What is best: longer column or smaller particles?

[Mattias]

Let´s say that I have a method that runs on a 2.1x50 mm column (1.7 µm particles) at 8000 psi. But I need more separation power without exceeding the 8000 psi.

These two options should give the same pressure (8000 psi) and retention time (?):

1. 100 mm column (1.7 µm) and reduce the flow by 50%
2. 150 mm column (2.6 µm) and reduce the flow by 33%

(the assumption here is that 1.7 µm gives twice the pressure of 2.6 µm)
Maybe it is obvious which gives the best plate count, but I am not sure!

[Rndirk]

Hi,

This link will be helpful.

The plate count is roughly equal to (Column length) / ( 2x particle diameter).


EDIT: not sure if this is a hypothetical question or a practical one. If practical, i'd first check if i can play with the gradient, flow and temperature to get the separation i need before considering buying a column with the same stationairy phase and other dimensions.

[Mattias]

Hi,

This link will be helpful.

The plate count is roughly equal to (Column length) / ( 2x particle diameter).


EDIT: not sure if this is a hypothetical question or a practical one. If practical, i'd first check if i can play with the gradient, flow and temperature to get the separation i need before considering buying a column with the same stationairy phase and other dimensions.

Thanks!

The question is hypothetical but has a practical background! From the formula I understand that the shorter column with the smaller particles would be the best choice in this case

But in a way I am still a bit confused. Let's say that I have a choice between a 100 mm (1.7 µm) column and a 200 mm (2.6 µm) column. According to the formulas these two options would roughly give the same pressure and plate count. But the run time be obviously be shorter on the 1.7 µm column.

So for any pressure limitation it would always be better to use a 1.7 µm column? Even if you have an old Alliance equipment, it would be better to use a 1.7 µm column (but maybe only 30 mm length) than a 150 mm column of 5 µm material.

It is a hell of a difference between 1.7^2 and 5^2!!

[Rndirk]

You're right the two columns you proposed have roughly the same plate count.

BUT smaller particles are better if your pump can handle the pressure. Check the Van Deemter theory, with smaller particles you can use higher velocities ~ without losing much efficiency.

This has to do with smaller diffusion lengths of your molecules inside the column with smaller particles. For the same length, flow, diameter of an LC column your peaks should be more narrow with smaller particles.

[DR]

Capacity can also be an issue as peaks tend to be narrower and taller with smaller bore column jackets. If you're doing an assay where your peak height threatens to get into non-linear absorbance ranges, you may have to adjust dilution or injection volumes to accommodate - or use a wider bore jacket, other things being equal.

[HPLCaddict]

Since this is a hypothetical question:
If your goal really is to get maximum efficiency with a given equipment (and, thus, a given pressure limit) then actually it's better, at least from theory, to use larger particles and maximize column length than using smaller particles. The reason has already been mentioned: Doubling column length yields double plate count and double backpressure, while halving particle size yields double plate count and 4-times the backpressure. Therefore, with smaller particle sizes you will sooner reach the pressure limit of your equipment. The downside, of course is, that longer column means longer run time. In practice, noone would use something like four 250 mm columns with 5µm particles in a row.

With your initial example, you introduced an additional parameter: the flow-rate. And since efficiency depends on the flow-rate (good old Van Deemter has already been mentioned) this leads to further complications. A 1.7 µm column may actually perform worse than a 2.6 µm column of the same length if it is used below the optimum flow-rate.

And just to complicate things further: When talking about UHPLC, we always have to consider the extra-column volume. A shorter 1.7 µm column may perform worse than a longer 2.6 µm column just because it is more sensitive towards extra-column band-spreading due to it's lower volume.

And with all this "plate counting" never forget: resolution includes the square root of efficiency - doubling the plate count only increases resolution ~40%.

[Mattias]

Thank you very much for your replies!!

It is not as easy as it may look in the column brochures

So the backpressure is so to say "the currency" here. You can use that to have longer columns or small particles. If run time is not an issue, longer columns will give more plates than smaller particles (for the same amount of backpressure)

In any case, it is a waste of resources to use a 600 bar pump at 50 bars. And these kinds of methods are still being developed...

The methods I usually develop try to use shorter columns (often 100 mm), small particles and a flow rate that gives about 75% of the max backpressure of the pump.

[HPLCaddict]

If one wants to go deeper into the theory, consider delving into the concept of kinetic plots. There's a very good introducing article by the late Uwe Neue available on LCGC:
http://www.chromatographyonline.com/kin ... ade-easy-1

[Rndirk]

Thank you very much for your replies!!

It is not as easy as it may look in the column brochures

So the backpressure is so to say "the currency" here. You can use that to have longer columns or small particles. If run time is not an issue, longer columns will give more plates than smaller particles (for the same amount of backpressure)

Yes and then there's also column internal diameter: smaller = more pressure needed, but less solvent.

The general trend with chromatography is miniaturization => less solvents, faster runs, miniaturized sample prep, more peak capacity,...