HPLC 2014, held May 11‒15 in New Orleans, LA, developed into a meeting where the lectures dominated the program. About a third of the 400+ posters were from vendors. The exhibition was comprised of less than 40 vendors. Previous HPLC meetings in North America have attracted over 100 vendors. In contrast, the action on HPLC at Pittcon was mostly associated with the exhibition. Yes, there was an exhaustive technical program, but according to my analytical balance, the exhibition was heavier.
The state of LC technology in 2014 and beyond
Prof. Alan Marshall (Florida State, Tallahassee, FL) opened HPLC 2014 with a review of his work in high-field nuclear magnetic resonance (NMR) and mass spectrometry (MS), particularly ion cyclotron mass spectrometry. Two decades ago, the question often heard was, “MS is so powerful, so why do we need HPLC as a front end?” My response was isomers, especially optical isomers. Prof. Marshall pointed out other examples. Even high-resolution MS has trouble with limited resolving power for peaks with different intensities. For example, if one has isotope peaks that are of equal height, they can be resolved with an instrument with a mass resolution of 350,000. This is available today with several high-field MS instruments. However, if one wants to pick up the trace peak with an intensity of only 1% of the neighbor, MS resolution of 3,500,000 is needed. Unfortunately, peaks with a relative intensity of 1:100 are frequent, but multimillion MS resolution is not.
UPLC celebrates its tenth birthday
Ten years ago, Waters (Milford, MA) introduced the ACQUITY® UPLC system, which set new standards for LC separations. Research into LC column technology had shown that columns packed with particles smaller than 2 μm were much more efficient than corresponding columns packed with 3-μm or larger particles. The problem was that the instrumentation to utilize the “sub-twos” was not routinely available. One needed higher pressure and much lower extracolumn band broadening. ACQUITY set the standard (15,000 psi) and the industry struggled to catch up. Today, all the major vendors have competitive premium-performance instruments. However, it is about time for next-generation instruments to appear.
What would these instruments look like? Emerging column technology is a good place to start, since in LC, column technology usually drives instrument design. Prof. Mary Wirth’s HPLC lecture (Purdue University, Lafayette, IN) showed progress made by her lab on long (a few meters) capillary columns (0.5 μm diam) packed with particles with even smaller diameter. These columns generate millions of plates and extraordinary peak counts (nearly 1000). Of course, such narrow capillaries are compatible with very small sample amounts. This rules out preparative applications. Since the flow rate is in the low nL/min range, compact and possibly even small portable analyzers will probably evolve.
Prof. Barry Karger (Northeastern University, Boston, MA) presented a related option: open tubular wall coated (PLOT) capillaries combined with MS detection. The column particles are bonded to the capillary wall. Typical columns are as long as 9 m. These can deliver peak capacities of over 1000 with an 8-hr gradient elution protocol. Prof. Karger reports the narrow i.d. has many advantages with MS detection, including reduced ionization suppression and closer coupling of the electrospray ionization (ESI) interface. Limit of detection (LOD) is about 10 zmol or about 6000 molecules. Single molecule analysis is still in the future, but the Northeastern lab is closing in. One target is circulating tumor cells. These are collected from whole blood with the aid of biospecific coatings on magnetic beads and then lysed with ultrasound.
Since the ultrahigh-performance liquid chromatography (UHPLC) revolution is a decade old, it is appropriate to ask what the future will hold. The most obvious question is: Will the pressure increase significantly? This was addressed in a poster by Ruben De Pauw (Vrije University of Brussels, Belgium). His team constructed a UHPLC with a Pmax of 2600 bar powered by pneumatic amplifier pumps. Joule (viscous) heating of the column limited column diameter to 2.1 mm i.d. or smaller. When multiple columns were run in series, the viscous heating was reduced, since the tubing used to couple the columns appeared to be a good radiator. Also, connecting columns in series reduces the flow maximum inversely, which also reduces heat buildup. However, heat transfer from narrow capillaries to the environment should be rapid, provided the wall is not too thick.
These reports illustrate that wall coated open tubular (WCOT) or packed submicron diameter capillary columns will enable separations of complex samples requiring high resolution from the long columns. However, the same technology could give rapid assays (subsecond peaks) with very short columns.
Column packing technology
Almost 50 years ago, Ron Majors, Ph.D. (retired, Agilent) developed Varian’s Micro Pack balanced density slurry technique for commercial packing of columns with small (10- and 5-μm diam) particles. Slurry packing involved dispersing with sonication of the column packing in a cocktail of brominated and heavier halocarbons. Varian’s Micro Pack technology was supported by protocols requiring empirical optimization for each column packing, tube diameter, etc.
Several lectures at HPLC 2014 showed that packing columns is still an empirical exercise. Yes, the particles are 80‒90% smaller, but Prof. Jim Jorgenson (University of North Carolina, Chapel Hill) showed that empiricism still rules. Interestingly, packing sub-twos(<2 μm diam) into long, narrow columns is favored by a Goldilocks formula on aggregation—too dispersed is not as good as thick slurry that appears to have moderate aggregation. But too big is bad. Further, in packing long capillaries, he broke the columns into thirds and tested each segment. The first segment was significantly more efficient than the middle or last thirds. In a bit of hand waving, he proposed that aggregates are effective in delivering close-packed assemblies to the column tube, where the high forces involved in the initial segment squish the particles into a close-packed arrangement. He also found that radial distribution of the particle size was not uniform when the column packings were not monodisperse.
