Effect of column i.d. on efficiency

[HW Mueller]

Darn, above I skimmed over some of the contributions and thought the discussion is about d in the mm range. But now I have another problem: Some experiments in this lab failed to show any evidence of laminar flow at 0.1 mm (PEEK or steel capillaries, no filling), while 0.8 mm steel or PEEK capillaries showed strong evidence for laminar flow. Thus, I don´t understand how plug flow (CE) is any diffrent, in prctical terms, than normal flow at or below 0.1 mm d.

[Uwe Neue]

DR: the loss in in these parameters is really small compared with the gain in analysis time. In addition, I can argue that you do not loose, but rather gain by the use of a small column. You can run a gradient at 2 mL/min on a 3 cm x 2.1 mm column over 1 minute and get a peak capacity of 60 to 80 with 3 micron particles, which is plenty for many separations. The peak width is under these circumstances around 30 microliter, so you want to use a microbore flow cell. This way you may sacrifice a bit in absolute (=detector) sensitivity, but you gain a factor of 25 or so compared to a 15 cm x 4.6 mm column just based on column volume. Thus you actually have a larger sensitivity ("on column") with the fast separation and the small column. (25-fold gain from the column volume with a maybe 4 fold loss from the detector makes it a 6-fold gain in "on-column" sensitivity.) Precolumn bandspreading goes away, since you are running a gradient. With modern instruments, you can get around the gradient delay volume with delayed injection, where you gain in total cycle time by using the pre-column volume of the instrument for column reequilibration.
So it can be fast and sensitive...

[DR]

In theory, this sounds good, but in practice (where I practice) - I think we would lose some range because either flow cells are too large or there is too much extra column volume - I think I'm stuck until I can afford "modern equipment".

Analyzing pharmaceutical stability samples requires a large dynamic range and capacity (load column w/ lots of API and look for <0.1% w/w related substances, while keeping the main peak on scale). Going to low ID columns with any of my current equipment would increase my propensity for overloading the column or I'd start missing smaller things due to extra column effects, relatively large bore flow cells etc.

In theory and in practice are two different things... I would hazard a guess that this is why UPLC is making a bigger splash than microbore focused equipment has.

[Uwe Neue]

Well it is not only theory, it depends also what you are trying to do. We first got into these ultra-fast separations for drugs in plasma with MS detection. But you can also do fairly decent work in 5 minutes with a short 4.6 mm i.d. column packed with small particles and a well designed gradient at a flow rate different from the standard 1 mL/min (= higher flow rate). You can now use standard equipment without any headaches.

[adam]

From a previous post:

"The one point that I find confusing is that the less dense packing should have a cost in terms of mobile phase mass transfer?"

Does anyone have an answer to this?

[Uwe Neue]

The packing density is lower than in a standard packed bed, therefore there is more mobile phase and therefore the diffusion distances are longer. I doubt though that this effect is large.