Van de Graaf Accelerator question

[broean01]

I'm using a model PN-250 positive ion accelerator from High Voltage Engineering Corporation, and I was wondering how the beam focus works.

The electric field used to accelerate our protons is generated by a stack of ten plates with voltages stepped down by 10 MOhm resistors. The focus control isconnected to the second plate from the top of the stack.

 

[litup]

My work involved the medium and high current ion implanters from Varian and Nova. These are junior accelerators, usually under an MEV of accel. Like you said, electrically driven acceleration rings but first lower energy extraction of ions from the source, usually coming out of the source between 25KEV and 50 KEV, in the implanter, before the main acceleration rings are encountered, they have a mass analyzer that can separate individual isotopes, not sure if you have that or need that step, it involves a 60 degree or 90 degree swing of the beam under a controlled magnetic field to analyze the beam for what isotope you want, like Boron 11 Vs Boron 10, (10 has a weaker beam so we use 11 for silicon wafer doping, 4 times stronger beam) So after the main acceleration ring, it goes to a series of magnets spaced circularly around the beam with focuses it much tighter. We also use self focusing involving a somewhat higher concentration of a gas in the beamline post accel, Do you use that method also?
Some ion implanters use electric focus methods but ours use mainly magnetic focusing, magnetic lines cause the beam to bend so alternating magnets can focus the beam. Here is a link to a picture of one my implanters, this appears to be a Varian 350D production implanter, just the control panel. I have had my hands on about a thousand of those guys:)
http://www.lanl.gov/mst/docs/LALP-06-036.pdf

 

[broean01]

To clarify, the focus I'm asking about modifies the electric field of the acceleration column in some way to affect beam spread directly from the accelerator. We use magnets to analyze, steer and focus further down the beamline, but I know how those work.

 

[litup]

To clarify, the focus I'm asking about modifies the electric field of the acceleration column in some way to affect beam spread directly from the accelerator. We use magnets to analyze, steer and focus further down the beamline, but I know how those work.

