They say it takes several years for an awards program to really get off the ground and start amassing a history, establishing a place for itself in the industry. If this is true, then I believe we can safely say that this year, the 

 Pittcon Awards have officially arrived. With past Lifetime Achievement Award winners such as Walt Jennings and Harold McNair, and bright young stars such as Gert Desmet and Kevin Schug claiming Emerging Leader honors over the past couple of years, there is a great deal of suspense building over who will win the 2010 awards, set to be presented at Pittcon 2010 in Orlando, Florida.

Past winners have gone on to make great contributions to 

, writing installments of our well-known "History of Chromatography" column, guest authoring installments of "Column Watch," and even acting as scientific advisor for the many online projects that are currently in development. In short, this year's panel of industry experts will have their hands full maintaining the tradition of the 

 Awards, and we have every confidence they will.

The nomination process for the Emerging Leader Award is now open, and in the coming months, the greatly anticipated online reader vote for Lifetime Achievement Award winner will take place, in which the readers of 

 have a direct say in who this year's winner will be. To nominate the candidate of your choice, please visit: 

 for the full nomination guidelines. We hope all of our readers will participate in these awards, whether through nominations or by casting a vote. These are truly your awards, so look for more information to arrive in the coming months, including information on how to attend this year's awards luncheon. We look forward to a successful 2010 

 Awards program, and we thank you as always for your continued support.

Articles

From the Editor

They say it takes several years for an awards program to really get off the ground and start amassing a history, establishing a place for itself in the industry. If this is true, then I believe we can safely say that this year, the LCGC Pittcon Awards have officially arrived. With past Lifetime Achievement Award winners such as Walt Jennings and Harold McNair, and bright young stars such as Gert Desmet and Kevin Schug claiming Emerging Leader honors over the past couple of years, there is a great deal of suspense building over who will win the 2010 awards, set to be presented at Pittcon 2010 in Orlando, Florida.

The Third Annual LCGC Pittcon Awards: Establishing a Tradition

I first discovered the promised potential of lab-on-a-chip (LOC) technology at a Gordon Conference in the mid-1970s, when gas chromatography (GC) columns were reported on glass microscope slides. Since then, the technology's potential has fueled much innovation and product development. Yet this potential has not for the most part transmuted into practical, commercially available products. Although the widespread appeal of LOC generates much publicity and excitement, manufacturing a robust, reliable, and user-friendly device that is as good as or, ideally, better than established technologies has proven elusive. The technology's challenge, it seems, is twofold. One factor is technical: the difficulty (so far) to efficiently transfer "real-world" samples to the chip or miniature device. The other factor is human: our inborn resistance to accepting change, particularly when the perceived advantages of doing so are but minimal. 

Certain separation techniques, including capillary electrophoresis (CE) and nano high performance liquid chromatography (HPLC), have been performed on a chip. These techniques use both on-chip and off-chip detection. However, when a mass spectrometer serves as the detector, additional practical challenges arise. It is relatively easy to focus a laser through a glass-microfabricated channel for on-chip, laser-induced, fluorescence detection (1). But it is quite another matter to remove the sample from the chip and thereafter ionize and transport the sample for mass spectrometry (MS) detection. Nevertheless, despite this difficulty, reports continue to suggest a future for chip-based analytical devices (2–4). The actual measure of the technology's success, however, will take the form of commercial LOC products that customers choose and use because they work better than established techniques. 

 To those who grapple daily with the challenges of conventional laboratory methods, the concept of integrated chemical and analytical processes that share a common substrate free of sample transfers and associated plumbing has much to recommend it. You can discard an LOC device after a single or otherwise limited use. Doing so obviates the need to wash glassware and frees you from concern about sample-to-sample carryover. It also makes possible low-cost, easy-to-follow procedures. Moreover, from the LOC–MS point of view (5), the volume of liquid handled is nicely compatible with mass spectrometer operation. For both electrospray ionization (ESI) and matrix-assisted laser desorption ionization (MALDI)-MS, the lower the flow, the better the analysis. Thus, MS, an easily scalable technique, exhibits an enhanced response when you decrease the sample size (5). Unfortunately, some chip-based commercial products currently fail to impart such advantages.  