I was reminded of Dr. Frank Yang’s work on packing capillary columns in the early 1980s. His patent (U.S. patent 4,483,773 ) describes a pressure programmed slurry process for packing capillary columns for LC and also for GC that produced high permeability and efficiency. Also, Dr. Chao Yan (Global Chromatography Co., Ltd., Shanghai, China) uses electrokinetic technology to produce very efficient LC capillaries (U.S. patent 5,453,163 ).
All of these reports seem rather empirical. I think that a complex factorial design study might elucidate the complex role of the many variables. But I also expect that the required skills probably reside outside the current HPLC community.
Protein compatibility with UHPLC
Some proteins change their higher ordered structure with pressure. Folding is one example where high pressure denatures the proteins, but some spontaneously refold as the pressure is reduced. Barofold (Aurora, CO) developed the PreEMT process for this purpose. Dr. Alexey Makarov (Merck Sharp & Dohme, Inc., Rahway, NJ) lectured on the effect of pressure on protein chromatography. The instrument was a conventional UHPLC with UV detection. A backpressure regulator that could add up to 4000 psi to the pressure of the column effluent was inserted between the column outlet and the detector. The effect on chromatographic retention ranged from nothing to large. In general, proteins that had helical segments showed the largest pressure dependence, possibly due to pressure-induced unraveling of the helix, which exposes hydrophobic patches that add to reversed-phase liquid chromatography (RPLC) retention.
Similar considerations may lead to practical multidimensional LC (LC×LC). In contrast to multidimensional GC, which uses trapping (and hence concentrating) devices between the dimensions, examples of LC×LC often just use a loop valve, or perhaps two. The lack of trapping for increasing concentration between the dimensions leads to excessive bandwidth and “dilution to oblivion.” Indeed, the most successful examples of LC×LC involve off-line interfaces between the two LC modes. This may be as extreme as drying the fractions collected from the first dimension, followed by dissolving the fraction in a second solvent that is more compatible with the second LC mode.
A team from the laboratory of Gert Desmet (Vrije University) described construction and evaluation of a low thermal mass modulator that is analogous to the thermal modulators (traps) used successfully in GC×GC. In LC, the thermal modulator cools the mobile phase to about +5 °C for the trap stage. For release, the trap is rapidly heated to drive off the analytes. This increases the concentration and hence improves detection in the second dimension by a factor of 18.
Valve module for LC×LC
Another approach to LC×LC was offered by Agilent (Palo Alto, CA). The Parking Deck Cluster (PDC) is an automated valve module for the Agilent 1200 series. In the simplest form, the PDC automates heartcutting with a two-position, six-port valve. This is suitable for a targeted separation with only one region of chromatogram requires higher resolution than is achievable with a normal 1-D run. At the other extreme, called the 12-loop PDC, the two-position, six-port valve feeds two 14-port distribution valves. Each pair of the six distribution positions is interconnected with a sample loop. The column effluent from the first dimension is parked sequentially in one of the 12 loops. After collection, the 12 segments are run individually on a second column. In addition to automation, the main advantage of the PDC configuration is that the second dimension is not constrained to very short run times, as is common in comprehensive LC. Applications examples included assay of impurities in pesticides and aggregates in antibody samples.
Several lectures focused on column optimization for a particular separation, usually with carefully selected constraints. Too often, speakers seemed to extrapolate the results of the constrained and tightly focused study to relevancy in the universe and beyond.
In one of the vendor seminars, Prof. Richard Henry pointed out that chromatography is an exercise in making tradeoffs to optimize conditions for the problem at hand. Each assay is really different, except possibly if one is developing a platform assay. These make the ultimate compromise to use one method that delivers adequate results for multiple assays. This is justified by avoiding the time in changing from one method to the next, and also in validation of several different assays.
Prof. Henry also explained that when the system pressure is measured, most assume that nearly all the pressure is used to force the liquid through the column. He reported a simple experiment to test for extracolumn pressure drop, and also band broadening. The column is replaced with a zero dead volume union. In one example, he recorded the instrument only pressure of 4000 psi at 5 mL/min.
Molnar-Institute’s (Berlin, Germany) DryLab has become even more powerful with introduction of version 4.2. The DryLab Cube provides the chromatographer with a visual image of the design space for Quality by Design (QbD). Now it is easy to compare columns and predict their response to changes in pH, gradient time and %B start and end, temperature and composition of ternary mobile phases, and column length and diameter, i.e., all the items that need to be considered in specifying, optimizing, and validating a method. A novel, robust tool can estimate the simultaneous influence of all these factors, with 729 runs in 20 sec. This includes a classification of which conditions will lead to OOS (out-of-specification) results. This is all presented in a single knowledge management document. In Europe, these studies are considered essential to documenting HPLC and LC methods. The net deliverable should be quicker and more rigorous method development, optimization, and validation.
Optimization of core-shell separations
What is the optimum core-shell particle? Well, this depends upon your application and instrument. Several lectures addressed optimization. For example, Dr. Joe DeStefano from Advanced Materials Technologies (Wilmington, DE) proposed that 2.0 μm with a 0.4-μm shell was optimum for chromatography of small molecules with legacy (~6000 psi Pmax) instruments. Small molecules diffuse rapidly, so one can have a thicker shell, which aids in column loading and wider elution window. Good points.
