LC–MS-based top-down proteomics involves the chromatographic separation and mass-spectrometric characterization of intact proteins without prior enzymatic cleavage into peptides. This article describes the preparation of monolithic reversed-phase (RP) columns with optimized porous structure and their performance separating intact proteins. Peak capacities of around 600 could be achieved within a 2 h gradient. The combination of high-resolution monolith chromatography with high-accuracy timeofflight mass spectrometry (TOF-MS) allowed intact proteins and protein isoforms that differ only in their oxidation state to be distinguished. The article also describes various MS fragmentation techniques for the identification and the more detailed characterization of intact proteins.

The concept of separation media based on monolithic materials can be traced back to the early 1950s, when Nobel Prize Laureates Martin and Synge discussed the benefits of these materials for different applications in chromatography (1). One of the first monolithic polymer gels for liquid chromatography (LC) was developed by Kubín et al. in 1967 (2). Highly swollen hydrogels, prepared in glass tubes from 2-hydroxyethyl methacrylate as the bulk monomer and ethylene dimethacrylate as the cross-linking monomer, were applied for the size-exclusion separations of water-soluble polymers. Rigid macroporous polymer-based monolithic columns were introduced in the early 1990s (3). This type of stationary phase is composed of a single monolithic unit with interconnected microglobules and macropores (flow-through pores), and is typically bonded to the column wall via a covalent bonding to enhance the robustness. 

Polymer monoliths are prepared from liquid precursors, that is, the bulk monomer and cross-linker dissolved in the porogen, allowing their 

preparation in virtually any format. The separation performance strongly depends on the porous structure of the monolithic material (4). It has been demonstrated that the composition of the polymerization mixture, including the type or ratio of the porogen solvents in the polymerization mixture (5), and the polymerization conditions, such as polymerization temperature and time (6), are key parameters that need to be controlled precisely. The surface chemistry and therefore selectivity can be tuned by incorporating 'functional' monomers in the polymer backbone (7). As a result of the absence of mesopores (stagnant zones inside polymer microglobules), the mass-transfer contribution to total band broadening is greatly reduced (8).  

The success of polymer monolithic columns for the reversedphase gradient-elution separation of peptides in a LC–MS bottom-up proteomics approach has been demonstrated on several occasions (9–11). Karger et al. explored the use of 20 µm i.d. monolithic columns and established an HPLC–ESI–MS method yielding low-attomol detection sensitivity (12). Recently, we demonstrated a LC–MS–MS separation of a tryptic

 digest on a 1 m monolithic column yielding a peak capacity in excess of 1000 (13).  

LC–MS-based top-down proteomics is based on the chromatographic separation of intact proteins followed by their mass spectrometric elucidation. Single-stage MS provides information on the molar mass of intact proteins. The mass alone can give information about chemical or post-translational modifications of proteins, provided that the identity of the protein is already known. Tandem mass spectrometry MS

  can provide structural information and is essential for the identification of unknown proteins. In MS

  the selected precursor ions are fragmented in the gas phase using neutral gas molecules for collision-induced dissociation (CID) (14) and higher-energy collisional dissociation (HCD) (15); electrons for electron capture dissociation (ECD) (16); or radical anions for electron transfer dissociation (ETD) (17). In top-down proteomics high-resolution separation technology is of vital importance to address two bottlenecks: ion suppression and the spectral complexity. 

Here, the potential of capillary polymer monolithic columns is discussed for the separation of intact proteins, including protein isoforms. The effects of column properties (length and morphology) and gradient conditions on the separation performance were evaluated. The potential to achieve highresolution protein separation is demonstrated with the liquid chromatography time-of-flight  mass spectrometry (LC–TOF-MS) analysis of a mixture containing 48 intact human proteins. In addition, the application of complementary fragmentation techniques (CID, HCD and ETD) is demonstrated for protein identification using a hybrid ion trap-orbitrap mass spectrometer. 

: Divinylbenzene (DVB, 65%) and styrene (S, > 99%) were purchased from Merck (Hohenbrunn, Germany). Sodium hydroxide pellets were obtained from Merck (Overijse, Belgium). Acetonitrile (ACN, HPLC supragradient quality), formic acid (FA, ULC/MS quality), tetrahydrofuran (THF, > 99.9 %), aluminium oxide (Al

O3), toluene (99.8 %), 3-(trimethoxysilyl)propyl methacrylate (98 %), methanol, 1-decanol (> 95.0 %, GC quality), trifluoroacetic acid (TFA, ULC/MS quality), carbonic anhydrase from bovine erythrocytes, ribonuclease A from bovine pancreas, lysozyme from gallus egg white, myoglobin from equine heart, trypsinogen from bovine pancreas and the "Universal Proteomics Standard 1" containing 48 human source or human sequence recombinant proteins (ABRF mixture) were purchased from Sigma-Aldrich (Bornem, Belgium). 

DVB and S monomers were purified by passing them over activated basic alumina. Water was purified in-house using a Millipore Simplicity system (Millipore, Bedford, MA, USA). Fused-silica capillary tubing (200 µm i.d. × 365 µm o.d.) was purchased from Aventes (Eerbeek, The Netherlands). 

HPLC experiments were performed using an UltiMate 3000 Proteomics MDLC system (Dionex Corporation, Germany) equipped with a membrane degasser, ×2 dual-gradient pump, a thermostatted flow manager and well-plate sampler applying a linear aqueous ACN gradient. For LC–TOF-MS analysis of intact proteins the HPLC system was coupled on-line with a maXis ultra-high-resolution qTOF mass spectrometer via a nano-electrospray interface (Bruker Daltonics, Bremen, Germany) equipped with a Proxeon steel emitter. Spectra were obtained in positive ion (PI) mode. LC–TOF-MS analysis of intact protein was performed on a 250 mm × 0.2 mm i.d. PepSwift monolithic capillary column purchased at Dionex Benelux. 

An Orbitrap Elite hybrid mass spectrometer (Thermo Fisher Scientific, Bremen, Germany) was used for MS–MS fragmentation experiments. Protein solutions were infused directly into the mass spectrometer. A solution of beta2microglobulin (0.5 µg/µL in 1:1 water/acetontrile containing 0.1% formic acid) was infused into the Orbitrap Elite mass spectrometer (Thermo Scientific) at 5 µL/min. The source voltage was set to 5 kV and source temperature to 275 °C. The AGC settings was 1E6 and 500 ms for MS and 5E4 and 250 ms for MS/MS. Five to 10 microscans were averaged for MS and MS–MS spectra. 

Scanning electron micrographs (SEMs) were obtained using a field emission scanning electron microscopy (FE-SEM) (Jeol JSM-7000F FE-SEM, Jeol Ltd., Tokyo, Japan) after sputtering a 20 µm coating on the monoliths with a Cressington 208 HR sputter coater.

To demonstrate the effect of porogen composition of morphology, monolithic columns were prepared 

in 200 µm i.d. fused-silica capillaries from styrene and divinylbenzene precursors through a thermally initiated free-radical copolymerization reaction. Prior to the polymerization reaction, surface modification of the inner wall of fused-silica capillaries was performed with 3-(trimethoxysilyl)propyl methacrylate following an approach described by Svec (18). Polymerization mixtures comprising 40 w% monomers (20% S as bulk monomer and 20% DVB as cross-linker), 60 w% porogen (THF and decanol),and initiator (AIBN, 1 w% with respect to monomers) were degassed in an ultrasonic bath for 10 min. Surface-modified capillaries were filled with the polymerization mixture and then sealed with a septum. After polymerization at 70 °C overnight in a water bath, the columns were flushed with methanol and mobile phase.  

LC–TOF-MS analysis of intact proteins was performed on a 250 mm × 0.2 mm i.d. PepSwift monolithic capillary columns (Dionex Benelux, Amsterdam, The Netherlands). 

