Noobie's corner

[GRovian]

I was wondering if I can make a topic strictly for beginners in MS technology.

Our (very small) lab just got our first GC/MS, a single quad and there's a bunch a questions one can't find answers in the doc provided. For example, how much is too much in air/moisture contamination? How stable is this kind of beasts? Can I calibrate today and use the same calibration for a week or two, as would happen with a FID (with QC monitors, of course) or should I start my daily routine with a calibration and pray for a low drift through the day?

I can visualise a lot of people rolling their eyes, now. Thanks in advance.

Rov

[James_Ball]

Some of those answers depend on the make and model of your GC/MS.

I use Agilent 5973/5975 instruments currently and I have used 5995/5970/5971/5972 instruments in the past.

There will always be a little air/water, but if you can get a ratio of m/z 28 versus 69(from tuning gas) below 5% you are doing good. Lower is better but you can actually analyze sample with it as high as 10%, but sensitivity will be a little worse and you will need to clean the source more often.

Stability is normally good, but I find the newer ones to seem less stable than the older models. I think it is difficult to increase sensitivity and maintain rock solid stability and Agilent has been pushing sensitivity to higher and higher limits with small losses in stability overall.

Calibrations depend on the samples you run. I have had calibration curves last months when running a clean matrix like volatiles by purge and trap. I actually had a 5970 that held calibration 18 months, you could run a calibration check and it would pass by all three of the last calibrations performed each 6 months. Currently I get about a month to two months on a calibration it seems. When running injected semi-volatile samples where you can have high hydrocarbon matrix effects I have seen a calibration blown in just a few samples, or last a month.

We all have to learn somewhere, and this is as good a place as any

[GRovian]

Excellent, thanks James.

The equipment it's a Thermo Scientific ISQ coupled with a GC Trace 1300, same maker. Agilent was way out of our price range.

Another thing that bothers me. Since the instalation day we have a annoying leak (I'll compare it Monday with the cal gas peak 69, as you said) and it's amplitude is changing if I change from splitless to split. Smaller in split, larger in splitless.

Something like splitless 1 mL/min will produce a 5E+7 peak 28 in Air Scan. A split 1 mL/50 will produce a 3E+6 peak.

To my limited understanding of the thing, that means the leak is BEFORE the column injection port, right ? No need to go crazy on the ferrules inside the oven, I reckon.

[Don_Hilton]

The leak is likely to be the inlet system! Diffusion runs both ways - helium out and air into the system. With the inlet set in split mode, there is a lot of gas coming into the inlet system and sweeping away air that diffuses into the system. In a splitless sytem you have only the helium that comes into the column, that which is vented through the septum purge, and leaks out if hte septum has not been chnaged and has become corred. And where the heilum passes into the air, you have some air coming into the system. With mass spec systems, you often run the inlet at a very slight pressure over atmospheric, so the flow of helium is slow enough that back diffusion of air can be a problem in checking for leaks.

A trick I do with my Agilent systems to reduce the exposure of my column to air, I use the gas saver with spitless injections and split at a high enough rate that the apparent leak goes away, so the only time my instruments are truly splitless is from the time they are ready for injection to when the purge valve opens in the injection cycle. I don't know if you can set up the gas saver the same way on the Trace GC, but you may want to look to see if you can do that.

The questions you have a a Noobie - are good in this discussion. I've seen some folks who have years of operating GC's behind them who should have asked some of these questions -- or at least learned the answers. I had operated GC/MS systems for many years before someone mentioned the diffusion of air into GC/MS systems through the inlet. I never let on that I had not considered this. A few weeks later I had an instrument that was showing a leak and happend to spray Dust-Off or a similar product across the GC septum and watched a peak for the hydrofluorocarobon rise and fall on the monitor. I did two things: 1) changed the seputm (wich took care of the air background) and 2) made a note that the young guy I was working with actually knew something.

[GRovian]

Great, don Hilton, many thanks. Since it never worked clean in the first place (but the trap looks clean) could also be a problem in the helium copper tubing, I suppose.

Speaking of septum, how many injection you allow before you preventively change it? 100? 500? I know at our GC/FID we changed once every 3 months, roughly 500 but I understand MS is a much more fussy detector.