Most interest focused on the optimum particle size for columns packed with core-shell particles. Core-shell particles generally offer more plates per column compared to columns packed with porous particles of the same size. Higher efficiency often translates into improved chromatographic resolution. Or, one can increase the flow rate to take advantage of the flatter van Deemter performance and increase sample throughput by 30‒50%.
An update on monolithic columns
After almost three decades of development, monolithic columns for HPLC still seem to be a niche segment. Monoliths have strong performance advantages over packed columns, especially speed enabled by low-pressure drop, but I did not notice any posters that reported the use of monolithic columns, except from vendors and developers. Why? What am I missing? I should also point out that Prof. Nobuo Tanaka (GL Sciences, Iruma, Japan) received the Martin Gold Medal, sponsored by the Chromatographic Society of the U.K., for his work developing HPLC columns with silica monoliths. These are now distributed by EMD Millipore (Billerica, MA). EMD has been developing silica rod monoliths for about two decades under the Chromolith® brand. This year the Chromolith diol, amino, nitrile, HILIC (hydrophilic interaction liquid chromatography), and phenyl surface chemistries were added. The columns are very stable with time and are reproducible.
3-D printing of metals
Prof. Brett Paul (University of Tasmania, Hobart) described 3-D printing of HPLC columns in metal. After deposition of the metal powder, a high-power laser fuses the new layer to the old to build complex shapes with creativity that we have seen for 3-D printing using plastics.
The finished metal piece is accurate to 25 μm in 100 mm. Mechanical strength is as hard and strong as one would expect from metal parts made by conventional technology. An early design used a serpentine column encased in a metal block. Cooling channels can be included, as can Peltier heaters/coolers. His team went on to pack columns with various conventional column packings and to prepare monoliths. Prof. Paul anticipates that this technology will find use in robust, compact, and portable LCs.
New LC instrumentation
With almost 50 years’ experience in gel permeation chromatography (GPC), Tosoh Bioscience (King of Prussia, PA) introduced its third-generation high-temperature GPC called the EcoSEC High Temperature GPC System, which replaces the HLC-8121 introduced in 1988. High-temperature GPC is required for analysis of polyolefins. Two pumps (one for analysis and the other for reference columns) meter mobile phase to a 24-vial sample injector and onto the column rack, and then to the differential refractive index detector, with a low dispersion (10-μL) flow cell. True dual flow provides superior baseline stability (low noise and drift). Reproducibility in GPC is especially important. One study shows the intraday %CV for retention time of 0.017% and area of 0.42% for six consecutive runs. This translates to a %CV of molar mass of 0.068%. Other industrial polymers with adequate solubility at room temperature can be analyzed with the EcoSEC GPC System.
Metrohm (Riverview, FL) introduced the 940 Series Professional Vario Ion Chromatograph, designed for flexibility and ease of use. Available detector modules include conductivity, UV/VIS, and electrochemical. A range of suppressors for ions and CO2 provide stable baselines and improve detection. Metrohm’s Dosino can be used for regeneration of the suppressor.
Mass directed flash chromatography
Several decades ago, flash chromatography used thin layer chromatography (TLC) to monitor the separation. At best, TLC is cheap, slow, and semiquantitative. Over the years, Biotage (Charlotte, NC), among others, upgraded flash chromatography with metering pumps and absorbance detectors, etc. However, there was and is concern about nondetectable reaction products. Are they mixed with the target peak?
MS is an obvious possible improvement, until one considers the aversion of most mass specs to high solvent flow. This year, Biotage introduced the Isolera™ Dalton, which uses a proprietary sampling interface that automatically provides a 20-nL aliquot to a flow stream of methanol going to the MS. This interface is fully compatible with RPLC and normal mode chromatography. The entire process is controlled by Isolera Spektra software.
For protein aggregates, it takes a family
At HPLC 2014, Shimadzu (Columbia, MD) introduced a family of three instruments for characterizing protein aggregates. At the smaller end, starting at about 10 kDa and extending to the low mega-Dalton range, Shimadzu developed steric exclusion chromatography (SEC) packages that facilitate discovering the kinetics and thermodynamics in dilute solution. For subvisible particles (100 nm to 10 μm diam, Shimadzu introduced the Aggregates Sizer Aggregation Analysis System. The Sizer uses the shape of the laser diffraction pattern to classify the size of individual particles and then counts the number to give an output of the concentration in μg/mL. Sample cells are 0.4 mL or 5 mL. The latter is recommended for studying the effect of stirring. Measurement time is as short as 1 sec. For visible aggregates (dp > 50 μm), the company recommends a particle counter that it has marketed for decades.
Method scouting system
Assay development for HPLC and UHPLC is a data-intensive operation. First one needs to scout for suitable conditions. After finding a good lead, one needs to evaluate assay robustness, usually using the QbD approach. Assay development labs should look at Shimadzu’s highly automated Nexera Method Scouting System. It can investigate up to 96 combinations of columns and mobile phases. Two quaternary pumps facilitate scouting of up to 16 mobile phase pairs on up to six different columns, hence 96 combinations. Once the target conditions are selected, the Nexera can help map the operating space using QbD variables for flow rate, small changes in mobile phase, column temperature, gradient time, and more.
