Veba,
SIELC columns are compatible with 100% organic and 100% water so you can do any type of gradient - single, double or triple (each variation is a parameter - amount of ACN, amount of buffer and buffer pH)
In all gradients you go from weak mobile phase composition to strong mobile phase composition. In some case you need to consider some of the unique qualities of mixed-mode columns when you create gradients. Since the column you tried have polar embedded groups and ionizable groups you need to evaluate the end point of gradient. It is easy to go from mixed-mode/anion-exchange (low organic) to HILIC/ion-exchange (high organic). When you move from RP/IE to HILIC/IE you mobile phase becomes weaker and elution of compounds slows down. You also need to consider the fact that when people use only two channels they don't think about two interactions - RP and ion-exchange. We saw a lot of cases when method on mixed-mode column called for gradient of 100% of water to 100% of ACN and only channel A had ions. So effectively by going along with your gradient mobile phase was depleted of ions and elution of ionic compounds slowed down or completely stopped and compounds got stocked on the column. For all practical purposes I would recommend to start with the following gradient:
A: 5% ACN with 5 mmol ammonium formate pH 3
B: 70% ACN with 30 mmol ammonium formate pH 3
go from 100% A to 100% B in 10 minutes, hold 5 minutes, equlibration of 8 minutes (for 2.1x150 mm column). method can be shorter for shorter column, higher flow rates and stronger mobile phases.
If you retention is too short decrease amount of ACN in both channels and amount of buffer in channel B. You can also explore various pH and acids (formic and acetic for LS/MS)
we have few tools which help you with buffer selection and method development:
http://www.sielc.com/upload/file/pdf/SI ... t_2009.pdf
http://www.sielc.com/Technology_MethodDevelopment.html
If you let me know exact nature of your compounds I will advice you on column selection and the gradient to try.
P.S. example above is for RP/ion-exchange case. If you are doing RP/anion-exclusion your buffer concentration in initial point is higher than at the end point of the gradient.
P.S.S. For UV you just use UV transparent mobile phase additives