May 1, 2014 (Vol. 34, No. 9)
Any analytical method must strike a balance between its intrinsic scientific potential its capacity for robustness and reliability, particularly in regulated industries.
In the case of HPLC, this balancing act is part of an initiative by the U.S. Pharmacopeia (USP), which is modernizing its pharmaceutical monographs.
A monograph is “the book” on a drug product. Besides specifying the active ingredient’s chemical composition and the product’s ingredients, packaging, storage, and labeling, a monograph describes the analytical methods required to ensure strength, quality, and purity.
“These tests need to be very selective, sensitive, and precise,” says Petra Lewitz, product manager for analytical chromatography at EMD Millipore. “In that context, HPLC is preeminent, but it is not the only method.”
In many USP monographs, the specified analytical methods are completely out of date. The USP is therefore recruiting industrial collaborators to provide sensitive, robust, current methods that improve overall understanding about pharmaceuticals and enable labs to operate more efficiently.
For example, many monographs rely on C18 columns, but as Lewitz notes many pharmaceuticals are polar or hydrophilic: “This stationary phase lacks selectivity and sensitivity for many pharmaceutical compounds unless ion-pairing agents are employed. But then, they become unsuitable for mass spectrometric analysis.”
In this case switching to a more selective column, such as a hydrophilic interaction liquid chromatography (HILIC) column, or a slightly smaller particle size, might be sufficient for method modernization. “You don’t have to change the instrument,” Lewitz adds.
Interestingly, Lewitz does not categorically endorse such modern alternatives as UHPLC or core shell columns. Column clogging, she observes, is too likely using these technologies, especially with complex matrices such as formulated drugs.
For that reason, Lewitz favors monolithic columns, which provide the backpressure advantages of core shell (and HPLC) but will stand up to harsh, complex samples. “It’s important to find the best compromise between speed, resolution, and robustness.”
Unlike conventional packed HPLC columns, Chromolith® HPLC Columns from EMD Millipore are not filled with minute silica particles. Instead, they consist of a single rod of high-purity monolithic silica with a bimodal pore structure—macropores reduce column back pressure, enabling faster flow rates; mesopores form a fine porous structure, providing a large active surface area.
Solving Niche Problems
Instrument vendors are constantly alert to their products’ potential for solving narrow scientific problems. For example, some vendors are using HPLC to detect aflatoxins in foods.
Aflatoxin B1 occurs most commonly in corn and peanuts, but aflatoxin M1 occurs in products derived from animals that consume tainted feed. European regulations require an upper aflatoxin M1 concentration of 50 parts per trillion. Analyzing the toxin at this concentration requires matrix cleanup, post-column derivatization, and highly sensitive fluorescence detection.
“Aflatoxin is difficult to detect at those levels, and milk is a difficult matrix,” says Jason Weisenseel, Ph.D., chromatography technical leader at PerkinElmer.
Post-column derivatization involves running eluents and reagents through a loop and activating linkage with pulses of light. The fluorescence method involves mixing reagent with column effluent, which increases dead volumes and reduces sensitivity. Reagents and derivatization kits add to assay costs and complexity. Pre-column derivatization, by which reagent is added to the entire sample through the autosampler, could potentially eliminate dispersion. “But we don’t have that luxury with aflatoxins,” Dr. Weisenseel explains.
Dr. Weisenseel and co-workers have developed a rapid, sensitive ultraviolet detection method that requires no derivatization. The keys are immunoaffinity solid-phase extraction to remove matrix, a photodiode array UV detector, a long fiberoptic flow cell, and fused core column technology. Fiber optics efficiently convey light to and from the flow cell while reducing noise. Together, instrument innovations reduce limits of quantitation to about 100 parts per trillion. Matrix removal gets detection limits down to the 10–20 parts per trillion range.
Mixed-Mode Analysis
GEN has reported on the potential for mixed-mode chromatography to crack key downstream purification issues in large-scale protein purification, particularly for capture. Typical mixed-mode resins include cation or anion exchangers plus reverse-phase or hydrophobic interactions. In biomanufacturing, however, mixed-mode HPLC has remained experimental for the usual reason: risk avoidance.
Fewer barriers exist for adopting mixed-mode analytical chromatography, a specialty of scientist Xiaodong Liu, Ph.D., manager of R&D, chromatography consumables at Thermo Fisher Scientific. A recent presentation by Dr. Liu notes the shortcomings of conventional HPLC for analytes that are diverse in terms of size, charge, hydrophobicity, etc.
Reverse-phase HPLC shows poor retention for hydrophilic analytes, is incompatible with some aqueous mobile phases, and has a limited selectivity range. HILIC suffers from analyte solubility and matrix effects and is limited to hydrophilic analytes.
Dr. Liu’s mixed-mode resins consist of hydrophilic or hydrophobic interaction combined with cationic or anionic exchange, plus a zwitterionic resin.
“Mixed-mode columns provide adjustable selectivity, where separations are optimized on a single stationary phase by adjusting the chromatographic conditions without the use of ion-pairing reagents,” Dr. Liu observes. This allows retention of hydrophilic analytes that reverse-phase HPLC fails to separate, which addresses what Dr. Liu calls “a critical gap in reverse-phase capabilities.”
Resolution in HPLC depends on selectivity, efficiency, and retention time, with selectivity having the greatest impact. Selectivity range in reverse-phase HPLC is limited, and relatively unaffected by pH and ionic strength. In mixed-mode HPLC, these variables allow fine-tuning selectivity for optimal results.
For example, in a separation of penicillin G potassium on Thermo Fisher’s Acclaim Trinity P1 column, elution of drug and counterion were reversed simply by switching the acetonitrile:buffer ratio from 80:20 to 60:40, at constant pH. “The column is switching from HILIC mode to ion exchange,” Dr. Liu explains.