LC without column packings
LC is usually compatible with small molecules, but when things get big, say, over a million Daltons, one needs to be concerned that the column is acting as a size filter, or causing shear-induced degradation. More than 20 years ago, Prof. Yoichiro Ito at NIH (Bethesda, MD) introduced a liquid/liquid partitioning instrument that was the size of a washing machine. Over the years, the rotating coil centrifuge has been refined and miniaturized. In 2014, D-Star Instruments (Rockville, MD) introduced the Rotify™ benchtop system for fractionation of immunoglobulins, polysaccharides, and carbon nanotubes. The rotor diameter has been reduced to 6 inches from more than 10. The desalted products are collected in order of their pI for further characterization by time of flight (TOF) MS or other techniques. Run times are still long, with separations requiring several hundred minutes.
Many drug leads are purified by HPLC. In a thoughtful advance, Shimadzu has combined preparative purification with powderization so that the product is easy for subsequent processing such as formulation or preparation of screening libraries. Shimadzu calls the system Crude2Pure, or C2P for short. The first stage is a trapping column (Shim-pack C2P-H) designed to trap the target from the reaction mixture while the other components are washed away. Next, the target compound is eluted and routed to a spray nebulizer for rapid drying. This reduces the sample processing time from more than a day to about 5 hr.
After more than seven decades, amino acid analysis is still significant, especially for analysis of meats and fermented foods. Shimadzu and Ajinomoto Co. (Tokyo, Japan) have cooperated in developing the UF-Amino Station, where UF stands for ultrafast. This LC/MS system can analyze 38 derivatized amino acids during a 9-min run. Special-purpose AmiNavi® software controls the instrument and assay with outputs in compliance with USP and EP standards.
Phenylthiohydantoin (PTH) amino acids
Before LC/MS of peptides and proteins, Edman degradation was used to sequence the amino acid chain. It gave convincing proof of the connectivity. Much of the early literature in protein structure came from the slow and tedious Edman sequencers. It was before the time of “proteomics.” While LC/MS/MS can provide amino acid sequences very rapidly, there is still the nagging question: How do the sequences compare? Regulators frequently request the data. Accordingly, Shimadzu introduced the PPSQ-31A/33A automated Edman sequencer at Pittcon 2014. This is probably the last commercially available PTH sequencer.
Detectors for HPLC and UHPLC
Waters nearly pulled off a double in the Editors’ Awards at Pittcon, with the introduction of two major new products. The first, which won the Silver Award, was the ACQUITY QDa MS detector. UV absorbance detectors have been the most popular since the beginning of HPLC in the mid-1960s. Fortunately, UV was applicable to a wide range of analytes, plus the detectors themselves were rugged and reliable. Sure, they are nearly useless for non-UV absorbers. Vacancy UV detection never graduated from curiosity.
Waters sensed an opportunity for a major advance by designing an MS detector that is as rugged and easy to use as a UV, but provided more information, especially for analytes that are invisible to UV. Particularly, Waters wanted to add an MS to the ACQUITY product line just like other detector options. Plus, the MS needed to target a new competitive price point.
The detector is designed for zero operator involvement during startup, calibration, and shutdown. Startup takes about 20 min from a cold start, even with UPLC, UPC2®, and legacy LC plus lab scale prep. No special training is required. Control and data processing are via Waters Empower or MassLynx software. Waters boasts that Empower has more than 350,000 users in 60 countries.
I expect that the QDa will be most useful in method development, since peak tracking is often laborious, but with mass and UV absorbance keeping track of the peaks as one scouts different columns and modes, peak tracking should be much easier and quicker.
The QDa is an entry-level MS for the masses. Waters is a respected vendor to the premium-performance end of the LC/MS/MS market also. Based on its previous experience with the cube, the company saw an opportunity to improve performance by closely integrating the separations unit to the MS, as hanging the separations unit on the MS inlet. This was introduced as the ionKey/MS system, which is built around the iKey separation module. The separation column and ESI interface are in the iKey that is inserted into the ion source. One benchmark comparing the ionKey systesm to an external column packed with sub-2-μm particles showed an eightfold improvement in detection sensitivity.
The iKey is the separations unit, and is about 6 in. long and 1.5 in. high. The iKey is inserted. Flipping a lever locks it in place while making electrical and mobile phase connections. Columns are usually packed with sub-2-μm packings in 150 μm × 50 mm capillaries. iKey life is expected to be more than 1000 samples, but is usually much more. Several factors contribute to the expected long life. The sample load is reduced by more than 99.5%. Mobile phase consumption for each sample is reduced by a similar factor. Mobile phase is stored in a small container with filters to remove particulates. The pump does not need to pump as many stokes/sample, which reduces seal wear.
A new format raises concerns about reproducibility. One benchmark is that single user/day reproducibility in retention time (tr) was 0.15%. A similar study with five users and five systems and 18 iKeys had a %CV of 1.64% for tr. The ionKey missed an Editor’s medal by one vote.
