12 August / September 2016

profiles using the mixed stream injection mode. In Figure 2b under identical conditions we injected the same sample volumes of a much more concentrated antipyrine solution (100 g/L). Figure 2a and b together show clearly that the ‘plug effect’ takes place only in the mixed-stream injection mode and not in the modifier stream injection mode. In the latter case (purple lines in Figure 2a and b) there will be no sample solvent – mobile phase solvent mismatch since the sample is introduced into the modifier- stream prior to the mixing point between carbon dioxide and modifier. These observed peak distortions will have the most pronounced effects at preparative-scale injection volumes [10]. In analytical situations smaller volumes are injected so much smaller distortion effects are expected.

Peak distortions due to viscous fingering

After completion of the experimental set described above, the sample band elution profiles were numerically modelled on a theoretical basis assuming both un- retained and retained co-solvent injection plugs, respectively [10]. These calculations quantitatively confirmed our first set of experiments but also pointed out that there might be an additional significant effect when accounting numerically for the plug effect for antipyrine injections using an eluent containing 7.2%v/v MeOH (Figure 5 in [10]). We suspected these extra contributions to peak distortion were due to sample – mobile phase viscosity mismatches, so called viscous effects [11]. In order to prove our hypothesis, in cooperation with Professor Shalliker, we used transparent columns and visualised with his cameras what happens inside the chromatographic separation with LC experiments ‘imitating’ the same viscosity contrast as was the case in the SFC experiment (Figure 5 in [10]).

The viscosity contrast between eluent (with 7.2% v/v MeOH) and injection solvent in the SFC experiment was calculated to be around 3.8 times. To experimentally visualise viscosity effects of this magnitude we used a 5 mm I.D. LC column packed with a 5µm C18 silica phase and equilibrated with 45/55% v/v dichloromethane/toluene, which correspond to a viscosity contrast 0.38 cP [10]. This mobile phase has the same refractive index as the C18 silica. Cyclohexanol has a very high viscosity, and the same refractive index as the stationary phase. Hence the viscosity of the mobile phase can be easily adjusted simply by adding cyclohexanol to the dichloromethane/toluene mixture. The

Figure 3a and b. The photographs illustrate the change in shape of elution bands as a function of the viscosity contrast. The refractive index system was matched with the stationary phase, so that the column becomes transparent. In both cases the viscosity of the injection plug was 0.38 cP. The injection volume was 5 µL and the flow rate was 0.5 mL/min. Flow direction is from left to right. In (a) mobile phase viscosity 0.38 cP, i.e. viscosity contrast around 0 and in (b) mobile phase viscosity 1.44 cP, i.e. viscosity contrast about 3.8 times. The figure was adapted from Figure 7a and b in [10].

injection was visualised by adding an un- retained coloured dye to the sample. Two LC experiments were conducted; the first with no viscosity contrast between the eluent and the sample solution, see Figure 3a. The second experiment was performed such that there was a viscosity contrast between the injection solvent and the mobile phase of approximately 3.8 times, see Figure 3b, thus imitating the SFC experiment where we had an additional deformation, Figure 5 in [10].

The sample zone in the column without viscosity contrast, Figure 3a, is more or less bullet-shaped whereas this is not the case when there was a viscosity contrast (cf. Figure 3b). From inspection of these images we can clearly see that the sample zone was distorted and severely tailing and this would drastically broaden the elution zone of the injection solvent.

Peak distortions due to co-solvent effects

Recent investigation has shown that the MeOH co-solvent adsorbs very strongly to silica [12] and diol [13] stationary phases. We have found that when the MeOH co-solvent adsorbs more strongly to the stationary phase than the solute, unusual peak deformations of the preparative bands can occur. The normal Langmuirian shaped bands may turn to anti-Langmuirian shapes when changing from neat (pure) carbon dioxide to an eluent containing co-solvent. Such strange overloaded elution profiles have previously only been investigated and explained in LC when a strongly adsorbing, often ionic, additive is in the eluent [14,15].

If such an effect takes place there can be a band transition from the overloaded solute band having a Langmuir band shapes to an anti-Langmuir band shapes depending on the fraction of the modifier in the eluent plateau: so far this effect has only been observed and described in rare LC systems [14,15]. However, two requirements must be fulfilled [15] for this phenomena to take

place; firstly, the mobile phase must contain an additive that adsorbs stronger to the stationary phase than the solute. This can be checked for by injecting small amounts of both additive and the respective solutes in a LC system. It is imperative that this LC system utilises a mobile phase lacking the additive being studied as well as possesses a detector capable of quantifying not only the solute but also the additive component. If the additive peak injected under such conditions has a greater retention time than that of the solute, that is a proof of that the additive adsorbs stronger than the solute. The second requirement for this strange phenomena to take place is related to the situation when the additive has been added to the mobile phase of a certain concentration plateau level and is actually acting as an additive. Under such circumstances we have a constant stream of additive along the column. This means that an equilibrium is established of the additive between the stationary and mobile phases for the particular plateau concentration level of additive. Now, in this situation, the injection of the sample containing large amount of solute, will disturb the established equilibria of the additive in the column generation a positive displacement additive zone which if detected is called perturbation peak by the chemical engineering community and system peak by the analytical separation community [15]. The second requirement for this band deformation effect due to a strong additive to take place is that this perturbation peak of the additive is made to elute before the actual solute peak. This seems at a first glance paradoxical regarding the fact that the first requirement is that additive component itself should have as a stronger degree of adsorption as compared to the solute in a mobile phase lacking additive (see requirement one above). But this can actually easily be achieved by increasing properly the plateau concentration level of the additive and the reason for that is that according to the separation theory the

Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50  |  Page 51  |  Page 52  |  Page 53  |  Page 54  |  Page 55  |  Page 56  |  Page 57  |  Page 58  |  Page 59  |  Page 60