Answer to Question #13832 Submitted to "Ask the Experts"

We are currently working on a metal-oxide semiconductor-based x-ray sensor. We placed it in a closed circuit and supplied it with certain voltage and current, applied different x-ray doses on the circuit, and observed a change in the current with each dose. The question is how can we map the change in current with the dose? Is there any rule that relates the change in current with the applied dose?

Unfortunately, while a linear response between sensor output and dose rate is ideal, it frequently is not realized. The x-ray dose (rate) at the sensor location depends on the x-ray energy distribution and the intensity, the energy fluence rate, e.g., MeV cm^-2 s^-1 being the particular quantity that is proportional to dose. Exposure time is important if you are concerned with integral dose.

For a fixed operating voltage, the expected energy intensity and dose rate are directly proportional to the x-ray tube current; for example, if the tube current increases from 5 mA to 10 mA, and all other parameters remain the same, the dose rate at the dose point will expectedly increase by a factor of two. Therefore, you can make assessments of the relative response of your detector by maintaining detector position and high voltage constant and varying the current and recording your detector output. You may not have absolute dose (rate) values, (unless you have another means of measuring/obtaining them), but you will be able to see if the detector is providing a linear response or if it might be suffering losses that might be associated with ion recombination, saturation effects, or other possible factors. If such losses are occurring, the response curve will show decreasing slope with increasing current. You state that you applied different x-ray doses (I assume you mean dose rates if you are measuring current) and observed currents produced in your detector/sensor. If you have maintained the same machine high voltage throughout your measurements, changing only the beam current of the machine (or possibly the distance from focal spot to dose point) to change the dose rates, then we would expect your measured currents from your device to be linear when plotted against dose rate if no charge was being lost.

If the high voltage changes, the effect on intensity is more dramatic than what is observed for a beam current change. Assuming a constant beam current, for an unfiltered x-ray output the energy intensity and associated dose rate at a given dose point will expectedly increase as about the square of the voltage. Because of internal, and possibly added, filtration in most machines, low energy photons get reduced, and the intensity dependence may approach a voltage cubed relationship. Thus, if you used changes in voltage to produce different dose rates it will be very difficult to make meaningful estimations of how the dose rates are related. If you know the delivered dose rates from other independent measurements or other sources then, for a sensor that was not losing charge, we would expect the measured current vs. dose rate to be linear if the device exhibited no significant energy-dependent response. Unfortunately, many semiconductor devices do exhibit energy dependent dose responses, often especially notable at lower energies. This can complicate interpretation of the sensor response. One way to get around this, if you want to evaluate how the sensor response is changing with energy, is to select a sequence of high voltages and at each voltage generate a response curve by varying the tube current and making a measurement at each current. Plotting the sensor current vs. dose rate (or tube current) and comparing the curves at different voltages will give you some insight into how responses are varying with energy, at least in the energy available via the x-ray machine.

I hope this is helpful to you.

George Chabot, PhD, CHP