Speaking of Waters, UPC2 needs column temperature control for reproducibility, since temperature affects density and pressure, and hence retention. Temperature has a larger effect on tr than in HPLC. Thus, Waters introduced a new heated column compartment specifically designed for columns ranging in length from 50 to 250 mm and from 2.1 to 8.0 in diameter. It uses recirculating air to control temperature with a Tmax of 90 °C.
Last year, Wyatt Technology Corp. (Santa Barbara, CA) joined Waters in upgrading its refractive index (RI) detectors for compatibility with the narrow peaks from UPLC separations. This year, Wyatt introduced the μDawn™ multiangle light scattering detector featuring reduced chromatographic dispersion required by UHPLC. Both RI and MALS are required for determination of polymer size. For the μDawn, the redesign decreased the flow cell volume from 63 to 7 μL. Also, interdetector band broadening between the RI and MALS was reduced to only 7 μL. Between the UV and μDawn, the band broadening is only 2 μL.
Assays of molar mass can be completed in less than 5 min. When combined with an RI detector, the molecular weight dispersion is easily displayed. Crossover studies comparing legacy HPLC technology to UHPLC show improved resolution of small peaks. When one adds a UV detector to the RI and μDAWN, one can get the molar extinction coefficient of a peak, which aids in identification of very large analytes that are probably too large for MS. This may be useful in characterizing drug antibody conjugates (ADCs).
Electrochemical detection (ECD)
At HPLC 2014, Thermo Fisher Scientific (Waltham, MA) introduced the updated pulsed amperometric detector (PAD) to the UltiMate 3000 ECD that includes a pulse potentiostat and interchangeable gold electrodes. This is ideal for detection of simple sugars including sweetened beverages. The Omni Coulometric Cell was a related introduction. This upgrade expands the PAD into general electrochemical detection. As the name implies, the Omni flow cell provides many options, as shown below.
- Electrochemical detection often suffers from contamination of the mobile phase, usually the water. Since the OmniRS is rated to 9000 psi, it can be inserted into the flow stream just ahead of the injector. It can be set to oxidize mobile phase contaminants, which greatly improves detection limits and reduces operator frustration.
- If the OmniRS is placed just after the column outlet, it can selectively oxidize electroactive active interferences. Again, detection limits are improved.
- The OmniRS can be set to oxidize all components as they elute from the column. Then the redox active analyte(s) can be selectively reduced in an amperometric flow cell. This often greatly improves LOD.
Descriptive material lists five more application options for the OmniRS. For more information, please see www.thermoscientific.com/ECDetection.
Charged aerosol detector
The Thermo Scientific Dionex Corona Veo charged aerosol detector (CAD) from Thermo Fisher Scientific has been upgraded to meet the requirements of UHPLC and micro LC. The CAD is still unique in having a universal calibration curve, which facilitates direct determination of relative amounts. Plus, it often responds to analytes missed by other detection modes. The CAD now comes in two models, the Veo and Veo RS (RS is optimized for rapid separations). Both models have an improved Focus Jet™ concentric nebulizer. The RS has additional features such as 2× faster data rate (200 Hz), electronic control of the gas, and selectable evaporation temperature from +5 to 100 °C. An electronic flow splitter from 1:1 to 1:20 is an interesting option, particularly for preparative work and possibly LC/MS.
Columns and column packings
HPLC columns are a well-served market with an amazing 47-year history of continuous innovation and development. Interest is always high. Why? Columns give chemists an opportunity to discuss chemistry. After all, we spent years in school studying books of descriptive material. It’s not surprising that we like to show that all the work has present value.
Column product lines
Aichrom HPLC columns from Abel Industries (Vancouver, Canada) were expanded from the AichromBond XB-1 series, which includes C18, C8, phenyl, CN, NH2, and bare silica with a dp of either 3 or 5 μm. The silica is rated for use in basic eluents up to pH 10. AichromBond X-2 has a lower surface area, 200 m2/g for larger molecules, and shorter retention times. Surface chemistries include C18, C8, and silica. The AichromBond AQ is designed for separation of moderately hydrophilic compounds. For even more polar analytes, a HILIC phase is also available. The AichromBond SB series is designed for chromatography at low pH, even as low as pH 1.0. The alkyl side chains shield the silica surface from the analyte, which reduces silanol interference. SAX and SCX ion exchangers round out the product line.
Diamonds are nearly nondestructible, so it is not surprising that Flare columns from Diamond Analytics (Orem, UT) are rated to pH 11.3 and 120 °C. Flare columns are 4-μm particles that provide about 100,000 plates/m chromatographic efficiency. The phase starts off with a carbon core that is coated with a thin layer of polyamine followed by a thin layer of nanodiamonds. This cycle is repeated layer by layer until the proper particle size is reached. Then, the entire composite is locked in place with a diperoxide crosslinker. Finally, the surface chemistry such as C18 is covalently attached. Phases include Flare C18; C18 MM (for mixed C18 and weak anion exchange); and Flare HILIC, which features an amino‒diol surface chemistry.
SpeedCore HPLC columns
Fortis™ Technologies (Cheshire, U.K.) is the latest firm to jump on the core-shell bandwagon with introduction of the SpeedCore HPLC column line. Speed core starts with a 1.8-μm solid silica core encased in a 0.4-μm silica shell with 80-Å micro pores. Surface chemistries include C18, diphenyl, PFP (pentaflurophenyl), and HILIC. Packings for RPLC are endcapped. The columns are stable from pH 2 to 9. The principal advantage is that column efficiency declines very slowly as the flow rate increases. This favors running the column at very high linear velocity to save time.
