HPLC columns C8 versus C18

[krummel]

Hello

I have to do a separation of two compounds with a eluent constisting of 90/10 H2O/Ethanol. The ratio of the eluent is very important for following experiments and I can not change it.
With the normal C18 column I got a separation with 70/30 H2O/Ethanol and I see the first peak after 17 min. When I decrease the Ethanol part to 10% I see the first peak after 30 min.
Can anyone assess if the separation is possible with a C8 column when I use a eluent ratio of 90/10 H2O/Ethanol???

For the first test I used a Bondclone C18, but this column is not available as C8. What can I use instead.

Thanks!

[Mark Tracy]

There are several things you can do to shorten the runtimes:
1. Raise the flow rate. The efficiency will decrease somewhat with a 10µ particle size stationary phase, but just try it and see what happens.
2. Raise the temperature. This effect is not so dramatic, but the solvent viscosity decreases, and you can combine it with option 1.
3. Switch to a C18 column with smaller particles, and shorter length. A 3µ column about 100 mm long will have similar resolution to a 250mm 10µ column.
4. Use a C18 column of similar dimensions, but with a lower surface area and/or larger pore size and/or lower % carbon.
5. Use a C8 column of similar dimensions
6. Consider using a polar-embedded or polar-endcapped column. Your mobile phase is 90% aqueous, and these type columns often work better than standard C18 under such conditions.

Many vendors sell fine C18 and C8 columns, even Dionex. You can research the technical specs for things like pore size, % carbon, surface area.

[dougs]

I would definitely look into using a C8 column. I recently began analyzing a compound that has a peak at around 28 minutes with a c18 column and 70/30 methanol/water mobile phase. By switching to a c8 column I've reduced the run time to 9 minutes. The mobile phase remains the same. As well, my analyte has a degradant peak that follows right after the analyte peak. The separation between these peaks remains fine. That's not to say that you won't have separation issues.

There is a fair amount of literature on the analyte I'm looking at, and they all use a c18 column. I suppose people prefer to use the old trusted c18, instead of using something more suitable.

DS

[Mark Tracy]

I agree. C8 columns are under appreciated. I find that they sometimes even have better base asymmetry than their C18 counterparts.

[tom jupille]

It's hard to generalize.

A couple of years ago we put together a paper for Pittcon based on selectivity and retention data for around 300 reversed-phase columns. No matter which way we looked at it, there was no statistically significant difference between C18 columns and C8 columns; the amount of variation within each type was larger than the difference between the average of the types.

If anyone's interested, e-mail me and I'll send a pdf of the presentation.