Control of monolithic structure and effect on gradient performance: 

Figure 1 shows scanning electron micrographs of the monolithic materials prepared. The close-up [Figure 1(b)] of the monolithic structure depicted in Figure 1(a) clearly shows the typical interconnected polymer microglobules and macropores in the polymeric structure. As a result of the absence of mesopores (stagnant zones in microglobules), polymer monolithic materials are conceptually very well-suited to perform large-molecule gradient separations because mass transfer is mainly driven by convection, rather than by diffusion. The monolith is covalently bonded to the fusedsilica capillary surface via a 3-(trimethoxysilyl)propyl methacrylate spacer to prevent channeling effects of analytes between the monolith and the wall. 

The porous structure of PS-DVB monoliths was systematically varied by changing the porogen ratio (THF/decanol) to maximize the peak capacity. With a decreasing amount of THF in the polymerization mixture, an increase in both the macropore and microglobule size can be observed (Figure 1). These findings can readily be explained by the solvating properties of the porogens: while THF is a thermodynamically "good" solvent for the polymer, decanol constitutes a "poor" solvent. Decreasing the amount of THF (thereby increasing the concentration of "poor" solvent) results in an earlier phase separation and results in the formation of a porous monolithic structure containing large macropores. Since the polymerization reaction is mainly continued within the nuclei, the size of the polymer microglobules increase. The resulting polymer monolith will accordingly exist of large polymer microglobules with large voids (macropores). THF is a "good" solvent for both monomers and polymers and increasing the THF content in the polymerization mixture means the phase separation is postponed. As a consequence, the monomer concentration in the nuclei is smaller, leading to the formation of a larger number of microglobules that are smaller in size [see the monolithic structure depicted in Figure 1(d)]. 

Columns of different length and morphology were evaluated in gradient LC mode by separating a standard mixture of proteins (ribonuclease A, lysozyme, trypsinogen, myoglobin and carbonic anhydrase) and varying the gradient duration. The peak capacity (

) was determined by averaging the 4σ peak width (

 the retention factor of the analyte at the moment of elution and 

 is the plate height. Figure 2 shows the effect of gradient time on peak capacity (a) and peak-production rate (peak capacity per unit time) (b) operating the columns at a flow rate of 2 µL/min. While longer gradient times and shallower gradients yield higher peak capacities, the peak capacity tends to reach a limit at longer gradient duration due to a linear increase in 

. However, with increasing gradient time, the peak-production rate dramatically decreases. Monoliths with smaller macropores and microglobules perform better in terms of peak capacity and in peak-production rate. Also, the optimum gradient time to achieve the maximum peakproduction rate shifts to slightly higher values [Figure 2(b)]. The longer monolithic column performs better in terms of peak capacity and peak-production rate when applying gradient times > 50 min, which is far from the optimal gradient duration required to reach the maximum peakproduction rate. However, this may be alleviated with the development of future monolithic nanomaterials that can be operated at ultra-high pressures. 

The potential of long monolithic columns for highly demanding separations is further demonstrated by a top-down LC–TOF-MS analysis of a complex mixture of intact proteins on a 250 mm monolithic column applying a 2 h gradient (Figure 3). The protein mixture contains 48 intact human proteins and the sample complexity is further increased by the presence of protein isoforms arising from co- and post-translational modifications, including oxidation, methylation and glycosylation (19, 20). The maximum peak capacity was around 600. The threedimensional representation of the LC–TOF-MS analysis is shown in Figure 3(b), depicting the retention time (

-axis). Figure 3(c) shows the charge envelope of a single protein with charge states ranging between 6+ and 10+ eluting between 46–48 minutes. After deconvolution of the mass spectrum the relative molecular weight (M

) was determined to be 8,381.4 Da, which corresponds to the theoretical molecular weight of interleukin-8 (M

 = 8,381 Da). Figure 3(d) shows a zoom spectrum of the isotopically resolved 8+ ion of this polypeptide (resolution ≈ 30,000 FWHM). 

Figure 4 clearly shows the high information content of the three-dimensional representation of proteins eluting between 88  min and 95 min. Whereas the TIC representation would only show co-eluting proteins, the three-dimensional representation allows for the deconvolution of co-eluting proteins, based on differences in 

 values. Protein identification, as shown in Table 1, was based on matching of experimental M

 values with theoretical masses obtained from the list of known proteins present in the 48 protein standard. The six charge envelopes that were baseline separated were consistent with their allocation to isoforms of peptidylprolyl 

 isomerase A. The mass difference of 16.0 Da could be indicative of the oxidation of methionine or cysteine residues. 

The three-dimensional representation in Figure 5(a) shows highly similar charge envelopes, indicative of the elution of protein isoforms, which are baseline separated. The average molecular weight of the most intense signal (signal 1) was determined to be 11,729 Da. The experimentally determined Mr values coincide well with the expected mass of beta-2-microglobulin (M

 = 11,731 Da). The mass difference of 2 Da can be explained by the formation of a disulfide bridge, which is a documented post-translational modification for this particular protein. The mass difference of 16 Da between peaks 1 and 2 could indicate protein oxidation. This post-translational modification makes the protein more polar and results in a decrease in retention time because of the reversed-phase (RP) character of the stationary phase. 

Application of complementary MS–MS fragmentation techniques: 

  experiments on the intact protein can add further confirmation to protein sequence and to the characterization of post-translational modification(s). MS–MS experiments were performed on beta-2-microglobulin (~12 kDa) using an Orbitrap Elite hybrid mass spectrometer which is capable of 240,000 resolution (at 

 = 400). Ion trap CID, HCD and ETD were performed on the +12 charge state of beta-2-micoglobulin and the HCD and ETD MS–MS spectra are shown in Figures 6(a) and (b), respectively. 

The multiply-charged ions in the ion trap CID and HCD spectra were assigned to b- and y-fragment ions, and sequence coverage with HCD fragmentation was found to be higher compared to ion trap CID. The ETD spectrum in Figure 6(b) revealed a homologous series of c- and z fragment ions. The sequence coverage of applying ion trap CID, HCD and ETD fragmentation is demonstrated in Figure 6(c). Applying complementary fragmentation techniques provided more reliable and detailed aminoacid sequence information. It is important to note that beta-2-microglobulin contains a disulfide bridge and fragmentation was suppressed in the disulfide bonded loop (between amino acid 25 and 80). More structural information may be obtained after reducing the disulfide bridges before the LC–MS–MS experiment (21). 

The separation performance of capillary polymer monolithic columns that differ in porous structure and column length was evaluated and the gradient duration was optimized for the LC–MS analysis of intact proteins. The separation performance of the columns was assessed in terms of peak capacity and peak-production rate. Peak capacity can be interpreted as a direct measure for separation power, while peak-production rate can be used as a measure for effectiveness. Longer gradient times and shallower gradients yielded the highest peak capacities. However, with increasing gradient time, the peak-production rate decreased. 

The analysis of intact proteins using LC–TOF-MS allowed the tentative identification of proteins based upon their molecular weight. As demonstrated, this may be more confidently achieved by the application of MS/MS fragmentation techniques. The application of highresolution monolith chromatography in combination with complementary fragmentation techniques such as CID, ETD and HCD provides a unique platform to analyse proteins, including isoforms. Reduction of disulfide brides prior to LC–MS–MS analysis might be beneficial to obtain higher sequence coverage. 