Hamilton (Reno, NV) continues to expand the PRP column line with introduction of the PRP-C-18. The C18 surface chemistry is bonded to cross-linked polystyrene, divinylbenzene copolymer beads that are much more stable toward strong acid or base and temperature (up to 100 °C) than corresponding silica. The particle size is 5 μm. Pmax is 5000 psi.
Kromosil UHPLC columns
AkzoNobel’s (Amsterdam) Kromosil UHPLC columns start with 1.8-μm-diam spherical silica with 100-Å pore. C4, C8, and C18 surface chemistries are bonded to the particles before packing into 2.1 × 50 or 100-mm columns. Thus, AkzoNobel’s product line of columns now includes 1.8-, 2.5-, 3.5-, and 5-μm-diam packings. Kromosil is respected for providing silica-based column packings with exceptional stability to extremes in pH. With the range of particle sizes, one can select the optimum particle size to fit the constraints of the method.
Titan and BIOshell™ columns
In the Supelco (Bellefonte, PA) booth, I found two new column lines for LC. Titan columns start with a 1.9-μm totally porous silica that has been produced by the Ecoporous™ manufacturing process. Test chromatograms for 2.1-, 3.0-, and 4.6-mm-i.d. columns (all 5 cm long) show that efficiency increases significantly with column diameter, up to 306,000 plates per meter. Initially, the surface chemistry is endcapped C18. In addition to outstanding performance, Supelco is promoting very competitive prices, which start at $300.
BIOshell columns from Supelco are packed with core-shell particles for proteins and peptides. For peptides the core is 1.7 μm diam. The shell is 0.5 μm thick with a 160-Å pore. Surface chemistry is either nitrile or C18. For proteins the core is 3.0 μm with 0.2 μm shell. Pore size is 400 Å, surface chemistry is C4, Pmax is 600 bar, and Tmax is 90‒100 °C. BIOshell columns offer about 50% higher peak capacity for peptides than is realized with similar column packed with totally porous particles.
Columns for HPLC, UHPLC, and SFC
The chemists at ES industries (West Berlin, NJ) have been busy extending their line of stationary phases, particularly in fluorophases. FluoroSep-RP HILIC is a fluorinated diol phase showing preferential retention of polar halogenated analytes using HILIC mode. FluoroSep-RP XP is a fluorophenyl surface chemistry for aromatic halogenated analytes including isomers. FluoroSep-RP R XF has a perfluorinated alkyl surface chemistry designed for general-purpose columns. FluoroSep-RP Octyl has a perfluorooctyl group bonded to silica. Compared to C18 and C8, the perfluoro C8 is less hydrophobic with different selectivity. In favorable cases, it provides isocratic separations when other phases require gradient elution. Epic HILIC-FL sub-2-μm is a fluorophase for UHPLC HILIC separations. It is especially favored for LC/MS since separations are achieved with higher organic content mobile phases.
ES also added several new stationary phases to its more traditional column lines. GreenSep Basic 3 μm is a new basic phase for SFC. GreenSep Basic PFP 3 μm offers unique selectivity for SFC due to the pentafluoro surface chemistry, again for SFC.
Chromaga Chiral CCC from ES Industries is a new column packing for chiral separations with 3-chloro-4-methylphenylcarbamate and 3,5-dichlorophenylcarbamate bonded to cellulose. This phase shows promise in separating previously unresolved mixtures.
MacroSep C4 columns from ES Industries have a 300-Å pore. They are designed for separation of biomolecules, including glycoproteins, hemoglobin variants, and membrane proteins.
RPLC covers a wide range of applications, niches, and conditions. Restek’s (Bellefonte, PA) new Raptor SSP column line started with a biphenyl surface chemistry to the core-shell base. At Pittcon 2014, Restek added the ARC-18 phase, which bonds steric protection groups to the base of the C18 brush to increase stability at the extremes of low and high pH.
SEC columns for antibodies and aggregates
For more than three decades, Tosoh’s SW series of columns have dominated the modern market for columns for steric exclusion chromatography columns. Initially, the particle size was a nominal 10 μm or larger. A few years ago, Tosoh introduced columns packed with nominal 4 or 5 μm diameter. This year the company focused on monoclonal antibodies with 3- or 4-μm particles. The TSKgel UltraSW Aggregate, 3 μm, is designed for high-resolution separation of antibody multimers and aggregates. The TSKgel SuperSW mAB HTP, 4 μm, is designed for high-throughput analysis of antibody monomer and dimer under UHPLC conditions. Run times are typically cut in half. TSKgel SuperSW mAB HR, 4 μm, is recommended for antibody monomers, dimers, and fragments under HPLC conditions. Tosoh has developed a pore control technology that gives a shallow molecular weight calibration curve in the region where antibody fragments to dimers elute.
CSH C18 columns
LC/MS of peptides with C18 columns often involves use of ion pairing agents such as trifluoroacetic acid (TFA), but TFA suppresses ionization with ESI interfaces. Waters has developed a charged-surface hybrid C18 (called CSH C18) that obviates the need for TFA or other ion pairing agents. One example was a peptide map of trastuzumab, a therapeutic antibody.