Secondly, I remember the guy who install it was adamant that never ever to turn the helium off, or use the Gas Saver option, the most he agreed with was to use a 1 mL splitless when idle. When I asked about the reason he said he doesn't know but it was what the factory people, from Texas told him. Since we run just a few analysis a week and I have some contaminated helium going through, I would rather shut it off, when idle. What's the reason to have a minimal flow through the column/detector all the time?

As usual, thanks in advance for the kind souls out there willing to help a noobie.

G.

[Don_Hilton]

Do not cut off the flow to the column because if you do, the vacuum in the mass spec will pull air through the column and you don't want that. Air, heat, and a polymer are a bad mix - (The polymer does not take it well. This includes stationary phases.) Air onto a hot ion source is not a good idea either.

I don't go beyond 100 samples on a septum. This is not just the detector. If you have air coming into your system, you degrade columns. And, if you core a septum you have introduced an uncontroled split into your injectin. So, how many injections can you do through a septurm? It depends on the particular instrument, the septum and the analytical method. For about everything I do, 100 seems to be a safe limit.

[GRovian]

Aha, makes sense.

But since I'll have 3 idle days between each run, isn't acceptable to run a instrument method that will cut off the helium flow and bring down the temperature of the column to 40*C, as well as the heaters for the ion source to something like 50..60*C? This way there might be some air going in but the column will be cold (a PAH 5Si Restek), the source will be cold, what harm can there be? At least that's how I understand it, in my simplistic view.

Assuming I'm wrong (which is most likely) what idle settings would you use to minimise the helium consumption without sacrificing expensive equipment?

You're a Godsend, Don.

[Peter Apps]

I am in complete agreement with Don; do not turn off the helium. Do a simple calculation; what is the cost of the volume of helium that you will use on standby in three days ? Compare that to the cost of a column, the cost in time of the extra cycles of conditioning that the column will need before you can run samples and the cost in time of the extra pump down on the MS to get rid of the air and moisture stuck to its internal surfaces.

If you really must shut off helium when on standby, then put a two-way valve on the carrier gas input that will allow you to switch to HP nitrogen on standby. The last time I did the costing, the saving in helium would not even pay for the two-way valve.

Peter

[MSCHemist]

I am normally not happy with N2 or H2O above 5%. I normally keep it below 3% with the oxygen to nitrogen ratio indicating no leaks ( 28:32 should be 4:1 if air is getting in). If it goes higher than 3% I usually break out the 50% IPA/water and use that on all the lines and look for bubbling and a can of compressed air and look for M/Z 52 on the seals and connection on the instrument itself .

http://www.sisweb.com/index/referenc/tip19.htm

[James_Ball]

I would also agree with those above on not cutting off the carrier flow when idle. If you cool the column and MS down and you have an air leak, moisture will build up on the column and MS surfaces and it takes a while to get rid of them. There was another thread here not long ago about saving energy by idling instruments and I believe the conclusion was you really don't save much by cooling them down versus leaving them hot. Same goes for carrier gasses too.

For an extended idle period you might want to set your column flow to 0.5-0.3ml/min if the pressure controller can handle that low of a flow, with the column oven set about 105C to keep moisture from gathering on the column. My instruments only sit idle from when the Friday run ends sometime on Saturday until we start again on Monday, so my idle temp is 150C for the oven just to make sure any trace hydrocarbons don't hang around and any high boilers cook off over time, since these are purge and trap instruments the high boilers really aren't that high boiling.

On our GCs we normally change the septa ever few days. I did manage to get 300 injections on a septa once without any leaks but you have to keep the pressure from the nut very low so not to compress the septa and yet high enough to make it seal. Also keeping it cool extends the life, so if you are running a fairly high temp injector change them often.

The Dustoff works great for finding leaks, try it in the oven on the MS interface nut to make sure there are no leaks there, then back up through the system to make sure nothing comes in anywhere before that. You can also use it around the analyzer seals to check for leaks.

[lmh]

Just wanted to thank original poster for this thread; may it continue to be used. There area lot of jolly knowledgeable people here, and for those of us new to GC, this is a brilliant site.

