Can one develop a method with a gradient procedure in which A is 100% water and B is 100% organic? From practical view, this is a easy and robust way since no operator needs to involved in the mobile phase preparation. I am not sure what the regulatory agency would think. Thanks for the comments.

No problem, unless you have buffers and you need to worry about solubility. Also, bugs grow in water. Therefore it may not be the best choice, unless you take care of this potential problem.

There can be difficulties which depend on your gradient. If you run a gradient from 5% to 10% organic, this may not work well because your mixer has to mix a very low flow with a normal flow. This is a problem for both a high-pressure mixer and a low-pressure mixer. For wide gradients, I don't see a problem.

I ran a validated method with water in bottle A, methanol in bottle B (no buffers in either), with a gradient from 5% to 95% B. The chromatography was stable, and the analyte was consistently separated from a closely eluting interferent. But, I was using a system comprised of two isocratic pumps designed for low flows (capable of flows down to 1 uL/min), configured as a gradient system with high pressure mixing. I would not attempt this on our conventional flow systems with low pressure mixing.

Although there isn't any "chromatography" problem going from 0% solvent to 100% solvent, there can certainly be significant operational problems with this. First, if you are using a low pressure gradient system it's not advisable to start at 0% of any component since proportioning accuracy is typically unreliable below 2% (regardless of what the manufacturer tells you). Secondly, the solubility of dissolved gases is significantly lower in water-solvent mixtures than it is in pure water or pure solvent so you will be much more likely to have outgassing problems and loss of prime if working under these conditions (even if the water and the solvent have each been degassed, off-line or online). If you really think you need to work over this range, you probably need to use three eluent containers with one containing a water solvent mixture.

I must disagree with this statement from the previous message "if you are using a low pressure gradient system it's not advisable to start at 0% of any component since proportioning accuracy is typically unreliable below 2% (regardless of what the manufacturer tells you)." I have Alliance systems (Waters) that proportion very well 0-100% (even 98), ask to see data from your Waters rep if you need proof. I do it all the time with no issues, although some columns dont do well under these conditions (100% water I mean). In my experience high pressure system are poorer at 0-100% at the low and high ends, particullally starting from 0%, pump wind up time under pressure is difficult to do. And as to what would a regulatory agency would think, they dont care, as long as the method is properly validated of course.