Zenix® SEC columns
Sepax Technologies (Newark, DE) introduced Zenix SEC columns packed with 3-μm-diam porous particles with a choice of 100-, 150-, and 300-Å nominal pore diameter. Sepax developed a proprietary surface chemistry to reduce nonspecific adsorption with biological and industrial water-soluble polymers.
GlycanPac AXR-1 columns
Glycan structure of proteins is very complex and even more important since it affects the safety and efficacy of protein therapeutics. Until now, determining the glyco heterogeneity involved many steps with specific enzymes that differentially responded to glycan structure. This year, Thermo Fisher Scientific introduced Glycan Pac AXR-1 columns for HPLC separation of free glycans by sialylation number (from 0 to 5) plus high resolution of hetero forms within the sialylation cluster. Further, the mobile phases are MS compatible, which has not been true with other separation modes.
The stationary phase has a proprietary mixed-mode surface chemistry (WAX and RPLC) bonded to nominal 1.9-μm or 3.0-μm-diam silica (175-Å pore diameter). With the 1.9-μm particles, a mixture of 135 glycans produced 73 peaks. MS detection was effective in resolving many coelutions. Most examples utilize a 150-mm-long column, but if more resolution is required, a 250 × 2.1 mm column is available. One example showed separation of 105 fluorescently labeled glycans in 70 min.
CarboPac SA 10-4 μm column
Thermo’s Dionex CarboPac SA10-4 μm is designed for rapid assay of mono- and disaccharides in food, beverages, and biofuels. The speed comes from reducing the particle size from 6 to 4 μm. Both are composite phases. The stationary phase starts with a macroporous polystyrene-divinyl benzene copolymer that has 6 μm diam agglomerated with 55-nm microbead difunctional quaternary ammonium ion. Run time is usually less than 8 min. The column and mobile phase are compatible with the Dionex pulsed amperometric detector (above).
Fast Protein A columns
Protein A is a classic surface chemistry for selective retention of IgGs. It is derived from the protein coat of bacteria and thus is often immunogenic. Daiso (Osaka, Japan) introduced HPLC stationary phases with Protein A covalently bond with epoxy chemistry to porous silica. Daiso introduced ADREPMA™ (advanced recombinant protein for monoclonal antibody) for affinity chromatography. Benchmark studies by Daiso show that ADREPMA binds more quickly with higher capacity and superior recovery compared to “agarose material.”
Despite the biospecific affinity of Protein A columns, there are other considerations. Contamination of the product with ligand from the column is a major concern in preparative applications. Ligand leakage is less of a concern with analytical columns, unless they are used for prep. Clean-in-place (CiP) is another issue. Traditionally, preparative columns using agarose for support were cleaned-in-place with washes of 1 molar NaOH. This usually destroyed column packings using silica as a support. The Daiso literature shows that 150 short washes with 0.1M NaOH do not alter performance.
HILIC rages on
It has taken 25 years, but Andrew Alpert’s HILIC, including electrostatic repulsion hydrophilic interaction chromatography (ERLIC), is now firmly entrenched as a very productive separation mode in LC, particularly for biosamples. PolyLC Inc. (Columbia, MD) is Dr. Alpert’s firm that promotes the modes and columns. One of the company’s notes compares the peptide coverage for several 2-D LC modes. 2-D LC with ERLIC + RPLC identified 711 peptides which were derived from 123 proteins. The peptide count was much more than other modes including SCX-RPLC. With ERLIC in the first dimension, the peptides elute in order of decreasing pI, which is useful in detecting deamidation products.
Another application brief addressed quality control of antibody drug conjugates (ADCs). PolyPROPYL A™ operating in HIC mode was able to rapidly separate homologs of ADCs. Further, the column provided high resolution of oxidation variants.
Antibodix columns from Sepax Technologies start with nonporous polystyrene/divinylbenzene beads with a choice of 1.7, 3, 5, or 10 μm. A 1-nm-thick hydrophilic polymer is grafted to cover the aromatic surface to reduce nonspecific hydrophobic retention. Next, weak cation exchange functional groups are covalently attached to provide a high-capacity ion exchange layer. The columns show good separation of antibodies, including fragments and low multimers.
AS-19 column with 4-μm particles
The AS-19 from Dionex, now part of Thermo Fisher Scientific, was recognized for assay of drinking water quality according to U.S. EPA Method 300.0. By reducing the particle size to 4 μm, the new column meets the requirements of U.S. EPA Methods 300.0 and 300.1. The narrow-bore columns scale from 4 mm down to 0.4 mm. Reducing the particle size to 4 μm improves the separation power at all column diameters. Peaks for bromate, etc., stand out from the background.
The May issue of American Laboratory reports on new products for sample processing (http://www.americanlaboratory.com/914-Application-Notes/160946-Sample-Processing-at-Pittcon-2014/), including SOLA, introduced at Pittcon. But Thermo introduced the SOLAμ at HPLC 2014. Chromatographers at Thermo recognized that Moore’s Law for SPE was suspended. Yes, improved detection limits allowed reduction in sample volume (<100 μL), and the number of samples was increasing. Plus, sample loss during concentration by blow-down wasted time and degraded some analytes.
