Bonded Phases - The Synthesis of Bonded Phases

The Synthesis of Bonded Phases

Although bonded phases are now sometimes prepared form polylmeric material the most poplar and commonly used bonded phases are based on silica. The subject of bonded phase synthesis will therefore deal with bonded phases based on silica.

The first matter to be considered in preparing a bonded phase is the selection of appropriate base silica. This will mean choosing a silica hat has a specific particle size and pore diameter. The appropriate particle size of the silica will be partly defined by the resolution required from the column that, in turn, will be determined by the efficiency required to effect the desired separation of the mixtures to be analyzed. However, the particle diameter will also be partly controlled by the instrument constraints, e.g. the maximum pressure the pump can provide or that the sample valve can tolerate. To illustrate the problem of particle size selection, let it be assumed that a certain high efficiency is required to realize the separation. High efficiencies can be obtained by using small particles but, as the particle size is reduced, the flow impedance of the column increases which requires higher inlet pressures to be employed. It is clear that the particle size can only be reduced to that point where the maximum pressure of the pump is required to achieve the optimum linear velocity of the mobile phase. At this point, to obtain more theoretical plates, the column must be lengthened and simultaneously the particle size must now be increased to allow greater column permeability so that it can operate at the column pressure range available from the pump. It is clear that some basic theoretical considerations will be necessary to arrive at the appropriate particle diameter. It follows that the choice of particle size is not simple and its discussion is outside the scope of this book; interested readers are referred to (12).

The choice of the pore size and surface area of the silica is also somewhat involved and also needs some detailed discussion. In general the surface area of a particular silica tends to vary inversely as the pore size, so the greater the pore size the less the surface area. It follows, that when the dimensions of one is selected, the other, to some extent, is also defined. The optimum bonded phase is one that has the minimum number of hydroxyl groups unreacted, as a result, the maximum surface coverage. It is understood, that due to steric hindrance afforded by the bonded moiety itself, only a fraction of the silanol groups can be bonded and little or nothing that can be done to avoid this problem. However, there are other reasons for incomplete silanization of the silica.

Some incomplete silanization always occurs as a result of the reagent molecule itself being excluded from the smaller pores of the silica. Exclusion can be a particular problem when bonding relatively large molecular weight materials such as long chain hydrocarbons on the silica surface. For example, the very popular C18 brush phase is made from dimethyloctadecylchlorsilane that is a relatively large molecule that will certainly be excluded from the very small pores of most silica gels. It is therefore, important to choose a silica gel that has a relatively large pore size (or large compared with the size of the bonding molecule), perhaps a silica with a mean pore diameter of 150 . Unfortunately, silica with a pore size of this diameter may limit the surface area of the silica to between 150 and 250 sq.m per gram that, in turn, will reduce the retentive capacity of the stationary phase. Nevertheless, such a choice of pore size will also ensure that the vast majority of the pores will be accessible to the silanizing reagent. In the general use of a packed column in LC, pores that are smaller in size than those of the reagent will be left available for water and the smaller polar to enter and cause erosion of the silica base. In addition, care should be taken, to be sure that the pore size is not chosen to be so large that the silica gel is mechanically weak as it may collapses under the high pressures used during packing or, perhaps, even during normal use.

An alternative approach to pore size control is to subject the silica to hydrothermal treatment. It will be seen, that this is particularly easy to carry out if a fluidized bed is employed in the bonded phase synthesis. Hydrothermal treatment removes (by solution) silica from the surface of large pores and subsequently deposits it in the very small pores which can eventually become blocked and, thus, impermeable. It is clear that this procedure achieves the object of removing the small pores, and consequently possible sites for future erosion and, at the same time, render the large pores even larger and more accessible to the silanizing reagent molecules. If a fluidized bed system is not available the hydrothermal treatment can be carried in an autoclave. In general, each silica will differ in its response to hydrothermal treatment and, as a consequence, the optimum temperature and reaction time period will need to be determined by experiment. This will initially entail a number of test runs that should be carried out over a range of operating conditions. From the surface area and porosity data from such test runs the optimum conditions can be identified.