Why microfabricated devices are difficult to use: 

Given the relatively large surface-to-volume ratio of microfabricated channels compared to that of conventional HPLC columns, analytes are far more likely to interact with the capillary wall. Indeed, in the trace quantities typically used, they can adsorb to the wall's surface or become otherwise lost, to unexpected chemistry or accelerated degradation. Another barrier to the straightforward use of microdevices lies in their limited sample capacity. An analysis of, say, a biological sample can fail because the trace analytes of interest are in the presence of high levels of matrix components. Thus, loading the "dirty" extract into a microchannel can overwhelm the intended processes — chromatographic selectivity, solid-phase extraction (SPE) cleanup, and so forth — such that the intended processes might not occur. Finally, that same limited sample capacity necessitates sensitive detection capabilities which, in some cases, challenge the detection limits of modern mass spectrometers.  

The dream for LOC–MS is a fully integrated device that integrates sample handling, sample preparation, and separation science (CE, HPLC, and so forth) coupled to a mass spectrometer as a sensitive, selective detector. Additional promises include the ability to handle toxic chemicals safely, economically, and in an environmentally responsible way. Low sample and reagent quantities and improved detection limits for concentration-sensitive detectors make these promises worth pursuing, as does the reduced requirement for operator intervention and the need to connect tubes, fittings, or accessories.  

In principle, the LOC device should be inexpensive, portable, disposable, and amenable to productive use by unskilled operators. Regrettably, LOC analytical devices do not couple well with much larger instruments such as mass spectrometers. So the promise implied by the devices is difficult to deliver on. Yet progress is ongoing, and hope remains for the future of LOC–MS and related technologies. 

Progress to Date: Coupling the "Real World" to Microtechnology

 Before most samples can be analyzed, they usually must be "cleaned up" via a process that often includes extracting and concentrating the targeted sample analytes. LOC technology is no different. You must load a crude sample (biological, environmental, industrial, and so forth) onto the LOC device for extraction and sample cleanup. This "real-world-to chip" interface is a simple concept, but it presents a considerable challenge. A few innovative approaches (see the following text) have been commercialized, but on this aspect of LOC development, more progress is needed.  

When you load a sample onto a chip, and then prepare it by SPE or other form of extraction, you must afterward transfer the extract to the next stage of the analysis process. Working in our laboratory at Cornell University (Ithaca, New York), my colleagues and I reported an early attempt to do so. We constructed a parallel array of eight microporous, monolithic, SPE sorbents in a polymer (Zeonor) chip to accommodate a serial, microfluidic, SPE sample cleanup of eight separate, biological samples before being directly coupled with electrospray MS (6) (Figure 1).  

From 10-μL aliquots of human urine fortified with imipramine and imipramine-

 (internal standard), we obtained acceptable recoveries, quantitative precision, and accuracy from ESI-MS measurements at levels ranging from 0.025 to 10 μg/mL. We cleaned up the biological samples, passing them through the integrated, microfabricated, porous, SPE monolith using a syringe pump followed by analyte elution of the SPE extract directly to the ESI-MS system. This approach precludes an HPLC step, which is, in some instances, possible when the SPE cleanup step is adequately selective for the targeted analytes. One such instance would be when the MS step affords the tremendous selectivity of a tandem mass spectrometer operated in the selected-reaction-monitoring (SRM) mode and targeting unique fragmentation transitions for the analytes of interest.  

In a related experiment, we reported on-chip P450 metabolism in a microfabricated, serpentine channel where reagent solutions were delivered and mixed on the chip via an external syringe pump (Figure 2). Thus, imipramine was converted to desipramine via a chip-based P450 biotransformation, and the IC 50 value for the chemical inhibitor was determined by ESI-MS (7). We cleaned up the enzymatic product mixture using a chip-based, integrated, porous, monolithic SPE cartridge followed by direct analysis by ESI-MS. Figure 2 shows a schematic diagram of the syringe inlet ports directed to the enzymatic reaction region, which couples to the monolithic SPE cartridge in the chip, where the product mixture is desalted and cleaned up. After eluting the unwanted substances from the SPE cartridge, syringe pump C (Figure 2) delivers the elution solvent–spray mixture for direct ESI-MS analysis.  