SPE technology was stuck in a rut. What to do? Thermo’s solution was to adopt a fritless monolith adsorbent bed in each well of a 96-well SPE plate. Thermo introduced the SOLAμ, which is designed to process small sample volumes (<25 μL), and concentrate during the elution stage as much as 20 times. Sample volumes are so small that blowdown is fast if it is required at all.
Since the SOLAμ is fritless, one applies the sample to the top of the active sorbent monolith. Bed weight is only 2 mg. There is no dead volume. Fabrication of the monolith is more uniform than can often be obtained with packed beds and retaining frits. This shows up as improved consistency in replicates.
Update on Chromeleon chromatography software
Thermo chose Pittcon for introduction of Chromeleon CDS, extending Chromeleon 7.2, which was introduced just the year before. The CDS version supports MS detection for Thermo’s ion, liquid, and gas chromatographs. The software is fully validated for the company’s single and triple quadrupoles, and benchtop Orbitraps. The staff seemed especially proud of the operational simplicity, including one-click workflow, which reduces training time. CDS is compatible with Microsoft’s “dot.net” architecture, which should assure integration and communication.
Sample loops for UHPLC
Optimize Technologies (Oregon City, OR) introduced several accessories to make working with UHPLC less of a challenge. The sample loops are 0.005 in. i.d. for 1 or 2 μL internal volume with 1/16-in. ends. The Pmax rating is 20,000 psi. For Shimadzu’s UHPLC, Optimize offers piston seals and check valves.
Polymer Char (Valencia, Spain) has a singular focus on characterizing polyolefins (POs). This often includes high-temperature GPC. However, traditional assays of POs use macro assays such as xylene solubles, which is a gravimetric wet chemistry method. This is out of date, plus it is tedious and laborious (~6 hr/sample). Polymer Char introduced the CRYSTEX® QC to replace xylene soluble. With a protocol called Temperature Rising Fractionation (TREF), the analyst seals the sample in a long tube and places it in the CRYSTEX. The instrument pumps solvent through the tube as the column temperature increases. Output is ethylene content by infrared, viscosity, plus % amorphous and % crystalline. Run time is reduced from 6 hr to 2, and labor from 6 hr to a few minutes.
For religious and cultural reasons, meat authenticity is a concern for about 23% of the global population. Many of the rest would like assurance that what they buy is accurately labeled. A new LC/MS/MS method from AB SCIEX (Concord, Ontario, Canada) is useful in detecting maker peptides for porcine and equine flesh in other meats using a QTRAP 5500 MS with multiple reaction monitoring (MRM). MRM is able to differentiate between horse and beef flesh due to differences in one or more characteristic amino acids. Detection sensitivity can be as small as 0.13% contamination.
I am impressed with Camag’s (Muttenz, Switzerland) TLC Scanner 4 for densitometry evaluation of thin layer chromatography (TLC) and HPTLC plates, using two wavelengths. Dual wavelengths facilitate background correction by subtraction. The optical range is 190‒900 nm. The optical resolution is sufficient to record 36 tracks with 100 bands each. Each track is scanned repeatedly with a small offset, which facilitates location of the peak maximum. Integration of peak areas is with automatic or manual baseline correction. Spectra for each peak are recorded for comparison with other plates. Unexpected peaks can be referenced to the main component as prescribed by European and American Pharmacopoeias. My impression is that if you are doing quantitative TLC, you will really appreciate the TLC Scanner 4.
In the early days of capillary electrophoresis, leading scientists felt that CE and MS should be a natural fit since they both excelled with low flow. This vision was difficult to reduce to practice. One had to independently control the flow of liquid carrier liquid and ions, while generating a high voltage at the exit to support electrospray. Two years ago, scientists at Beckman (Fullerton, CA) seemed to have an improved interface for CE/MS that used a porous conductive capillary and a conductive liquid that provided ion flow without disturbing the flow in the CE capillary, yet provides stable electrospray ionization. At the low flow rates, problematic ion suppression is avoided, improving quantitation and method robustness.
The devil is in the details, so the CE team at Beckman was joined with a similar team from AB SCIEX to create the CESI 8000 System for Biologics Characterization. The 8000 will be supported by a new business unit called SCIEX Separations. This group will focus on developing products for therapeutic biologics.
HPLC 2014’s technical program was organized around three parallel tracks. Tutorial sessions were also available at the same time. Thus, anyone could partake of only about a quarter of the program. At Pittcon, the number of active rings in the circus was even higher.
Separation technology: an interesting decade
A decade ago, we separation scientists were giddy with the results of sequencing the first human genome, which was enabled by massively parallel HPCE. With the confidence built upon success, we set off to develop new separation technology using LC/MS to discover biomarkers of disease. The goal was to find unique markers that correlated with specific diseases. Patient pools were small, so apparent correlations were frequent, and mostly useless.
Today, the operating hypothesis is that most diseases express several to many markers, confounded by natural variation in patient response. Any biomarker discovery study requires sifting through GB or TB of data, often in disparate files. Biomarker teams need experts in chromatography, mass spectrometry, statistics, and informatics. Nevertheless, the emphasis has changed from developing high-peak-capacity separations technology to one showing that the tools for ’omics really deliver the promised benefits to society. It’s about time.
Robert L. Stevenson, Ph.D., is Editor, American Laboratory/Labcompare; e-mail: firstname.lastname@example.org