This work was supported by a grant from the Research Foundation Flanders (G.0919.09) and is gratefully acknowledged. Bert Wouters acknowledges the financial support Institute for the Promotion of Innovation through Science and Technology in Flanders (IWT-Flanders). NEPAF is a CELS project funded by ONE NorthEast and partly financed by the European Regional Development Fund (ERDF). Franck van Veen (Dionex Benelux, Amsterdam, The Netherlands) is acknowledged for donation of the PepSwift monolithic columns. 

 obtained his Masters degree in Applied Engineering (Chemistry) in 2007 from the Artesis College University of Antwerp, and his Masters degree in Biomolecular Science in 2010 from the Vrije Universiteit Brussel in Brussel. He is a PhD student at the Department of Chemical Engineering at the Vrije Universiteit Brussel and focuses on the development of novel spatial multi-dimensional chromatography. 

 is a PhD student at the Department of Chemical Engineering at the Vrije Universiteit Brussel and focuses on the development and characterization of novel polymer monolithic structures for application in chromatography. He obtained a Masters degree in Bioengineering, medical biotechnology (Cellular & Genetic Biotechnology) with great distinction in 2009 from the Vrije Universiteit Brussel.  

 is Scientific Manager at NEPAF and is a visiting Research Professor at Northumbria University. After his post-doctoral research at the Fujita Health University in Toyoake, he was appointed Director of Mass Spectrometry in the Department of Clinical Pharmacology at the Royal College of Surgeons (RCSI), Dublin. He has performed extensive work in the application of NMR and mass spectrometry to solve the structure of lipoarabinomannan at both the University of St Andrews and Leeds University. Achim's key areas of expertise include analytical biochemistry, protein chemistry, liquid chromatography, mass spectrometry and bioinformatics. 

 is co-founder of LC Packings and is currently Managing Director of Dionex Benelux (Amsterdam, The Netherlands). His interests include column technology for nano- and bio-LC and LC–MS application development. 

 received her PhD degree in Chemistry (specializing in biological mass spectrometry) from the University of Manchester, United Kingdom. After postdoctoral research positions at the National Institutes of Health and California Institute of Technology she joined Thermo Fisher Scientific as a Proteomics Applications Specialist working on the Orbitrap mass analyser. 

 is the mass spectrometry Marketing Manager for Thermo Fisher Scientific in Europe. He has over 20 years of experience working with biological mass spectrometry. 

 is a professor in the Materials and Chemistry department of Vrije Universiteit Brussel and is responsible for the surface analysis facilities. He is also based part-time as a professor at TUDelft where he is cluster leader of the research on durability of materials within M2i.  

received his PhD degree in Chemistry (specializing in Analytical Chemistry) in 2005 from the University of Amsterdam, The Netherlands. After his postdoctoral research at the University of California in Berkeley, he joined Dionex and conducted research on packed and monolith column technology for ultrahighpressure LC, two-dimensional LC and nano-LC. In 2009, the National Fund for Scientific Research (FWO, Belgium) awarded him an Odysseus grant. He currently holds a position as assistant professor at the Department of Chemical Engineering at the Vrije Universiteit Brussel. 

Articles

Describes the preparation of a reversed-phase monolithic column with an optimized porous structure to separate intact proteins using LC–MS.

The Potential of Polymer Monolithic Capillary Columns for the LC–MS Analysis of Intact Proteins

One of the privileges of teaching liquid chromatography (LC) classes for a living is that I get to meet chromatographers from around the world during my travels. As I write this instalment, I'm just finishing six weeks of classes, where I have had a chance to discuss problems that attendees were having with a variety of LC applications. Although the topics in this column originate in Poland, Turkey and Argentina, they are no different than those that arise in Singapore, Galway, Charlottetown or Huntsville. Let's look at a couple of problems that students had related to both conventional LC and ultrahigh-pressure LC (UHPLC). These problems apply equally to both techniques. 

One of my students from Poland, who was just starting out using UHPLC, asked if she could inject 5 µL of sample onto her UHPLC column without any problems. Her conditions comprised a 75 mm × 2.1 mm column packed with 1.8-µm diameter particles operated at a flow rate of 0.8 mL/min. It is fairly easy to perform a few simple calculations and answer her question.  

A good rule of thumb (1) is that you can inject ≈ 15% of the volume of the first peak of interest without causing a more than ≈ 5% increase in peak broadening, as long as you use mobile phase for the injection solvent. We need to determine the volume of the first peak of interest to see how this applies in the present case. We will assume a retention factor, 

, of 1 for the first peak; you can make similar calculations for other retention factors.  

First, we need to estimate a column plate number (efficiency), 

 is the column length in millimetres and 

 is the particle diameter in micrometres. A wellpacked column tested with an ideal solute will give larger values of 

, but equation 1 is a reasonable estimate for real samples. With our 

Next, it is necessary to estimate the column volume, 

is the column diameter in millimetres. For the 75 mm × 2.1 mm column we get 0.165 mL. From this we can get the column dead time, 

, by dividing by the flow rate, 0.8 mL/ min: 0.21 min. (Note: If you are trying to repeat these calculations, carry along several extra decimal places of precision — I'm rounding the numbers for simplicity of presentation). 

With the knowledge of the column dead time, we can calculate the retention time, 

= 1 peak by rearranging the equation for 

The next to last step is to estimate the peak volume of the first peak. We estimated 

 with equation 1, but if we were measuring 

 from an isocratic chromatogram, we would use

 is the baseline width of the peak measured between tangents drawn along the sides of the peak. Solving equation 4 for the peak width, we get

Using our 15% rule of thumb from the beginning of our discussion, an injection of (0.15 × 11.8 µL) = 1.8 µL, or ≈ 2 µL would be allowed in the mobile phase. This is less than half the desired 5-µL injection, so it is expected that excess peak broadening would be observed with a 5-µL injection. Armed with this knowledge, I would perform a series of injections of 1, 2, 3, 4 and 5 µL and observe the effect on the peak width of the first peak of interest. I agree wholeheartedly with Izaak Kolthoff, the father of analytical chemistry, who is credited with the quote, "Theory guides, experiment decides." So, use the results of the injection-volume experiment for the final decision. 

The above example illustrates one of the challenges of UHPLC — reducing the sample volume sufficiently to avoid band broadening. Perhaps this could be accomplished if any preceding samplepreparation steps involved reconstitution of a dried extract. For example, if an evaporation-to-dryness step was followed by reconstitution in 50 µL of mobile phase, this volume could be reduced to 10 µL and now a 1-µL injection would contain the same mass of sample as 5 µL of the 50-µL volume. 

Because the band broadening effect is worse for the first peak in the chromatogram, it may be possible to mitigate the problem by increasing the retention of the first peak. If you repeat the above calculations with several different 

-values for the first peak, you'll see that a 5-µL injection would be allowed with 

 = 4–5. This would likely be too long to wait (≈ 1 min) for the first peak for most impatient chromatographers — after all, you bought that UHPLC system so you could have short run times! 

A third option to solve this problem might be to use on-column concentration. When the injection solvent is more than 20% weaker than the mobile phase, it is possible to inject large volumes of sample because the weak solvent causes the sample to accumulate at the head of the column until the injection solvent washes through. For example, if the mobile phase were 65% acetonitrile and 35% buffer, injection in ≤45% acetonitrile would be expected to allow injection volumes much larger than the 2-µL injection calculated above. Thus, if the desired injection of 5 µL were in mobile phase, you could dilute the sample two-fold with buffer and inject 20 µL of the sample (now in ≈ 33% acetonitrile) and put the desired sample mass on the column with less chance of unwanted band broadening.  

As with the conclusion of the previous section, try one or more of these alternative solutions and see what works. 

A student from Buenos Aires was using LC for the analysis of a drug product. He was having problems with carryover, so he was experimenting with different autosampler wash solvents to see if he could mitigate the problem. With some wash solvents he noticed significant peak broadening for his sample, whereas others did not cause problems. Unfortunately, when he used the most effective wash solvent, 100% methanol containing 0.1% formic acid, he observed the worst peak shape. He wondered how he could select a wash solvent that would reduce carryover, yet maintain good peak shape. 

One of the most popular autosamplers in use today is the needle-in-loop design. This configuration typically uses a 100-µL sample loop that has the injection needle as a part of the loop. To inject, the needle is connected to the LC mobile-phase flow path with a highpressure seal. The nice thing about this design is that, because the needle is in the flow path, 100% of the sample (up to the loop volume) is injected. However, the entire loop contents of 100 µL are injected every time, even with a small sample injection. For example, if a 1-µL injection is made, the loop contains 1 µL of sample plus 99 µL of solvent. Usually the loop is backflushed onto the column so that the sample goes onto the column first, followed by the remainder of the loop contents. If the loop is filled with mobile phase (or a weaker solvent), there is no problem; it is as if a 1-µL sample loop were used. However, if the solvent in the loop is stronger than the mobile phase, the injection plug can be distorted and cause band broadening. Obviously, it is desirable to avoid injecting an excess of strong solvent to avoid unwanted band broadening.  

