Method scaling to smaller ID columns

[Mike H.]

Simple question, but I can't find the answer anywhere. When transferring a method from a 4.6 ID to a 3.0 ID column and you calculate the injection volume for the 3.0 ID method, does this apply to the sample load or just the injection volume?

[tom jupille]

Assuming the sample concentration is kept constant, decreasing the volume will also decrease the mass load.

[Multidimensional]

To maintain linear scale, "the smaller the internal volume of the column, the smaller the injection volume. A change in the column dimension may require a change to the injection volume (note: "volume" and concentration are two different things. If the solution concentration remains the same and you inject less, the on-column sample concentration will also be less)".

Perhaps this will help? The article was written to address scaling issues with changing to smaller volume or narrower ID columns, but injection volume scaling is related to that too.

"HPLC to UHPLC Conversion Notes (Column Dimensions, Flow Rate, Injection Volume & System Dispersion)"

Found at this link - https://hplctips.blogspot.com/2011/07/h ... notes.html

[lmh]

to maintain identical method, I'd keep the concentration the same and reduce the volume in proportion to cross-sectional area. Many detectors detect concentration, not amount, so they'll give a similar signal this way.
But amount of analyte and injection volume might be limited by different things. Particularly if your injection solvent is strong enough to elute the sample then the injection volume is limited by the amount of solvent you can get away with putting through the column, and must be scaled down in proportion to the column size. The amount of analyte you can inject depends on the capacity of the column to bind the analyte, which is often enormously excessive, in which case there is no reason why, when you change from a 4.6 to 3mm column, you shouldn't reduce the volume but increase the concentration, and thereby increase the signal (in a typical concentration-dependent detector). [Of course there are exceptions where this doesn't work, for example when you're analysing something with ridiculously low solubility, which promptly drops out of solution in the cooled autosampler... but this shouldn't really be happening in analytical chromatography]