Has anybody built their own mass spectrometer before?

[ikrase]

I'm working towards building a mass spectrometer or RGA. I already have high vacuum and thermionic electron emission (for electron impact ionization) working. Once I get my  hands on an electrometer and some more DIN rail power supplies, I will try building my own ionization gauge.

I'm curious if anybody here has ever done such a thing. The only one I know of is this guy: http://www.rapp-instruments.de/Radioaktivitaet/div/quadropol/quadro.htm

(Also check out his pipe tee mass spec)


The biggest issue I foresee is building the RF generator (if taking the quadrupole route). I know a bit about RF, but not very much. And this is an odd application, being among other things voltage based rather than power based. At least it's low frequency.


Another big question is whether DIY electron multipliers are viable. I have not had much luck finding out what the coating on the dynodes is made of.

[JonM]

I built two mass spectrometers back in the 1980s. The first was primarily a photoelectron spectrometer but was also used as a decent linear time of flight (ToF) mass spectrometer. It was built on a fairly low budget, but it was in a university with access to a lot of surplus equipment. The ionization source was an expensive UV laser but ToF MS can also be done with pulsed acceleration voltages on grids for electron impact (EI) ionization. The detector was a stack of two microchannel plates (rejects from Varian that failed image tests, we just needed a time resolved total signal). Data acquisition was with a "waveform digitizer", essentially an early headless digital oscilloscope.

The second mass spectrometer I built was in industry with a 1985 US$700,000 budget just for parts and components which included a 150mm bore 7 Tesla superconducting magnet and an ultra-high vacuum system with three deferentially pumped chambers. Obviously not a garage project, however the electronics and computing could be done at a much, much, lower cost today. In fact I have much of what is needed sitting behind me right now.

The "front-end" was a RF Quadrupole and ion source that were purchased as a module from Extrel (which is still in the business).

I was going to respond to ikrase's electrometer thread since I initially used a Keithley 614 to measure the ion current that reached the detection cell inside the magnet. No electron multiplication was needed, a beam stopping plate was just connected to the electrometer. Of course sensitivity would have been limited and we usually ran the RFQ front end as a high pass filter (not passing some set threshold of low mass ions, but everything above that mass) so the current could have been from ions of a wide mass range.

Later I switched to a Keithley 617 electrometer to use the GPIB interface as part of an ion optics auto-tune system (there were multiple ion lenses in the injection path).

If I were to build a mass spectrometer in my garage I would probably look for a surplus RF quadrupole system rather than build that part from scratch, but maybe it's not that hard. The quadrupole elements can probably be made at any machine shop that does precision center-less grinding. A pulsed ToF MS would also be considered.

I'll also note that ion trap mass spectrometers, like the Orbitrap, are hot these days. Could be a DIY possibility, likely requiring some precision machining.

ikrase: What mass range are you interested in? Will you need high sensitivity? Will there be a lot of masses in a single spectrum?

Here are a couple of resources that might be of interest (inspiration more than as a supplier due to cost). I used both of them and once lived a few miles from one (but we both moved).

 http://www.kimballphysics.com
 http://www.sisweb.com

[LaserSteve]

What kind of multiple are you interested in making, a PMT style dynode stack, or a channeltron?

Its not much of a problem for me to look that up. As I have before..

Steve

[chickenHeadKnob]

Utilizing this, I can build a parallel tunnel, apply a certain waveform with a DC bias, so only particles with a narrow range of mass to friction ration can pass since they deflect equally to both sides under that particular waveform.
By scanning DC bias voltage, I can sort particles, then an electron meter can read it out.

Looking up FAIMS now. Damn that's clever and elegant.  Making lemonaid from lemons, thanks for pointing to it.

[T3sl4co1l]

I don't have anything to add at this time, but shall watch with great interest.  I shall leave the proverbial:

.

Tim

[ikrase]

I am glad to hear that with a good electrometer I likely don't need multipliers so much. Was planning, like Thomas Rapp, to mostly rely on faraday cups.