So, how do we avoid injecting too much of a strong solvent? One way is to perform the wash cycle while the sample loop is in the flow path. With most LC systems, when the injection is made, the injection valve stays in the inject position during the run. If the autosampler wash cycle is performed while the injection valve is in the inject position, the remainder of the autosampler is washed with the wash solvent, but the sample loop is not.  

The sample loop is thoroughly flushed with mobile phase during the run, which is usually sufficient to clean it before the next injection — after all, the sample is soluble in the mobile phase, so the mobile phase should be good enough to flush the loop. If, however, the loop is washed in the load position, the wash solvent will remain in the loop when the next sample is aspirated. Thus, any remaining wash solvent will get injected with the sample. If this wash solvent is too strong, peak distortion is likely. Some autosamplers provide for the use of two wash solvents. If this is the case, the solution to the problem is simple. First, wash with the strong wash solvent in the load position to remove any residues from the sample loop, then wash with mobile phase (or a weaker solvent). Now when the sample is picked up, the remainder of the loop will be filled with a non-band-broadening solvent, so peak distortion should not occur. A final option would be to use a smaller-volume sample loop, if this is an option with your autosampler. Using a filled-loop injection with a 1-µL loop volume, for example, would allow injection of 1 µL of sample and because none of the wash solvent would remain in the loop, the choice of wash solvent composition would be of little concern. 

In Turkey, a student complained that he had purchased a reference standard that was claimed to be >99% pure in the certificate of analysis, but when he ran it on his LC system, he found impurity peaks that greatly exceeded the <1% claim. He showed me a chromatogram that he obtained with a large analyte peak and a smaller impurity peak that indeed was more than 1% of the reference standard. He also had a copy of the certificate of analysis that contained a chromatogram showing a single peak that looked similar to the one I've recreated in Figure 1. The mobile phase was 90% acetonitrile, 10% water on a C18 column for the certificate of analysis; unfortunately, I did not note the conditions for his analysis.

 Figure 1: Reconstructed chromatogram for a certificate of analysis. The arrow marks the column dead time.

However, even with this limited amount of information, it is possible to make an educated guess as to the source of the problem. If you recall from previous instalments of this column, I've stated many times that the ideal isocratic chromatogram will have peaks in the 2 < 

 < 10 range, or if this is not possible, 1 < 

 < 20. This gives the analytes sufficient time to interact with the column to achieve "good" chromatography, yet keeps the run time from being excessive. Normally, we calculate the retention factor, 

 are the retention time and column dead time as noted in equation 3. I did not include a time axis in Figure 1, because I can't remember what it was in the certificate of analysis, but the units in equation 6 cancel, so we can use a ruler to calculate 

 from the chromatogram. The first disturbance in the baseline (arrow in Figure 1) marks 

 ≈ 0.3. This is much less than the desired minimum 

 << 1, there is little time for the analyte to interact with the column, and the chances of having unresolved peaks present is increased. This is especially important when using a chromatogram to certify peak purity. The small tail on the peak may be peak tailing that is normal with many peaks or it may be a subtle indication that an impurity is present. There is no excuse for producing a chromatogram like this for a certificate of analysis, except laziness, ignorance or impatience. It would be easy to increase 

 < 10 by changing to a weaker mobile phase, such as 70% acetonitrile, 30% water. At that point, a more convincing case could be made to show that the peak was indeed >99% pure. Although I don't remember the conditions of the user's chromatogram, I do remember that the analyte peak was significantly broader than that in the certificate of analysis, suggesting that the retention time was much larger. My answer to the question was that the certificate of analysis didn't convince me of the purity of the reference standard, so I would be more likely to trust the analyst's chromatogram in which 

It doesn't matter where we live; LC problems have no international boundaries. In fact, I've noticed that chromatographic terminology is often adopted directly into many different languages. In some ways, chromatography really is a universal language. It reminds me of a time when a visitor from one of the Asian countries visited my son's classroom when my son was about 12 years old. The visitor asked if the kids wanted to learn some of his language. All were excited about the prospect. The visitor carefully recited "computer", "Coca-Cola" and "McDonalds." Maybe he should have included "chromatography." 

, is vice president of LC Resources in Walnut Creek, California, USA. He is also a member of 

 editorial advisory board. Direct correspondence about this column should go to "LC Troubleshooting", 

, 4A Bridgegate Pavilion, Chester Business Park, Wrexham Road, Chester, CH4 9QH, UK or e-mail the editor, Alasdair Matheson, at 

(1) L.R. Snyder, J.J. Kirkland and J.W. Dolan, 

Introduction to Modern Liquid Chromatography, 3rd Ed.

 (Wiley, Hoboken, New Jersey), 120 (2010). 

LC Troubleshooting

No matter where LC is used, it is never immune to problems.

LC Trouble Everywhere

The previous instalment of this column on headspace sampling for gas chromatography (GC) (1) discussed the basic physicochemical processes that determine the contents of the gaseous headspace above a sample in a sealed vial. Although it is possible — and sometimes even desirable — to sample and inject headspace directly into a GC system with a hand-held syringe, most of the time analysts will call upon an automated headspace sampler to acquire and transfer appropriately sized fractions of the headspace into a GC system's inlet. Aside from providing the benefits of better accuracy and repeatability similar to what's gained with a liquid autosampler compared to manual injections, automated headspace samplers control other crucial variables beyond timing. Temperature, pressure and flow rates in addition to the timing parameters, strongly influence the composition of the headspace gas as it appears at the GC column or inlet. Headspace autosamplers and the associated sampling methods are more complex than their liquid autosampler counterparts; they comprise a richer set of choices that chromatographers should understand and manage to obtain the best possible results.  

Most analysts are familiar with the inner workings of a liquid autosampler. Basically, the device is a robotic arm that actuates a microsyringe with better precision and accuracy than a mere human operator could achieve. Peering underneath the covers, it's not difficult to see and understand the motions and gyrations that a liquid autosampler undergoes as it processes sample after sample. 

A headspace autosampler is different. The mechanical tasks of vial positioning and needle movement in and out of the vial septum are familiar, but there is much more going on than meets the eye. The inner workings of a headspace autosampler are hidden underneath insulation and internal housings. It's not easy to observe the device's pressures and temperatures without attaching thermocouples and gauges at inconvenient locations. Of course, the designers and engineers who created the autosampler would have done so as part of the development process. For analysts, these physical properties are abstracted to the list of settings and timings that make up the sampling method. What, then, are the relationships between the sampler settings and the physical processes that result in sample injection? 

Modern autosamplers that perform equilibrium headspace sampling enact three fundamental steps for sample injection: equilibration, pressurization and sample transfer. Each sampler model supplements these three with additional nuances and enhancements. Certainly, there is more to discuss about these details, but that extends beyond the scope of this article. The three principal steps can be delineated as follows. 

Equilibration is the most important headspace sampling step. Careful attention toward choosing appropriate equilibration temperatures and times during method development helps ensure a robust and longlived procedure. 

 is the period between when a vial is filled with sample and sealed and when the sampler needle enters the vial to commence sample acquisition. During equilibration, sample components partition between sample and headspace, and equilibrium is reached after their concentrations attain constant values. Each component migrates at its own temperaturedependent rate, so the slowest-moving component of interest determines the minimum equilibration time period. 

Very volatile components start to migrate out of the sample phase immediately during sample preparation, and significant amounts can be lost to the atmosphere with concomitant reduction in sample recovery along with poor repeatability. Sample handling and preparation at reduced temperatures can help reduce such pre-vial losses. Careful attention and consistent sample handling procedures aid in improving repeatability. 

In addition to different migration rates across the sample-headspace interface, each component has its own temperature-dependent solubility coefficient, 

. For most commonly encountered solute-solvent systems the solubility coefficient decreases as the temperature increases, thereby raising analyte concentrations in the headspace at higher temperatures. Thus, careful control of vial temperatures during equilibration is important for obtaining both consistent quantitative results and equilibration times. 

