We are trying to separate two amines. One is a secondary amine and the other is a tertianary one. The only difference is that the tertianary amine has a methyl group on the Amino-group. Everything else is identical and both amines are very hydrophobic (at buffer pH 2.5 and 28% MeCN as mobile phase, RT was about 6 minutes on Zorbax SB C18, 4.6 x 75 mm, 3.5 um at 1.6 ml/mL). The resolution was about 1.7. It seems to be well resolved. However, it is a GLP project. I am afraid that the resolution could decrease during the long sample set (column contamination etc) or on columns of different lots. So I am trying to get a resolution of more that 2.5. So far I have tried Zorbax SB C18, CN, PH, Zorbax XDB PH and Zorbax Bonus. REsults: Resolutions were not better that 1.7. (MeOH, MeCN and THF have been tried combined with pH 2.5 phosphate buffer).

Has anyone got a solution for me?

Thank you very much for your help.

The standard approach that I am advocating in such a case is a drastic change in pH. I would go at least to pH 7, or even to pH 10. The best thing is to try both pH values. We have demonstrated exactly the kind of selectivity changes that you are after. Of course, you need a column that can be run under these circumstances and which gives you good peak shapes as well.

Please try with methanol/water
gradient methanol 60 to 80 % in 15 min.
column c 18 standart
write if all is ok.
gl.

I agree with Uwe on this one, at pH 2.5 the amino group will be ionised, not a good thing. Go to as high an alkaline pH as you can: CARE! silica starts to dissolve above pH7. You could try another manufacturer's column, Zorbax-based silica columns are very strange, as their manufacture is different to all others, sometimes they work better, sometimes not...y p y m...

no affiliation with any manufacturer etc.

ps tertiary
also, flow rate for 4.6mm diameter columns is optimum around 0.8 - 1 ml/min.
also.. calculate the k' of the peaks of interest, I think you may find they are a little high, aim for between 1 and 5, and not higher than 10
good luck!

Try reducing the flow. Two reasons: Improve resolution and increase plate count (efficiency). I guess at f=1.6 ml/min for Zorbax 4.6 x 75mm column, the plate should be less than 3000. I personally think it's too low.

Uwe's right, increase pH, protonate the amino group (more hydrophobic) and increase the % of ACN. Good luck.

to genhousun

gee, and all this time I thought you had to decrease pH to protonate a molecule.

Try a Kromasil C18 (or C8) column and adjust the organic level so k'<10. Kromasil can be run at pH of 9.5 (0.005% heptylamine) for extended periods without loss of resolution.

I've also had good luck analyzing amines using nonafluoropentanoic acid (0.02%, pH~2.5). This is an ion-pairing reagent however, so beware (volatile, so compatible with MS & ELS). If the column is dedicated to this one analysis, it may give you the increased resolution you are looking for.

Perhaps all you need is a longer column, say 15 cm?

Thank you all for the excellent suggestions. We have got excellent resolution at pH 10.

The original coumn is 3.5 um particle. We could not use a longer column. Otherwise the pressure would be very high. We are in the industry. We had to compromise between plate number (high flow) and run time.

Again, thank you very much.

We had sucessfully separated very structurally closed amines by using racemic camphorsulphonic acid as ion-pair reagent at pH~ 2 on HypersilC18/Luna C8/Inertsil ODS-2 columns.Feel free to contact me if you need any additional information.

Dr.Alex Weisman
Chemagis,
Israel