This was indeed to be a bit of a "toy" project, though maybe also useful as an RGA. I had envisioned this to broadly target gases and vapors and ionize with EI, and imagined that I might look for it to top out at about 150 AMU perhaps? That covers some easily-vaporized hydrocarbons like hexane and also sulfur hexafluoride.

I don't really have requirements as much as dreams to strive towards. But not being able to distinguish the different components of air would be embarrassing. It would be cool if I could, say, pick out the components of gasoline -- or maybe detect impurities in water. (this sounds hard given that so many of them are solids).

I had thought about the possibility of ToF, but mostly rejected it due to the apparent complexity of reflectors and the need for a high-speed electrometer. Should probably reconsider. I do at least have an oscilloscope -- an ancient 15mhz Tek one.


I know very little about ion traps.


Do quadrupoles actually need to be ground? I figured I would make mine out of inexpensive copper tubes or maybe drill rod.



I would be very interested to hear your critique of the Rapp Instruments quadrupole and pipe tee designs. (Linked in the OP).

I did get my hands on a very burned out Varian leak detector filament / ion source. ended up taking it apart for the octal feedthrough.

[JonM]

I am glad to hear that with a good electrometer I likely don't need multipliers so much. Was planning, like Thomas Rapp, to mostly rely on faraday cups.

This was indeed to be a bit of a "toy" project, though maybe also useful as an RGA. I had envisioned this to broadly target gases and vapors and ionize with EI, and imagined that I might look for it to top out at about 150 AMU perhaps? That covers some easily-vaporized hydrocarbons like hexane and also sulfur hexafluoride.

I don't really have requirements as much as dreams to strive towards. But not being able to distinguish the different components of air would be embarrassing. It would be cool if I could, say, pick out the components of gasoline -- or maybe detect impurities in water. (this sounds hard given that so many of them are solids).

I had thought about the possibility of ToF, but mostly rejected it due to the apparent complexity of reflectors and the need for a high-speed electrometer. Should probably reconsider. I do at least have an oscilloscope -- an ancient 15mhz Tek one.

Even a "reflectron" ToF MS seems easier to construct to me than a RFQ. For ToF the detector might be the harder part, but HV pulses, and HV across grids, are also required. You would certainly need a digital oscilloscope, or some other sort of digitizer, though. It has been many years since I looked anything related to ToF and so may be missing important details.

Do quadrupoles actually need to be ground? I figured I would make mine out of inexpensive copper tubes or maybe drill rod.

Thomas Rapp does not seem to mention anything about the rod finish or precision, but his instrument looks like it is very well built and required a lot of machine shop work.

I would be very interested to hear your critique of the Rapp Instruments quadrupole and pipe tee designs. (Linked in the OP).

I did get my hands on a very burned out Varian leak detector filament / ion source. ended up taking it apart for the octal feedthrough.

Depending on what you goal (or maybe I should say "the path you want to take") is, it may make a lot of sense to buy a surplus RGA or analytical RFQ system. I see a Leybold Inficon RGA head (apparently self contained, "just add computer") for $269.99  These were probably used as "smart ion gauges", basically pressure measurement that also indicates what gasses are left behind as a diagnostic.

That reminds me of the vacuum leak detectors popular in the 1970's (from Veeco, I think). They were the size of a desk and would be wheeled up to a vacuum system, connected, pumped down, and then you would squirt Helium out of a wand. When you squirted near the leak a tone indicating the detection of He would increase in frequency. I don't know what the mass spec technology inside that box was but they were probably replaced by the RGA which provide a lot more information in a small package.

Yes, it was Veeco, I see a MS-9 model on eBay for $309.92 + freight. Other models are there as well.

If you want the joy of building the RFQ MS from scratch I'm sure that you would learn a lot but it could be expensive and time consuming.

What type of vacuum system do you have? How large is the chamber?

[ikrase]

required a lot of machine shop work.

