Column Care for the Long Haul—Considerations for Column Storage

Dwight Stoll, LC Troubleshooting Editor

Several factors influence the useful lifetime of high performance liquid chromatography (HPLC) columns. In this instalment we consider some of the details associated with preparing a column for storage, with an eye towards choices that will pay dividends in future use of the column.

It seems simple. We finish our work with a particular high performance liquid chromatography (HPLC) column, put it in the drawer for safe keeping, and move on to the next column for the next project, or the next step in method development. But, what exactly should we do with that column before it goes in the drawer? Stopping to think about the details for a bit, we recognize that there is actually quite a lot to consider. In this instalment, I summarize some definite do’s and don’ts for column storage, and try to make sense of the variety of advice that is available on the topic.

Backing up a bit to think about column care in a broader sense, following best practices for the way we treat HPLC columns can have big effects on the performance of these columns in our work, especially over time. John Dolan has addressed various aspects of column care in his instalments of “LC Troubleshooting” over the years. Given their importance to column lifetime, I’ve summarized some of the important ones briefly again here. Readers interested in more detailed discussions of these topics can follow the references to previous issues of “LC Troubleshooting” (1).

• Avoid mobile phases that will cause chemical damage to the column. For silica-based columns with bonded stationary phases, this advice means two things: avoid very acidic conditions (<<pH 2) that cause hydrolysis of siloxane bonds between the silica and stationary phase ligand, and avoid alkaline conditions (>>pH 8) that can cause dissolution of the silica particle material itself (2). New silica-based materials introduced in the last decade have made chemical damage less of an issue (3), but users need to be aware of the limits of the particular columns they are working with (4).

• “Don’t inject junk” (5). Injecting things into the HPLC column that don’t come out of the column during the analysis is generally bad for performance. This can be particulate debris that gets stuck at the column inlet, increasing pressure drop across the column and causing uneven flow distribution, or chemical constituents of the sample that tend to be very strongly retained and cause changes in column chemistry as they accumulate on the stationary phase. The problem with particulates can be minimized by filtering the sample before injecting it into the HPLC system. Particulates can also originate from the column or the mobile phase itself, or by shedding of various parts of the instrument. Using an in-line filter upstream from the HPLC column can significantly reduce the impact of these particulates on the column as well (6). Finally, using guard columns can minimize the impact of sample constituents that tend to adsorb strongly to the column under the conditions of the analysis. The role of the guard column is to “catch” these components, and the guard is simply thrown away and replaced after a specified number of injections, ultimately extending the life of the analytical column.

In the Beginning: Establishing a Baseline

A common problem in troubleshooting the behaviour of HPLC columns is that we don’t have a good reference point to help us understand how and when the behaviour of a column has changed. For example, in the course of method development we might observe that the resolution of a critical pair of analytes has decreased. This situation leads to a bunch of questions-When did the change start to happen? What are the likely causes for the change? How did the column behave when it was brand new relative to its behaviour now? Likewise, if we start work with a column that has been in the drawer for a month, how do we know that column will behave like it did when it was new? One easy thing to do in this situation is to try to reproduce the separation indicated on the quality control (QC) sheet that comes with the column inside the box. For reversed-phase columns this QC sample is typically a simple mixture of small neutral molecules separated in a simple organic solvent–water mobile phase. For columns designed for separations of biomolecules, this QC separation might involve a standard mixture of proteins that are readily available (for example, myoglobin). If we can reproduce the separation on the QC sheet with retention factors, selectivities, and plate numbers that are similar to what was obtained by the manufacturer, that most certainly should increase our confidence that the column is working like it was when it was new. The problem, however, is that it is very unlikely that the molecules used in the mixtures for these QC separations will interact with the stationary phase in exactly the same ways that the molecules in our analytical samples do. The only way to really address this problem is to establish the baseline performance of the column by injecting a mixture of compounds that is relevant to the separation we are using or trying to develop, and separating the mixture using conditions that are relevant to the conditions we plan to use. In my laboratory this approach is our standard practice-although we may never use that baseline information again in the life of the column, when we do need it, it is invaluable. For example, if we are working with a new column that we hope to use for separations of antibody proteins, the very first thing we do with that column is inject an antibody standard under conditions that are likely to be similar to the final operating conditions we use for that column. Then, if we suspect that something is not right with the HPLC column or instrument, we can always check things out by repeating this separation and comparing results to what we obtained with the column when it was new out of the box. 

