Fusor Forums

This morning, while I was evacuating my fusor chamber and differentially pumping the RGA for high-pressure measurement, I cracked the needle valve on the still-pressurized deuterium gas line and immediately noticed a large peak at m/z = 6. The pressure in the main chamber was around 10mtorr and that in the RGA chamber was ~1e-5torr, and there were two peaks of roughly equal height (~9e-6) at m/z = 4 (deuterium) and m/z = 6. In less that ten seconds the peak at m/z = 6 shrank to nothing (literally nothing down to even e-11) and the one at m/z = 4 increased to roughly the sum of two initial partial pressures. I guarantee it was no instrumentation fluke as the RGA has been 100% reliable and the peak corresponded perfectly with opening the valve. When no deuterium was in the line nothing happened, so that rules out anything in the construction (drierite, hydrocarbon valve grease, stainless, etc; all are way above m/z = 6). Something new was definitely there at m/z = 6.

I'm reminded of this conversation had on the forums way back in 2004: http://fusor.net/board/viewtopic.php?f= ... ium#p36510. Frank Sans mentions that the ultra-high-purity deuterium he obtained had as much as 0.1mCi/L of tritium. Could this be the source? Some math (order of magnitude, nothing ultra-precise):

0.1mCi tritium (s.a. = 9650Ci/g) corresponds to 1.04-8g, or, at STP, 3.87e-8L. This corresponds to 1.29ppb in a 50L lecture bottle. The deuterium line is 1m of 4.5mm ID stainless tubing at roughly atmosphere, so we have V = 0.0159L of deuterium and thus 1.23e-11L (STP) of tritium. The total chamber volume is ~1.44L, so the expected partial pressure is 760torr * 1.23e-11L/1.44L = 6.5e-9torr. This is easily measurable by the RGA but 1000x the observed value. My deuterium is not UHP, but 99.8% or 99.9% if I remember correctly, so it may be possible that more tritium found it's way past the refinement steps. So... maybe?

All the RGA spectrum tables I have found online yielded nothing but doubly ionized 12C at m/z = 6 (http://ytionline.com/technical-informat ... ion-guide/) which is in no way the culprit. Since the peak disappeared so quickly and I was rushing to get a picture I did not measure the roughing pump exhaust with a Geiger counter.

Thoughts? I'm skeptical, but those far more experienced may have better explanations. The only other isotope would be 3He, but like I mentioned this happened only with deuterium in the line. While I guess 3He in the deuterium is possible, I just don't know. My neutron detector isn't exactly leaking...


The spectrum is clear... Apologies for the poor quality.

The m/z 6 ion is D3+. Google “H3 factor in hydrogen isotope ratio mass spectrometry” for papers that explain the source of the H3+ ion.

Ah, as per usual the explanation is more nuanced. I am still surprised that 50% of the deuterium pressure (only in the RGA, of course) was D3+. According to one paper (http://www.forensic-isotopes.org/assets ... .1_Web.pdf), "The H3+ factor should not exceed 10ppm/nA [...]" which is far, far below what I measured. It also says that "The H3+ factor is determined by measuring the intensity of m/z 3 as a linear function of m/z 2, usually performed with the working gas", implying the effect is persistent. The peak at m/z = 6 disappeared in seconds, however. Perhaps these discrepancies were the result of some transient in the RGA filament supply at startup or something of the sort. I will be running the fusor again soon and will see if the effect reemerges.

Thank you for clearing that up.

As I have never used, but only been around working RGAs. I am unable to comment, but have a question. Does your RGA have an electron multiplier on it. To go down as low as you claim, older RGA systems demanded an optional electron multiplier at significant extra cost above and beyond the base level RGA.

Do you own the RGA as part of a personal system ($$$$) or borrowing one or perhaps it is at a university system?

My history on RGAs is back into the 1990's. Modern systems may cost a bit less and include the multiplier, i just don't know. If you can, educate us if you care to elaborate.

If you wish to expound at length in a rather involved treatise, feel free to generate a FAQ in the vacuum forum. I doubt if the RGA with adorn many fusor systems, but some foreknowledge by a user might create a most informative FAQ. I really hate being the only FAQ writer when there are others who know far more than I do on many useful subjects touched on across the vast panorama of many of our forums.

Richard Hull

My RGA, which is the Spectra Vacscan Plus (late 90's to early 2000's) with a 100amu range, does have a multiplier, though the maximum operating pressure for it is 2e-6. For all purposes but measuring the base pressure of my system (9e-8torr) I use the Faraday cup.

I managed to score the RGA and all associated cabling, RF head, and control box on Ebay a couple years ago. Got it all for ~$200, I think. It was "pulled from a working system" and not guaranteed to work, but it was too good a deal to pass up. I figured that the most common points of failure are the filaments, and sure enough one filament arrived open circuit but the second works just fine. All modes but one (some gas library comparison feature) work on the controller, so all-in-all I got a really good deal.

The RGA has been a valuable tool in my learning of vacuum systems, particularly in the regime below 1mtorr. Water vapor constitutes >90% of the remaining gas (assuming a real-leak-free system) and can only be removed by long bakes at >100C. Various hydrocarbon peaks in the m/z = 40 range indicate I should probably stop using o-rings and vacuum grease for some conflat connections, especially close to the fusor core where things get toasty and electrons fly. Setting it up as a differentially pumped system with two high-vacuum turbo or diff pumps allow one to measure approximate gas ratios in the main chamber at pressures several orders of magnitude higher than usual. I've had the fusor chamber at over 100mtorr deuterium, or over 1000x the max pressure the RGA can be operated at, since the turbo pump maintained the RGA chamber at 5e-6torr. Doing so has allowed me to verify that >98% of the pressure is deuterium and not water, nitrogen, or other contaminants. Eventually, if I can establish a good enough noise floor for the D3+ ions, and if (when) I achieve much higher neutron counts on the order of several e6, I may try detecting tritium. The RGA has a minimum measurable pressure of 1e-11torr, so it might just be possible given good experimental methods and high fusion rates.

As for an FAQ, I could start one, but for it to be of any worth, those with far more experience than I would need to pitch in. I can discuss stuff like fusor-specific applications, basic high vacuum contaminant analysis, and operating modes/procedures, but as this very thread demonstrates, I am far from an expert in the matter and likely on the downward slope of Dunning-Kruger.

A quick follow-up to the question stated in the thread title: I have measured a small peak at m/z = 6 again, so the matter is clear; D3+ ions are formed in the RGA ion source.