Influence of post-packing conditioning on the performance of the column There are no general rules about on choice of solvent systems for post-packing column

conditioning [26]. Usually, the purpose of such conditioning is to compact the stationary phase and increase the column efficiency, as well as to provide a smooth transition between slurry solvent and the mobile phase in HPLC. Therefore, four conditioning systems were tested, as presented in Table 3.5.

91 Table 3.4. Performance of several HPHT diamond columns packed at different pressures and temperatures. All columns were packed from 10 mM NaOH diamond slurry without further conditioning/flushing. Column number 2 3 6 8 Dimensions, length × ID, mm 100 × 4.6 100 × 4.6 50 × 4.6 50 × 4.6 Packing temperature, °C 25 90 25 25

Packing pressure, psi 15000 15000 10000 15000

Void time, min 1.39 1.43 0.73 0.73

k, benzophenone 3.32 3.33 2.77 3.38 Efficiency, plates·m-1, benzophenone 55229 43547 5211 36942 Asymmetry@10%, benzophenone 1.52 1.64 1.80 1.78 Pressure at 0.1 mL·min- 1

of IPA and 25 °C, bar 303 296 139 148

Time, min 0 2 4 6 m A U 0 5 10 15 20

Fig. 3.13. Elution profile of benzophenone for four columns packed as described in Table 3.4: column 2 (red), 3 (green), 6 (black), 8 (blue). Mobile phase – 0.5 mL·min-1 of 0.1% IPA in n- hexane, 1 µL injection of 50 mg·mL-1 of propiophenone in mobile phase, 25 °C, UV detection at 254 nm.

92 Table 3.5. Performance of several HPHT diamond 50 × 4.6 mm ID columns packed from 10 mM NaOH (aq.) and conditioned using different solvents. All columns were packed at 15000 psi and temperature 25 °C.

Column number

04 05 11 12 13 14

1st flushing Water Water IPA 10 mM

HNO3

1 M aq. NH4Ac

Form- amide

2nd flushing - - - water water IPA

Void time, min 0.94 0.94 1.11 0.95 0.96 0.94

k, propiophenone 1.03 1.05 0.85 0.88 0.84 0.88 Efficiency, plates·m-1, propiophenone 72580 70660 6340 57709 6259 20784 Asymmetry@10%, propiophenone 1.13 1.12 0.47 1.95 0.55 0.65 Van Deemter terms A × 10-4, cm 5.73±0.14 - - 8.18±0.18 146±27 24.0±1.3 B × 10-4, cm2·s-1 0.51±0.02 - - 0.40±0.02 0.61±0.11 0.96±0.07 C × 10-3, s 2.80±0.08 - - 6.56±0.13 16.2±2.2 11.9±0.8

R2 for van Deemter

model 0.997 - - 0.993 0.858 0.981 Optimal linear velocity, mm·s-1 1.11 - 0.33 0.78 0.61 0.90 HETP at optimal velocity, µm 13.2 - 213 18.4 166 45.5 Pressure at 0.1 mL·min-1

93 IPA (column 11) is compatible with both polar and non-polar solvents in HPLC, so it can provide a smooth transition between packing conditions and conditions of HPLC experiments. 10 mM HNO3 (column 12) will protonate carboxylic groups on the diamond surface and decrease repulsion between particles, increasing packing density. A similar effect may be achieved using high ionic strength solutions (1 M NH4Ac, column 13), which can shield the surface charge of diamond particles and improve the packed bed. Column conditioning with formamide followed by flushing with IPA (column 14) can also provide a transition between aqueous and NP-HPLC solvents, but more smoothly than in the case of column 11, due to the use of polar formamide as an intermediate solvent. Finally, columns 4 and 5 were conditioned with DIW only and used for comparison. Table 3.5 summarises the parameters of column packing and their performance after conditioning with various solvents. Useful information about the quality of packed columns can be obtained through van Deemter plots (Fig. 3.14 and Eqn. 3.3):

HETP = A+B/u+(Cs + Cm)·u (Equation 3.3)

where A is eddy-diffusion term, B is longitudinal diffusion term, and Cs and Cm reflect resistance to mass transfer in stationary and mobile phases, respectively. According to the data provided in Fig. 3.14 and Table 3.5, conditioning did not improve column efficiency, as compared to column 4. The highest performance of 72580 plates·m-1 was achieved using water for conditioning, while columns 11 and 13 exhibited low efficiencies. Column 11 also produced an unusually shaped van Deemter curve, where efficiency at very low flow rates and very high flow rates was higher than at intermediate flow rates. Therefore, the van Deemter plot for this column is not presented in Fig. 3.14. Due to the use of an intermediate conditioning reagent with high polarity (formamide), column 14 has shown improved efficiency, as compared to column 11, but performance of this column was still far below that of column 4.

The only conditioning approach which provided results comparable to using water was the conditioning with 10 mM HNO3. Column 12 has only 20% lower efficiency as compared to column 4, which is mostly due to a higher A and C terms in van Deemter equation (Eqn. 3.3). In contrast, the B-term is better for column 12 (0.40 versus 0.51), which indicates that protonation of carboxylic groups at the diamond surface can affect longitudinal diffusion. However, the packing density for column 12 is lower as compared to column 4, perhaps due to the aggregation of particles in the presence of nitric acid. This resulted in a higher A-term parameter and reduced HETP value for column 12. Though it was expected that use of 1 M NH4Ac will have similar effect on the packed bed, column 13 was very

94 poorly packed and only an had efficiency of 6260 plates·m-1. The differences observed between columns 12 and 13 could be attributed to adsorption of NH4+ on the negatively charged diamond surface

Further comparison of column performance, depending on the conditioning, can be made based on the peak shape. As shown in Fig. 3.15, columns 11 and 13 have problems with packing, which resulted in the appearance of shoulders in the chromatographic peaks. Column 14 produced asymmetric peaks with extensive fronting. This is possibly caused by the strong adsorption of formamide on the surface of diamond, and difficulties in the complete removal of formamide via washing with IPA. This adsorbed formamide can result in the occurrence of multiple retention mechanisms in NP-HPLC and, consequently, in bad peak shapes. In the case of column 4, which has the best efficiency, the peak shape for propiophenone exhibited noticeable fronting. In contrast, the peak shape for column 12 was more symmetrical, but this column’s overall efficiency was lower.

In conclusion, the best columns with HPHT diamond are produced by using 10 mM NaOH as slurry solvent, packing at room temperature and the highest possible pressure, whilst avoiding shock waves. Conditioning can be done either with water, or with 10 mM HNO3, which helps to compact the bed and reduces peak fronting. This approach was used for the preparation of columns for further investigation in NP-HPLC and HILIC. While such packing procedures seem to provide columns of satisfactory quality, it would be interesting to compare the properties of HPHT diamond columns with the properties of columns packed with other carbon and silica sorbents.

3.3.3.3. Comparison of HPHT diamond column performance with other