The contents of headspace vials may attain an initial equilibrium state at room temperature in an external sample tray while waiting for the actual headspace process to begin, but they will cease to be in equilibrium once inserted into a heated carousel before sampling. As soon as the temperature changes, the contents of the headspace will shift in response to the changes in solubilities. In certain cases, the prethermostating wait time may influence the overall equilibration time; method developers may need to take this into account and allow sufficient equilibration time for the shortest anticipated wait time between vial preparation and the start of equilibration. 

Vial heating is one of the most important aspects of a headspace sampler. Early samplers used water or even oil baths for consistent results, but modern headspace devices employ either a solid metal heating carousel or, in some cases, a carefully controlled air bath oven arrangement. A vialtovial temperature control precision of about ±1–2 °C should yield acceptable results consistently. Absolute temperature accuracy influences inter and intralaboratory reproducibility of results. Headspace samplers should be calibrated consistently to a standard reference thermometer about every six months if sample calibration procedures cannot establish sufficient results comparability.  

Headspace sampling temperature dependencies lead to the general use of elevated equilibration temperatures. Higher equilibration temperatures are desirable for both greater headspace sensitivity and shorter equilibration times. Choosing an equilibration temperature of at least 15 °C above room temperature will ensure good thermal control, but there are limits to how high equilibration temperatures can be pushed. 

Elevated temperatures can accentuate the breakdown of thermolabile materials, and a careful study of such effects should be performed when there is any question of analyte stability. In addition, high temperatures may degrade the performance of headspace vial septa that are not designed for hightemperature applications and cause both leaks and sample contamination with septum bleed components. 

Many headspace samplers stir or agitate the vial to speed up equilibration. As components leave the liquid phase at the gas–liquid interface, the reduced solute concentrations in the interfacial liquid layer are replenished relatively slowly by molecular diffusion from the bulk of the liquid. Convective stirring occurs naturally to some extent, but mechanical mixing is more effective toward replenishing solute that has entered the gas headspace phase. 

Another side effect of elevated equilibration temperatures is an increase in the internal or "natural" vial pressure before the actual sampling step. Most of this pressure is created by the vapour pressure of the solvent. Water as a headspace solvent has a vapour pressure of close to 1 atm (101 kPa) at 100 °C, and other more volatile solvents can create similar or higher pressures if present in significant quantities. Internal vial pressures increase exponentially as temperatures increase, and it is possible to generate pressures high enough to breach the vial septum seal or even burst the vial itself. 

The natural vial pressure can be measured with a needle-mounted pressure gauge while a test vial with sample is heated in a small external chamber. Most chromatography suppliers and manufacturers can provide such equipment. When high pressures are anticipated, use a selfventing septum safety cap to prevent equipment damage. It is better to have to repeat analyses than try and remove broken glass from a delicate piece of equipment. 

Besides the danger of broken glass or lost sample, natural vial pressures that are too high can affect the pressurization stage of headspace sampling by causing sample vapours to depart from the vial too early in the sampling process. Such an occurrence causes doubled peaks, poor repeatability or both. 

After the contents of a headspace vial have reached equilibrium, the headspace gas is ready to be transferred into the GC inlet. The most common mechanism to effect this transfer involves pressurization of the headspace in the vial with inert gas via a hollow heated needle, followed by release of the pressure back into the sampler pneumatics via the same needle. This approach is different than the usual manual headspace sampling procedure, in which a gas-tight syringe withdraws a portion of the headspace by outward displacement of the syringe plunger. Of course, it is possible to fill a manual syringe with inert gas, displace the gas into the headspace vial, and then withdraw it.  

The automated pressurization method establishes a consistent vial pressure. In the sample transfer phase, this pressure drives sample through the autosampler internal pneumatic tubing directly into the inlet or column in a balanced pressure system, as presented in Figure 1. In a loop-type system (see Figure 2), the vial pressure moves sample gas into an internal sample loop from which it enters the GC inlet upon switching a sampling valve. The majority of applications use carrier gas directly for pressurization, although it is possible to utilize an independent gas source for pressurization and transfer in certain situations, such as when the carriergas inlet pressure is too high for the headspace vials or septa to withstand or when inlet split flow is supplied from a separate flow controller. 

Pressurization should establish a vial pressure that is greater than the natural vial pressure. If not, then sample headspace gases will flow out of the vial into the needle upon the needle's entry into the vial. This initial extraneous reverse sample pulse can either enter the column directly and cause double peaks or can be lost in the subsequent sample transfer step. Excessive vial pressurization can open septum safety seals or even burst the vial, so the pressurization set-point is subject to the same limitations as the natural vial pressure. 

A short pressurization delay time before sample transfer out of the vial allows the introduced gas to mix thoroughly with the gaseous headspace contents. An interval on the order of 30 s or so is suitable. The added gas will act as a sample diluent in a loop-type system when the pressure of the sampled gases is vented before column introduction. The higher the pressurization level, the more dilution will occur upon venting. 

After pressurization, sample gas from the vial flow is allowed to flow back into the autosampler pneumatics via the vial needle, driven by the pressure gradient that was established during pressurization. If the sample needle is connected directly to the column, as in some configurations of a balanced pressure system, then the gas in the vial effectively becomes the carrier gas source during the sample transfer interval. The vial pressure decays during sampling. Longer sampling intervals will yield larger injected amounts, up to the point where the column flow decreases more than is desirable. 

With proper calibration and precise timing, this arrangement will yield very reproducible results as long as the vial pressure does not decay excessively. For short sampling intervals with column flows under about 2–3 mL/min, solute amounts entering the column or inlet are roughly proportional to the sampling interval. Long sampling intervals will produce incrementally larger peaks, but the column and inlet system's requirements for injected peak widths may impose a practical upper limit on the sampling interval, especially for early, weakly retained peaks. In some cases, an intermediate solute trap can adsorb and, upon rapid heating, release solutes in a narrow band that helps extend the maximum sampling interval. 

For headspace samplers with a sample loop, pressurized sample gas from the vial is directed through the sample loop to vent through a restrictor. A minimum sampling interval is required to flush the loop with sample gas. The total sampling interval and initial pressurization determine the residual loop pressure at injection. Longer sampling intervals result in lower loop pressures, which then yield lower injected peak areas as an increasing fraction of the headspace vial gas is vented to the surroundings. As with the direct injection type of system, careful control of pressurization and sample interval timing is required for the best results. 

Upon injection, the loop contents are switched into the carrier gas stream and carried along to the GC inlet or column directly. 

Variability in the bulk headspace gas composition — such as the presence or absence of significant amounts of solvent in samples compared to standards — can affect the flow rate of sample gas from a headspace vial either directly into the column or through a sample loop and into a restrictor. Added solvents may change the viscosity of the gas mixture, which then affects the flow rates accordingly. Thus, sample and standard headspace vials should contain solvent systems that are as closely matched as possible. 

There are a few different mechanical arrangements employed in headspace autosampling systems. Headspace samplers are available that recreate hand motions with a captive syringe, similar to liquid autosamplers. These systems are relatively simple to understand by mimicking laboratory technicians. The two other types of headspace samplers are more pertinent for this discussion: balancedpressure and loop-sampling. Figures 1 and 2 illustrate the salient features of these two designs, albeit in a simplified fashion that omits a number of nuances that go beyond the limits of this column. Despite these simplifications, the diagrams should provide a better understanding of what's hidden beneath the covers of many automated headspace sampling systems. 

Figure 1 shows how a direct-injection or pressure-balanced headspace sampler works. The vial transport and thermostating details are not included: Readers can assume that the vial as shown is at its equilibrium temperature.  

