Use of the kinetic plot method to analyze commercial high-temperature liquid chromatography systems: I: Intrinsic performance comparison
Introduction
During the past years, the exploitation of high temperatures as a means to speed up liquid chromatography analysis has come more and more into focus [1], [2], [3], [4], [5], [6], [7]. This vivid interest in high-temperature HPLC (HT-HPLC) is essentially due to the considerable improvements that have been made in the development of stationary phases that are sufficiently robust to withstand the elevated temperatures. Commonly used silica-based C18 phases are easily degraded using aqueous organic mixtures at temperatures above 60 °C. Hydrolysis of silica leads to column failure [4], [8]. However, some silica-based columns are stable up to 100 °C at low or medium pH and are now available. All of them have a high protection of the silanols that reduce the dissolution of the silica. They include hybrid silicas, highly C18-bonded silicas, polydentate bonded silicas [9] and sterically protected bonded silicas [10] with isobutyl groups. Zirconia-based phases have very good thermal stability at elevated temperatures [11]. The most popular zirconia-based packing is polybutadiene-coated zirconia (Zirchrom-PBD) and is stable up to 150 °C [12]. Other zirconia-based stationary phases are polystyrene-coated, carbon-clad, phosphate-coated and polyethyleneimine-coated zirconia. Phases based on polymers, such as polystyrene-divinyl benzene can withstand temperatures up to 200 °C [12]. Graphitic carbon columns are very robust and can resist temperatures above 200 °C [12], [13].
The potential advantages of the possibility to operate the columns at these elevated temperatures are considerable [14]. From the Wilke–Chang equation [15], the diffusivity of the analytes is expected to be roughly proportional to the absolute temperature divided by the solvent viscosity (T/η) and so will be the optimum linear velocity which has been proved to be proportional to the diffusion coefficient while the minimal plate height is independent of it [16]. The pressure drop across the column is proportional to both the solvent viscosity and the linear velocity according to Darcy's law. As a result, when working at the optimum linear velocity, the analysis will be faster at elevated temperature but the column pressure drop will be higher. Alternatively, for a given linear velocity, the column pressure drop is obviously lower at elevated temperatures and a longer column can be used.
With respect to the general band broadening characteristics, the increased diffusivity can be expected to have a different influence on each of the three main contributions (eddy diffusion, longitudinal diffusion of the solute and mass transfer contributions), traditionally referred to as the A-, B- and C-term contributions in the Van Deemter equation. While the effect of the temperature on the A-term is uncertain [17], the B-term contribution undoubtedly will increase with increasing temperature, and will hence become more significant in the low linear flow-rate range. The C-term contribution on the other hand should decrease with increasing temperature because the mass transfer resistance decreases: diffusion coefficients in the mobile phase and in both the stationary phase and the stagnant regions of the mobile phase, together with the stationary phase desorption rate, increase with increasing temperature [18]. It is therefore generally predicted [18] that an increase in column temperature should decrease the plate height, particularly if the mass transfer resistance in the stationary phase is dominant in determining H. At high reduced velocity, where the C-term dominates, the increasing temperature should therefore have a beneficial effect on the speed without compromising the resolution. Thus, in general, it can be highly advantageous to operate columns at elevated temperatures [17], [18], [19].
In addition to this, it can also be expected that the asymmetric peak shapes that are occasionally seen with silica-based columns for compounds with basic groups will be improved. Peak asymmetry is usually caused by secondary interactions with some active sites such as silanol groups providing slower kinetics of transfer [3], and are reduced at elevated temperature, because of the improved kinetic effects associated with desorption from these active sites [20], [21]. Another potential cause of peak asymmetry could be the occurrence of non-linear adsorption isotherms (e.g., caused by the saturation of high-energy sites).
The above mentioned potential benefits of HT-HPLC are traditionally presented in a fragmented way. The reduced pressure drop and the reduced C-term band broadening emanating from the reduced mobile phase viscosity, as well as the influence of temperature on the retention factors are usually reported in separate graphs or tables. A much clearer view on the kinetic potential of HT-HPLC would be obtained if all this information could be combined into one single picture.
In the present study, this is realized using a simple data transformation, referred to as the kinetic plot method, and directly applicable to any set of experimental Van Deemter data [22]. In the first part of the study, this method is used to map and compare the ultimate performance limits of the tested supports, devoid of any practical constraints. In the second part, the adopted data transformation method is extended to account for the existence of practical instrument constraints such as the pressure drop and band broadening in the heating and cooling capillaries, the limited detector speed, etc. In addition, some extrapolations are made of the potential beneficial influence of combining high temperatures with high pressures.
Three commercial columns have been investigated. For each column/temperature combination, the mobile phase was always adapted to compensate for differences in retention factor. Using alkylbenzenes with a very similar molecular diffusivity as test components, the differences in band broadening between the test components were essentially due to retention factor effects and are only to a much lesser extent influenced by differences in diffusivity. This was done intentionally, because part of the aim of the present study was to find a good criterion for the selection of the retention conditions that yield the fairest comparison between the different systems and the possible operating temperatures.