Despite the apparent success of the reported results, relatively few additional reports on this concept have come to light. Hence, we have seen little commercial development of this approach even though in a "perfect world" it would seem logical to develop such an integrated micro-sample cleanup analysis system.  

A truly successful micro-sample preparation system that is also user friendly could preclude sample losses, analyte degradation, and excessive waste of solvents. It also could reduce the use of expensive reagents and enzymes and the time required to prepare samples for analysis. As to why a commercially viable, integrated approach has not emerged, the most likely reasons are a lack of confidence on the part of manufacturers in a strong commercial market and a lack of convincing evidence that LOC concepts are faster, better, and cheaper than conventional procedures. 

 Most of the reports for chip-based separation science involve just one or two operations actually performed on the chip. These operations can include CE or HPLC separation of a chemical mixture, most often with spectroscopic detection performed on the chip. In practice, it is far easier to effect a CE separation-detection operation on a chip because the only external connections required are electrodes, which attach to the chip's microfluidic reservoirs. Under an applied high voltage, the electrophoretic separation occurs with or without electroosmotic flow, which depends upon the pH of the medium and the surface chemistry of the microchannel. The small volumes of fluids needed are well suited for the CE technique and detection by electrospray MS.  

An early report established that one can couple rapid, chip-based, CE separations of acylcarnitines with electrospray MS (8). Nevertheless, little significant progress has been made toward developing a commercial CE–MS product that adopts LOC technology (Figure 3).  

In contrast to the relative ease of CE separation, chip-based HPLC separation is far more challenging because of the technology involved in microfluidic pumping of HPLC mobile phases. Tempting though it might be to simply connect an external nano-HPLC pump via capillary tube connections and fittings, doing so would mean losing the portability and simplicity of a true chip-based system. At the recent ASMS Conference on Mass Spectrometry, a poster presentation by Eksigent Technologies, Inc. (Livermore, California) (9) described such an approach with its recently introduced commercial cHPipLC Nanoflex system. This glass chip contains etched, microfabricated channels packed with C18 reversed-phase packing, which accommodates a direct coupling from a nano-HPLC pump, delivering mobile phase to the chip. This commercial system reportedly makes nano-LC–MS easier and more reproducible. For nano-LC–MS applications, the exit of the glass, chip-based packed capillary column connects via a fused-silica capillary to a nonintegrated nanoelectrospray emitter, which is directed to an atmospheric pressure ionization (API) mass spectrometer. High-quality LC–MS data are then presented for an enzymatic digest (Figure 4). Yet it remains to be seen whether routine LC–MS users will embrace this commercial idea for either small- or large-molecule applications.  

When she was with Barry Karger's research group at Northeastern University (Boston, Massachusetts), Prof. Iulila Lazar successfully confronted the challenge of on-chip pumping (10). Two sets of multiparallel nanochannels produced a hydrodynamic-gradient, HPLC flow directly on the chip, effecting the separation and analysis of enzymatic digests. Such work represents the kind of innovation needed to produce independent, chip-based, HPLC technology. However, the technology must pass the practicality test of commercial manufacturing and end-user success for routine problem solving. Prof. Lazar's research continues at the Virginia Polytechnic Institute (Blacksburg, Virginia) in the Virginia Bioinformatics Institute (11). To date, though, there appears to be little effort to commercialize this approach to LC–MS-on-a-chip for ESI-MS. 

More recent commercial developments in the area of chip-based LC–MS originated from Agilent Technologies (Santa Clara, California) several years ago. The Agilent HPLC Chip Cube is a novel, integrated, two-dimensional HPLC column system formed from a disposable chip of polyimide substrate that includes an integrated electrospray emitter at the terminus of the HPLC packing material. This nicely integrated chip-based device is easily introduced manually into the center of a special HPLC injector. The injector accommodates sample loading and injecting as well as gradient formation of mobile phase fed from a modern external nano-HPLC pumping system and its associated autosampler. This arrangement precludes any plumbing connections by the user and nicely provides an interface for transferring sample between the real world and the chip world via the autosampler. The Chip Cube, commercially available for the past several years, is currently usable only with the Agilent electrospray mass spectrometer product line. Convincing reports of this system's performance have appeared in the recent literature (12). Figure 5 denotes those portions of the Agilent chip that correspond to conventional non-chip components of a conventional nano-LC–MS experiment.  

