Tekmar Velocity XPT?

[bigwalt]

Hello all,

Has anybody had any problems running a Velocity XPT? Specifically, linearity/active site issues in GCMS? Just wondering if I'm the only one...

BigWalt

[Bigbear]

Yes. Do you have the glass lined "blank" trap? Mine went active after a year or so. Swap it for the silcosteel one. Also check the FFC, they say they can be cleaned, but........ I did have luck bypassing the FFC and blank trap with 1/16" silcosteel tubing from the trap straight to the 6 port valve.
What method are you running? What GC system? If Agilent they are working with me and 2 other sites to get EPA 524.2 working better.
Are you very sure your problem isn't water. I chased those very same problems for about a year and swapped my Velocity for a Stratum. I was getting the same activity problems with the Stratum. Turns out it was all the water trnsfered to the ms activating the source. Are your problem children the substituted benzenes plus naphthalene? Try cleaning your source and re-try your linearity.

[bigwalt]

thanks for the advice, BigBear, but I've narrowed the problem down considerably...it may still be water, but I doubt it.

step one, do manual injections into the front of the Velocity XPT, to rule out problems with the Solatek. Not as easy as it sounds, at first, because the stupid thing is rigged only for automated sampling. So...tell the solatek to shoot 0mls of sample from vial #1, with 0uLs of standards 1 and 2. It goes thru the motions, and dutifully dispenses no sample, allowing me to manually inject the sparger old-school style. I reproduced the non-linearity problems, and the Solatek was ruled out.

step two, manual injections at the GC (6890) injection port. The expansion volume of methanol at 200 C is about 481 uL, so I can safely inject 1 uL of high concentration standards. Beautiful peaks, great linearity, problem must be upstream of the GC injection port...logical, right? That rules out the GC and the MS.

Therefore, it is somewhere between the fritted sparger and the GC injection port, most likely the transfer line, the sample mount, or the FFC...(or all three of them)

step three, install a Top of Trap injection port on the Velocity, and inject standards during the desorb step and during the drypurge step. Results were surprising, but I'm pretty sure I figured it out. The TOT port has a rubber septum, which bleeds like any other septum, and is apparently bleeding enough so that it actually covers the active sites. The septum bleed compounds are heavy enough that they don't interfere with the analytes, and the linearity of the system is quite acceptable, for now. In the meantime, a new transfer line is on the way, and I can flush the sample mount and related tubing with hot methanol...


more on this conundrum as it unfolds...


On the other hand, please do tell me how to bypass the Forward Focusing Chamber - that thing just seems like a bad idea to me, creating more problems than it solves

[Bigbear]

I still say it's water. I have done everything you are doing plus replaced my Velocity with a Stratum and the only thing that fixed the problem was a source cleaning. When yoou run out of options do a source cleaning.
On my system I built a Teklink method without an autosampler. This allows you to purge manually.
Try shooting the same amount of your IS and surrogates through the injection port. You will have to make a more concentrated solution so you get the same mass in 1 ul as you transfer into your Velocity. You may be masking the active site if you shoot all your compounds.
I feel your injection port is to hot. i use 150 ( the same as on my P&T). I'm contemplating lowering those temps to 120.
I bybassed the FFC by replacing it with a short piece of 1/16" silcosteel tubing.

[bigwalt]

BigBear said; "Try shooting the same amount of your IS and surrogates through the injection port. You will have to make a more concentrated solution so you get the same mass in 1 ul as you transfer into your Velocity. You may be masking the active site if you shoot all your compounds."

did exactly that, and the responses were beautiful for both test solutions, the standards only sol'n and the sol'n with all the analytes...but you may still be right. it may be a water problem. The test injections at the GC injection port were both free of water, so if water is the problem, it would not have shown during those test shots.

the injection port temps have varied over time, with no effects good or bad. I feel like it isn't really an issue at this point, and if you replaced the Velocity with a Stratum, and you still have the same problems, then I can hold off on that purchase order. You just saved us 15 G...thanks!

as for the source, the analyst before me would break down the source in a heartbeat, and it would seem to work a little better, but we shouldn't have to clean the source twice a week...that just ain't right. On my SemiVols instrument, (I know, its a different system for a different method) my source will look like it has been dipped in chocolate pudding and it will still run just fine, tune checks are great, responses are linear and repeatable. I just don't think we should have to clean the source after every sequence...but I have been wrong before on more than one occasion

I dunno, we'll see how it works out. I miss the old days tho, when a 16-place OI purge and trap system was considered cutting edge...

