Context: Monte Carlo simulation of a linear accelerator photon beam.

The energy spectrum for photons as calculated from the phase space files found in here has a peak somewhere near $500\,\text{keV}$. The characteristic x-rays on the other hand for tungsten fall roughly in the range $60-70\,\text{keV}$. These photons represent a beam just after the flattening filter and just before the secondary collimators.

Question 1: Does anyone have any idea why the peak is so off? Or perhaps it represents something entirely different?

Question 2: Since we are here, one more clarification: The term "characteristic x-rays" is used in electrons energy spectrum, right ? Isn't it wrong to use it the way I do here ?

enter image description here


1 Answer 1


That could be a positron annihilation peak.

If the halo of your beam is scraping somewhere there will be a variety of interactions between the particles and matter. If it is an electron beam then many of the interactions will be electromagnetic showers which will produce (energy allowing) some $e^+$--$e^-$ pairs. When the positrons annihilate they will mostly produce pairs of 511 keV photons.

  • $\begingroup$ Nice idea; let me measure precisely the value of the peak! $\endgroup$
    – stathisk
    Mar 12, 2014 at 17:37
  • $\begingroup$ Since this is MC, the other way to check is to look at the spectrum of particles produced. Positrons are diagnostic. $\endgroup$ Mar 12, 2014 at 17:40
  • $\begingroup$ Yep, the peak is at $E=0.511\text{MeV}$. The explanation seems plausible. Thanks @dmckee! $\endgroup$
    – stathisk
    Mar 12, 2014 at 17:47
  • $\begingroup$ Damn, your ESP is working well today ;-) $\endgroup$ Mar 12, 2014 at 17:49
  • 1
    $\begingroup$ @JohnRennie More experience than superpowers. The smooth background rules out a lot of complicated scenarios and I have done enough medium and low energy work to have seen that peak a lot: it's sharpness and location are very (heh!) characteristic. $\endgroup$ Mar 12, 2014 at 18:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.