rather hydrophilic materials. As both normal-phase (NP) and HILIC modes of LC are based on the use of polar adsorbents and non-polar mobile phases, many types of diamond (prepared without additional chemical modification) can potentially be applied in these areas.
In 2002, Patel et al. [97] were the first to publish a report on the use of MSND in NP- HPLC of organic com-pounds. The study used a 100 × 4.6 mm ID chromatographic column packed with 3.8 ± 0.1 µm MSND particles, with a surface area of 216 m2·g−1, using heptane as the mobile phase for the separation of ortho- and para-xylenes. Despite very low column efficiency (only 1300 N·m−1) and incomplete resolution of solute peaks due to tailing, interesting separation selectivity (α = kortho/kpara = 2.63/2.29 = 1.15) was observed for these positional isomers. No such selectivity could be seen with a bare silica column for these analytes in heptane.
Emelina et al. [98] used NP-HPLC for comparison of the adsorption properties of two MSND phases: the first purified with CrO3/H2SO4 mixture and the second gas-phase purified with ozone. Two columns (75–80 mm long, 2.0 mm ID) were dry packed with 3–4 µm sintered DND particles. The authors did not find any substantial difference in separation selectivity between the two adsorbents, but slightly stronger retention selected analytes was found for ozone treated MSND, which is in good agreement with data showing a higher
29 concentration of polar groups for this phase (see discussion in Section 1.4.1). The retention order for the following solutes was obtained, toluene < bromobenzene < o-dichlorobenzene < benzophenone < anisole. However, no actual separations were reported due to very low column efficiencies, of the order 630–790 N·m−1. Later, the same group used NP-HPLC for characterisation of the adsorption properties of hydrogenated MSND, although once more no separations were demonstrated [25]. Using benzylamine as a probe, the authors identified the presence of hydroxyl groups of varied acidity at the surface of hydrogenated MSND.
The first successful NP-HPLC separation was reported by Nesterenko et al. [7,26,61] in 2007, who used a 150 × 4.0 mm ID column, packed with carefully fractionated 3–6 µm MSND particles, and a bimodal pore size distribution, with a surface area of 153 m2·g−1. Separations of 8 alkylbenzenes and 10 dialkylphthalates were achieved using n-hexane/IPA (95:5, v/v) or pure pentane as mobile phases (see Fig. 1.7). Nesterenko and co-authors compared the selectivity of their diamond phase with that of PGC and conventional NP adsorbents, such as silica and alumina [26]. Obviously, hydrogen bonding plays a very important role in retention of polar molecules under NP-HPLC conditions, however results indicated the presence of a graphitic layer on the surface of the sintered ND, resulting in selectivity close to that of the planar PGC. As expected, selectivity of silica and alumina were very different to that of the MSND substrate (see Fig. 1.8).
The high concentration of very polar groups on the surface of MSND could result in an adsorbed water layer and, hence, be applicable in “aqueous normal-phase chromatography”, more commonly known as hydrophilic interaction liquid chromatography (HILIC). Fedyanina and Nesterenko [23,24] studied the retention mechanism of substituted phenols and benzoic acids on an MSND column using both methanol-water and acetonitrile-water mobile phases. In both cases the retention of solutes decreased with increases of organic solvent, up to 75–80%. Beyond this, any further increase of organic solvent caused an increase in the retention time for all solutes, as shown in Fig. 1.9, left. Interestingly, it was noted that the retention order of ionogenic solutes did not depend on the concentration of organic solvent in the mobile phase. Linear correlations for logk – pKa and logk – Cmethanol were obtained for both phenols and benzoic acids. The corresponding plot of logk – pKa for phenols is in Fig. 1.9, right. These observations indicate that the main retention mechanism in this system is in fact hydrogen bonding, though some hydrophobic interactions were also present. However, both selectivity and efficiency did depend on the nature of the organic solvent. In the case of water–methanol, the column efficiency was on average 1.3 times higher than in the case of a water–acetonitrile based mobile phase.
30 Fig. 1.7. Separation of 8 alkylbenzenes (a) and 10 dialkyl phthalates (b) on MSDN. Conditions: n-pentane, 1.0 mL·min-1, detection UV 254 nm (a); n-hexane/IPA (95:5,v/v), 0.6 mL·min-1, 30 °C, detection UV 254 nm (b). Analytes, (a): (1) 1,3,5-triisopropylbenzene, (2) 1,3-diisopropylbenzene, (3) tert-butylbenzene, (4) isopropylbenzene, (5) benzene, (6) toluene, (7) n-amylbenzene, (8) n-nonylbenzene. Analytes, (b): in order from di-n-decyl- (C10) to di- n-methyl-phthalate (C1) [61].
Fig. 1.8. Comparison of selectivity of MSND with silica gel, alumina and PGC [26]. The left side of the plot presents data obtained for polymethylbenzenes (data points 10, 11, 13, 16 and 18) and right side presents data for monoalkylbenzenes with differing chain length of n-alkyl groups.
31 Fig. 1.9. Indication of typical HILIC retention behaviour for derivatives of phenol on MSND column with an increase of methanol content (left) and correlations between logk and pKa obtained for phenols (right) on porous graphite (1) and MSND (2,3) columns in mobile phases containing 90% methanol (1, 3) and acetonitrile (2). [23].
32 This comparative study also showed how DND exhibits a high separation selectivity for structural isomers, such as 2,4- and 2,5-dichlorophenols, which can be separated on MSND, but cannot be separated on the Hypercarb PGC phase. Some indications of the impact of
hydrophobic and π–π interactions on the retention of phenols and benzoic acids were also
observed [23,24].