As far as I know, the only thing that can happen in the accel column is to constrict the beam to a smaller diameter like a shutter in a camera. I don't think there is any effect on the actual beam width since each accel ring just gives a kick to the velocity of the beam, it does not effect focus. The only problem with using the accel column to clamp down the beam size is the beam would eat into the accel ring and destroy it eventually. That's why in the analysis section of the ion implanter, it is lined with graphite because the other ions get turned at various angles by the analyzer field, the graphite gets eaten out and has to be replaced because the beam contaminants drill holes over time in the waveguide if it wasn't there and that would do wonders to the beamline vacuum system:) In the accel columns I use, the discs have holes that are narrower at the beginning of the accel column and are somewhat larger in diameter at the end, the last ring having a diameter maybe 50% larger than the beginning, precisely because if they were all the same size, since there is no effect on focus, beam spread would eat into the ring which would heat up that ring, perhaps causing the epoxy holding the assembly together to split apart, again ruining the vacuum. As far as I know, accel columns can only effect the velocity of the beam not the focus. Perhaps you could incorporate small permanent magnets around the opening of each ring to combine the action of the beam focus area downwind and do focus at the same time but that would complicate the construction of the accel column and increase the weight, in ours, the accel column has lead power in the epoxy to reduce the radiation coming off the column for safety reasons so the columns we use are already pretty darn heavy. One of the high voltage machines I worked on, a one megavolt job, had an accel column that was about 6 feet long, surprised the heck out of me when once I saw someone working on it where the column was supported only at one end and it was hanging out in midair, I couldn't believe it would hold together so well since there must have been 50 ring sections all just epoxied together, so maybe incorporating magnetic structures in the accel rings would work. I think that is the only way you are going to be able to get focus AND accel out of the accel column. We had to replace a lot of accel columns on our implanters because of beamspread from one end of the column to the other, the rings got really pitted with beam products attacking them and we did not have any way of dealing with that since the only thing we had to prevent that was the graphite beam slit just before the accel rings. We had one we called at Varian the 'Callahan tunnel' because our chief scientist, Ray Callahan, who I believe started at High Voltage Engineering in the late 50's, made one beam slit that was designed to allow the separation of one of the higher AMU beams where the isotopes had to be separated but the actual mass difference was very small. Trying to remember what isotope that was, not Arsenic at 75, it didn't have any rivals there, maybe antimony at 130 v 131? Something like that, one isotope would be very close physically after mass analysis because the masses were such a small percentage difference so he made this quite long graphite slit, maybe 5 cm deep with a very narrow slit width which seems to have done the job but part of the reason was accelerating such high AMU isotopes together made them bore out the accel rings because of beamspread and the contamination of mixed isotopes in the beam when it came to the scanning electrodes or magnets (depending on what series of implanter it was in, the high current machines used magnetic scanning in the X-Y direction and vertical scanning via a spinning wafer wheel).
The "Callahan tunnel" did a good job of saving the accel column rings from being eaten out by poorly contained isotopes. Of course that meant the slit itself had to be replaced periodically but that was a lot cheaper and a lot less time consuming than replacing the whole accel column which meant the whole frigging beamline had to be re adjusted, a pain in the butt if you asked me:) But we never tried to focus and accelerate in the same mechanism and I suspect it was because of the weight problem of having at least four magnets near the inner disc opening would be needed on several rings at least. I don't know for certain but I do know it would certainly complicate the construction of the individual rings. Maybe you could make each ring electrode magnetic with slits separating each n/s pole to make the electrode do double duty without having to install a separate electrode, but that would be a tricky thing in itself, if the inner ring got hot enough from continuous use, the magnetic material of the rings could go past the curie point and lose its magnetic field and then the beam spread would go to hell in a handbasket. I am just speculating here, that would entail the development of a whole new kind of accel column for sure, but it might be possible. The only thing I know fore sure is all the beam focus assemblies WE used were magnetic in nature not using any form of electric field since magnets bend, electric fields deflect only, like a mirror, so it would be very tricky to turn the deflection of scanning electrodes, for instance, into a focus effect. So for whatever it's worth, you could theoretically incorporate magnetic structures into an accel column, it would at least be an engineering challenge and not something you could just slap together with spit and polish:)

 

[broean01]

This isn't something I'm thinking about installing. It's already there.

Quoting directly from an undergraduate thesis written regarding this accelerator:

"Also connected to the second column plate from the top is the focus control. The focus control allows the operator to change the voltage between the plates and change the focus of the beam so that it either diverges or converges more."

 

[litup]

This isn't something I'm thinking about installing. It's already there.

Quoting directly from an undergraduate thesis written regarding this accelerator:

"Also connected to the second column plate from the top is the focus control. The focus control allows the operator to change the voltage between the plates and change the focus of the beam so that it either diverges or converges more."

Do you have a link to the paper or image I could see? You can do the same with electromagnets or movable fixed magnets which I saw on an antique implanter I had to get functioning again. What a mess, yeech, I am glad I only had to do that one, had to convert from liquid freon cooling which was all the rage 30 years ago to DI water cooling after freon cooling was outlawed. Problem there was all the stainless fittings had to be retrofitted with teflon because of the habit of DI water to eat out metallic fittings. And that was just the minor problems:) It had a movable permanent magnet for varying the focus, which I did with the beam on the lowest energy setting, just what came out of the source at about 25 Kev and an optical port right over the top of the beam by the focus magnet, I straddled the beamline just by the exit of the mass analysis line while it was running and was able to see the beam enough to focus it down to a nice pencil thick beam from a fanned out piece of crap doing nobody any good, in fact would have eaten out the o-rings on the perifery (that happened to one of my machines in Israel) Does wonders for the vacuum:)

One thing I am thinking about an electronic focus, it would be more energy intensive, fixed magnets we used at Varian take no energy to run.