More recently, Waters Corporation (Milford, Massachusetts) introduced its Trizaic ceramic, or "nanoTile", chip. Although the chip parallels Agilent's Chip Cube concept, it nevertheless does so with some interesting differences. The company's UPLC technology device uses a ceramic microfluidic device packed with Waters' sub-2-μm chemistries to provide high-peak-capacity separations and increased usability. The nanoTile substrate is available in scalable formats whose internal diameters range from 75 μm to 300 μm and make them ideal for use in proteomics, biopharma, and small molecule pharma separations. Figure 6 shows a comparison between a modern nano-UHPLC column and the nanoTile chip for the LC–MS analysis of a tryptic digest of 

. This device can withstand sustained use. Moreover, it is easy to use, this according to its developers who claim they eliminated the steep learning curve that affects new users vis-à-vis adjusting fittings and affixing emitters on the nano-scale. As of this writing, Waters has not yet launched the technology. In any event, whether it receives widespread acceptance and commercial success remains to be seen.  

Prof. M. Ramsey and colleagues (University of North Carolina, Chapel Hill) (13) reports an elegant, chip-based, research approach that describes a two-dimensional nano-HPLC–CE combination on a single chip. It incorporates capillary LC coupled to an integrated lossless sample transfer for electrophoretic separation and subsequent ESI via nanoelectrospray. The report describes nano-LC coupled serially to CE with ESI-MS detection(13). The experiments are conducted by rapid, repetitive, CE sampling from the relatively longer nano-LC chromatogram into a much shorter CE channel (for rapid separation). The process effects an electrophoretic separation within seconds, followed by the next CE sampling, taken from the nano-LC chromatogram as the latter progresses. The effluent from the CE capillary is directed continuously to the nanoelectrospray emitter, which is microfabricated from the corner of the glass chip to provide real-time ESI-MS data from this multidimensional system. Given the tremendous analytical horsepower of truly orthogonal separation mechanisms, such as HPLC and CE coupled with ESI-MS detection, the chip-based integration of these techniques would command high merit were they ever commercialized and made easy-to-use.  

Such a commercially successful product would necessarily be fully automated and easy to use before scientists working on real-world problems would embrace it. Other issues remain, such as those involving proper loading of complex or dirty samples onto the nano-LC column and acceptable CE running buffers that are "MS-friendly," but for each a solution exists, making this approach potentially viable and competitive in the future. 

 Although chip-based mass spectrometers have been reported, this paper will not focus on this topic because commercial products based upon analytical applications employing condensed phase separation sciences appear to be quite a ways off in the future. The current systems employ electron ionization (EI) and sample inlets via GC. Readers interested in chip-based mass spectrometers should refer to recent reports (14–16). 

 Advion BioSciences (Ithaca, New York) addresses the real-world-to-chip issue in a simple and novel way (Figure 7). The company developed what it calls the ESI-Chip from a silicon substrate and commercialized it paired with a specially designed robot that delivers samples via a common disposable pipette tip from a microtiter plate to the chip (17). The robot can access as many as 400 samples serially during unattended ESI-MS analysis.  

Because most modern laboratories routinely use either 96-well or 384-well plates as sample holding devices for LC–MS and related applications, Advion developed the robotic TriVersa NanoMate system to accommodate the multiwell plate format. Instead of using an HPLC autosampler to access samples in a multiwell plate, the NanoMate system is programmed to access each of the wells in succession using a disposable pipette tip for each sample. The sample selection step involves aspirating an aliquot of sample solution from a targeted well in the plate and then delivering the solution to the back of an emitter microfabricated in the ESI-Chip, where it undergoes automated nanoelectrospray-MS analysis. The chip consists of a 20 × 20 array of microfabricated nano-ESI emitters (18) placed in sequence close to the ion-sampling orifice of an API mass spectrometer. The emitter and the pipette tip are single-use items that preclude the possibility of sample-to-sample carryover. If the samples within the multiwell plate are sufficiently cleaned up by a selective sample extraction technique, this infusion-MS process can provide a rapid (30–45 s/sample) analysis of each sample in the plate, without chromatography. The strengths of this system include automated nanoelectrospray, zero carryover between samples, and increased sensitivity with reduced matrix suppression, afforded by nanoelectrospray (19). In addition to the original simple infusion experiments, the system also can perform an LC–MS analysis, using a special LC coupler to the microfabricated chip emitter, while simultaneously collecting fractions into a multiwell plate using the TriVersa NanoMate system. Following an LC–MS-MS analytical run, you can program the system to return to targeted collected fractions (wells) for additional infusion MS-MS experiments, as frequently done in drug metabolite identification or peptide–protein identifications. 