[GRovian]

I am in complete agreement with Don; do not turn off the helium. Do a simple calculation; what is the cost of the volume of helium that you will use on standby in three days ? Compare that to the cost of a column, the cost in time of the extra cycles of conditioning that the column will need before you can run samples and the cost in time of the extra pump down on the MS to get rid of the air and moisture stuck to its internal surfaces.
Peter

Thanks, Peter. Point taken, the gas will stay on. Can I at least move to a splitless flow, when not in use?

I am normally not happy with N2 or H2O above 5%. I normally keep it below 3% with the oxygen to nitrogen ratio indicating no leaks ( 28:32 should be 4:1 if air is getting in). If it goes higher than 3% I usually break out the 50% IPA/water and use that on all the lines and look for bubbling and a can of compressed air and look for M/Z 52 on the seals and connection on the instrument itself .

5% compared to peak 69 of cal gas, right? I'm afraid I don't quite get what means to "break out the 50% IPA/water and use that on all the lines". Our installing tech was adamant never to use Snoope or similar liquids to check for leaks in the GC/MS pipeworks.

For an extended idle period you might want to set your column flow to 0.5-0.3ml/min if the pressure controller can handle that low of a flow, with the column oven set about 105C to keep moisture from gathering on the column. My instruments only sit idle from when the Friday run ends sometime on Saturday until we start again on Monday, so my idle temp is 150C for the oven just to make sure any trace hydrocarbons don't hang around and any high boilers cook off over time, since these are purge and trap instruments the high boilers really aren't that high boiling.

On our GCs we normally change the septa ever few days. I did manage to get 300 injections on a septa once without any leaks but you have to keep the pressure from the nut very low so not to compress the septa and yet high enough to make it seal. Also keeping it cool extends the life, so if you are running a fairly high temp injector change them often.

The Dustoff works great for finding leaks, try it in the oven on the MS interface nut to make sure there are no leaks there, then back up through the system to make sure nothing comes in anywhere before that. You can also use it around the analyzer seals to check for leaks.

Invaluable informations, James, I owe you my first born. A rather nosy pre-teen, btw. I think -based on what you say- I can devise a safe idle method, that can move the flow through the column down to 0.2 mL. What about the temp in the injection port? You suggest that low temp extend its life. Can I keep it safely at 105 when not in use?

2 more questions. There is a purge flow setting in the injection port settings in the method that can also be set as low as 0.5 mL, default set at 5mL. What would you suggest about it, on idle? Secondly, the Dust Off was also suggested by our tech, what peaks should I look for? Some websites suggest a value of 62 for dichloroethane, is that correct?

Gabriel

[Peter Apps]

Setting the inlet to splitless makes no difference to the total flow of gas through the inlet - with back-pressure control the only difference between split and splitless is the route that the gas takes through the inlet and its plumbing. Set up your gas saver to give a flow of 20 ml/min (any lower and you risk a leak robbing the inlet of pressure) and put the gas save on time (i.e. the time after the end of a run that the gas saver kicks in) to whatever you are happy with, and let the electronics take care of it.

I actually prefer to have the inlet on split during standby, then less septum bleed and air gets onto the column.

Trying to get all your gas flows down to a bare minimum risks having a small leak taking the total flow above what you have set the EPC to supply. The pressure in the inlet then drops and the GC goes into fault mode. Even with the high cost of helium fiddling about trying to save a few mls here and there is not productive.

I am a fervent advocate and hand held electronic leak detectors for finding leaks in pressurized systems, and for Dust-off or something similar (even dichloromethane vapour) for finding leaks into vacuums but looking for the ions on the MS.

Peter

[Don_Hilton]

For masses in Dust-Off or similar products. I never remember them. But the CAS number for the contents of the can is written on the can. And you can look up the spectrum by CAS number.

For low flows on standby mode in GC/MS systems - remember that point about diffusion working both ways. With my instruments, I keep a low flow and have about 35 mL/min going out throught the split vent ust to avoid getting the diffusion of air into the GC column. I've had an instrument waiting for me to get back to it for over a year. I've changed the helium cylinder once or twice.

[BMU_VMW]

Dust-off has mass 69 and 83 (at least the one we use here in Belgium)

The Thermo 1300-series GC's can be equiped with a He-saver module.
It is a very recent feature (about a month) that uses N2 for all split/purge flows and He for the injection and column-flow.
Since you are likely to have your system on standby for a long time it might be worth considering this option.