The standby configuration [Figure 1(a)] is in use at times other than when actively sampling. In this configuration, carrier gas flows into the column or inlet at the normal column flow rate through a heated transfer line that encases inert deactivated metal or fused-silica transfer tubing, or even the first meter or so of a fused silica column itself. If valve V2 is opened then a secondary flow of carrier gas transits the integrated needle carrier and exits to vent through a restrictor (not shown). The carrier supply is normally pressure controlled for capillary or packed columns. If an inlet splitter is employed, then the split flow is supplied separately at the GC inlet in this type of system. 

Figure 1: Reconstructed chromatogram for a certificate of analysis. The arrow marks the column dead time.			

When the system is ready for pressurization, the needle is lowered by a motorized mechanism (omitted) [see Figure 1(b)]. As the needle moves downward past the lower seal, carrier gas begins to flow through it. Carrier gas flows into the vial once the needle has penetrated the septum. The carrier gas will continue to flow into the vial until the internal vial pressure is equal to the carrier gas pressure. During the pressurization interval, carrier gas continues to flow out through the transfer line and column as well. 

At the end of the pressurization time, after the vial has reached the carrier gas pressure and the sample and carrier gases in the vial have mixed sufficiently, valves V1 and V2 are turned off [see Figure 1(c)]. The incoming carrier-gas supply and the outgoing vent are cut off, and gas from the headspace vial, shown here in dark blue, is directed straight to the column or inlet via the heated transfer line. Sample injection occurs at this moment. 

Carrier gas flow to the column is re-established after the sampling interval has elapsed by activating valve V1 again [see Figure 1(c)]. Remaining pressure in the vial may be optionally vented by lowering the needle until its top hole clears the lower seal, as shown in Figure 1(d). Finally, as sample begins to be separated on the column, valve V2 is opened and the needle is withdrawn from the headspace vial. Sample gas that remains in the needle is flushed out by momentary outgoing carrier flow as the needle moves upward, until the lower hole enters the sealed area inside the needle carrier. The system has returned to the standby position and remains there until the next sampling cycle begins. 

Simplified loop-sampling headspace system pneumatics are shown in Figure 2. A six-port sampling valve selects whether sample from the headspace vial or carrier gas to the column flows through the sample loop. A separate auxiliary input provides gas for headspace vial pressurization, although in practice carrier gas usually is connected to both inputs. The entire pneumatic path and valving are heated. In the standby position, shown in Figure 2(a), carrier gas flows directly to a heated transfer line and onward to the column or inlet. The sample loop and headspace needle can be purged with auxiliary gas by activating valve V1 as shown in the figure. 

Figure 2: Loop-type headspace sampling. (a) standby; (b) pressurization; (c) loop filling; and (d) injection. Orange connecting lines denote inert gas flows; blue lines denote sample gas.			

When the system is ready for vial pressurization, the needle is lowered into the vial; in some designs, the vial is raised up over the needle. With valve V1 open and V2 closed, pressurization gas flows into the vial until the set pressurization level is reached [see Figure 2(b)]. After the pressurization time has elapsed, the contents of the vial headspace are ready for transfer into the sample loop. During this time, carrier gas continues to flow directly to the column. 

To initiate sample transfer into the sample loop, valve V1 is closed and valve V2 is opened [see Figure 2(c)]. Now, the pressure in the headspace vial drives sample gas into the sample loop and outward through valve V2 and a vent restrictor (not shown). When sufficient time has elapsed to fill the loop with a representative sample, valve V2 is also closed. A brief loop equilibration or settling time ensues. 

At this point, the system is prepared for injection. The six-port valve position is switched, which inserts the sample loop into the carrier gas flow and sends sample onward to the column or inlet, as illustrated in Figure 2(d). The duration of the injection period should be sufficient to transfer all of the loop contents into the transfer line. 

The sample amount can be controlled by changing the size of the sample loop or reduced by using higher pressurization levels along with longer sample loop transfer times, which will dilute the sample with more auxiliary or carrier gas. It is also possible to set up a short injection time that causes partial sample loop transfer, but this may impact results repeatability in an unfavourable way. 

Valve V2 can be left open during the sampling interval to vent the remaining headspace gases from the vial, or it can be closed when the six-port valve is actuated. After sampling has finished, the six-port valve is returned to its original position and the system returns to its standby state while the sample components are separated. If valve V2 is open at this point, any sample remaining in the loop will be vented out through the needle. 

GC Connections

The article discusses the fundamentals of instrumentation for headspace sampling, including vial thermostating, pressurization and transfer to the GC inlet.

Headspace Sampling: Instrumentation

In this month's instalment, we look at the results of a recent survey on high performance liquid chromatography (HPLC) columns, examining the trends in analytical and preparative column use and purchasing patterns since the last survey. Factors investigated in this study were mode and stationary phase usage, particle sizes and column dimensions, column budgets and the factors influencing buying decisions. Column lifetime and use of guard columns is also considered. Future column buying plans were explored. 

surveys its readers to obtain a current profile of users of high performance liquid chromatography (HPLC). The last survey on HPLC columns was conducted in 2009 (1). In the past, I have used results from these surveys to chart trends in column technology and in the practice of HPLC. In late 2010, a web-based survey was sent to subscribers whose primary chromatography technique was HPLC. A total of 324 readers responded and this number was statistically sufficient to allow comparisons to previous survey data and to investigate trends.  

To ensure that the results of the current survey were compatible with those of previous surveys, I used the same methodology to report the results. Because many of the questions allowed respondents to give more than one answer, in some cases, I normalized response totals. Normalizing the results to a base of 100% makes it easier to compare the results of previous surveys with those of the present survey and to identify trends in the use of HPLC columns, modes and packings. Questions pertaining to mode usage, column lifetime, particle size usage, purchasing considerations and possible future needs were explored to understand selection criteria. 

To understand the current usage rate of instruments, a question was asked pertaining to number and type of instruments personally used per respondent per week. Questions in previous surveys were less specific and laboratory managers therefore occasionally responded with estimates for their entire laboratory, which tended to skew the results. Table 1 shows the average response for each category of instrument. The numbers reported should not be construed to mean that every respondent uses every type of instrument; rather, they serve to provide an idea of the relative number of the various types of systems in general use. Later, I will look at the numbers of columns used by these instruments. 

  Table 1:  Types of HPLC instruments used by survey respondents. 

According to the survey results, a user of conventional HPLC instruments is responsible for a weighted average of 2.6 units. Because each respondent may have been responsible for a single instrument or for multiple instruments, this weighted average was derived from looking at the total number of instruments of this type identified by all respondents and then dividing this number by the total number of respondents. For example, some respondents reported that they have only one conventional HPLC system while others reported that they have five instruments for which they are personally responsible. Based on the relative numbers of Table 1, for every one of these respondents that used a conventional LC system, only 1 in 16 would possess a capillary or nanoLC system; 1 in 4 might have an instrument dedicated to microbore columns; and 1 in 7 could have a preparative instrument in his or her laboratory. These numbers are slightly different from the 2009 survey (1), especially for conventional users, which was down from 3.6 per respondent (this was probably skewed by the laboratory manager contribution discussed above). 

Most liquid chromatographers use more than one chromatographic mode to solve specific separation problems. Because respondents were allowed to choose multiple modes, the total number of techniques used by respondents revealed that each used 3.2 modes on the average, in agreement with the last survey (1). This result doesn't mean that everybody is using all modes all the time, but that occasionally another mode may be used to solve a difficult separation problem. If one looks at the normalized data in Table 2, which compares the relative mode usage in 1997, 2007, 2009 and 2011, it should be no surprise that the results again show that reversed-phase chromatography was, by far, the most popular mode; according to the raw data, more than 93% of respondents use this mode and normal bonded phase usage was a distant second with 31%. These data have been corrected for those using ion pair chromatography (a subset of reversed-phase chromatography) and cyano columns for reversedphase work. In earlier surveys, cyano columns were presumed to be mainly used in normal-phase separations, but this survey revealed that the predominant use of CN columns was in reversedphase chromatography where this phase serves well as a short chain phase with unique selectivity. 

Table 2:  Analytical HPLC mode and stationary phase usage. 

