Biphenylpropyl Modified Silica: An Interesting Choice for RP Chromatography

Alkyl especially octadecyl modified silicas are the workhorses in modern RP chromatography. Their retention mechanism is mainly due to hydrophobic van der Waals interactions. Steric or polar interactions tend to play a minor role. However, some separations are difficult or impossible to achieve with these stationary phases. Therefore, surface modifications with orthogonal selectivity are of great interest. Sorbents for reversed-phase chromatography must have sufficient hydrophobicity and they have to be chemically inert. These restrictions exclude many functionalities. However, aryl ligands as hydrocarbons are hydrophobic and chemically inert [1]. Their ability to undergo p-p interactions affects their selectivity over conventional alkyl phases [2,3]. Silicas modified with phenylpropyl or phenylhexyl silanes show lower hydrophobic interactions and thus shorter retention times compared to octadexyl modified silica. To increase the hydrophobicity of aryl modified silicas biphenylpropyl ligands have been chosen for surface modification. As expected this stationary phase shows enlarged retention times and different selectivities compared to octadecyl phases. They seem to be an interesting alternative to conventional alkyl modified silicas.

Experimental

HPLC grade solvents were used for the preparation of the eluents. In the case of acidic analytes 0.2% trifluoroacetic acid was added to the eluent. The various test compounds were of reagent grade or higher purity and were purchased from several sources.
Depending on the polarity of the analyte, different eluent compositions are chosen, but for each compound only one composition is used on all columns. For comparison of the solvent systems with acetonitrile and methanol the proportion of methanol was chosen so that the eluent systems have a comparable elution force on the reference phase.
The HPLC equipment used in this work was a Vanquish and an Ultimate 3000 UHPLC system (both Thermo Scientific). The dead volume was determined by injection of a solution of uracil in water.
The following stainless-steel columns (125x4 mm) filled with NUCLEODUR® C18 Gravity, NUCLEODUR® C8 Gravity, NUCLEODUR® C18 ec, NUCLEODUR® Phenyl-Hexyl, NUCLEODUR® Sphinx RP and NUCLEODUR® p² are commercially available (Macherey-Nagel). NUCLEODUR® itself is a totally porous, spherical and high-purity silica gel with an average pore size of 110 Å and a surface area of 340 m²/g.

Results and Discussion

In this work, the behaviour of 86 organic compounds is investigated on different RP phases in comparison to the reference NUCLEODUR® C18 Gravity [4]. The presentation of the specific data and detailed HPLC conditions would be beyond the scope of this publication, but they are available from the author. The analytes are organic compounds from various substance classes with a molecular weight below 500 daltons. Only neutral compounds were used, but ionised forms of e.g. anilines or acids could be formed by protonation and deprotonation reactions during chromatography.
The capacity factors k’(i,c) of the individual compounds i determined on the investigated column c are compared with those of the reference phase NUCLEODUR® C18 Gravity (k’(i,ref)). This column is filled with a sorbent covered with a dense layer of octadecyl ligands. In Table 1 the used formulas are summarised.
Figure 1 shows a plot of the capacity factors of an octadecyl (NUCLEODUR® C18 ec) and an octyl (NUCLEODUR® C8 Gravity) phase with acetonitrile as organic modifier. Using these sorbents, the analytes interact mainly with alkyl groups of different lengths. The similarity of the interactions results in a nearly linear course but with different slopes.
Figure 2 shows the measured values of two phases modified with aryl groups using acetonitrile/water. The modification of NUCLEODUR® p² consists of biphenylpropyl groups, while phenylhexyl ligands are used for the second sorbent.
Figures 3 and 4 show the results of the same stationary phases using the eluent system methanol/water. The scattering of the measured points shows that the biphenylpropyl modification differs markedly from the selectivity of octadecyl and other alkyl modified silicas. To quantify this behaviour, it is necessary to convert the capacity factors (see Table 1). A measure of the differences in the interaction strength of a compound i on column c is the relative capacity factor k’rel(i,c). The average of this value over all compounds chromatographed on a column is a measure of the interaction differences between the investigated sorbent and the reference. The relative standard deviation RSD describes the variation of the relative capacity factors around the mean value. It is a numeric measure of the scattering of points in Figures 1-4 and describes selectivity differences compared to the reference phase. The calculated values are summarised in Table 2.