Poor Peak Area RSD

[wrighty]

Hi there,

I am currently working on a project to determine the purity of glycerine (glycerol). I am a bit confused with some data that i am collating.

I am running a sequence of 6 samples which comprise of a 0.5mg/ml solution of glycerol and DEG in water. 0.5ul is then injected onto a ZB624 column.

the method that i am following says that the RSD between injections must be less than 15. My RSD's are ranging from 16 upto 50 which is terrible!

i checked the instrument over and i other work carried out on it produced an RSD of less than 1 so my problem lies in the sample prep.

is there anyway that the concentration and injection volume would be responsible for such poor RSD's? Also in the run of six samples, i get one poor result which throws the set out, is this indicative of any common problme?

thanks in advance for your help.

steve

[Peter Apps]

Hi Steve

When you say 0.5 ul is injected onto the column, do you mean that you do an on-column injection, or are you doing an ordinary split injection ?

Split injections of water samples are notorious for poor repeatability, mainly because the water produces such a large volume of vapour that there is a huge pressure surge inthe inlet and sample goes in all directions (i.e. out of the septum purge, back down the gas lines and out through the split). Is there any other solvent you could use ?

Presuming that your repeatability is for replicate injections from one vial it cannot be due to sample prep. If it is not replicates from one vial please run a set and post the results - that will tell us whether the problem lies with the GC.

Peter

[Schmitty]

This sounds like the USP test requirement for Glycerol. Are you using the inverted cup injection port liner, as recommended? Is your split at the specified flow?

[chromatographer1]

Steve,

I don't know if you are making manual injections or not.

If you are, then inject slowly. If you are using an autosampler, use the slow injection speed. If possible use a small plug of water or methanol behind the sample plug in the syringe.

Your problem is probably due to flash as Peter described. If your pneumatics are contaminated it may take some cleaning and drying out to fix your problem.

Active sites will absorb and desorb your analyte causing the large RSDs you describe. Get everything clean, inject your sample slowly to minimize flash, and try to use a wash plug behind your sample plug in your syringe and everything will work assuming your column is inert and is installed in the correct location in your injection liner.

Make sure you have the proper injection temperature too.

best wishes,

Rod

[wrighty]

Hi guys,

yes it is the USP method. Everything is set as detailed in the method. I ran six injections form the same vial and these are the obtained peak areas:

DEG GLYCEROL
546.04431 500.11264
557.05927 512.81818
570.03204 527.08008
368.09897 367.81671
543.38495 504.49762
599.5589 567.85663
mean 530.69640 496.6969767
stdev 82.24 67.72
RSD 15.50 13.63

the split is at 10:1, the only variation from the method is that the carrier gas is hydrogen rather than helium.

cheers

steve

[chromatographer1]

Steve,

There is a big drop in area in that one injection. Later, there is a larger area injection. That looks like a flash problem. However,........

Could you not be getting a good sample with your autosampler each and every time?

Are your caps so tight you could be pulling a partial vacuum during sampling?

Could you be using a worn needle in your autosampler?

Could you be getting a partial blockage of the needle due to particles of septum?

Do you need to replace your septum?

Good luck,

Rod

[wrighty]

thanks for the pointers rod.

the injections are all from the same vial so i would eliminate the cap.

the septum has been replaced (this is done regularly as general housekeeping) and the syringe is new. i have used varying syringe sizes and the problem keeps occuring.

this flash problem is interesting though. im about to prep up samples using compatible samples and allow them run over night. hopefully i should have interwsting results in the morning.

Will post the results.

Thanks all

steve

[AICMM]

Wrighty,

Does glycerol and DEG elute fairly late or fairly early on your ZB624 column?

Best regards

[Peter Apps]

If I understand right, you are diluting the glycerol sample in water. Try making the solution 10 times stronger, and make the split ratio 100:1. This can sometimes help to make injections more robust to vapour volume.

On the partial vacuum in the vial - I had exactly this problem yesterday (or rather the trainee did !) - six injections from one vial (with a lot of sample used for washing the syringe) caused a vacuum in the vial so the syringe was not picking up the same voluem each time. rsds of peak area and area ratios were ten times worse ! If you do multiple injections from one vial put the top on loosely so that air can get in as sample is removed. With you samples you will have no problem with evaporation.

Peter

[Peter Apps]

If I understand right, you are diluting the glycerol sample in water. Try making the solution 10 times stronger, and make the split ratio 100:1. This can sometimes help to make injections more robust to vapour volume.

On the partial vacuum in the vial - I had exactly this problem yesterday (or rather the trainee did !) - six injections from one vial (with a lot of sample used for washing the syringe) caused a vacuum in the vial so the syringe was not picking up the same voluem each time. rsds of peak area and area ratios were ten times worse ! If you do multiple injections from one vial put the top on loosely so that air can get in as sample is removed. With you samples you will have no problem with evaporation.

Peter

[wrighty]

well i've had some interesting results with this work. Firstly, by increasing the sample concentration and split ratio, the RSD dropped to 10 (against a maximum of 15) which was encouraging.

However, if i used methanol in the original concentrations/ split etc., i achieved a RSD of 5.

Thanks for all your help-it does look as though the water vapour issue was the cause of my troubles and using methanol is the way around it.

cheers

steve