Ah. I should have mentioned that I'm a mechanical engineer first and an electrical person second, and the Rapp quadrupole looked simple to me compared to the typical commercial mass spec designs.

For ToF the detector might be the harder part

I'd be interested in more information about this -- are electron multipliers more favorable? Or do you need to make some kind of specialized high-speed electrometer? Most electrometers dont' seem to go above audio frequency.

I don't know what the mass spec technology inside that box was but they were probably replaced by the RGA which provide a lot more information in a small package.

I know that many Varian leak detectors used a magnetic sector.

What type of vacuum system do you have? How large is the chamber?

I don't really have any *large* chambers, would acquire them as needed, esp. as many mass spec designs would go in a long KF40 or KF50 nipple. Or sector spectrometer in a pipe tee My system reliably gets down to 5E-6 torr, probably lower if I cleaned it for once. Its mostly held together with KF fittings, and has a needle valve gas admittance passage.

[CopperCone]

what is the 3db point on the electrometers required?

[Kleinstein]

The TOF solution needs a fast ion detector and this kind of implies using an electron multiplier.

The quadrupole route can work quite well with just a faraday cup and sensitive electrometer circuit, which is not that hard anymore. Here the difficulty lies in the mechanics and the power RF part. That RAP plan does not look that much different from older commercial instruments.

I am not sure one would need to go so far to higher AMUs. I used a RGA some time ago and AFAIR I hardly looked at above 50 AMU. An upper limit of 32 is a little low, as it would be nice to still see argon.

As a complete system consists of computer and detector head, the chances are not that bad to find a reasonable cheap used detector head from systems where the computer part went dead / obsolete. So this would only need something like a µC for control with DAC's and maybe a repair of the electronics.

[JonM]

required a lot of machine shop work.

Ah. I should have mentioned that I'm a mechanical engineer first and an electrical person second, and the Rapp quadrupole looked simple to me compared to the typical commercial mass spec designs.

For ToF the detector might be the harder part

I'd be interested in more information about this -- are electron multipliers more favorable? Or do you need to make some kind of specialized high-speed electrometer? Most electrometers dont' seem to go above audio frequency.

As Kleinstein said, an electrometer will not work. You will need an electron multiplier of some sort. It looks like anything new is expensive. Getting a detector from surplus equipment, or possibly making one from scratch, should be possible. ToF will also require a high voltage pulse to accelerate the ions, or maybe you could pulse the ionizing electron beam and use a DC accelerating voltage.

I don't know what the mass spec technology inside that box was but they were probably replaced by the RGA which provide a lot more information in a small package.

I know that many Varian leak detectors used a magnetic sector.

Yes, probably a small magnet with fixed slits for the single mass/charge value.

What type of vacuum system do you have? How large is the chamber?

I don't really have any *large* chambers, would acquire them as needed, esp. as many mass spec designs would go in a long KF40 or KF50 nipple. Or sector spectrometer in a pipe tee My system reliably gets down to 5E-6 torr, probably lower if I cleaned it for once. Its mostly held together with KF fittings, and has a needle valve gas admittance passage.

That is a good vacuum for a home system. What type of pump(s) are you using?

I like KF fittings and used them in the higher pressure sections of my MS. It looks like this will be a lost bit of history, can't find anything on-line, but I believe that KF fittings were derived from the dairy industry. If you lookup "milk line fittings" they look familiar. The story I remember is that Nor-Cal (now part of Pfeiffer Vacuum Technology AG) was originally in the dairy equipment business. They don't mention that on their Web site but they do say that they worked with Lawrence Livermore Laboratory to design standardized vacuum fittings.

My ultimate pressure goal was 1E-10 Torr so CF flanges with Copper (and Aluminum at the SS - Aluminum transition, knife edge on Al had TiN ion-embedded hardening) were required. I still have a pile of used Copper gaskets in my garage.

I once wanted a home high vacuum system, but now I will have to have a really good application (or run into a deal that can't be passed up!).