[bigwalt]

Hey Bear,

I have an idea. let's set up a sequence of DI water samples, to be run like any other samples, but no internal standards or surrogates. Just purging and trapping DI water...and when the system hits "desorb", we do a direct injection into the GC injection port. The analytes bypass all of the potential active sites in the purge and trap system, but the normal moisture load is still present. Comparing these shots to direct injections without purging will show the effects of water vapor being pushed into the MS. That way, we will know if the water is directly screwing up the detector, or if it is playing havoc with the transfer lines, or if it has no effect at all...

I think I'll try this tonight.

[Bigbear]

How did it go? You can use your IS/surr solution to try different paths in your Velocity. Install the injection port between the trap and 6 port valve. Shoot during desorb, then shoot during dry purge. The first shot tests the transfer line and 6 port, the next tests the trap and all plumbing to the trap. Traps can go active ( its happened to me).
What condition is your sample mount? I would change mine every 12-18 months and it would be very "rusty" above where the sparge vessel sits.
Are you doing 524? What are your Velocity and GC parameters? Are you using a Vocarb 3000 trap?

[bigwalt]

Hey Bear,

My brilliant idea went absolutely nowhere. When you do a direct injection, no purge and trap, you get one set of pressures and flows to deal with. When you inject during the desorb phase, you get a completely different set of pressures and flows...the inject-during-desorb peaks were about 10 times bigger than the regular-injection peaks - oh, well, it was worth a shot! (get it? sorry, bad puns are just part of my charm).

I inherited this mess from 2 other analysts. I offered suggestions when they asked; they would ask, I would offer suggestions, they would ignore them and keep doing what they were doing, getting the same results over and over. Now it's my turn.

methods 524 and 8260

Agilent 6890N, running a 5975b, with solatek and Velocity XPT

Vocarb 3000 (we have swapped out all sorts of different traps - no luck)
11 minute purge, ambient temp
drypurge from 0 minutes to 6 minutes...no real changes.
4 minute desorb at 260C

I forget the rest, cuz I've only had one cup of coffee so far, but we've tried all sorts of temp variations, flow variations, etc...no luck.

I also tried injecting into the top of trap port I installed, but there was so much septum bleed that it made the results inconclusive. Now that the septum has had a week to bleed off and settle down, I will try it again.

here's a question for you, Bear. Of all the things you've tried, what has been the most helpful, and what has helped not-at-all?

another question - how big is your rough pump? The very first thing I noticed about this system when they installed it was the puny looking rough pump. I can't help but wonder if there is something going on there. with fluctuations in the vacuum state during desorb, causing unexpected results. I think we have an old rough pump somewhere, maybe I'll hook that sucker up and see what happens...

[Bigbear]

All the rough pumps for the 5973/75 are the same Edwards 1.5's.

The best luck I've had is to limit the water getting to the source. If I were you I would try cleaning your source and then try a calibration.

My parameters are:

EPA 524.4 using a db-624 20X 0.18X 1.0. I use 0.6 cc/min and an a 75:1 split ( purge 25 ml). I desorb for 3 min and at 1.5 min I set the gas saver to go to 150 cc/min ( to dry the transfer line and injection port) . My GC program goes to 230C and I hold it there for 10 min ( again to help get the water out).
On my tekmar I desorb@250, bake it for 6 min @ 260. I dry purge for 6 min @ 40cc/min. I use a Vocarb 3000 trap and have never used any of Tekmar's water management traps! Just their blank traps. As I said earlier I ran a 1/16' silcosteel line from my trap directly to the 6 port valve. I ran it through the blank trap to keep it heated.
To answer your last question, what has helped me most was cleaning my source! I now have to clean it every 100-200 samples. I just put in a 6mm draw out plate that my Agilent experts say helps. They are also working on an alternate manual tune in which the emission current is maximized. It seems to work but I don't feel up to typing it all here. When they are satisfied the water management issue is dealt with there will be an ap. note published.
Different flows with direct injection? How is your Velocity plumbed? Do you have it in series with the gas supply to the GC inlet? If you do there should only be a short pressure blip as a result of the 6 port valve cycling.

