I have heard that normal phase gradients are best performed on appropriate bonded silicas (diol, nitrile etc)rather than bare underivatised silicas.

Has anyone experience of running normal phase gradients and can verify this, or provide any links to literature supporting (or not) this notion.

Thanks