Importance of Filtering MS Foreline Pump Exhaust ?

[Brian1]

Hi folks,

I have read the few other posts that I found addressing this, but would value any opinions on my specific circumstance (methods run, freq. of runs).

I was gradually promoted in my lab about 5-6 months ago to operating our GC (5890), GC/MS (5890/5971 & 6890/5973), and HPLC. I have been really enjoying it, learning a great deal, and excited versatility of these instruments in terms of career opportunities (I graduated w/ my B.S. one year ago). Aside from the HPLC, the instruments seem to have been somewhat neglected the last few years. I have been using this forum frequently in combination with a few books and operating manuals to assist in troubleshooting and have been successful in correcting simple issues that were causing serious inefficiencies - thank you all particularly Peter Apps and Don Hilton!

The foreline pump (old Edwards model?) on our 5971 is not vented to a fume hood and does not have an oil mist filter. We run methods 525 and 548 once a month as well as 624 and 625 probably twice a month. The oil has not been changed in probably 3 years or more. I requested that the company install an oil mist filter, but was denied as to them it is a waste of money unless you are running dioxins for a superfund site. After writing a lengthy email explaining my reasoning, citing sources and the specific chemicals of concern, I was told ......

"It is my opinion that the amount of chemicals released into the air from the am is "di minimis". It is only one 1ul injection per hour and anything that does make it through the ms spec is immediately diluted in the lab air system. In terms of risk you probably inhale 1000x more benzene and other chemicals ever time you fill your car up with gas. But if it makes you happy we can get a filter for the 71"

And as far as my efforts to develop and follow a routine maintenance schedule (changing septa and liners, clipping front end of column, updating manual tune), I was told we follow an "if its not broken lets not fix it approach".

Am I being unreasonable and overly cautious?

[Don_Hilton]

My personal thoughts on this are: If it will condense in the oil, it is not going to come out into the air at any appreciable level. So in most cases, I don't tend to have a great concern about chemicals injected into the GC coming into the air from the pump. But, you may have noticed a bit of oil mist when you pump down the system. That goes into the air. And, breathing oil mist is not going to make you grow big, strong and tall! And, if the material will volatilize from the oil, I don't expect a particle trap to catch it.

I worked in a lab where we did not have mist filters on some of the pumps, but we did run tubing to a nearby hood or other exhaust. Where the tubing would be strung across pipes in the ceiling there would be dips. And, over time, one would notice oil collecting - not a lot - but it would collect in the dips. (It was inside the tubing – not the laboratory air!)

And, if there is a sufficient leak in the system, air moving through the pump will lift oil into the air in an aerosol. I have seen many GC/MS systems, I have seen a number of oil mist filters at instrument PM time. And in most cases, they have been essentially dry. But, there is the occasional system where the oil needs to be drained out of the filter. In that case there has been a small air leak – sometime a crack or poor connection in aging foreline vacuum tubing – and the filter did save the laboratory air from contamination and probably an oil film on equipment and furniture nearby.

And, if there is a bit of an oil film on the furniture or walls near the instrument, you can demonstrate that you have a way to avoid the cost of cleaning that up repeatedly. (I have noticed that there are some managers who will accept a person who likes to work in a clean laboratory, but safety concerns make people expensive crybabies. Not that the manger will tell anyone that. -- just my impression.)

The fact that pump oil has not been changed in three years tells me something. There is an expression "penny wise and pound foolish." Without even looking at the pumps, I can safely bet you a dinner out - some place nice - that the oil is dark in color. The oil has changed over time - and, you have chemical degradation products of something! And, if those degradation products are like degradation products in other oil that has been heated, you can expect that they are not relatively inert, like new, clean pump oil. This is eating at parts on the pumps and they will wear and fail (so some of that crud is degraded pump seal and perhaps metal worn from pump parts). And, a pump cannot pull to a lower vacuum than the vapor pressure of the pump oil. All those pump oil molecules that have broken down from large molecules to small molecules are now part of the vapor pressure in the foreline. While not great, they are working against what you are trying to do and are making the turbo pump or diffusion pump work harder. I’ve replaced turbo pumps before. Not a difficult task – except for getting the boss to agree to spend for it. And diffusion pumps must have oil changed on a regular basis. You are counting on the mass of the molecules in the diffusion pump oil to sweep the lighter molecules down the pump and away. This includes residual analyte molecules as well as air - and degraded molecules of diffusion pump oil. The vacuum reading may be OK, but how is the noise level in the instrument as compared to three years ago? I’d hate to replace a detector to get rid of noise if diffusion pump oil was getting old. And, if you check the LOD and get the value you need, perhaps discussion of instrument background is, perhaps, only academic. But that gunk in the oil is part of any aerosol generated by the pump.

The "economy" of not changing oil will result in premature failure of the pump - and replacement, probably with a rebuild unit. Cost of down time? Oil can be changed when an instrument is vented for source cleaning. It adds half an hour to an hour to change the oil. (I've done it many times.) And time taken in troubleshooting a vacuum system that will not give proper vacuum in the analyzer section – because seals are failing in the rough pump – will eat up all that saved time – and then some. (I’ve done that too.) So unless a jug of oil and the disposal fees for getting rid of old oil cost more than a rebuilt pump, I think the economy is working against you on neglecting changing oil. A diffusion pump takes a bit longer, as that thick oil is not easy to clean out.