[ikrase]

I believe that KF fittings were derived from the dairy industry. If you lookup "milk line fittings" they look familiar. The story I remember is that Nor-Cal (now part of Pfeiffer Vacuum Technology AG) was originally in the dairy equipment business. They don't mention that on their Web site but they do say that they worked with Lawrence Livermore Laboratory to design standardized vacuum fittings.

Interesting. The fittings are somewhat different (KF have much larger flanges for example, and centering rings rather than gaskets with ridges).

I have seen some vacuum foreline traps held together with milk line fittings ("Sanitary Flange" or tri-clamp).



One design I have seen, which was used in the original patent for the ToF mass spec, used neither an ionization beam pulse nor an accelerator pulse, but rather a sort of swept sideways deflector waveform combined with a slit.


How high is the high voltage? My calculations suggest you don't want too  high.

[cowasaki]

I've moved here from mainly posting on a microprocessor forum where a large part of the members clearly don't even know what a transistor is  ........

"Has anybody built their own mass spectrometer before?"

......

"I've built two....."

Absolutely wonderful. I'm loving this forum.  Sorry I've nothing to add other than admiration.

[mikeselectricstuff]

Could you cannibalise a photomultiplier to use as an EM ? These are pretty readily available surplus

[JonM]

I dug out a copy of the paper related to the laser ionization ToF MS. There are certainly better references somewhere, but this provides an idea of the voltages, etc. for a simple linear ToF MS.

 Flight tube length: 61 cm
 Four grids of 90% transmission Nickel mesh
 Grid spacings: 1.1, 1.0, 1.2 cm
 Grid Voltages: +105.2, -105.2, -400, -2000
 UV Excimer laser wavelengths: 248.5 nm & 308.0 nm
 Laser beam width along flight path: 1 mm
 Laser pulse width: 10 ns

At least the dimensions and voltages have some relevance, but note that the laser beam being focused to 1 mm width and having a 10 ns width in time contributes a lot to the final mass resolution. There are probably tricks for electron impact ionization and pulsed extraction.

For reference, the paper is listed below. Note that it is old, and this was a proof of concept. The group went on to build more sophisticated ToF mass spectrometers while I concentrated on Laser Photoelectron Spectroscopy (before going to industry to build a Fourier Transform Ion Cyclotron MS).

 "Sensitive and Selective Detection of Molecules with Ultraviolet
  Laser Ionization Gas Chromatography/Mass Spectrometry", Gerald
  Rhodes, Richard B. Opsal, Jon T. Meek, and James P. Reilly,
   Anal. Chem. 55, 280 (1983).

For a home project, I would still lean toward rehabilitating a surplus RF Quadrupole MS and use a Faraday cup + electrometer as the initial detector.

[ikrase]

Oooh, cool, thank you!

Well, I got my ionization gauge working -- with less voltage than I thought would be needed. A typical ionization gauge has a coaxial grid-collector assembly with a radius of about 10mm and uses a +30 volt cathode bias and a +150 volt grid bias.

Mine has a grid radius of only 4 mm (estimate) and uses about +2 volt cathode bias and +28 volt grid bias. With a cathode emission of about 70 microamps, I was easily able to see the response on my electrometer track the pressure reading on my commercial ion gauge as I swept the pressure in my vacuum system from 3E-5 to the upper limit of ion gauge range. (This was all using the nanoampere or 10-nanoampere range.)

I've attached the picture of my ion gauge electrodes.


Obviously there is a very long way between a bayard-alpert guage and any kind of mass spec, but making a basic ion tube (an improvement over my junktacular triode vacuum tube) is a step towards it.



I am mostly astonished that this thing worked the first time!

[JonM]

Excellent! It looks rugged and you can replace the filament if it ever burns out!

Now to calibrate against the commercial ion gauge...

[ale500]

 Yeah, I'm a chemist, my ph.d I did it in mass spectrometry and the postdoc too. And I am an electronics person, too. And I still want to build my own MS.