Initial commercial acceptance of this technology was slow, but several novel applications have evolved. These include metabolite identification, protein sequence coverage (measuring biospecific noncovalent interactions), monitoring for chemical terrorism agents, profiling crude oil chemical composition (petroleomics), and characterizing and quantifying lipids (lipidomics). Although one could imagine potential for this technology in quantitative bioanalytical MS-MS applications, the relatively high cost of the additional hardware, as well as the consumable cost of the single-use chip, have contributed to the technology's lack of "market pull."  

Modern bioanalytical SRM LC–MS techniques are difficult to displace because the established technology works so well. In addition, most practitioners in small molecule measurements have been slow to adopt nano-ESI and other "nano" techniques, so a reluctance to adopt new technologies prevails absent truly compelling benefits. In contrast, the future potential for LC–MS applications in clinical diagnostics could prove significant. Rapid, infusion-MS techniques may well satisfy the need to analyze very high sample numbers (millions per year) for certain tests, which are now performed using LC–MS techniques. Supporting that prediction is a report on chip-based, infusion-MS determination of carnitines prefractionated off-line as dried blood spots for newborn screening using the TriVersa NanoMate system described previously (20). 

Innovative, simple, and practical LOC technology: 

A recent creative and simple approach to miniaturizing analytical chemistry for diagnostic purposes tends not toward LOC but "lab on a piece of paper and a cell phone." Prof. George Whitesides and his colleagues at Harvard University (Cambridge, Massachusetts) recently reported a variety of very simple devices using common materials rather than sophisticated chips microfabricated from silicon, polymer, or glass substrates (21). The do-it-yourself diagnostic kit does not yet include integration with MS, but it can be used in developing countries or in remote locations where placing specially trained personnel is difficult and costly. The Harvard investigators have described patterning Whatman paper with a photoresist to create a microfludic device with a central channel and four side channels leading to wells for different colorimetric assays (Figure 8). Whatman paper wicks fluids well, so it can soak them up without an external pumping source, and it filters out particulate material as well. Such features readily and simply solve some of the problems associated with on-chip movement of fluids and minimizing particulate blockages in microchannels. Reagents for glucose and protein assays are embedded within the wells. You can run both tests using visible oxidation reactions (wherein a camera-phone photographs the test results) and then send the images to a central facility where experts interpret the data and recommend treatment. Although not yet thoroughly field-tested, this is a very creative and reasonable approach to implementing miniaturized chemistry or biochemistry with simple materials. Regarding integrating mass spectrometry with this technology, you could possibly use direct analysis in real time (DART) (22), desorption electrospray ionization (DESI) (23) or analysis of samples at atmospheric pressure (ASAP) (24), if a low cost mass spectrometer were to become available.  

The previously described approach does not (for obvious reasons) employ a mass spectrometer. A developing, parallel strategy in the form of small, dried blood spots on Guthrie cards or high-quality filter paper, however, does. Dried blood spots have been used to screen newborns for inborn errors of metabolism (25) for more than a decade. Now this concept has rekindled interest within the pharmaceutical industry for pharmacokinetic measurements in drug development (26). Among the attributes of dried blood spots, which involve a form of miniaturization, are simplified sample collection, transport, and analysis. Following, usually, a one-step solvent extraction of the dried blood spot and, sometimes, a derivatization step, flow injection analysis–MS (FIA–MS) is performed to monitor a panel of targeted analytes. A recent novel report (27) showed direct extraction of analytes from small, dried blood spots using the TriVersa NanoMate system (vide supra) and surface sampling followed by chip-based nanoelectrospray MS. Other, related developments in this simple approach to sample collection and analysis, are possible in the near future, and those could be amenable to LOC techniques.  

Although the latter two innovations do not directly involve chip-based technologies, it is easy to imagine an evolutionary process leading to such a capability. Key elements that they share are simplicity, low cost, and ease-of-use — all of which will in coming years drive wider acceptance of LOC and miniaturization (28). 

