Poor Precision when using Poroshell column...

[Mike H.]

...that improves after 5-6 standard injections. Has anyone else experienced this? What could be the mechanism?

I'm seeing an RSD of 0.8% for the first 6 standard injections, but subsequent sys suit sets (6- std inj each) within the same run improve to 0.1% rsd. I've never experienced this in LC and would like to know if the solid core column require longer equilibration times? My typical run is 2 blanks, imp marker, blank, then 6 std injections.

[mattmullaney]

Hi Mike,

Do the peak areas steadily increase in value when the precision (% RSD) is higher in value? If so, it may simply be just a "need" for priming injections to be made.

[M Farooq]

Core shell particles have relatively low surface area so they should equilibrate faster than their fully porous counterparts. It seems that this observation is due to injector's and column's history. Is it a gradient method?

[danko]

The problem is likely to be as MattM indicates in his post.
Need for saturation the column with sample relevant substance/s is related to the stationary phase chemistry, degree of derivatisation and the nature/chemistry of the sample. It has nothing to do with the particle size and shape.

Best Regards

[Mike H.]

It is a gradient method. The peak are appeared random. At first I thought maybe the peak was so sharp that I wasn't capturing it all when using a 5Hz sampling rate. This isn't the case however. I've never had to prime an LC column. How common is it to have to do?

[danko]

It's common with some substances and stationary phases and not so common with others.
Do you see increase of the areas for each injection - as MattM suggested?

[M Farooq]

It is a gradient method. The peak are appeared random. At first I thought maybe the peak was so sharp that I wasn't capturing it all when using a 5Hz sampling rate. This isn't the case however. I've never had to prime an LC column. How common is it to have to do?

I am glad that you mentioned the sampling rate. Sampling rate will affect your precision of retention times for ultra-high efficiency core-shell columns. 5 Hz is typically below the Shannon's criterion for sampling data for a high efficiency chromatogram (e.g. 10,000 plates). In gradient, I am expecting even higher "plates". Do you have an Agilent instrument?

We have recently published on the nuances of sampling rates for fast poroshell columns. See figure 3 here. The article is open access
http://pubs.acs.org/doi/abs/10.1021/acs ... em.5b00715

[mattmullaney]

Good Paper, Thank You M. Farooq!

This link is one I've used, similar to the paper in a few respects only, but discusses proper sampling rates and time constant selection for high efficiency stationary phases (such as the core shell type)...as well as making plumbing modifications to the LC if needed:

http://mac-mod.com/pdf/technical-report ... uceECV.pdf

Hey Mike H., did your problem get cured?