The Heart of Chromatography
The column is still the heart of chromatography. While in recent years it has been possible to reproducibly manufacture ever smaller particles down to a diameter of 1 µm, packing an HPLC column is still one of the biggest challenges.
On August 6, 2012, the NASA rover Curiosity landed on Mars. It was one of the most challenging missions to another celestial body. Due to the enormous weight of the Mars rover of about one ton, a new method for safe landing had to be tried out. Although the procedure had never been used before, everything went as planned. In contrast, packing small particles for HPLC separations on Earth in 300 μm ID columns to obtain a high batch-to-batch reproducibility is still a difficult task. Therefore, the commercial availability of miniaturized HPLC columns is still limited.
Many users consider the separation efficiency and robustness of micro LC columns to be lower compared to columns with an ID between 2.1 mm and 4.6 mm. We have examined whether these prejudices are justified on the basis of our own investigations. The assessment of separation efficiency was based on the van-Deemter analysis, which is briefly explained below.
Figure 1 shows the typical plot of a van-Deemter curve. Here, the plate height H is plotted against the linear flow velocity u0. The van-Deemter curve shown in light blue in Figure 1 results from the addition of several terms. The A-term, also known as Eddy diffusion, describes the dependence of the plate height on the particle diameter and is a measure of the homogeneity of the packing. In general, smaller particles always lead to a smaller A-term. The B-term, also referred to as longitudinal diffusion, leads to a strong band broadening at very low flow rates (B-term region). For practical applications, the longitudinal diffusion usually plays no role because the set flow rate is well above the critical range. At high flow rates, the C-term dominates, which comprises the mass transfer between analyte, stationary and mobile phase (C-term region). For particles with a diameter of 3 µm and larger, a steep increase in the C-term is usually observed with an increase of the flow rate, whereas for sub 2 µm particles only a flat increase is observed.
Small particles thus have a positive effect on the A- and C-term and lead to a low plate height. For each separation column, an optimal flow velocity (uopt) results, at which the minimum plate height and thus the highest efficiency is achieved. This flow velocity is not a constant for a specific column but depends on the properties of the analytes and the chromatographic conditions like e.g. temperature or viscosity of the mobile phase.
Figure 2 shows the results of a van- Deemter analysis for three micro-LC columns with an ID of 300 µm. Two analytes with different molecular weight were chosen in order to demonstrate their influence on the van-Deemter curve.
Based on the data from Figure 2a for the analyte etoposide, it can be seen that the reduction of the particle diameter leads to a decrease of the plate height H, which is due to a higher number of theoretical plates N. The smaller the particle diameter, the lower the slope of the C-term. A different picture is observed when instead of etoposide with a molar mass of 588.57 g mol-1 naphthalene with a molar mass of 128.17 g mol-1 is used. In this case, the column packed with the sub 2 µm material has a lower plate height than the column packed with 2.7 µm superficially porous particles only if the linear flow velocity is higher than 4 mm s-1. In general, the optimum flow velocity for all the columns studied shifts to higher linear flow velocities, which is mainly due to the faster diffusion of naphthalene compared to etoposide.
Packing a Column
Hetzel et al. were able to show on the basis of further evaluations that the column packed with the sub 2 µm material has a similar packing quality as analytical columns with an inner diameter between 2.1 mm and 4.6 mm. The quality of the packing was assessed on the basis of the reduced plate height hmin. This dimensionless parameter is obtained by dividing the plate height at the van-Deemter minimum (Hmin) by the particle diameter dp:
For very well packed columns, hmin should have a value of approximately two, which means that the chromatographic equilibrium will be obtained across two layers of the packing. The use of the reduced plate height has the further advantage that different particle sizes and particle technologies and also separation columns of different lengths and inner diameters can be directly compared.
The discussion about the efficiency of micro-LC columns is very controversial in the current scientific literature. As we already explained in the article “Micro-LC Basics”, all system volumes leading to a band broadening must be minimized. This applies in particular to the injection volume and the volumes of the transfer capillaries from the injector to the column as well as from the column to the detector and the volume of the detection cell. Therefore, if an HPLC system is used whose system volumes significantly contribute to the overall band broadening, the conclusions about the packing quality or efficiency of micro-LC columns are misleading. The seemingly inferior efficiency of micro-LC columns in these cases is due to the influence of extra-column volumes on band broadening.
In addition, the particle diameter and particle size distribution play a crucial role in the packing process. Narrow particle size distributions lead to a reduction of the A-term, indicating a more homogeneous packing. This assumption is underlined if the so-called core shell materials are considered, which usually have a smaller particle size distribution than fully porous particles. It is often assumed that smaller particles are more difficult to pack than larger particles. In the study carried out by Hetzel et al. it was observed that the reduced plate height hmin for a column with an ID of 300 µm containing particles with a diameter of 1.9 µm is smaller than for particles with a diameter of 3.0 µm. One explanation for this phenomenon is the ratio of the inner diameter of the column to the particle diameter of the stationary phase, which is called “aspect ratio”. This is 158 or 100 for an inner diameter of 300 µm if particles with a particle size of 1.9 µm or 3.0 µm are considered. Due to the fact that wall effects play an increasingly important role with a reduction of the column’s inner diameter, a better packing quality can be achieved for micro-LC columns with sub 2 µm particles. This in turn would mean that very small particles with a diameter of around 1.0 µm, which are currently the subject of research, in combination with micro-LC columns should have a very high separation efficiency. This applies, as already mentioned above, only under the prerequisite of the consistent minimization of all extra-column volumes. However, the basic prerequisite for this is that the packing process is standardized.
Even though we are embarking on increasingly complex missions to explore the universe, packing HPLC columns will probably remain a challenging task for some time to come. With the data presented here, we were able to show that a high packing quality that is comparable to analytical columns can also be achieved for micro-LC columns. However, the intrinsic separation efficiency of micro-LC columns can only be fully utilized if micro-LC systems with very low extra-column volumes are used.
Thorsten Teutenberg1, Terence Hetzel2, Juri Leonhardt1
1Institute for Energy an Environmental Technology e. V., IUTA, Duisburg, Germany
2Bayer AG, Wuppertal, Germany