. Jul 10, 2014

Staff Emeritus
None of those are near the half life of Ga-68. Are you sure you are measuring what you think you are?

11. Jul 10, 2014

### sakkoyun

I was choosen the numbers randomly for an illustration. Actually the half lives that I measured are 65 min, 67 min and 66 min.

12. Jul 10, 2014

Staff Emeritus
I'm afraid by asking people to help you with made up, inconsistent numbers, you ended up wasting people's time. Perhaps it would be better to show us the real numbers.

13. Jul 10, 2014

### sakkoyun

My aim is to know the technique with a basic examples. It is not do an article all together. I have no realistic data. Again, I would like to know the technique. Because of someone, I made up an element, a realistic peaks and a realistic half life.

14. Jul 10, 2014

Staff Emeritus

15. Jul 10, 2014

### Staff: Mentor

The right analysis method depends on all those details you refuse to give.