 

[broean01]

I've scanned the relevant page from the manual

 

[litup]

Ok, thanks. I see it better now. This is an assembly that includes the ion source and extraction electrode in Varian terms. So this presumably is not the entire acceleration, right? The close spacing of the pieces and the number of them seems to me to exclude and acceleration voltage of more than 30 or 40 Kev I think. Isn't there another acceleration tube after all this assembly? It would seem you would need that if you want 100 or 200 Kev of acceleration. You clearly would not get to 200 Kev with just that source assembly. It appears to be an ionization chamber source, one which Varian gave up after the early years, going instead to an actively heated source, with a tungsten filament about 0.01 in diameter fed with something like 400 watts, couple of volts and up to 200 amps which brute force ionized the gas fed into the chamber followed by extraction electrodes made of graphite with a narrow slit about a half inch long that gave out a ribbon shaped ion beam which was focused down to a beam about the size of a #2 pencil and then various scan techniques to allow implanting of said ion under the surface of a silicon wafer, or sometimes other substrates like sapphire or glass or GaAs and the like. We implanted mainly phos, boron, or arsenic to allow doping of the silicon into conduction with tight depth control via a variable acceleration voltage, sometimes even slowing down the extraction Kev to something like 5 Kev for very shallow implants. That meant reversing the acceleration voltage potential and decelerating the ion beam instead. It is interesting to me the smaller and smaller features of modern computer and memory chips require smaller and smaller accel voltages, some down to a few hundred volts because as the chip features get smaller so does the needed depth of implant so the newest implanters are more like desktop machines not needing the huge 200 Kev power supplies or even the 40 or so odd Kev of the source extraction supply. The beam current is still tied to the extraction voltage so I imagine if they want a heavy implant at say, 400 volts (0.4 Kev) They would have to have a decel mode PS to buck the higher extraction voltage. Of course they may have evolved even better extraction ion sources, haven't kept up with the details of the latest low voltage machines.
So anyway, do your machines have more acceleration downwind from the ion source assembly you show in the diagram? I know the graphite extraction electrode does do focus duty by the shape of the electrode, carefully sculpting electric lines of force. So the surface shape is hyperbolic, kind of like a pair of eye glass lenses, one curve on back surface and another on the front. I don't think that works at 200 Kev energy levels however which is why I think Varian went with PM magnet focus assemblies which of course uses no power, just bends the beam inwards effectively. It has a null field in the center of the beamline and any stray ions gets nudged back into place when they arrive at the focus assembly out of the beamline center. Anyway, I guess there is more than one way to focus a beam.

 

[broean01]

Yeah, there's a total of ten acceleration plates. Nominal beam energy is 250 keV and we can only get that using a buffer gas. I understand how the protons we accelerate are generated and accelerate, I'm just not sure how this focus electrode affects the beam. We don't do ion implantation. This is an old accelerator which we intend to use as a teaching apparatus. Experiments performed include rutherford scatter/backscatter spectroscopy and particle-induced x-ray emission spectroscopy.

The focus electrode is highly effective as a rough focus, though we do use quadrupole electromagnets to adjust beam geometry just before the target chamber.

 

[AJ Bentley]

Google einzel lens

 

[AJ Bentley]

You could do with rather more than 250keV for RBS and PIXIE. I used to run around 2MeV if memory serves. (1984 was a long while back). We could just about get through the layer of tin on a can.

 

[litup]

You could do with rather more than 250keV for RBS and PIXIE. I used to run around 2MeV if memory serves. (1984 was a long while back). We could just about get through the layer of tin on a can.

Best we had was 800 Kev double ionized. The einzel lens works because of the curved lines of electric force acting just like a glass lens with light. Thanks for the link. Don.

 

[broean01]

Yeah well as an intern I can't exactly go procure a more powerful accelerator. We've been doing successful RBS. Waiting on an x-ray detector feedthrough to try PIXE.

 

[AJ Bentley]

You might be able to hire time on one quite cheaply though.

Things may have changed but when the interesting physics moved up to higher energies a few decades ago, lots of Universities found themselves stuck with relatively low energy particle accelerators and when I left they were scratching around for work and even mothballing or decommissioning them.
There may be a few still struggling on although I guess most are gone...

