The reason for the variability in retention time in normal phase HPLC is the strong dependence of the retention on the content of highly polar mobile phase constituents, especially water. Even if no water is added to the mobile phase on purpose, small quantities will always be dissolved even in very apolar solvents. Furthermore, bare silica is extremely hygroscopic, thus water will be adsorbed on the surface. Hence, the water content in a normal phase system can vary significantly and is seldom controlled. Best results, e. g. reproducible retention times, are obtained when working with half-saturated mobile phases. A straight forward way to obtain a half saturated mobile phase is to divide the mobile phase in to equal parts. To one part (e.g. 500 mL), one adds 2-3 mL water and allows the mixture to stir for about 30 min. Thereafter the excess water is removed (separating funnel), and the dry part (500 mL) is added to the saturated portion.
Even if a half saturated mobile phase will reduce the time to reach equilibrium in the column to a large extent. The equilibration of a normal phase column, especially when dealing with bare silica, can still take hours. Furthermore, polar modifier stationary phases, such as cyano or diol, are generally much less prone to variations in the water content of the mobile phase and can be equilibrated faster.
Some examples of solubility data of water in organic solvents at 25°C.:
Ajax content requested but not retrieved...
Something went wrong. This is not a picture.