Question about proportional counters for gamma rays

[neanderthalphysics]

TL;DR

We work back to estimate the energy of one incoming photon from the degree of ionization it causes.
What if there is more than one? Do the signals overlap?

I understand the concept that a more energetic photon would cause more ionization events and therefore we can estimate the original energy of the incident photon.
But what I don't get is that in practice, you don't get one photon entering the detector at a time, you get a whole bunch of them, of perhaps unknown intensity and energy spectra.
How do we know that multiple weak ionisation events from X-rays are not caused by one high energy gamma ray?

 

[DEvens]

It depends on the specific situation and what you are trying to detect. And the nature of the detector.

In some situations there can be overlap of multiple photons. One example is, some kinds of detectors have a brief "dead time" after they detect a photon. Ionization in the detector has to relax and give up its energy, and the detector has to "reset" so to speak. This means that, if you have a monochromatic photon source of variable intensity, the signal from the detector will saturate as you increase the intensity. Eventually, extra photons coming in will only see the dead time for the detector. This was a keen lab in 4th year undergrad. There was some lovely math to calculate the dead time based on the relationship between distance to the source and detector signal.

For other kinds of detectors, the signal is proportional to some such thing as total energy deposited, and this relationship holds over a very wide range of signal strengths. But it's only weakly dependent on the energy of the photon producing the signal. For such detectors you might not be able to distinguish between a photon of 5MeV and two photons of 2.5MeV each. Actually, it's a bit more complicated than that. The 5MeV photon might penetrate much deeper, possibly a larger fraction of its energy going right through the detector. Where the lower energy photon might have a larger cross section to get absorbed outright. Or to scatter multiple times instead of one time only. Or, the high energy photon might scatter and produce multiple lower energy photons that then deposit a lot of energy, so making the higher energy photon look very "bright." So, in the case of such a detector, it is very difficult to be confident exactly what produced the signal.

Some detectors operate more along the lines of a photographic plate. They can detect where a photon impacted, and give some energy information about the impact. So they can tell red from green, for example. Or 1MeV from 5MeV. Sometimes such detectors are based on something similar to a photographic emulsion, so can detect only one photon per pixel, then need a hard reset. Sometimes they have an array of scintillation crystals, and can reset. Then you get back into dead times per pixel, which could be a function of the energy of the crystal.

https://en.wikipedia.org/wiki/Wire_chamber
There is a lovely thing called a drift chamber that has many fine wires. These will detect the particle moving by, and give a variety of information about energy and direction. Analyzing the results of such things is quite the complicated task.

One method of understanding the signal from a given detector is the Monte Carlo method. You model the behavior of the detector. Then you randomly generate particles to match the source you expect to put the detector next to. The you predict the kind of signal you will see. If you get something similar to real measurements, then you have some confidence you have a handle on the specific nature of the physics.

 

[neanderthalphysics]

Thank you for your informative reply.

Given the limitations of detector dead time and problems differentiating one high energy photon vs more lower energy ones, I'm curious about how people get nice spectral plots of "photon count vs photon energy" for photons in the X-Ray and Gamma Ray regime, from non-monochromatic sources such as X-Ray tubes, particle accelerators, free electron lasers, etc. Especially if the sources are of high brilliance and short duration, like femtosecond lasers.

For high brilliance/short duration X-Ray pulses, are we talking about many repeat experiments with the detector very far away from the source (1/R2 distance law, so that you don't get overlapping ionization events), or trying to match Monte Carlo simulations to experimental observations?

 

[gleem]

One could use a curved crystal spectrometer. X-rays hitting a crystal such as quartz are reflected at an angle that depends on the energy similar to an optical grating spectrometer. Placing some sort of detector that responds in a known way to the intensity of the x-ray at a particular angle can produce an energy spectrum. Such a device can analyze x-ray pulses with any pulse width or intensity (maybe with some limitations) since there is no energy signal to process .

 

[DEvens]

Mmmmmm! Femto second pulses!

But seriously. With a pulse that short, light moves about 0.3 microns during the pulse. So the entire pulse is not just inside the detector at the same time. It's within a very small part of the detector. Like, within the thickness of the paint on the cabinet that holds the electronics. So you would absolutely need some such scheme as suggested by gleem. You could never resolve one photon from the next otherwise.

It's still a very interesting problem. You would need to worry about lots of things like diffraction and reflection efficiency at different energy. And lots of other things. Probably you could calibrate it with monochromatic sources.