Just a thought.

 

[litup]

Yeah well as an intern I can't exactly go procure a more powerful accelerator. We've been doing successful RBS. Waiting on an x-ray detector feedthrough to try PIXE.

I hope you have an already existing port for the detector! SS walls are very hard to drill through, especially if you have to make a 2 or 3 cm wide hole, ugh:)

I had the privilege of seeing the very first EDX for the electron microscope, I was working at Goddard Space Flight Center on the Apollo (Timing and Tracking), long story, but almost got a job as an assistant to this scientist, I forget his name now, but he hand made an xray probe mount and used the xray output to modulate the Z axis of the scope monitor, brightening the image where the target AMU was set. He had a manual adjustment of the probe angle marked off in AMU #, and I saw it work on some of the moonrocks brought back by the astronauts. I also held a moonrock in my hand, quite a thrill for a nerd for sure! So he would adjust the detector for the exact angle of say Aluminum, and the image would brighten wherever aluminum would show up. Great work! His budget was cut 20,000 bucks and that would have been my salary in 1970 so he couldn't hire me. I would have worked for him for free if I had been living alone:)

 

[broean01]

Buying time on another would defeat the entire purpose of my work. I'm developing this accelerator as a laboratory tool for students to use.

Things may have changed but when the interesting physics moved up to higher energies a few decades ago, lots of Universities found themselves stuck with relatively low energy particle accelerators and when I left they were scratching around for work and even mothballing or decommissioning them.
There may be a few still struggling on although I guess most are gone...

Just a thought.

That's actually where this one came from. It was found in a barn at Duke University by one of our professors in the late 90s. We got it for free but the school has been installing, refurbishing, calibrating and upgrading it for the past decade.

 

[litup]

Buying time on another would defeat the entire purpose of my work. I'm developing this accelerator as a laboratory tool for students to use.



That's actually where this one came from. It was found in a barn at Duke University by one of our professors in the late 90s. We got it for free but the school has been installing, refurbishing, calibrating and upgrading it for the past decade.

What school? For students, such a machine can still be valuable for the basics for sure, you can study vacuum technology, power supplies, plasma, well in your case protons, ionization chambers, beam scan, etc., all these things would be needed to get into the big leagues at any rate. Keep up the good work!

 

[broean01]

What school? For students, such a machine can still be valuable for the basics for sure, you can study vacuum technology, power supplies, plasma, well in your case protons, ionization chambers, beam scan, etc., all these things would be needed to get into the big leagues at any rate. Keep up the good work!

Gettysburg College, and that's why I'm working on it. I want to go into theoretical particle physics. Possibly QG, but the standard model seems like a good stepping stone.

 

[litup]

Gettysburg College, and that's why I'm working on it. I want to go into theoretical particle physics. Possibly QG, but the standard model seems like a good stepping stone.

Well good luck on your journey! I see on the school website, do you know the physics prof Tim Good? He is shown with the student Matt working on plasma simulations of upper atmosphere. I found this link to quantum gravity workshops and homework problems if you want:
http://math.ucr.edu/home/baez/qg-spring2003/

 

[DMC101]

I realize this is an old topic, but electrostatic lensing and focusing has been a reality for quite some time. The extraction electrode itself generally has an electrostatic focusing property, a can of sorts, that serves to begin the beam shaping prior to source exit. I have never seen an electrostatic analyzer at an angle however, generally RF and DC are used to oscillate ions, and the ones at the right weight pass through the quads or hexapoles and either get scanned electrically or resolved and conditioned further along. Those outside the m/z range oscillate beyond the poles and are lost. Some of these applications use collision induced dissociation (mass spectrometry) to create daughter ions, but that's a different game.

It's good to see some implant talk though. I have also worked on the Varian 350D and DF4 along with many other names for over 20 years. Nothing like smacking a load lock to retrieve a wafer on a gravity fed endstation.