Chromatography Techniques

LABNEWS Banner

 Facebook  YouTube Twitter   tumblr 





Site Sponser






Home > Magazine > Liquid Chromatography

Read/Post Comments

Sizing Up Columns

Computerized column packing optimizes preparative purification for increased purity, yield and productivity.

by Peter Rahn, Gareth Friedlander, Phil Koerner, et al., Phenomenex, Inc., Torrance, CA

Introduction

In recent years the trend to maximize throughput has resulted in the adoption of shorter bed length preparative columns (50 to 100 mm) operated on open access systems using generic gradients as a standard approach to purify intermediates and final products. As these compounds continue their development cycle, greater quantities of purified material are required, resulting in the need for larger scale preparative separations. Depending on the quantity of material required, either a laboratory chemist or the process chemist performs the purification. When purifying larger quantities, each lab has different requirements for purity and yield, as well as different limitations in their preparative HPLC equipment that influences their decisions to change the column length and inner diameter (ID). This technical note explores the impact column length and ID have on purity and yield when purifying larger quantities of these compounds.

Materials and methods

Figure
Figure 1. Traditional HPLC column slurry-based packing results in decreased performance as media is extruded from the packed bed. Click here to enlarge.
Figure 2
Figure 2. Axial compression is incorporated into manufacturing pre-packed short prep columns. The piston is left in place before the pneumatic ram is removed. Click to enlarge.
Figure 3
Figure 3. Results of packing four types of media via the Axia packing process results in >15% higher efficiency in 21.2, 30- and 50-mm dia columns. Click to enlarge.
All preparative separations were performed on a Gilson GX281 Preparative HPLC system that includes the Binary 333 and 334 HPLC pumps and a UV-VIS 155 variable wavelength detector. Trilution LC software version 1.4 was used for the data analysis. The analytical analyses were performed using an Agilent 1100 LC system equipped with a quaternary pump, in-line degasser, multi-wavelength detector, and autosampler. HP Chemstation software (version A.09.01) was used for the data analysis. The HPLC columns used were Luna 5-μm C18(2) 50 x 4.6 mm, 50 x 21.2 mm, 50 x 30 mm, 50 x 50 mm (Phenomenex). The mobile phase used in these separations was aqueous mobile phase 0.5 % TFA in water and organic mobile phase was 0.5 % TFA in acetonitrile. The gradient used was 5% to 95% B over 5 min with flow rates of 1.5 mL/min on 4.6-mm ID column, 30 mL/min on 21.2-mm ID column, 60 mL/min on 30 mm-ID column, and 150 mL/min on 50-mm ID column. The UV detector was set at 254 nm. HPLC grade acetonitrile and water were obtained from Fisher Scientific. For purification samples a Suzuki reaction mixture was provided by Dr. Shahnaz Ghassemi from Biotage. Another sample was propranolol and diphenhydramine (Sigma Chemical). Fractions from preparative separations were analyzed using a Luna C18(2) 5-μm 50- x 4.6-mm column.

Results and discussion

Historically, HPLC column performance has decreased as the column ID increased even though the same media was packed into the columns. Analytical columns always exhibited high efficiency and good peak symmetry but the preparative columns generally showed lower efficiency and greater peak tailing. This loss in performance is inherent in all slurry packed columns and was caused by the media extruding from the packed bed while the final hardware was assembled (Figure 1).

A major improvement in preparative column performance has been achieved by adapting axial compression to manufacture laboratory scale preparative HPLC columns. A computerized mechanical process packs the column bed and the media is never allowed to expand or extrude from the column and the internal packing force is maintained on the column (Figure 2). This new technology produces higher performance preparative columns yielding the same plates per meter independent of length and ID. This new technology was granted the R&D 100 Award in 2006 for the innovation resulting in the Axia product line. Axia has recently been extended to include additional column lengths resulting in all 50-, 100-, 150- and 250-mm length columns being made in 21.2-, 30- and 50-mm IDs.

Figure 4
Figure 4. Chromatograms reveal continued high performance as column diameter is increased. Click to enlarge.
Figure 5
Figure 5. While maximum load for each column must be adjusted, each column provides the same purification capability. Click to enlarge.
The results from packing four different types of media via the Axia packing process are shown in Figure 3. The improved performance for 21.2-, 30- and 50-mm ID columns graphically illustrates the increased efficiency (>15 %), improved reproducibility in efficiency from column to column (% RSD decreased by 4x), with peak asymmetry reduced by 2 to 6 fold.

To illustrate that the column performance is independent of the column ID, the same Suzuki reaction mixture was separated on 4.6-, 21.2-, 30- and 50-mm ID columns (Figure 4). These chromatograms show the same high efficiency separation is achieved on the analytical (4.6 mm ID) and on the Axia packed 21.2-, 30- and 50-mm ID columns. No loss in performance is observed as column ID is increased. Since these preparative columns have the same plates/meter (efficiency and asymmetry factors) independent of ID and length, the chemist has more options to scale up a separation without sacrificing purity or yield.