Column Storage: Avoiding Major Pitfalls

There is a short list of definite must do’s when preparing a column for storage, all of which involve flushing the most recently used mobile phase out of the column and replacing it with a solvent suitable for storage. What constitutes a suitable storage solvent is discussed in more detail below, because these solutions are stationary phase specific.

• Do flush strongly acidic or alkaline mobile phases from the column (<<pH 2 or >>pH 7). This step will minimize the possibility of chemical degradation of the stationary phase during storage (see above for the mechanisms of degradation).

• Do flush mobile phases containing high concentrations of salt (for example, >30 mM sodium chloride) or ion-pairing reagents (for example, octanesulfonate) from the column. Chloride salts in particular are very corrosive to stainless steel, and will attack the column wall, and inlet–outlet frits (7). The metal ions that are released when the metal surfaces corrode or erode can lead to numerous problems, including contamination of the stationary‑phase material and interference with analyte detection (8). Although high concentrations of salts are not so commonly used in reversed-phase separations, they are very commonly used in ion-exchange separations, and are essential in hydrophobic interaction chromatography (HIC), which is becoming widely used for protein separations. The other major concern with high salt concentrations is that if the column begins to dry out during storage, the column can be turned into a giant salt crystal, which is impossible to recover from.

Storing Reversed-Phase Columns

Now, when we think about how to care for specific types of phases, things become a bit more nuanced. One of the first things to consider in the case of reversed-phase columns is how to flush the mobile phase from the column to prepare for the storage solvent. One problem we want to avoid in this step is precipitation of buffer salts. Although acetonitrile is the most commonly used organic solvent in reversed-phase separations, and phosphate salts are among the most commonly used buffers, they are not compatible at high levels of acetonitrile. A study by Schellinger and Carr (9) mapped out the solubility of different buffer systems (for example, ammonium phosphate, potassium phosphate, and so forth) in different organic solvents, including acetonitrile. For example, they found that a 30 mM potassium phosphate buffer at pH 3 was soluble in mixtures of buffer and acetonitrile only when the acetonitrile level was less than 75%. For most reversed-phase separations this solubility issue is not a problem because compounds of interest are typically eluted at percentages less than this level. However, this means that when flushing the buffer to prepare the column for storage we should not use high acetonitrile levels, and certainly not 100% acetonitrile, as it will cause the buffer to precipitate both in the pump and connecting tubing, as well as the column.

In preparing for this column instalment, I informally surveyed the recommendations of about 30 manufacturers of reversed-phase columns by going through column boxes in my laboratory and reading the column care sheets provided by the manufacturers. This informal survey was interesting because on one hand the advice in those sheets is more varied than I would have expected given that these were all reversed‑phase columns. On the other hand, it seems that some of the advice has been handed down through different generations of manufacturers, kind of like an old family recipe. The first thing that surprised me is that several of the care sheets recommend flushing the column first with pure water to remove buffers in preparation for storage. Although this step will undoubtedly be effective in removing the buffering agents, it may also cause the stationary phase to “dewet” (10,11). Here, dewetting means that the water is expelled from the bonded phase, and sometimes entirely from the pores of the particle, leading to a dramatic loss in retention if the column is used in this state, simply because analytes cannot enter the vacated stationary phase. The good news is that retention can usually be fully restored by reconditioning the column with a mobile phase containing more than about 50% organic solvent. But, it is probably best to avoid this situation altogether when possible. Thus, it is generally advisable to first flush buffering agents from the column with about 10 column volumes of mobile phase containing about 10% organic solvent in water. This approach will be effective and avoid both the precipitation and dewetting problems.

