7


exhibited substantially higher back pressure. This lower pressure drop is probably more important with a Waters SFC system where the maximum pump pressure is only 400 bar.


Berger Again


In several brief communications Berger [18,19] recently used an Agilent 1260 Series II (different instrument from earlier reports). This system has an unusual autosampler design. There is no injection ‘loop’. A 0.1 µL sample injection was pushed into the flowing mobile phase with a high pressure syringe. However, the content of a needle seat capillary (75 µm, 0.46 µL) filled with IPA preceded the sample and a small plug of IPA followed the sample. This possibly focuses the injection [20]. The column was 3x20 mm, packed with 1.8 µm RX-Sil. The injection valve was connected to the column with 385 mm of 125 µm tubing and 105 mm of 75 µm tubing (in an ‘ultra-low dispersion’ HX) in front of column, and 31 cm of 125 µm after the column without a HX, The same 2 µL flow cell was used. Efficiency was somewhat disappointing [18] with hmin


= 2.4. The extra-


column volume was ≈11 µL. The injection volume was extremely important. Increasing injection volume of the sample dissolved in methanol resulted in fairly rapid loss of efficiency.


In a follow-up study [19], the same column was used but the plumbing was changed. A 270 mm long stainless steel tube of 75 µm ID connected the injection valve directly to the column inlet without a heat exchanger. A special 2 µL, 3 mm path length flow cell with a 380 mm long, stainless steel 75 µm ID inlet tube was fabricated by Agilent R&D, Waldbronn, DE, and was connected directly to the column outlet. There was no thermal control. The extra-column volume was ≈ 5.75 µL. The hmin


was a remarkable 1.65, (k = 3.2, n = 7).


The equation [21] for dispersion in tubing with laminar flow is:


σ2 t = π r4 L F/24 D


The radius was 0.00375 cm. The length, L, was 65 cm. Flow, F = 0.0333 cm3/sec. The diffusion coefficient was estimated as 7x10-5 cm2


/sec. The calculated dispersion of the tubing was ≈ 0.800 µL2 . Any turbulence in the flow would decrease this value.


The detector flow cell nominally acts as a mixing chamber, with an equation: σ2


d =V2 /12 It is sometimes suggested that the


denominator should be smaller at 6 or even 4. Calculated dispersion of the detector cell was 0.333 µL2


. Thus, calculations of


the combination of tubing and detector dispersion was as low as 1.13 µL2 unlikely to be > 2.


and


The column had a theoretical dispersion of ≈ 14.2 µL2


dispersion was 17.1 µL2


observed efficiency was quite surprising, particularly with such a short column and sub-2 µm particles. Two superimposed injections with an average h = 1.56 are presented in Figure 1.


The fastest peaks had w1/2 = 4.86e-3 min (292


ms)(120 Hz detector). The sample injection volume was 0.1 µL. The feed speed was 1000 µL/min or 16.7 µL/sec. Thus, the injection time was ≈ 60 ms.


The dispersion of the 3x20 mm column should compare favourably with the nominal column dispersion of a 2.1x100 mm column, packed with the same 1.8 µm particles, which is 17.7 µL2


(k = 2). However, attempts


to use a number of 2.1 mm ID columns resulted in poor efficiency, even with 0.1 µL injections.


Although the use of 75 µm ID tubes have significant negative aspects, operating them near optimum flow rates on either 3 mm ID columns (≈2 mL-min-1 mm ID columns (≈ 1 mL-min-1


) or on 2.1 ) mostly


results in laminar flow in the tubing [17]. While turbulent flow is good for minimising dispersion in the tubing, it also results in major increases in extra-column pressure drops.


Other Factors Affecting Observed Efficiency


Many of the difficulties in obtaining high efficiency with small ID columns packed with sub-2 µm particles involve injection volume and injection solvent [22]. This is particularly true with smaller ID columns packed with sub-2 µm columns. In SFC one cannot dissolve the sample in the mobile phase, since it is essentially a compressed gas, that cannot be metered quantitatively into a vial, at least with current technology.


With larger columns (3-4.6 mm ID), the tendency in SFC, is to dissolve the sample in the modifier. However, this tends to make the ‘strong sample solvent effect’ a problem, by overwhelming the local mobile phase solvent strength on the column just after injection. It is particularly true when the injected sample volume is large, or at low


(h =2 @ k = 2). The observed total at k = 3.43. The high


modifier concentrations. Here, on small ID columns, ‘large’ can be 0.1 µL.


Decreasing the sample solvent polarity is desirable, as suggested by Broeckhoven [8], but many polar solutes are not soluble, or poorly soluble, in non-polar solvents like hexane/heptane. Adding IPA or EtOH obviously increases polarity/solubility but Guillarme [23] suggests that mixed sample solvents are always a problem and usually result in peak distortions. He also suggested a number of other pure aprotic sample solvents that should diminish the strong sample solvent effect, such as acetonitrile, methyl t-butyl ether (MTBE) and dichloromethane (DCM) as sample solvents but they do not seem very ‘green’.


Conclusions


Current commercial SFC’s are not plumbed by the manufacturers for use with sub-2 µm packings even though some characterise them as UHPSFC. However, it was demonstrated that true UHPSFC performance can be achieved with 3 mm and perhaps even 2.1 mm ID columns by minimising tubing length, tubing ID, and using smaller detector flow cells. It also required optimised injections in terms of sample solvent composition and injection volume. This ultimately requires columns that are actually well packed. The efficiency improvements in several of the works reported here are equivalent to at least the early UHPLC’s but with significantly faster speeds.


Claims that column technology has outstripped instrument development in SFC doesn’t necessarily mean that all columns are as efficient as expected. The lack of certainty in the inherent efficiency of the columns used in these reports has led to considerable confusion about the effect of tubing ID, detector flow cell effects, and injection issues. The modifications reported by most groups should logically have dramatically improved extra-column dispersion but in most cased didn’t appear to. It appears that most of the polar columns of 3 or 2.1 mm ID packed with sub-2 µm polar particles used in these studies were relatively poorly packed or exhibited mixed retention behaviour to the point where the column dispersion was large enough to make the extra-column dispersion fairly irrelevant .


Statements that column technology has outstripped instrumentation is only partly true with regard to sub-2 µm particles in 2.1 and 3 mm columns.


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  |  Page 61  |  Page 62  |  Page 63  |  Page 64  |  Page 65  |  Page 66  |  Page 67  |  Page 68