What I wanted to say is: I like your idea of building one. And from my experience with some commercial machines from Thermo (Finnigan) they are not as complicated as one may think. We always had problems with EI, it is normally used to produce fragmentation products. You may want to look into electrospray, you need a solution of ions for that, acidic water would also do . As you see protons with variable number of bonded water molecules.
At least in theory, the rods of the -multipole should have a hyperbolic surface, round ones are a not the best ones but they will work, you get more ion loss and less confinement.
What I would do is to research the papers form the 50s 60s 70s regarding self-built machines in universities. If you had access to them of course.
RF is not that difficult to get: a multivibrator with two tubes is what we had on our high-pole count multipole, it worked at around 1.2 to 2.4 MHz and some 250 V. With higher frequency you get lighter iones through. A very simple design. A variable capacitor in parallel with the primary winding of the output "transformer" made of thick wire and some 20 turns. It can also be done with transistors, the primary is then just one turn and the secondary has like 200 .

[John Heath]

Yeah, I'm a chemist, my ph.d I did it in mass spectrometry and the postdoc too. And I am an electronics person, too. And I still want to build my own MS.

What I wanted to say is: I like your idea of building one. And from my experience with some commercial machines from Thermo (Finnigan) they are not as complicated as one may think. We always had problems with EI, it is normally used to produce fragmentation products. You may want to look into electrospray, you need a solution of ions for that, acidic water would also do . As you see protons with variable number of bonded water molecules.
At least in theory, the rods of the -multipole should have a hyperbolic surface, round ones are a not the best ones but they will work, you get more ion loss and less confinement.
What I would do is to research the papers form the 50s 60s 70s regarding self-built machines in universities. If you had access to them of course.
RF is not that difficult to get: a multivibrator with two tubes is what we had on our high-pole count multipole, it worked at around 1.2 to 2.4 MHz and some 250 V. With higher frequency you get lighter iones through. A very simple design. A variable capacitor in parallel with the primary winding of the output "transformer" made of thick wire and some 20 turns. It can also be done with transistors, the primary is then just one turn and the secondary has like 200 .

I am not getting the multipole approach ? Why not just accelerate the charged particle in a straight line then move it off center with a magnetic field. This way the velocity is not relevant as it will always bend over to the same target regardless of velocity .1 , .9 or .99 c , Thomson exeriment. The offset target becomes the mass period. However If a Coulomb force used such as in the multipole approach then the target is no longer reliable indication of the particle mass. Not a phd so please be easy in the response.

[T3sl4co1l]

I am not getting the multipole approach ? Why not just accelerate the charged particle in a straight line then move it off center with a magnetic field. This way the velocity is not relevant as it will always bend over to the same target regardless of velocity .1 , .9 or .99 c , Thomson exeriment. The offset target becomes the mass period. However If a Coulomb force used such as in the multipole approach then the target is no longer reliable indication of the particle mass. Not a phd so please be easy in the response.

In short (and also as I understand merely the basics):

A multipole grid sets up for a sort of ion resonance.  The mass/charge ratio is selected by the amplitude (and other parameters, I forget which ones are independent and which are dependent on apparatus geometry, scale and such).

If you do a static magnetic field, you also select mass/charge ratio, but you need to sweep the magnetic field to target a single detector (may be very inconvenient, subject to errors in magnetic field strength due to materials and shimming), or you need a huge array of detectors (which used to be impractical, but an etched or photolithographic array -- a linear CCD in short -- would be the modern solution).  And probably you don't get as good selectivity (I think the quadruple acts to focus the ion beam, when at resonance?).

Cylinders are traditional, because centerless grinding is a manufacturing process capable of astonishingly regular roundness and straightness.  A hyperbolic (or related) curve might be ideal, but far, far harder to machine anywhere near as accurately.