After we have flushed the most recently used mobile phase from the column, we must decide what solvent will be used for actual storage of the column. Going again back to the column care sheets, I found that in the 30 sheets I surveyed about 35% of them recommended storing the column in pure acetonitrile or methanol, and the other 65% recommended storing the column in a mixture of organic solvent and water, where the recommended ratio ranged from 50:50 to 80:20 organic–water. A study by Mowery (12) of the rates of erosion and corrosion of stainless steel components for HPLC in reversed‑phase mobile phases showed that acetonitrile and methanol were far more erosive when used as pure solvents compared to when they were mixed with water. Even adding a few percent of water slowed the erosion by at least a factor of 10. Given the large surface area of the porous stainless steel frits that are typically used to retain the particles in the column bed, even a small amount of erosion or corrosion can lead to contamination of the stationary phase metal ions liberated upon oxidation of the bulk metal surface. Indeed, Euerby, Tennekon, and colleagues (13) showed that contamination of the reversed‑phase stationary phases with metal ions seemed to promote epimerization of the molecule tipredane on-column. Furthermore, the amount of metal liberated from the column hardware during storage in pure organic solvents was enough to increase the rate of on-column epimerization, and lead to very bad peak shapes for molecules having chelating moieties. 

In the end, the “right” choice of storage conditions is dictated by the application at hand. There undoubtedly are applications where column performance is unaffected by storage conditions, so long as the major pitfalls described above are avoided. However, a little work on the front end of method development to see if column storage conditions affect the selectivity of the column for the analytes at hand may well save a lot of trouble (and troubleshooting) later on in the life of the method.

Storing Other Columns: Ion-Exchange, Mixed-Mode, and HILIC Columns

With other column types the potential major problems discussed above (that is, chemical attack, corrosion, and precipitation of salts) still apply. Here, I briefly discuss some details to be aware of that are specific to ion-exchange, mixed‑mode, and hydrophilic-interaction chromatography (HILIC) phases. After dealing with the complications that can arise from the use of very salty mobile phases in ion-exchange columns, the next biggest issue is that ion-exchange mobile phases are very often entirely aqueous, and can be very friendly environments for microbes. Steps should be taken to minimize growth of these bugs during storage. As discussed in my most recent column on filtration (6), adding a small amount of organic solvent (on the order of 10%), or adding sodium azide at a low concentration (for example, 0.05%) in the storage solvent can be sufficient to prevent microbial growth. I’ve seen both approaches recommended in the manufacturers column care sheets.

With these columns, the rate of reequilibration after storage may depend on the storage conditions. For example, if an ion-exchange column is stored with a solution containing counterions that are strongly retained by the stationary phase, then it will take a long time or a high concentration of the counterion in the mobile phase used for the separation method to reequilibrate the stationary phase. Likewise, storing a HILIC column in an acetonitrile–water mixture may take a long time to reequilibrate if a low ionic strength buffer (for example, 5 mM ammonium acetate) is used for the analytical method. Some manufacturers of HILIC columns recommend storage in solvents containing 80–90% acetonitrile, and buffers containing 5–10 mM ammonium acetate or ammonium formate.

The last consideration I’ll mention here is that some stationary-phase chemistries may be susceptible to chemical modification by the storage solution that is different from the types of chemical attacks discussed above for reversed-phase columns. One notable example of this is the potential for esterification of ion-exchange and mixed-mode phases containing carboxylic acid functional groups (for example, weak cation-exchange phases) by alcohols. Although this esterification will be very slow at room temperature, it can lead to significant changes in separation selectivity. For this reason some manufacturers of these phases explicitly advise against storage of these phases in solutions containing alcohols.

How Long Is Too Long?

Most sources of advice on the topic of column storage suggest that the column can be stored in mobile phase for two to four days without any major ill effects. Beyond four days, the column should be flushed and prepared for long-term storage as discussed above. Anecdotally, many users I know report storing their columns for long periods in mobile phase (specifically, the initial mobile phase when gradient elution is used), especially if weakly acidic mobile phases are used (for example, 0.1% phosphoric acid), without any known problems. Thinking towards the other extreme, though, raises the following question: How long is too long? Or perhaps, what happens when the column dries out? As a final step before storage the column should be sealed tightly by screwing in the endplugs supplied by the manufacturer in the column box. Given enough time, or if this is not done, the solvent will eventually evaporate. I am not aware of any published long‑term studies of column storage, but I can say that we have inadvertently collected some data on this point in my laboratory as part of our work with the Product Quality Research Institute (PQRI) column selectivity database that is built upon the hydrophobic subtraction model (14). In a few cases, we have reevaluated the selectivity of columns that have been sitting on the shelf for more than five years, and observed no statistically significant changes in selectivity over this time period. We assumed that the columns had dried out, and rewetted them by first flushing with 100% acetonitrile, and then equilibrating in mobile phase for about 1 h before making any selectivity measurements.

