Flow-rate modification in HPLC

[akshpolae]

Dear all,

There is a HPLC method in which Zorbax SB-C18, 2.1X50 mm, 5-micron column with Eclipse XDB-C18 4.6X12.5mm, 5-micron guard column was used. But, I am going to use Kinetex Phenomenex 4.6X250 mm, C18, 100 A column, with UHPLC C18 for 4.6 mm cartridge.

In the method for 2.1X50 mm column the flow-rate was 0.1mL/min, what should be the new flow-rate in 4.6X250 mm column? Is there a way to calculate this theoretically or should I do determine it experimentally through trial and error?

Thank you for your help and inputs.

Best,
Akshaya

[HPLC chemist]

You are going from a UPLC to HPLC column, so there is no way to adjust the flowrate but to develop a new analytical method!

[tom jupille]

Actually, simply scaling from an HPLC column to a UHPLC column of the same stationary phase chemisty is easy. There are a number on calculators available on the web, including one on our web site (http://www.lcresources.com/sandbox/UPLC-HPLC.xls).

When moving to a different stationary phase chemistry, HPLC chemist is correct; you will probably have to modify / redevelop the method.

[dap]

Actually, simply scaling from an HPLC column to a UHPLC column of the same stationary phase chemisty is easy.

Only for isocratic mode and not from UHPLC to HPLC!

[Rndirk]

Using a simple rule of thumb

(ID column 1 / ID column 2)² * flow rate column 2 = flow rate column 1

I would start developing using a flow around 500µL/min

[tom jupille]

Only for isocratic mode and not from UHPLC to HPLC!

Actually, the scaling is almost as easy for gradient as for isocratic; the "wild cards" for gradients are dwell volume and mixing washout. The flow rate scaling works equally well in either direction.

[lmh]

In this case I wouldn't bother doing a formal scaling.

I assume from the change of column that the OP needs high resolution and doesn't care about run-length. They haven't given the particle size of the new column, but switching to big particles in a long column is typical for high-resolution methods(*).

If the aim is to get the highest resolution, obviously we want the lowest point on the van Deemter curve, and we're balancing band-broadening effect of general diffusion (worse at low flow) against the band-broadening effect of diffusion in and out of pores (worse at high-flow). Kinetex is a solid-core column, so diffusion in and out of pores is reduced, meaning that high-flow is generally good. 4.6mm Kinetex columns aren't rated to high pressure (2mm columns are), so it's likely the limiting factor will be pressure; I'd go for a fairly high flow, maybe 1mL/min, and see what it looks like...

There's certainly little point in scaling directly between these two columns, as you wouldn't make that column change unless you wanted to change the performance of your chromatography, and you may as well take advantage of the change to optimise for what you need.

(* small particles give short run-time, not high resolution. Halving the particle size doubles N but quadruples pressure, so if you're limited by the pressure your system can handle, to maintain constant (maximum) pressure, you need to reduce the column length 4-fold, so you halve N overall; the advantage is that the method gets 4-times quicker at a cost of only root-2 loss of resolution).

[akshpolae]

Dear all,

Thank you for all your inputs and suggestions. I started out with 1mL/min and I am currently running at 0.8mL/min (operating pressure is around 100 bar).
I have set up method conditions same as the original method(temp 40 deg C, injection volume: 20microLit, Wavelength: 220nm), only with a different column, longer run time. I have a gradient that starts off with 10% organic and goes up to 50% organic (Acetonitrile).
Due to some reason, the solid-core column is not giving a smooth baseline. I have used solid-core columns before, and when there is a increase or decrease in the percentage organic, the base line is not smooth. Is it general when solid-core columns are used in HPLC? Also, there is approx. 3 min lag in the response time. Meaning, when the gradient increases from 10% to 20% at say 3 mins as per the method, I am able to see that response only after 6 mins.

[HPLC chemist]

The delay time is a function of your 'void' volume or in particular the amount of tubing between the injector and detector. Mine was 2 minutes.

An impurity from a previous analysis can be eluting off as you increase the amount of organic. You can reverse and wash the column with 95% IPA (no buffers) then return to your mobile phase. It's not peculiar to solid core columns, but they do having more mass than other columns.

[akshpolae]

Thank you. I will try washing the column and check if improves the baseline.