The needs and requirements of a laboratory chemist purifying 50 to 100 mg compared to the process chemist producing 500 mg to multiple grams of a purified product are very different. The laboratory chemist is always under a restraint to produce multiple products or compounds and cannot spend the time developing and optimizing purification methods. The laboratory chemist has a large number of samples that must be purified and will continue to use generic gradients and UV- or mass-based collection system. Process chemists purify fewer compounds per year but the purified compounds must be well above 95% pure in 100 g to kg quantities. To achieve these higher mass and purity requirements the separations must be optimized and repetitive runs and UV-based fraction collection are common. Since the requirements of the two labs are very different, their choices and approaches in scaling up a separation are also very different. How the choices of column length and ID affect these different labs are shown in the following examples.

 Figure 6
Figure 6. When load is kept constant, increasing column length improves resolution and provides higher purity. Click to enlarge.
 Figure 7
Figure 7. Increasing column length and sample load proportionally to the length reduces the number of runs required. Click to enlarge.
Initially a mixture of two compounds (propranolol and diphenhydramine) were separated on a 50- x 4.6-mm column using a generic gradient, the maximum load was determined and the separation was directly scaled up to the larger ID Axia packed columns—21.2, 30 and 50 mm (Figure 5). By increasing only the column ID, the sample load increases exponentially [SL (D2/D1)2]. For the smaller 21.2- and 30-mm ID columns, the load was only 16 mg and 32 mg, respectively, but the purity of each fraction was the same as achieved on the larger 50-mm ID column for which the load was 78 mg. For the laboratory chemist who needs higher yields and still needs to separate a large number of compounds, increasing the column ID is recommended since the separation time and gradient conditions do not change.

If higher purity product is required, then increasing the column length will achieve this goal as shown in Figure 6. The sample load was maintained as the column length increased. Increasing the column length improves resolution but the purification time and backpressure increases directly proportional to column length (Time = L2/L1 and Pressure = L2/L1). Since the gradient time must change proportionally to the increase in column length this results in an increase in the separation time for each sample. Increasing the column length may be a hindrance for the laboratory chemist as the instrumentation may not be capable of operating at the required higher backpressure and high flow rate. 

For the kilo lab where purity and overall yield are critical requirements, optimizing gradient conditions along with increasing the column length allows higher sample load per run and reduces the number of cycles required to produce the needed material (Figure 7). In addition, the combination of fewer runs and higher sample loads results in fewer fractions to evaluate and pool. Although throughput (g/hr) remains the same with increased column length, the more important requirements of very high purity and overall yield are achieved. Increasing the column length has the additional advantage to allow larger sample volumes before the dissolution solvent degrades the separation.

Conclusion

Table 1
Table 1. Purification needs comparison for different laboratories. Click to enlarge.
Table 1 summarizes the impact of changing column ID and length for the two different laboratories. Given the limitations typical in a chemist’s lab environment it is always better to increase the column ID to achieve higher purity and higher yield. With the same length column, the separation time, gradient conditions and backpressure do not change and a large number of samples can still be processed per day. For the process chemist performing fewer separations, optimizing the gradient conditions is well worth the time and effort. Multiple runs are always required to produce the larger quantities of material and yield and higher purity are critical success factors. Increasing the column ID and length will help achieve the higher purity and yield while decreasing the number of runs required. This higher purity and yield can now be achieved without sacrificing performance as the columns increase in length or ID. 

The Axia technology has become the new industry standard for consistency in preparative columns with the same performance achieved from 4.6 mm analytical columns to 50-mm ID preparative columns. The high level of process control has resulted in columns with the same performance characteristics (plates per meter) for columns between 50 and 250 mm in length. There has been a significant improvement in the asymmetry and efficiency across all lengths and IDs allowing chemists more flexibility to choose the appropriate column size to achieve their goals for increased purity and yield for their preparative purifications.


Contact Editor Print Article Email To Friend

Comments
Type Your Comment...
Name:   
E-mail (optional) *E-mail addresses do not publish to the site.
Comment:   


Chrom Notebook

michelle
Michelle Longo
Asst.
Managing Editor

Slow and Steady Wins the Race
In the 100 years since Marie Curie won the Nobel Prize in Chemistry, women have made leaps and bounds in science and engineering. Following the extreme example of Madame Curie, women have exponentially raised their position in S&E in the last few decades. In a field previously (and traditionally) dominated by men, women seem to be slowly but surely taking over—or at least leveling the playing field. ...continue



Application Notes
  • Heparin Characterization
  • Sensitive, Ultrafast HPLC Fluorescence Detectors
  • Water’s Impact on Ion Chromatography
More Application Notes

Most Viewed Content

Advantage Business Media © Copyright 2011 Advantage Business Media
Privacy Policy |  Terms & Conditions |  Advertise with Us