I don't know that I have been explaining myself clearly enough from re-reading my earlier posts but.....the problem with water is , it seems to activate the source. The only "fix" is to clean the source.
My fingers hurt!
Later
Bear

[bigwalt]

muchas gracias, Bear

It seems to me that you have done pretty much everything that an analyst could be expected to do, trouble-shooting wise. I am coming to the conclusion that there is simply a design flaw somewhere in this system, and when (if) Agilent ever figures it out, then we'll get some relief. Its funny, but I never seemed to have these problems way back when in the dark ages when we manually shot our standards into a 5 mL syringe and injected them into the 16 position purge system...ahh, the good old days! Oh well.

As for the plumbing, it is as the install guy set it up; the transfer line from the P & T goes into the injection port on the GC...I honestly haven't studied up on it enough to speak intelligently. I made the possibly fatal mistake of assuming the install dude knew his business I'm relatively sure there is a pressure pulse on desorb, and that is all that would be needed to change the injection characteristics - how else to explain a 10X increase in peak size, shooting the same amount of the same standard?

I dunno, I'm treating this as a learning and growth opportunity, and paying the light bill while I'm at it.

<whine> I guess the previous operator was right. the only thing that is gonna help is to clean the source, but I don't wanna clean the source all the time! I'm too busy, and the thing tunes just fine! I shouldn't need to clean the source! It's not FAIR!</whine>

hey, do me a favor....send me a hello email at TwoCents at cableone dot net. I'd like to email you some chromatograms, and get your opinions, and pick your brain, etc...thanks!

[bigwalt]

BigBear said, "You can use your IS/surr solution to try different paths in your Velocity. Install the injection port between the trap and 6 port valve. Shoot during desorb, then shoot during dry purge. The first shot tests the transfer line and 6 port, the next tests the trap and all plumbing to the trap. Traps can go active ( its happened to me). "

did this, and the results show that the purge and trap system is working as it should - no active sites, good recoveries for blank shots and standard shots. It is looking more and more like you are right, Bear. It is a water problem. The MS works great as long as there is no moisture present. Too bad there is always moisture present when you are running volatiles on waters...

One thing bugs me, though. Correct me if I'm wrong, but Agilent has been working on this problem for a while now? They know there is a problem with water in the MS, and they have known this for some time? Because we've had 2 analysts go absolutely bonkers over the last couple of years! We've had linearity problems since we bought the stupid thing, and they never once told us it may be a design flaw...

[Bigbear]

Don't know if it's a design flaw or not, I suspect it's a result of our wanting more sensitivity. The manufacturers make the sources smaller to do so. but there are unexpected trade offs. No I don't work for Agilent.

Hey Walt did you get my e-mail?

[Ron]

The fact that Agilent came out with a different optional source to increase the linear range is an indictation there is a design flaw in the original ion source in my opinion.

[bigwalt]

The fact that Agilent came out with a different optional source to increase the linear range is an indictation there is a design flaw in the original ion source in my opinion.

well Ron, that may be a valid way of looking at the situation, but if the old design would work for months with no downtime (and it does), and the new-and-improved design needs to be cleaned every 4 days, I'll stick with the old design if at all possible. Right now, I'm trying to figure out a way to hook up the new purge and trap gear to my old 5973 MS, and use the new 5975b for semivolatiles, which would not present a moisture control problem.

Just my $0.02

[Ron]

You're not the first persion I know of that bought a new Agilent GCMS for volatiles and after fighting it for a while gave up and switched the new instrument to semivolatiles and used an older instrument for volatiles analysis. It isn't a good situation, but I'm not sure if the volatiles market is big enough for Agilent to pay that much attention to, especially with the new focus on life sciences and LCMS.