Section snippets
Chemicals and columns
Propyl (C3), butyl (C4) and pentyl (C5) benzene were purchased from Sigma–Aldrich (St. Quentin Fallavier, France). The tested columns were a Discovery zirconium polybutadiene (ZR-PBD) 3 μm, 50 mm × 4.6 mm from Supelco (Bellefonte, PA, USA), a Discovery zirconium carbon (ZR-CARB) C18 3 μm, 50 mm × 4.6 mm from Supelco and a Nucleodur C18 Gravity 3 μm, 70 mm × 4.6 mm from Macherey-Nagel (Düren, Germany). The ZR-PBD and the ZR-CARB were tested at 30, 90, 120 and 150 °C. For the Nucleodur column, the measurements
Results and discussion
Fig. 1 shows the Van Deemter curves for all considered conditions and components. The T = 150 °C measurements performed on the ZR-PDB and the ZR-CARB are not represented because they showed an extremely large scatter and obscured the 90 and 120 °C data. The exact origin of this scatter is unclear, but could be related to both the difficulty of the pump system to deliver high flow-rates at high pressure and the relatively low frequency of the employed detector.
Most of the observations that can be
Conclusions
Three different commercial HT-HPLC supports have been tested at various elevated temperatures for their ability to yield the fastest critical pair separation. For the considered RP-LC alkylbenzene separations, it was found that the gain in critical pair separation speed that can be expected from a temperature increase between T = 30 and 120 °C is typically of the order of a factor of 3 to 4, more or less independently of the required plate number. This gain factor relates to the case wherein the
Nomenclature
- A, B, C
constants in Van Deemter equation (sn/mn−1, m2/s, s)
- AF
asymmetry factor
- dp
particle size (m)
- dref
characteristic length (m)
- dtube
diameter of tubing (m)
- Dm
molecular diffusion coefficient (m2/s)
- E0
t0-based separation impedance (E0 = H2/Kv0)
- F
flow rate (m3/s)
- h
reduced plate height, h = H/dref
- H
height equivalent of a theoretical plate (m)
- k
phase retention factor
- Kv0
u0-based column permeability (m2)
- L
length (m)
- M
molecular weight (g/mol)
- n
exponent in free-exponent Knox equation, see Eq. (8)
- N
plate number
- Neff
Acknowledgements
D. Ca. is supported through a specialization grant from the Instituut voor Wetenschap en Technologie (IWT) of the Flanders Region (grant no. SB/1279/00). D. Cl. is a Postdoctoral Fellow of the Research Foundation, Flanders (FWO Vlaanderen).
References (36)
- et al.
J. Chromatogr. A
(1995) - et al.
J. Chromatogr. A
(2003) - et al.
J. Chromatogr. A
(2005) - et al.
J. Chromatogr. A
(2006) - et al.
J. Chromatogr. A
(2004) J. Chromatogr. A
(2004)- et al.
J. Pharm. Biomed. Anal.
(1995) J. Chromatogr. A
(1996)- et al.
J. Chromatogr. A
(2006) - et al.
J. Chromatogr. A
(2002)
J. Chromatogr. A
J. Chromatogr.
J. Chromatogr. A
J. Chromatogr. A
J. Chromatogr. A
Anal. Chem.
Anal. Chem.
J. Sep. Sci.
Cited by (49)
The state-of-the-art and future perspectives for SFC
2022, Separation Science and Technology (New York)Fast-generic HPLC methods
2020, Handbook of Analytical SeparationsA high performance system to study the influence of temperature in on-line solid-phase extraction capillary electrophoresis
2015, Analytica Chimica ActaCitation Excerpt :It is well known that in LC backpressure can be reduced at high temperatures due to the decrease on the mobile phase viscosity, and several home-made thermostatization systems have been reported to work with columns of reduced particle size [12]. Temperature has additional effects: as it is increased the analysis time is reduced, the diffusion coefficients also increase and mass transfer kinetics are improved, the chromatographic efficiencies are better, etc. [13]. In some cases, a change of temperature may promote simultaneous effects, on retention and ionization equilibrium [14], on retention and conformation [15], or other [16].
Kinetic performance comparison of a capillary monolithic and a fused-core column in micro-scale liquid chromatography
2012, Journal of Chromatography ACitation Excerpt :Fanigliulo et al. [10] compared experimentally determined and theoretically predicted performance at higher pressures. Cabooter et al. [24] used the kinetic plot method to study the gain in separation speed offered by HTLC and indicated the range of plate numbers where HTLC and UPLC are beneficial. The effects of increasing temperature and pressure are limited.
Kinetic plot based comparison of the efficiency and peak capacity of high-performance liquid chromatography columns: Theoretical background and selected examples
2012, Journal of Chromatography ACitation Excerpt :Lestremau et al. [4] also considered larger temperature increases: considering a temperature increase from 30 °C to 120 °C, their kinetic-plot analysis predicted a gain factor of about 3–4. Similar data were obtained in [65] studying three different high-temperature stable columns (Zirchrom-PBD, Zirchrom-Carb and Nucleodur Gravity), adjusting the mobile-phase composition such that the retention coefficients did not vary significantly from temperature to temperature. In this paper, it was concluded that the kinetic gain factor that can be expected over most of the C-term dominated range largely parallels the decrease in viscosity of the mobile phase induced by the temperature increase, agreeing very well with the above mentioned gain factors of 2–2.5 (for operations at 80–90 °C) and 3–4 (for operations at 120 °C).