Preferred/Best Diffusion Pump Oils

[bunnspecial]

I'm in a pump oil changing mood/maintenance cycle at work, and am wondering if anyone here knows anything about diffusion pump oil.

This is for mass spec use, and I need-at a minimum-a vacuum in the low 10^-4 torr range, and preferably into 10^-5 torr. I'm dealing with small pumps as such things go-basically pumps for benchtop sized mass specs, which tranlates into a pump about the size of a large coffee cup(typically 20-30mL of oil).

I did one last week with Santovac 5P, which is what Agilent now specs across the board. This was on an older instrument(HP 5971) that said Santovac 5, but the local Agilent FSE said to use 5P anywhere that 5 was previously speced. I originally bought it based just on an Agilent p/n, and paid the frightfully expensive Agilent price of $175 or so for an 18.5mL bottle.

I still need to do an Agilent 5975 and a Varian Saturn 2220, and don't have enough on hand to do both. That means I'm going to need to order SOMETHING. Of course, Agilent is going to follow the company line and tell me to use 6040-0809, which is 5P in an Agilent bottle, and that includes on the Varian instrument also. If I do order more 5P, it will be from a supplier other than Agilent, where I can find the same oil for about half the price

I also need to do a few more foreline(rough) pumps. There again, the recommendation is for Inland 45, which seems as good as anything. The pumps in question are two of the super Varian DS 102s, and the grossly underspecced for the application(but small and quiet) Edwards E2M1.5.

 

[Sunnyinhollister]

I have not had to deal with diffusion pumps in almost 20 years. We had larger pumps - 8"+ ANSI flanges and I cannot honestly remember what we used for the fluid. We were also pumping on a lot of nasties such as BF3, AsH3, PH3, ect. If you cannot convert over to a turbo/dry pump setup, I would call Kurt J. Lesker Company and talk to them. They have been in business forever and should set you in the right direction.

 

[Kestas]

With the diffusion pumps I'm familiar with, 35 years ago they needed complete overhaul every 6 mos or so. The ones nowadays use a special oil that doesn't require regular service. I believe the oil is crazy expensive. Sorry I can't help further.

We use Inland 45 for our roughing pump.

 

[kschachn]

Why, mercury of course, especially if the pump is glass. Only kidding, but we did have those in college.

We used a perfluorocarbon of some sort, you could buy it reclaimed for less than virgin material. But if you aren't pumping anything with oxygen you can get buy with something cheaper.

You say the one is expensive, but how often are you replacing it? If it is a lot then maybe you need something more resistant.

 

[bunnspecial]

Thanks guys.

The "official" service interval is every year. Realistically, the one I just changed has a sticker on the front that it was last done in 2000(and it was unused from March 2010-March 2019). The Agilent scheduled for changing is 9 years old has never been changed, while the Varian is closer to 20 and also has never been changed. The one I just did was "just because", while I'm changing the others because I can see that the resting chamber pressures have been rising slowly but steadily over the last several years-I try to note that stuff at least weekly in the paper logbook kept next to each one.

These honestly live a pretty easy life. Most of the time they're pumping helium off the gas chromatography column with an occasional solvent(methanol and dichloromethane are probably the big two, although there are others) and also tiny quantities of analytes. Even the solvents aren't in huge quantities-we usually inject 1µL into the gas chromatograph, and that often gets split anywhere from 5:1 to 40:1(meaning somewhere between 1/6 and 1/41st the amount injected actually goes onto the column). Plus, most of that doesn't even linger around in the diffusion pump if my other readings are any indication-I can usually see the foreline pressure jump 50-100 mTorr when the solvent comes off the column meaning that a lot ends up either in the rough pump oil(which I do faithfully change every 6 months), the vapor trap on the rough pump exhaust, or in the room air.

Two of these-the Agilent 5975 and the Varian-could have been ordered with turbopumps when new, but it was a ~$30K option that we didn't get, and AFAIK it's not realistic to fit one. Agilent wants to pretend that the Varian mass specs never existed since they were a competitor, while I suspect that the between the pump itself and a controller Agilent would want enough that we'd be close to the cost of just buying a 5977 with a turbo. The HP 5971 was never fitted with a turbopump(interestingly, its predecessor, the 5970 that was discontinued in the late 80s, always came with one, but the 5971 was basically a complete rethinking that tried to make "benchtop" mass specs more affordable and require less maintenance).

 

[bunnspecial]

I should also mention that turbos aren't really a great option because I'm planning to wean off helium completely in the next year and switch mostly if not entirely over to hydrogen.

Turbos become less efficient as the gas size decreases. They are workable with helium, but IIRC Agilent(who is pushing hydrogen across the board on all their GC systems, not just GC-MS, given the rising cost and reduced availability of helium) specifically advises those who intend to run hydrogen to spec diffusion pumps wherever possible. You still need turbos, regardless of the carrier, if your'e running a triple quad, TOF, or something else with large vacuum chamber, but for a benchtop single quad they(Agilent) still like diffusion pumps with hydrogen.

 

[bsmithwins]

I did some diffusion pumps in a previous career and the pump oils (never serviced Hg based pumps, thank the Maker) I worked with generally fell into two types: hydrocarbon based and silicone based.

My take is that hydrocarbon oils are cheaper but crack and leave carbon deposits behind and back stream more. The silicone oils were more stable (and expensive) and had increased service life times.

I'd think you'd want to completely strip and clean the pump before switching.

BSW