A KISS — keep it simple, stupid! — product strategy, combined with a user-friendly interface and effective results may be what developers must first and foremost consider when pursuing new LOC concepts. To date, we have seen numerous reports of clever, microfabricated devices, and applications that purport to provide the promise of LOC benefits. Yet for the most part, we have seen only limited commercial success on the part of any company that has been bold enough to venture into the market offering LOC products. Although exciting to contemplate, "chips" have not been widely accepted yet in the commercial arena. Customers become enthusiastic about them at meetings and from novel literature reports of LOC developments. Yet LOC techniques have not demonstrated the compelling, cost-effective benefits required to displace current technology or workflow. Nor has it shown the ease of use needed to induce users to change. The major benefit claims of minimum sample requirement and solvent and reagent savings from LOC approaches seem encouraging. Ultimately, however, they are not so high a priority for most practitioners, who are not so constrained by considerations of sample waste that the benefit of acquiring LOC technology would clearly outweigh the burden of doing so. As for other waste, in most cases, scaling-down approaches such as micro- and nano-LC easily reduce solvent and reagent consumption to acceptable levels. Perhaps the mission to "go green" will help support LOC efforts, but for the foreseeable future, environmental considerations do not constitute a compelling reason for most to switch to chip-based approaches.  

Good reasons nonetheless abound for embracing LOC developments for future implementation, commercialization, and success. Despite the merits of our current sample analysis capabilities, many deficiencies in our analytical strategies and workflow continue to plague us: sample transfer losses, extra-column variance in chromatographic separations, expenses related to excess solvent and reagent disposal, and detection limitations associate with sample dilution — to name but a few. Although our modern sample preparation-analysis procedures coupled with MS work fairly well, there is always room for improvement. Finally, for many laboratories, acquiring personnel whose skill levels are equal to performing experiments using complex equipment is challenging. LOC techniques could lessen the challenge by integrating some procedure steps on an automated chip.  

Given that human nature disposes us toward risk averseness or laziness, we often find insufficient reason to change our ways without compelling grounds to do so. Thus, commercial success of LOC products that integrate MS will await a day when such products prove themselves easy to use, robust, reliable, and economically viable.  

The author wishes to thank Dr. Kathryn H. Henion for her valuable assistance in editing this manuscript and reproducing some of the figures. 

Dr. Jack Henion is Emeritus Professor of Toxicology at Cornell University, Ithaca New York, where he was a member of the College of Veterinary Medicine from 1976 to 2006. He is also a co-founder of Advion BioSciences, where he serves as Chairman and Chief Scientific Officer. From 1993 to 2006, he served as Chairman, President, and CEO. 

 "MS — The Practical Art" Editor Michael P. Balogh is principal scientist, MS technology development, at Waters Corp. (Milford, Massachusetts); a former adjunct professorand visiting scientist at Roger Williams University (Bristol, Rhode Island); cofounder and current president of the Society for Small Molecule Science (CoSMoS) and a member of LCGC's editorial advisory board.

MS–The Practical Art

In this installment of "MS-The Practical Art," guest columnist Jack Henion presents an historical and practical overview of lab-on-a-chip (LOC) research and the commercial achievement associated with it so far.

The Reality of Lab-on-a-Chip Technology for the Mass Spectrometry Laboratory

Ionization suppression is a complex phenomenon associated with atmospheric pressure-based ionization techniques such as electrospray ionization (ESI) that affects mass spectrometry (MS) detection (1–4). The differences in ionization response in ESI have been commonly attributed to deprotonation, increase in surface tension, increase in droplet size, presence of nonvolatile components, and in complex samples, large amounts of competing species that take some of the available ions. Here we summarize, in chronological order, some observations on ionization suppression with these reagents. Approximately one third of the reviewed publications mention ionization suppression, and of these, about 55% observed significant ionization suppression, while approximately 45% reported no significant ionization suppression. For ease of comparison, Tables I and II list some key parameters of the experimental and instrument setup of these investigators, but note that differences in sample matrix also could affect ionization suppression. 