Mixed-mode HPLC has been around for 30 years, but it has experienced a renaissance during the last decade. Yet adoption, as in bioprocessing, has been slow. “Part of the reason is that reverse-phase HPLC is so easy to use,” Dr. Liu says. Mixed-mode HPLC is perceived to be more difficult because there are at least two separation modes to deal with, and relatively minor mobile-phase alterations significantly affect selectivity.
“We realize that reverse-phase HPLC is the norm, and recommend that people begin with that,” Dr. Liu admits. “But where reverse-phase doesn’t do the job, mixed-mode can be a good alternative.”
On this basis, Dr. Liu is developing a technique for specific applications that remain problematic for conventional HPLC. Such applications include the analysis of glycans and surfactants and the detection of contaminants in drinking water.
Scaling Up Method Transfer
During the mid-2000s competitors of Waters’ UPLC® instrumentation pointed to difficulty in transferring methods from HPLC to UHPLC. First Waters and then new entries into UHPLC eventually published equations and formulae to facilitate method migration.
“A lot of knowledge exists on this subject, and the process has become fairly straightforward,” states Helmut Schulenberg-Schell, Ph.D., director of business development, liquid-phase separations at Agilent Technologies. “Much less is known about applying analytical HPLC techniques to preparative LC.”
Preparative LC is rapidly becoming the go-to method for purifying milligrams to several grams of synthesis products in chemical and pharmaceutical laboratories. Yet organic chemists, says Dr. Shulenberg-Schell, are too busy with synthesis to learn yet another technique for compound purification. Prep LC is too involved for chemists to undertake routinely, but it is not quite at the level where a chromatography specialist needs to get involved.
Traditionally, chemists have used flash chromatography to purify milligram to gram quantities. While still a powerful technique, flash chromatography demands some degree of hands-on time from the chemist in terms of development and execution. More significantly, it is incapable of applying what is already known from analytical HPLC to large-scale purification.
Agilent’s solution is the 1260 Infinity Automated Purification System, which eliminates nearly all the effort entailed in preparative LC. After the operator selects a peak to purify from the analytical LC/MS run, proprietary software automatically sets conditions, including gradients, for the preparative run, thus eliminating guesswork and human error during method development. The 1260 also includes automated fraction collecting.
“This instrument will drive the industrialization of compound purification,” declares Dr. Schulenberg-Schell. “Most chemists, once they reach 80–90% purity, cannot afford the additional time to optimize separation. Our system calculates gradients on the fly that are specific for target compounds, without having to turn the method over to a specialist.
Role of LC in MS
The battle for preeminence between the components of the hyphenated LC-MS method has led to amusing claims, where chromatographers refer to MS as “just another detector,” while spectroscopists view LC as nothing more than a fancy sample-preparation tool.
“It goes back and forth every few years, where the spectroscopists think their instrument is so good they no longer need HPLC,” comments Curtis Campbell, Ph.D., HPLC business development manager at Shimadzu Scientific Instruments. “Eventually they admit they can’t do LC-MS without LC because they need to separate what goes into the mass spectrometer.”
Proponents of mild techniques like LAESI (laser-ablation electrospray ionization) tout the technique’s ability to introduce samples into a mass spectrometer without prior manipulation, but they will grudgingly admit that MS works better if components are separated before injection.
The significance of complex mixtures and matrices, some containing several chemically related compounds or even structural isomers, makes LC all the more significant for getting the most out of mass spectrometry. “Analysts need the separation selectivity of LC to move these compounds around and confirm their identities by MS,” Dr. Campbell adds.
Even a partial LC separation improves MS results. Campbell cites the time savings available to analysts bold enough to push both analytic components to their limits. “You see people doing ballistic 20–30-second gradients to finish runs in under a minute where the MS collects all the data it can and software makes sense of the results. Or you can get a beautiful chromatogram [separating all components] through a 15-minute LC run, and use the MS to merely confirm identity.”
Which side dominates the discussion today? According to Dr. Campbell, mutual respect has been building between the techniques. GEN readers are aware of advances on both sides, and the continuing “democratization” of both GC and MS with accompanying growth in capabilities. With LC separation modes and platforms so numerous, and MS and analytic software capabilities so varied, it seems certain that these techniques will be used together for a long time.
The average analytical scientist may not be able to design robust LC-MS methods that fully exploit the capabilities of both modalities. At the least, however, scientists of ordinary skill are capable of running existing advanced methods and analyzing the results with the help of software.
Pushing the Limits of Core Shell Performance
The most-cited advantage of core shell/fused core columns is their ability to deliver UHPLC performance at HPLC backpressures. That does not prevent innovative vendors from creating fused-core particles in sub-2-micron formats.
Phenomenex started this trend with a 1.7-micron fused core particle. “The value proposition for these columns is that they improve performance by 25–35% compared with their fully porous counterparts,” explains Michael McGinley, senior product manager, bioseparations. “Top-of-the-line UHPLC systems provide 250,000 theoretical plates per meter, which we achieve with 2.6-micron core shell columns.” The company’s 1.3-micron product, which debuted in 2013, reaches more than 400,000 plates.
According to McGinley, Phenomenex continues to push the limits of core shell technology in small particles to take advantage of customers switching to UHPLC as they retire older systems.
Core shell columns are difficult to make and pack, so they are presumably costlier than the corresponding fully porous material columns. Another pitfall is assuming that simply switching columns is enough to provide all core shell benefits. “If you don’t optimize and adapt methods properly, for example, with respect to dead volumes, you won’t see a lot of improvement,” McGinley cautions.
