Internal Standard fluctuating response

[alpha]

For 3 weeks, I have been having trouble with the responose of heavy moleuclar weight compounds, in particular chrysene-d12 and perylene-d12. All my samples are spiked withe the same concentrations of the analytes, but the response can very tremendously from run to run.

For example: Run 1, 613782 response for perylene-d12.
Run 2 , 267107
Run 3, 241671
Run 4, 488199.

Does anyone know why this might occur? The MS source was cleaned and the filaments replace prior to this problem.

thanks for the input.

[alpha]

forgot to mention.....I have a total of 6 internal standards in the samples, but only chrysene-d12 and perylene d-12 seem to change from run to run by anywhere from 30-110%. The other 4 will stay relatively constant with no more than 10% drift.

[AICMM]

Alpha,

Are these clean solvent blanks or samples that you are having the variability with? What is your injector temp and type, what is your liner, and what is your column and flow?

Best regards.

[alpha]

These are samples that have variability, but i have had this happen while running my calibrations.

Method conditions have not changed. I am using a 6890N GC with a 6973 inert MS. Inject temperature changes with the method I am runnig, but it happens at both settings....275C and 300C. I'm using a J&W 30m DB-5MS column with a constant flow rate of 1.8 with a Hamilton silinized double gooseneck liner.

All conditions, liners, column, etc., types have not been changed. This problem only started in the past 1 1/2 months.

[Peter Apps]

If you are running PAH samples and you have not changed the liner for a month and half you will probably find that it is not as clean as it should be. Try a clean inlet liner.

Peter

[AICMM]

Alpha,

I would suggest you try a single taper liner with a little bit of deactivated wool to enhance the flash of the PAH's. A wisp of wool will make a huge difference and if you use too much some of your acid compounds will crater.

I would also recommend that you lower your column flow rate by about 1/2. I realize you will pay the price in time but it may make your reproducibility much better. Fast, useless, runs don't save any time.

Best regards.

[alpha]

Peter, by saying conditions have not changed, I only meant no new products have been introduced. For example, I have always used Hamilton liners and didn't switch from Restek. I do weekly, if not daily, maintenance on my instrument. Liners and septa get changed almost daily....column gets clipped ~ weekly, if not sooner. And gold seal is changed monthly, if not sooner.

AICMM, we have noticed that using wool can increase the tailing of some compounds. This becomes a problem because of Method 8270 witch has tailing criteria that has to be met. As for the flow rate, that is not an option at the moment. If we do that, it will put us weeks behind since new MDLs, calibration, etc will have to be done. that is a possiblity for the future when business slows down.

Thanks for the helpful advice.

[AICMM]

Alpha,

First, wool can wipe out some of the acids like nitrophenol but I have never seen it increase tailing. The key is only a wisp of wool.

Second, changing the flow rate should have nothing to do with the MDL.

Third, you have been trimming column, have you been updating column length in the method so the EPC can decrease the head pressure with decreasing column length?

Is your MS a turbo or diff-stack? If turbo, which turbo, the small one or the large one?

Best regards.

[alpha]

AICMM,

The previous analyst had used wool, and it did increase tailing as well as affect some acids. The tailing is probably partly to do with the amount of wool, he would use ~.25cm of wool in the liner. I may try that soon.

Changing the flow rate will not affect the MDL. But we do have clients such as DOD that state, if method conditions for analysis have changed, then a new MDL must be completed with the new conditions.

Column length is updated as it is clipped. For a while, I did forget to update the length, but it never really seemed to affect analysis.

Finally, the MS does have a turbo. As to large or small, I'm not sure. I'm still fairly new to the position (with limited training, since the previous analyst did not give notice) and learning on the fly.

Again, thank you.

[Peter Apps]

Hi Alpha

Sorry, I misinterpreted. I presume that you are working with an Agilent autosampler, what are the autosampler settings wrt injection speed, pre and post injection dwell etc. These can make a huge difference. Unfortunately the Agilents do not do solvent flush injections. What solvent are you using ? Am I right to presume splitless injections ? and that you have the ordinary (non PTV) inlet ?.

I have had similar problems with heavy PAHs when using a Gerstel PTV inlet in a 6890. With toluene as solvent I had to cool the inlet to 60 C and reduce the gas flow rate during injection to stop MW discrimination due to sample evaporating off the end of the needle.

Peter

[AICMM]

Alpha,


Since it is a turbo but you are probably doing EI instead of CI
, I am going to guess that your are using the small turbo. I would still bank on your having way too much flow into your MS. If it is the small turbo, it has about the same capacity as the diff stack and my experience has been that the diff stack started to choke at anything significantly above 1 ml/min, especially for semi-VOA. Basically, I think your heavy compounds are not being ionized well within the source since there is too much pressure in the source.

To prove me right or wrong, take a failed sample, drop the flow in 1/2, double the run time, and shoot it again.

Best regards.

[alpha]

AICMM,

I will definately try your suggestion when I get back from vacation.

Just one last odd observation: I analyzed 8 water extracts and had no significant variation with the d-12 chrysene IS. As soon as I put some soil extracts on, the IS started to drift again. The soils are extracted with a Dionex ASE 200, with 90:10 DCM/Acetone and then dried with Na2SO4 (this has been protocol for manyyears).


Peter,

It is an Agilent autosampler with a splitless inject. Our solven is DCM. Our analysis conditions have not changed in years. This problem just suddenly started after cleaning the source and replacing the filaments.


Thanks again for the input. I'll check back when vacation is over.

[AICMM]

Alpha,

One other thing that is easy to try, flip filaments and try again. It may simply be that the filament you are using is not aligned very well. Doesn't explain the difference between waters and soils though.... Soils are probably dirtier though, arent't they?

Best regards.