Setting Up for Success

With most things in analytical laboratories, simplicity leads to consistency of execution. It will be more likely that columns are stored properly if we have a plan for doing so that is straightforward and easy to implement. In my laboratory we have an old pumping system from a retired HPLC instrument that we’ve dedicated to the purpose of flushing columns and preparing them for storage. Several laboratory managers I know have related that they do the same thing. Taking this approach one step further, one can set up a series of methods on a pumping system that is dedicated for the purpose of flushing, or on each instrument so that these methods can simply be run at the end of a series of analyses if it is expected that the column will be taken out of use after the run. For example, a method could involve an initial flush with something like 10:90 acetonitrile–water for 10 column volumes to remove buffer salts, a flush at high organic solvent to remove strongly adsorbed compounds and ion-pairing reagents that had accumulated during the run, and finally a switch-over to the actual storage solution (for example, 50:50 acetonitrile–water for reversed‑phases). This strategy will increase the likelihood that columns are properly prepared for storage, extend the lives of columns, and reduce the amount of troubleshooting needed later on in the life of the column. 

Acknowledgements

I want to thank Tony Taylor of Crawford Scientific for some insightful discussion around the topic of column storage.

References

1. J.W. Dolan, LCGC North America23(11), 1174–1181 (2005).
2. H.A. Claessens, M.A. van Straten, and J.J. Kirkland, J. Chromatogr. A 728, 259–270 (1996). DOI: 10.1016/0021-9673(95)00904-3.
3. E.M. Borges and M.R. Euerby, J. Pharm. Biomed. Anal.77, 100–115 (2013). DOI:10.1016/j.jpba.2013.01.013.
4. J.W. Dolan, LCGC Europe29(2), 82–84 (2016).
5. J.W. Dolan, LCGC Europe27(12), 640–644 (2014).
6. D.R. Stoll, LCGC Europe30(2), 80–83 (2017).
7. K. Collins, C.H. Collins, and C. Bertran, LCGC North Am. 18(6), 600–608 (2000).
8. S. Liu, C. Zhang, J.L. Campbell, H. Zhang, K.K.-C. Yeung, V.K.M. Han, and G.A. Lajoie, Rapid Commun. Mass Spectrom. 19, 2747–2756 (2005). DOI: 10.1002/rcm.2105.
9. A.P. Schellinger and P.W. Carr, LCGC North Am.22(6), 544–548 (2004).
10. T.H. Walter, P. Iraneta, and M. Capparella, J. Chromatogr. A1075, 177–183 (2005). DOI:10.1016/j.chroma.2005.04.039.
11. M.R. Schure, J.L. Rafferty, J.I. Siepmann, and L. Zhang, LCGC Europe27(1), 18–27 (2014).
12. R.A. Mowery, J. Chromatogr. Sci. 23, 22–29 (1985). DOI:10.1093/chromsci/23.1.22.
13. M.R. Euerby, C.M. Johnson, I.D. Rushin, and D.A.S.S. Tennekoon, J. Chromatogr. A705, 229–245 (1995). DOI:10.1016/0021-9673(95)00273-P.
14. L.R. Snyder, J.W. Dolan, and P.W. Carr, Anal. Chem. 79, 3254–3262 (2007). DOI:10.1021/ac071905z.

Dwight Stoll is Associate Professor and Co-Chair of Chemistry at Gustavus Adolphus College in St. Peter, Minnesota, USA. He has authored or coauthored 48 peer-reviewed publications in separation science and more than 90 conference presentations. His primary research focus is on the development of two-dimensional liquid chromatography (2D-LC) for both targeted and untargeted analyses. He has made contributions on the topics of stationary-phase characterization, new 2D-LC methodologies and instrumentation, and fundamental aspects including reequilibration in gradient elution reversed-phase LC and analyte focusing. He is the 2009 recipient of the John B. Phillips Award for contributions to multidimensional gas chromatography, the 2011 recipient of LCGC’s Emerging Leader in Chromatography Award, and the 2015 recipient of the American Chemical Society Division of Analytical Chemistry Award for Young Investigators in Separation Science. Direct correspondence about this column via e-mail to LCGCedit@ubm.com