DIY Time-Of-Flight Mass Spectrometer Ion source

[ChristofferB]

Oh boy this is one of the more involved projects.

DIY mass spectrometry is not massively popular. Aside from Ben Krasnow and two others I know of, there aren't a lot of people trying.

Ben and one other, the two most successful ones, are both magnetic sector spectrometers (1940's style).

I'd like to try a more modern technique, simply because it seems easier, and I had almost all the signal chain available in NIM modules (Nuclear Instrumentation)

The very short story: you make an ion and time it as it flies between two detectors; as its energy is well defined, its speed is proportional to its mass.


Image 1: Electron multiplier 1 generates timed electron pulses that ionizes molecules in the chamber. These are accelerated through a drift tube towards the detector, electron multiplier 2. The time differential is proportional to the ion mass m/z

Ok but creating ion bunches with start times is difficult. Many instruments use a high voltage pulse on a grid to let bunches through.


Ingeniously, Burle made a TOF MS in the 1960's that uses a channel electron multiplier (CEM) as ion source: the collector is removed so the electrons shoot out in bunches triggered by photoelectrons through a window. The CEM/"Channeltron" then also gives a start signal for the TOF measurements.
A channeltron /CEM is really just like the multiplier part of a photomultiplier, but made of a resistive material so it's a continuous voltage divider for itself.


Image 2: Ionization principle details

So in practice, I took a standard mass spectrometer detector CEM, removed the anode cap that usually collects the electrons, mounted it in a high voltage feedthrough flange. It has an integrated resistor to make sure the output is a bit above zero potential, to make the electrons attracted to the anode. I have currently shorted this to simplify this.




The feedthrough, one pole, connected to the cone of the CEM, is connected via an RC filter to a high voltage source.
The other terminated to ground via 50 ohm, and a signal coupled out to a preamp/scope etc.





I am not sure whether I should have a larger resistance from the CEM output to GND, to make the electrons coming out actually be attracted to something grounded, like the chamber wall opposite it. I am also considering using a higher voltage on it, so far 2800V is the highest I've tried, my new 7.5kV 1mA variable psu is in the mail.

The Ion source is housed in a KF40 cross with a quartz viewport opposite the channeltron feedthrough, and one port towards a turbomolecular pump.
Operating pressure around 10^-6 and 5*10^-7 Torr.
So far I haven't seen an output from this setup.

It is also good to note that this is not my idea; A few papers in Review of Scientific Instrumentation touched on the idea around 1969-72.

I believe Burle commecerialized the CEM and was just hunting for applications, but it seems a cool concept.

Hope you found this interesting! More to come as it happens!