Although the data in Table 2 shows that reversed-phase chromatography has dropped on a relative basis, this drop is not caused by its decrease in popularity (which remains extremely high), but rather results from the increased use of other modes such as chiral and hydrophilic interaction chromatography (HILIC). HILIC, where respondents indicated a doubling in usage since 2007, is a separation technique for highly polar analytes that gets around some of the problems associated with reversed-phase chromatography, such as low retention or phase collapse (dewetting). HILIC uses a polar stationary phase such as bare silica gel, polar-bonded phases (for example, diol) or specialty phases with dual functionality and requires a high percentage of a nonpolar mobile phase, usually acetonitrile, similar to the requirements for normal-phase chromatography. However, unlike normal-phase chromatography, which uses nonpolar solvents like hexane and methylene chloride and tries to exclude water from the mobile phase, HILIC requires some water in the mobile phase to maintain a stagnant enriched water layer on the packing surface into which analytes may selectively partition. In addition, water-miscible organic solvents are used. Under HILIC, polar analytes are well retained and are eluted in order of increasing hydrophilicity. Bare silica, specified for adsorption chromatography in the survey, is still a widely used HILIC phase, but the current survey did not explore its use in HILIC applications. 

Supercritical fluid chromatography (SFC), which employs columns similar to those used in HPLC, has also seen renewed interest, especially for preparative chiral separations. The use of chiral columns in SFC was not explored in the present survey, although a few respondents "wrote in" their use of HPLC columns in this increasingly popular chromatographic mode. 

The use of ion-exchange chromatography showed a slight upward swing in usage and is back to its historical levels. Included in this usage is ion chromatography, which is sometimes treated as a separate separation technique but in reality is ion-exchange chromatography because it employs the same separation principles and uses columns with ionic or ionizable functionalities. Ion-exchange chromatography is frequently used in proteomics where a large number of tryptic peptides are fractionated by 2D LC first with ion exchange followed by reversedphase chromatography–mass spectrometry (MS) for separation and detection. It also finds use in amino acid separations and protein purification protocols. Although ionpair chromatography also separates ionic and ionizable compounds, it is generally lumped with reversedphase chromatography because it employs the same packing materials. Many chromatographers familiar with reversed-phase chromatography prefer to use ion-pair chromatography for ionic and ionizable compounds because the columns are more familiar to them and the principles of reversed-phase chromatography can be generally applied in method development. 

The normalized data in Table 3 further breaks down the reversed phases into individual organic moieties and clearly show that C18 (octadecylsilane) was most popular (the raw data showed that 92% of all users use this phase at some time) followed by C8 (octylsilane). The use of phenyl-bonded phase has showed steady usage as chromatographers exploit its unique selectivity compared to alkyl phases. The phenyl-hexyl and diphenyl phases that have seen some interest were not explored this time, but they will be included in future surveys. For the first time, I expanded the survey to include fluorinated phases that 14% of respondents mentioned that they have tried for separations where the conventional reversedphase chromatography materials didn't do the job. In addition, graphitized carbon was added in this year's survey; only 4% of respondents indicated that they had tried this specialized material that is one of the more rugged reversedphase packings. Again, just as in 2009, cyano phases used in aqueous solution showed quite strongly in the survey as a bonafide reversedphase packing. 

Table 3: Stationary phases used in reversed-phase chromatography. 

Particle Sizes in Analytical and Preparative LC

Since the beginning of HPLC in the late 1970s, the trend has been to use columns packed with smaller spherical particles. Throughout most of the 1970s, 10-µm irregular particles were the norm because they were commercially available and slurry packing column efficiency was quite good. In the late 1970s, 5-µm spherical particles gave better performance because of improved mass transfer and better packing ability. Through occasional 

 column surveys, I started tracking particle size usage in the mid-1980s. Table 4 provides data of the use of various particle sizes from 1985 until 2011 (current survey). As one can see from the first survey, columns packed with 10-µm particles had already given way to 5-µm particles. These particles maintained their dominance through the 1990s and are still in widespread use today. There are many validated methods that have been developed on 5-µm particles, and users don't want to take the time to revalidate their method unless there is a strong driving force to do so. Nevertheless, as can be seen in Table 4, the use of even smaller particles (<2–4 µm) began in earnest in the mid-90s and now these particles have caught up and surpassed the 5-µm particles. With the reduction in particle size, column efficiency improved further allowing a reduction in column length. Shorter columns are now in vogue. Not only does a shorter column provide faster separations but solvent use is reduced. Recently, the use of sub-2-µm particles has become fashionable in even shorter column lengths. In just a short time, the use of the sub-2-µm columns has doubled. Columns of 30–50 mm in length packed with these particles now provide plate counts formerly obtained on 15- and 25-cm columns packed with larger particles. 

Table 4: Analytical HPLC particle size trends 1985â2011*. 

In the current survey, we queried the use of 2–3 µm porous particles that have attracted the interest of some users. These columns are intermediate between the popular 3.0–3.5 µm columns and the newer sub-2-µm columns. They show lower pressure drops and efficiency than the sub-2-µm particles but better efficiency and higher pressure drop than the 3.0–3.5-µm particles. Since the last survey, the increasing use of the superficially porous particles (SPP) (also referred to as solid core, core–shell and the like) has taken place. Even though the question for the survey was based on use of porous particles, I suspect that the respondents also took into account their increased use of SPP particles because the number reported in Table 4 was quite high for the relative usage of 2–2.9 µm particle columns. The SPP columns (with particle sizes of 2.6–2.7 µm) have quickly become a partial substitute for the sub-2-µm columns because they offer similar efficiency at half the pressure drop. Hence, users of conventional HPLC equipment can take advantage of ultrahigh-pressure liquid chromatography (UHPLC) separation capability. 

In comparing the use of preparative chromatography to analytical HPLC, 62% of those performing analytical separations do not use preparative chromatography. Not surprisingly, those that perform preparative separations tend to prefer larger particles as depicted in Table 5. Reasons for the preference of larger particles (10-µm and higher) are lower cost, lower pressure drop and loadability. For the latter consideration, the higher efficiency of smaller particles may be lost when overloaded conditions are employed. Even so, in the present survey, there has been an increase in respondents using smaller particles for preparative work (Table 5). For example, 18% of respondents use 5-µm particles for their preparative work, up from 9.7% in 2007. 

Table 5: Particle sizes used in preparative chromatography. 

Column Lengths and Diameters in Analytical HPLC

In this year's survey, we included a question about column lengths and diameters used in analytical HPLC. For many years, when 5-µm particles were dominating analytical work, 25 cm × 4.6 mm columns were the standard but now the 15 cm × 4.6 mm column is the "workhorse" dimension. Table 6 provides the top five column dimensions chosen by survey respondents. The rise of 2.0–2.1mm i.d. columns in 50- and 100-mm lengths undoubtedly results from the widespread acceptance of LC–MS, in which flow rate requirements of MS inlets require smaller inner diameters. The 3.0-mm i.d. columns are a good intermediate size that allows the use of conventional LC equipment, yet saves solvent because of smaller flow rate requirements compared to 4.6-mm i.d. columns. 

Table 6: Top five column dimensions used in analytical HPLC. 

Several questions in the survey related to the purchase of analytical HPLC columns on a per instrument basis. Most of the respondents purchase prepacked columns and very few purchase empty columns and bulk packings, presumably to pack their own. More than 90% of respondents have no budget for empty columns or bulk packings. After prepacked analytical columns, the largest expenditure is for guard columns. About a third of respondents have no budget for guard columns, but for those that do, the average guard column expenditures are about half of those of the average analytical column budget.  

The average annual purchase budget for prepacked analytical HPLC columns per instrument was approximately $2890, about 5% lower than the last survey results (1). For the respondents that do purchase guard columns, the average budget was about $1650. The average cost of a guard column is $80, so this would amount to approximately 20 guard columns per instrument per year or about three guard columns per analytical column (again, not an unreasonable number). Guard columns extend the life of analytical columns; therefore, without these expenditures, users would have to allocate a higher percentage of their budget for prepacked columns. 

