Chromatography Forum

considering the following parameters for controlling selectivity on a diol column under normal phase conditions, which would be recommended and in what order?
1) mobile phase composition
2) stationary phase
3) elution technique (isocratic vs. gradient)
4) temperature
Here is some information regarding my specific application.
A component of the sample matrix contributes several peaks to the 'landscape'. Some of these contributions lie between or among the analyte peaks. The separation has already been optimized for the analytes independent of the matrix contributions. Options to improve selectivity would appear to be few since the twelve analyte peaks elute between 12 and 52 minutes. The separation could be extended by weakening the mobile phase, although the current composition is 2% ethanol in cyclohexane and the runtime is about an hour on a 4.6 x 250mm 5µm column. The most significant interference occurs early in the chromatogram (14 minutes) suggesting that gradient elution is in favor. I've heard that normal phase gradients are prohibitive. I've obtained several diol columns and so far have only used one. I'm leaning against the gradient option and toward stationary phase and temperature. Please advise.

It seems as if the interfering components are of a single type, for example an oligomer of something from your matrix. Have you considered eliminating it using a sample prep technique (SPE, SEC etc.)?

The interfering component is an additive to the product. It is similar to the analytes in that it contains the same functionality and is similar in molecular weight. It also exibits similar solubility characteristics. The interfering component is a mixture containing roughly 70% of one compound and an array of other related compounds. It is these related compounds that have added an additional group of peaks that elute in the region of the analytes of interest. The primary component of the product additive elutes at 5 minutes, well before the analytes of interest. The additive is derived from plants and refined to achieve the 70% purity level. It is possible that the related compounds have different properties lending to SPE or SEC. Based on the additional information, would you suggest exploring sample cleanup further?

which would be recommended and in what order?
1) mobile phase composition
2) stationary phase
3) elution technique (isocratic vs. gradient)
4) temperature

No single answer. Pros and cons:

- mobile phase composition. EtOH is very strong. Biggest differences in selectivity will come from substituting a chlorinated solvent (e.g., CH2Cl2), an, ether (e.g., THF or MTBE), or an ester (e.g., ethyl acetate). You will obviously have to adjust the % strong solvent to get equivalent retention. The exact equivalence will vary from column to column; check with the column manufacturer to see if they have equivalence tables for your column. If selectivity changes so much that different sets of peaks overlap, you can always "interpolate" with a ternary blend. The major disadvantage is that many of the alternative solvents have a higher UV cutoff; whether or not this is a factor obviously depends on your particular sample.

- stationary phase. You could try a cyano column or (gasp!) bare silica. The latter is likely to give you the biggest selectivity change, but is likely to be very touchy about trace water content. Unlike mobile phase changes, there is no good way to "interpolate" if you see big changes, so this would rank near the bottom of my list.

- isocratic versus gradient. In general, a gradient will not improve selectivity for a specific set of peaks (unless the selectivity is k'-dependent). Since you do seem to be getting better results with a very weak eluant , a gradient may help combine good results and reasonable run time. The major catch to normal-phase gradients is remembering that equal-percent changes do not mean equal-strength changes. Roughly speaking (very roughly!)you get about the same strength change from 0-1% as from 1-10% as from 10-100%. That means you probably should run a multi-ramp gradient (shallow at the beginning, getting steeper with time).

- temperature. If you have a good temperature-control system, this is probably the first thing to try. It's easy to change, and you can interpolate if necessary. Biggest potential problem is transfer from one instrument to another, especially if they have different column oven designs.

OK, this suggests that SPE will be no good for you.

Tom Jupille's suggestions cover most of your options. I would also consider an amino phase as something drastically different from what you have.

As it is normal phase and the retention times are already long I guess some optmisation has already been made and I would try two 250mm 3µm column in series. That doesn#t solve the selectivity problem but increases separation power.

Alex

Alex,
Can you provide a source for 3µm diol columns?
Don

If you search back on this forum using "Diol", somebody asked the same question in 2005 and got the name of a firm that was distributed by Mac-Mod.

I'd be slightly wary of the pressures that coupling two such columns could generate. Pressure would be OK as long as mobile phase viscosity was kept relatively low and nothing dropped out. Having said that, if 500 mm of 3 um is all that does the job, then go for it!.

Bruce Hamilton

Just some more thoughts:

- Try selectivity by spectroscopic properties. Is it possible to use a wavelength that selectively monitors yuor active(s)?
- Solvent blends. A number of different solvent s is availible. additionally to the above mentioned I had good experiences with AC//iPrOH as modifier.
- There is also a wide range of NP colum chemistries: Silica, Diol, Cyano, NO2, amino....
- As you seem to have problem at short retention times, gradient elution could be useful (starting e. g from 1%). Solvents should be premixed.

There is no problem with back pressure when using 500 mm 3µm in NP, however pump seals need attention due to the high vapor pressure of the solvent (they are running essentially dry).

Alex