Tim

[John Heath]

I am not getting the multipole approach ? Why not just accelerate the charged particle in a straight line then move it off center with a magnetic field. This way the velocity is not relevant as it will always bend over to the same target regardless of velocity .1 , .9 or .99 c , Thomson exeriment. The offset target becomes the mass period. However If a Coulomb force used such as in the multipole approach then the target is no longer reliable indication of the particle mass. Not a phd so please be easy in the response.

In short (and also as I understand merely the basics):

A multipole grid sets up for a sort of ion resonance.  The mass/charge ratio is selected by the amplitude (and other parameters, I forget which ones are independent and which are dependent on apparatus geometry, scale and such).

If you do a static magnetic field, you also select mass/charge ratio, but you need to sweep the magnetic field to target a single detector (may be very inconvenient, subject to errors in magnetic field strength due to materials and shimming), or you need a huge array of detectors (which used to be impractical, but an etched or photolithographic array -- a linear CCD in short -- would be the modern solution).  And probably you don't get as good selectivity (I think the quadruple acts to focus the ion beam, when at resonance?).

Cylinders are traditional, because centerless grinding is a manufacturing process capable of astonishingly regular roundness and straightness.  A hyperbolic (or related) curve might be ideal, but far, far harder to machine anywhere near as accurately.

Tim

Yes manufacturing is something else. So many bright and skilled minds out there. The B probe come to mind. It was a test satellite that went from the north pole to the south pole. They needed 3 vary precise round balls for gyroscopes. The manufacturing process produced 3 round balls for NASA that were so round that if they were the size of the earth the roundness would be better than 3 feet. How did they do this? And lets not forget RAM chips with separations that are 50 atoms? That is small. If a meteorite were to wipe us out the micro chips could survive so that those in the future would know we were here and we new our onions when it came to manufacturing.

[ale500]

The multipole approach...

The machine is called a triple quadrupole machine, it has actually two quadropoles and one octopole in the middle. The first one is used to get a beam of ions either one particular mass to charge ratio or all ions, the second to do ion-molecule collision experiments, and as an octopole has better confinement as a quadrupole it is used here, the third one is the ion selector used for detection. The multipole I mentioned, with the tube RF generator was used before the first quadrupole for not specific ion-molecule reactions. I used it to attach olefines to metallic ions and later to do dissociation by collision with a monoatomic gas.
Regarding the academics, I just wanted to mention that I'd like to build such a machine, too, because I know a bit of the applications field. Technically speaking, I find it a very nice mixture of several disciplines beyond the possible uses. It doesn't matter from which side you look at it you can get something usable, you can learn lots of stuff.. it hardly gets better than that, (maybe satellite building, or NMR machines), and so on.

[John Heath]

CRTs for TV were mass produced cheap. This is more than half of what is needed for a spectromass meter. It does not have to be expensive. Those in marketing could be underestimating the number of amateur weekend scientists who would shell out 500 bucks for a crude spectromass meter. I would pay 500 and I do not think I am alone judging from this thread.

[ikrase]

Magnetic sector mass specs were the first to be invented iirc. They are not as popular for several reasons, but the bulkiness of the magnets doesn't help.

Pipe tee MS is very cheap, uses only neodymium magnets, but has abysmal range and resolution.

I think the biggest expenses are the vacuum system and the ion source as long as no electron multipliers are used.

[John Heath]

Magnetic sector mass specs were the first to be invented iirc. They are not as popular for several reasons, but the bulkiness of the magnets doesn't help.

Pipe tee MS is very cheap, uses only neodymium magnets, but has abysmal range and resolution.

I think the biggest expenses are the vacuum system and the ion source as long as no electron multipliers are used.

Low resolution Hmmm. It is a given that the quad pole is much better for finely separating the masses with high resolution. The question in my mind is why. Or for that matter why is a magnetic separator low resolution. The magnet separator has a the advantage of not being effected by ion velocity. The quad system does not have this advantage. One would think that this in itself would give the magnetic separator an advantage. Apparently not as the quad pole is producing the best results.