Performance of 2.7 um vs 3.0/3.5 Columns as f(flowrate)?

[HW Mueller]

Why "the superficially porous stuff should win"? Due to the shorter diffusion path? What about the higher area of the fully porous material? Even with the examples given by Merlin it is difficult for me to see which is better, all around. If this shoulder is a 5th compound then the solid core is a deceiver here.

[mbicking]

I will let the theoreticians weigh in on exactly why, but I believe the shorter diffusion path is the reason for the better efficiency (smaller C term in the VD equation). And remember this comparison was only looking at efficiency, not selectivity. I do not know the source of silica for each column, so if the particles are made from different silica (which is likely) a difference in selectivity would not be surprising.

[Uwe Neue]

Merlin, I did not want to offend you, and I understand your limitations with respect to equipment. Since there are many different people that read the posts here, I felt that I needed to point out the weakness of this comparison.

[unmgvar]

mbicking

I have done some of the maths,
except for the longer column here, and the fact that we are talking about a 2.7u verses a 3u there is absolutely no added value in the separation power that could be attributed to the fact that we have a fused core or a fully porous silica. the difference is so slight that it could be within the error margin of the experiment

Also as HW Mueller said if you have done a 3D we would love to know if peak 2 is pure or not.

I have done some experiments on kinetex column and saw that the selectivity is different from porous columns in many cases but that works sometimes in your favor and sometimes against, as what we care in the end anyway is the resolution and not N and especially not N/m.
Also me and other colleagues are troubled by the fact that we were unable to do a robustness test with 3 different batches of kinetex.
this reminds me of the Merck column RP-select B. great column great selectivity, very poor batch per batch reproducibility

[Tom Waeghe]

I did a few calculations and am giving a try to posting an image of the PPT slide I made from Excel.

http://tinypic.com/view.php?pic=jzbm1s&s=6

Click on the image, click Raw Image link, and then adjust your IE to 150%.

If you'd like to get a copy of the PPT file (with references for inserts), let me know.

Occasionally, we get calls from customers who don't understand the concept that a highly efficient, low volume column is more susceptible to losses in efficiency due to extracolumn volume and dispersion. Waters has used the term "bandspreading"; other use band broadening, instrument bandwidth (IBL, Michael Dong, Supelco), etc.

All columns such as 1.7 um, 1.8 um, 1.9 um, and superficially porous columns need to be run on instruments with low extracolumn dispersion, and the smaller the column and/or the faster the flow rate, the smaller the ECV (uL) and ECD (uL2) must be, in order to get close(r) to the theoretical performance of those columns.

I believe Uwe has made the comments and reminded folks that you can't see the full performance of the UPLC columns compared to other columns unless you use an instrument with adequately low ECD. That is the true intrinsic value of the Acquity system and it's excellent design. All other instrument brands (even the brand new 1290) must be modified from their "as-shipped" state to get full performance.

Oddly enough, when I joined MAC-MOD Analytical and we were exploring how to get more people to be successful with Fused-Core columns, we only had an Agilent 1100 Quaternary system to work with. However, it can be retrofit nicely (even better if you can add on a faster than 14-20 Hz detector) following one of our http://www.mac-mod.com/tr/tr.html technical reports. We later discovered that, aside from the large delay volume (which can be reduced significant on binary systems), 1100 and 1200 systems with VWDs/MWDs can be plumbed with shorter lengths of 0.005" ID tubing more readily than those with diode array detectors.

If you'd like to get a copy of the PPT file (with references for inserts), let me know.

[Tom Waeghe]

I apologize for the misstatement ("misremembered") about the 1290 ECD in my previous post. Guiochon, Farkas et al. claimed that the modified 1200 was slightly better than the standard (as-is) 1290 they compared, but that even the standard 1290 could be optimized for lower ECD.

Journal of Chromatography A
Volume 1217, Issue 18, 30 April 2010, Pages 3000-3012

"Note that the extra-column peak variances measured for the optimized Agilent 1200 system are smaller than those observed for the Agilent 1290 Infinity system because this latter system was not optimized [in their studies] to deliver minimum extra-column band broadening."

"Strikingly, the data measured with the optimized instrument are very close to or even better than those measured withthe reference instruments such as the Agilent 1290 Infinity and the Acquity chromatographs. In conclusion, it seems a priori that the optimized Agilent 1200 instrument should provide the highest
column efficiency and should serve as the reference standard for assessing column performance."