Another question broke down the purchases of different dimensions of LC columns. Users were asked about their yearly purchases of conventional analytical columns (3.0–4.6-mm i.d.), microbore columns (2.0–2.1 mm i.d.), narrow-bore columns (1.0-mm) and capillary columns (less than 1-mm i.d.) on a per instrument basis. Table 7 shows the column units used based on a weighted average. For conventional HPLC column users (3.0–4.6 mm i.d.), an average of 7.4 columns per instrument per year are used, up from 5.8 in 2009 (1). Microbore columns that can also be used on some conventional analytical instruments and all UHPLC instruments and are the mainstay of LC–MS users averaged 5.3 columns/year per instrument, up double from the 2009 survey. UHPLC favours the use of microbore columns because of frictional heating effects displayed by large internal diameters and pump flow rate limitations at very high pressures (1000 bar) for some UHPLC systems. In addition, the rise in the use of LC–MS, which favours microbore columns, has helped to spur their growth. One would expect narrow-bore and capillary columns to have lower usage based on the smaller number of available instruments that accommodate these smaller internal diameter columns (see Table 1). 

Table 7: Types of analytical columns used by respondents (weighted average). 

One of the questions asked chromatographers from which source they prefer to purchase prepacked HPLC columns. According to the survey (see Table 8), the chromatography supply houses have increased their lead, mainly driven by the stronger presence of one manufacturer. A continued drop-off in share was noted for the general laboratory suppliers. This year 18% of the users surveyed purchase their columns from this source, a drop of 25% from 2009 (1). General laboratory distributors usually provide less technical assistance than the dedicated chromatography suppliers or instrument manufacturers, but they promote themselves as onestop shopping where users can get solvents, chemicals and laboratory glassware, as well as HPLC columns. The general laboratory distributors sometimes provide additional services such as electronic ordering, local stocking, volume discounts and on-site services. The percentage of users loyal to their own instrument manufacturer has slipped a bit in favour of consideration of purchasing columns from any instrument manufacturer, whose ranking is flat compared to earlier surveys (Table 8). According to the present survey, a further drop-off in share was noted for the independent column manufacturers who had seen a substantial rise in their share of the HPLC columns' market in the last survey (1). Surprisingly, a few more people are packing their own columns, perhaps driven by the faltering economy. 

Table 8: Sources for HPLC column procurement. 

Table 9 lists, on a normalized basis, the factors that respondents considered important when purchasing an HPLC column. Respondents were asked to name the top three factors that influenced their buying decisions. The most important factor — ranked the same in earlier surveys — is column-tocolumn reproducibility. Nobody wants to develop a method only to find out that the column that they purchase later for the same application does not perform in the same manner. However, column reproducibility has definitely improved over the years, driven by competitive pressure, quality initiatives such as ISO and 6-Sigma and a better understanding of the manufacturing processes. The next factor was column lifetime, an issue that will be explored in more depth in the next section. In this survey, taken during one of the worst economic global turndowns in the Western world, price was rated the third most important factor in considering a column supplier, while the company reputation remained in fourth place. The remainder of the factors showed that some variations with column hardware design are considered less important than previously reported, but a variety of phases, speed of delivery and technical assistance of the supplier gained in importance. 

Table 9: Factors considered when selecting an HPLC column supplier. 

As mentioned earlier, column lifetime is an important selection and operating parameter in HPLC. This survey specifically explored column lifetime from two standpoints — lifetime in months and by the number of injections. Although some modern liquid chromatographs track the number of injections, many chromatographers do not. Usually, one will install an HPLC column and keep it in the instrument, performing one or more methods until the column shows a loss of resolution or shows significant tailing that may affect quantitation. Chromatographers may track the time of usage in a logbook. 

The first question that was explored was as follows: What does one see as the failure mode when a column is considered "dead"? Respondents were given a choice of eight column problems that are normally considered to be signs of a failing column. Figure 1 shows a bar graph that depicts the responses for each symptom. Of course, some of the symptoms are related. For example, broadened, tailing peaks manifest themselves as loss of column efficiency, but the most important warning that a column may be failing is a high back pressure followed closely by loss of efficiency. High back pressure may be a problem with the column packing but sometimes it is associated with plugging the inlet frit with foreign matter. Retention time changes can sometimes be related to a loss of stationary phase, an upset of the column equilibrium or pump problems. Rarely does a column show an erratic flow rate when failing; flow rate changes are most often associated with the pumping system. Other written-in responses for characterizing a "dead" column were peak splitting (doublets); results outside of method specification, standard mix doesn't pass, system suitability requirements not met (which are all related); loss of resolution; appearance of ghost peaks; and high RSDs of replication injections. 

As shown in Table 10, since the earliest days the observed analytical column lifetimes have continually increased through the years. I first began asking about lifetime in column surveys in 1985, 21% of the respondents had columns that lasted less than three months. Now only 13% of respondents' columns have lifetimes of less than three months. If one looks carefully at the data of Table 9, nearly 19% of the respondents now experience column lifetimes of at least a year and a whopping 43% have columns that last at least up to 18 months, some beyond. About 17% of the users don't track column lifetime by month at all. It is surprising that with such long lifetimes now being achieved that users are still requiring more than seven new columns a year for each of their instruments so other factors must be influencing their purchase, such as an increased volume of samples or procurement of improved columns. 

Table 10: Average analytical column lifetime for those who track changes. 

When the question was asked about column lifetime based on the number of injections, a larger number (30%) of the respondents reported that they do not track their injections. Table 11 shows the data of those who do monitor column lifetime based on the number of injections. There has been a definite trend in columns surviving more injections over the last few years, especially comparing the data from 1997 to 2009. Some of this increase is because of chromatographers becoming more aware of the limitations of silica-based HPLC columns, using specialty high and low pH columns and exploring polymeric and other column types that are more rugged. In Table 11, only 20% of the respondents get less than 500 injections from a typical analytical column while 30% get up to 1000 injections and 20% get more than 1500 injections. 

Table 11: Percentage of respondents who reported on column lifetimes by the number of sample injections for analytical columns. 

Although the data are not shown in the table, a large percentage of preparative column users do not track monthly column lifetime (62%) nor do they track the number of total samples injected (74%) before discarding a column. Intuitively, one would think that preparative columns would have shorter lifetimes because they tend to be used for dirty samples, at high flow rates and with higher sample loads than analytical columns. Surprisingly, those few respondents using preparative columns that do track lifetime reported that more than 60% of their preparative columns last longer than a year. This greater lifetime perhaps indicates that preparative columns are used less frequently than analytical columns and therefore tend to be stored and used at a later date when the need for preparative applications arises. Another possibility for longer lifetimes would be that preparative columns are substantially more expensive than analytical columns and users may take better care of them. A third possibility is that the larger particles used in preparative chromatography do not generate as high a back pressure as obtained with the smaller particle analytical columns. Higher back pressure tends to limit the column lifetime because more stress is placed on the column materials. 

One way to increase the lifetime of an analytical or preparative column is to use a guard column to protect it from chemical or mechanical interferences. I have explored the use of guard columns in the present survey. Although guard columns have proven to cut the overall costs of analytical columns, 36% of respondents do not budget for guard columns and 31% do not use guard columns at all. As reported earlier, on average, about 20 guard columns are used per instrument per year. Table 12 shows the breakdown of responses for those who use guard columns, when asked how often they change their guard column. Based on these responses, one can estimate that an average user changes the guard column after 280 injections. Depending on the sample cleanliness, this may or may not be often enough. Normally, it is somewhat of a "trial and error" procedure to determine when a guard column should be replaced. 

Table 12: Number of injections before changing guard column. 

Column Watch

The results of a recent survery on stationary phases used in high performance liquid chromatography (HPLC) are revealed.

LCGC Asia Pacific-03-01-2012

The Potential of Polymer Monolithic Capillary Columns for the LC–MS Analysis of Intact Proteins

LC Trouble Everywhere

Headspace Sampling: Instrumentation

Current Trends in HPLC Column Usage