How to simulate ion signal induced charge for SeF6

Garfield++
1

Dear all,
I want to simulate induced signal for SeF5- fly in SeF6. Following the suggestion, I use SF6 instead. My first question is that is it necessary to build gas file using Magboltz if the gas had been defined? I tried to generate gas file about SF6, but it mentioned me “PROGRAM STOPPED TOO MANY PRIMARIES IPRIM”. Although I enhanced the range of electric field.
And the model was built by COMSOL, with also the weight potential.

        ComponentComsol fm;
        fm.Initialise("mesh_induced.mphtxt", "dielectric.dat", "Potential.txt", "mm");
        fm.PrintRange();

It was successful to load it. Next I want to load the mobility for SeF5- in SeF6 which was calculated by myself and load SF6 gas

        MediumMagboltz gas("sf6");
        gas.SetTemperature(293.15);
        gas.SetPressure(760.);
        gas.Initialise(true);

        gas.LoadIonMobility("/Users/tianyu/InducedCharge/SeF5-_SeF6.txt");
        fm.SetGas(&gas);
        fm.PrintMaterials();

And calculate the signal

        fm.SetWeightingField("WeightPotential.txt","anode");
        Sensor sensor(&fm);
        sensor.AddElectrode(&fm,"anode");
        const unsigned int nBins = 10000;
        sensor.SetTimeWindow(0, 2000, nBins);


        AvalancheMC drift;
        drift.SetSensor(&sensor);
        drift.UseWeightingPotential(true);
        drift.SetDistanceSteps(1e-4);
        drift.EnableSignalCalculation();


        drift.DriftIon(0,0,9.5,0);

My result is as shown.

Screenshot 2025-03-19 at 14.34.36
Screenshot 2025-03-19 at 14.34.36581×568 17.1 KB

Of course it seems abnormal. So my second question is what’s wrong to my code?
Another question is that the ion is negative ion, so when using drift.DriftNegativeIon(0,0,9.5,0), the induced charge is 0. Why such thing happens?
Thanks
Induced.C (3.1 KB)
Potential.txt (291.0 KB)
SeF5-_SeF6.txt (309 Bytes)
WeightPotential.txt (293.1 KB)

2

I’m sure @hschindl can help you

3

BTW, I calculated the induced charge using COMSOL and Matlab, I guess that should be the expected result.

d88d9ee2e91ad385d9f63018a17decfd
d88d9ee2e91ad385d9f63018a17decfd1002×487 25.3 KB

4

Hi,
at what electric field did you get the “PROGRAM STOPPED…” error message from Magboltz? In pure SF6, electron attachment happens very quickly and you might be running the simulation at conditions where you don’t get meaningful macroscopic transport parameters (drift velocity etc.).

The plot you attached seems to be the induced charge as function of time (as opposed to the induced current). Can you explain in more detail why it looks abnormal?

Regarding the negative ion drift: where is the starting point (0, 0, 9.5) with respect to the readout electrode? Can you make a plot of the drift line?

5

Hello,
Here is the model and electric field generated by COMSOL.

bf9396be153b36d91d1e97b1a76d32c0
bf9396be153b36d91d1e97b1a76d32c0751×635 19.9 KB

It is consisted by cathode (top), focus plane (a hole in the middle of plane) and anode (hexagon area). The voltage of cathode is -2000 V and the voltage of focusing plane is -1500 V.

Starting point (0, 0, 9.5) means the ion should be at the center and z = 95 mm.
What I expected is that the induced charge should back to 0 as the increasing drift time. Besides, the drift velocity of ions is too slow, the time should be longer.
Here is the result of induced current vs time. Looks like white noise.

Screenshot 2025-03-19 at 17.10.48
Screenshot 2025-03-19 at 17.10.48581×568 15.8 KB

6

Dear @dm-leo can you provide also your field files (mesh_induced.mphtxt, dielectric.dat, Potential.txt) such that we can run your example code?
Thanks
Piet

7

Can you do some sanity checks?

  • What is the electric field at the starting point of your ion? Assuming the region between the cathode and the focusing plane can be approximate by a parallel plate chamber, I guess the field will be around 500 V / 10 cm = 50 V / cm, which is quite low.
  • What is the drift velocity that you expect at that field given the mobility value you are using? And what would be the drift time of the ion?
8

InducedCharge.zip (1.2 MB)
All the file was packaged in attaching file.
Thanks

9

Hi, of course the electric field is quite low. The reduced mobility I calculated is about 0.444, so the velocity is about 2 cm/s. The time should be larger than 5 s. So at least us is not suitable.

10

Hi,
on the last plot you’ve posted (the one showing the induced current as function of time), the duration of the signal is ~9 ms, and it’s for the positive ion (so over a distance of 5 mm), so the order of magnitude doesn’t seem completely wrong. But I agree that the current looks a bit small. Can you try to switch off diffusion (function DisableDiffusion in AvalancheMC)? You might also want to increase the step size (1 μm seems a bit small for the scale of the device you are simulating).

11

Hi, @hschindl
After switching off the diffusion and change the step size into 0.01 μm, I could get both the signal of negative and positive signal.

Screenshot 2025-03-20 at 14.50.42
Screenshot 2025-03-20 at 14.50.42646×568 20 KB

Screenshot 2025-03-20 at 14.51.46
Screenshot 2025-03-20 at 14.51.46567×570 15.9 KB

But the induced charge is still 0 when using

sensor.IntegrateSignals();
sensor.GetIonSignal("anode", nBins - 1) / ElementaryCharge;

for negative ions. Besides, the time seems not correct if the electric field is so small. Do you have any suggestions?

12

Ah, I see what you mean. GetIonSignal returns the signal induced by positive charge carriers, GetElectronSignal returns the signal induced by negative charge carriers, i. e. electrons and negative ions.

13

Thanks :grinning:I think the problem was solved