Chromatography Forum

I've just noticed that Agilent sells columns with very small particle diameters (1.8 um), but where the column has a diameter of 4.6 mm.

I always thought this was not possible. I thought with those small particle diameters you had to use smaller column diameters - due to the viscous heating.

Can anyone provide any insight on this.

Thanks

Adam- I am sure you are correct about viscous heating effects which are a negative point for larger diameter columns of small particles.

However, you also have to think that the effect of instrumental dead volume is much higher for small diameter columns- a positive point for larger diameter columns. Further, small diameter columns often have lower efficiencies than 4.6mm columns; I am not certain whether this is due to packing difficulties, or whether there is really some sort of "wall effect" which is operating. I'd like to hear any opinions on this subject.

I guess it is just a matter of which effects weigh heaviest in the particular situation (e.g. what instrument you have). I could see a situation where a 4.6mm column could be the best choice, although I doubt this would be the normal case.

Victor, the lower efficiency you’ve observed is most probably due to a sample and/or volume overload. I’ve seen it many times: People switch to a smaller diameter column, but keep injecting the same amount and/or volume.
Another possibility is that the flow rate is not adjusted in accordance to the column diameter, thus altering the linear velocity – here you might like to look at van Deemter’s plots for further clarification.

Adam, these particular columns are packed with a larger size- distribution particles, and this means that the interparticle- distances are not as small as in the case of Acquity columns for instance. Another consequence of both the larger size distribution and the column’s inner diameter is, that the pressure is roughly a third of what other columns in the class will generate, thus allowing the user to utilize them on traditional HPLC systems.
I’ve tested a couple of these columns and found them quite good.

Best Regards

The viscous heating indeed deteriorates the performance of larger diameter columns. The effect increases roughly with the square of the internal diameter (Mazzeo, Neue, Kele, Plumb, Anal. Chem. 77 (2005), 460A-467A). This effect increases with increasing pressure. If you stay at low pressure, you do not see this effect, but you may end up on the wrong side of the van-Deemter curve, depending on the analytes that you are using.

Today, with all the efforts that went into the development of the UPLC system, there is not an issue about an inferior performance of small diameter columns any more. In the old times, some of this effect was due to extra-column bandspreading, some to the surface finish of the smaller i.d. columns.

For standard small molecules, the minimum of the van-Deemter curve for sub-2-micron particles is close to 5000 PSI, and the thermal effects become measurable.

So if I run these 1.8µm 150 x 4.6 mm configuration on a "standard" HPLC I'm likely to go over pressure at flow rates between 0.5 to 2.0ml/min?

What pressures would I be roughly looking at? Would it be greater than 600 bar (Agilent RR)? If so it seems a bit strange for Agilent to produce a column that won't run adequately on their system even?

PS. Why can't the BEH Acquity columns be packed in a 4.6 mm id column format?

For a sub-2-micron column of this length, the minimum of the van-Deemter curve is at or above 6000 psi or 400 bar for most pharmaceutical molecules. In water, this is about 1 mL/min.

I have mentioned above that the thermal effects increase with the square of the diameter. We designed the columns and the system together to avoid excessive thermal effects. After all, we do not stop at 400 bar, but the UPLC system goes to 1000 bar.

With respect to your PS: To run a 4.6 mm column on an UPLC system would be the same as running a 10 mm column on an HPLC system. It can be done, but very few people do so.

I don't think these columns are strictly been marketed as "Fast LC " columns and as such can be run on standard LC's such as Agilent 1100 or 1200's or even the 1200 Rapid Resolutions.

However, to run sub 2µm colums with 2mm ID's will really require the Rapid Res system with the correct plumbing configuration to eliminiate as much dead volume as possible. The mobile phase compressability is also required for the binary pump of the RR to reduce baseline noise.

The standard ID sub 2µm colums are really a half way house to fast LC.

And dont even get started on the differences between "Fast LC" and "UPLC"!