Will small amounts of oil aerosol hurt you? Probably not significantly – if the oil is clean, like from a properly maintained instrument. I’ve not checked the MSDS for pump oil. But the questions I can’t answer are: what is a small amount of oil aerosol or what is significant hurt.

My suggestion: get a fitting and some tubing (Even garden hose – but have a clear section between it and the pump so you do not have degraded garden hose falling into the pump or any trapped oil running into the pump either!) and vent the pump to someplace where the exhaust is vented to the outside, away from people.

I am part of that generation who uses gloves for dangerous chemicals - liquid bromine might classify for that. But through the years, I have noticed that we have found that many things that we once considered to be safe to humans - like PCBs and BFRs - have a detrimental effect on people, apparently even to unexposed children of people who are exposed. What was once excessive caution may today even fall into the range of inadequate caution. Thus we analyze the environment for chemicals that we “safely” dumped into rivers and landfills 75 years ago. (But, I don't think I am unjustified in grinning when I ask a young chemist to put some deionized water in a beaker so I can rinse some parts. And, the first thing they do is put on gloves. It is a matter of balance – and that is the hard part.)

And the usual caveat: The views expressed mine and are not necessarily the views of my present or any previous employer.

[bisnettrj2]

Feel free to not respond to me, I just wanted to throw in my two cents.

Be cautious as to how aggressive you are in trying to change the way things are done at your current lab. They've been operating as they have for a while and it seems to have worked for them so far. There are 'right' ways to do things, 'wrong' ways, and 'lazy/frugal/cheap' ways (read - don't fix it if it isn't broken). The lazy/frugal/cheap ways aren't right or wrong, necessarily. They just aren't perfect, usually. But they do often save money . Trying to tell them you're right and they're wrong may not work out so well for you. Especially since they've made it this far, without you.

Should you change rough pump oil every 3-6 months per the manufacturer? Depends on what you're running, I would think. But just like in your car, oil gets dirty as bits of contaminants, metal, pump seals, etc make their way into the oil - and the longer you let them accumulate, the more expensive the 'fix' might be. Start small with your suggestions - say you want to change the rough pump oil on an annual basis, at least on the instruments running the more frequent tests. And I thought the point of an oil mist filter was to trap oil from escaping the pump's exhaust? Wouldn't it be beneficial to keep the pump and its surrounding area oil free?

Preventative maintenance is good when you're required to do it (I'm under the impression some industries have to do this per their regulatory authority, like labs doing USP work, but I could be wrong), when you run instruments so hard that you can't accept a down instrument except when it's on your terms, or when an instrument is new and under contract by a service provider (because an annual PM is usually included in the service contract). Otherwise, you're throwing money at instruments when it might not be necessary, especially with relatively low-price, low-volume tests like what you're running.

Ask to purchase common replacement parts for the GCs and MSs - ferrules, a couple column nuts, a couple transfer line nuts, septa, liners, gold split seals, filaments for the MS, and a backup column for each instrument (you can purchase them month-by-month over the year, I'm relatively sure GC columns don't have a shelf life), and replace them when it is warranted, not just on a blind "we ran 100 samples, time for a new liner" basis. That way you're prepared to fix it when it is broken, and you've saved down time. Clip column when performance dictates - loss of active compounds, excessive tailing, etc. Re-tune when you have to, not just because you want to - having to means you can't pass a tune check or you're getting decreased sensitivity over time (which may also mean it's time to clean the source, which would dictate a re-tune anyway).

[Don_Hilton]

On changing septa, liners etc.

Septa can last to what some people might consider a surprisingly long time - over 100 injections. (While other people keep driving beyond that and suddenly have sensitivity issues as the septum begins to leak.) If you are running ten samples a week - there is no point in changing the septum weekly.

Liners can have a very long life in an analytical method - or a very short life.

The evaluation of the data from the runs can dictate the needs for some of these preventative maintance issues. And not fixing it until it is broken may be a matter of watching data until it tells you that something is reaching the end of its useful life - not bad yet, but getting there. So, there is a part of the picture that may be missing.

And for some methods, retuning requires recalibration or at least more extensive checkign than an instrument that has not been retuned - which becomes a lot more work that runing a check standard.

[James_Ball]

For those EPA methods(we run them here also) just do the tune checks listed in the methods (BFB or DFTPP), as long as those pass and are not pushing the limits I would not bother with retuning. I usually try to avoid opening the tune gas valve as much as possible since it seems to take some time for it all to be removed from the analyzer and the vacuum returned to normal.

No more samples than you are running I would say you are ok as far as chemical contaminates coming out of the vacuum end. Usually you will get more coming from the split vent than from the vacuum pump, you can place a small charcoal filter on the split vent to take care of that. We probably run on average 20-40 samples per day on each instrument and we usually do oil changes annually.

For years in the volatiles department we didn't have mist filters, seems they did not even ship them with the E2M2 pumps like you have on the 5971, the first mist filters we had came on our first 5973 using the E2M1.5 pumps. Those run much hotter and higher RPMs and generate more mist than the old slow E2M2. We have replaced all but one of the small pumps with old E2M2s, just wire them up to plug directly into and outlet not the back of the instrument. Those little pumps I would definitely check the oil more often, the higher temps will break it down faster, the old ones are good for much longer with the small load you seem to be running on them. I have had instruments running purge and trap that don't get vented for 18 months with no problem, oil is still clear as it was when new, as long as I have a moisture trap on the pump inlet.