Roughly 70% of the publications that reported significant ionization suppression used HFBA, three used NFPA, and one each used TDFHA and PFPA. Among the publications that reported no significant ionization suppression, only 43% used HFBA; 36% used PDFOA or TDFHA; two used NFPA; and one used PFPA. Amino acids and aminoglycosides accounted for a total of 28% of the significant ionization suppression group, versus 50% of the nonsignificant ionization suppression group. Proteins and peptides accounted for 28% of both groups. Diverse analytes accounted for 44% and 21% of the significant ionization suppression and nonsignificant ionization suppression groups, respectively. 

Table I: Publications reporting significant ionization suppression			

Keever and colleagues (15) noted that HFBA produced some signal suppression, but postcolumn addition of propionic acid enhanced response 10-fold for a cephalosporin antibiotic. Venkateshwaran and colleagues (25) noted that peptides and proteins are very susceptible to denaturation at extremes of pH, which leads to unfolding of the biomolecule and a change in the ESI charge state intensities towards lower 

 values due to an increase in the number of exposed amino acids, which increases the number of ionizable sites in the molecule. Chaimbault and colleagues (26), in a study of underivatized amino acids, using 0.25–5 mM PDFOA, reported that an increase in the 

 the intensity of the positive ion MS signal; that is, no suppression, but ionization enhancement. Watt and colleagues (16) observed that buffered 1 mM HFBA (pH 7.4) gave lower sensitivity, possibly due to reduced protonation of epibatidine. Petritis and colleagues (10) observed intense background signals at 

Table II: Publications reporting minimal ionization suppression			

Ishihama and colleagues (23) reported that perfluorocarboxylic acids have low surface energy, which is associated with weak intermolecular interaction and high volatility. However, PDFOA, but not TFA, is a surface active agent. In this case, surface tension had little effect on signal suppression. For myoglobin, PDFOA yielded almost the same signal as formic acid and did not change the charge distribution significantly, which reflects the conformation of a protein. Heller and colleagues (30) reported ESI is more sensitive for aminoglycosides than thermospray, and found ionization suppression to be far less severe with HFBA than with TFA. Gustavsson and colleagues (5) studied signal suppression using microflow injection analysis with 3 mM HFBA and 3 mM TDFHA on eight amines. The ESI–MS response decreased to only 30% for HFBA and to 15% for TDFHA, compared with formic acid, with no degradation of the ESI interface performance after 24 h of continuous infusion. With constant pH (3.4–3.7) and ionic strength, there was no correlation between the analytes' p

 values and their ESI responses. Kwon and colleagues (31) compared sensitivity of detection of amino acids with 6.5 mM NFPA and HFBA, and NFPA provided less background chemical noise. Shoji and colleagues (36) obtained satisfactory results with a C30 column and 20 mM HFBA versus C18 and 60 mM, which had significantly suppressed the ion intensity in tetrodotoxin analysis. 

Garcia and colleagues (6) obtained very small responses with 15.4 mM HFBA in flow injection analysis experiments with intact proteins, compared with formic acid. Petritis and colleagues (24) compared HFBA, NFPA, TDFHA, and PDFOA and found that 10 mM perfluorocarboxylic acids with ESI–MS created an ionization suppression of 1.5–2 fold in underivatized amino acids, and was almost the same in small peptides. However, in the 0.5–2 mM range, it was not significant (less than 1.4-fold). Lindemann and colleagues (35) found 7 mM HFBA produced no significant signal suppression in the ESI mode. McCormack and colleagues (17) reported 7.7 mM HFBA suppressed the response of all the siderophores studied, relative to formic acid. 

Piraud and colleagues (27) reported that under neutral conditions, amino acids are delivered to the source as zwitterions, which probably limits their ionization, but in 0.5 mM TDFHA, they 

 that was molecule-dependent. Keevil and colleagues (13) used a step gradient with a decreasing concentration of HFBA (starting with 1 mM), where tobramycin and sisomycin were eluted into the mass spectrometer at nearly zero HFBA concentration with some loss of sensitivity due to signal suppression. Teerlink and colleagues (29) analyzed carboxymethyl lysine with a gradient of 5 mM NFPA as solvent A and acetonitrile as solvent B and reported no significant ionization suppression.  

Dillon and colleagues (37) analyzed recombinant antibodies and observed MS signal reduction with the combination of 0.2% TFA and 2.3 mM HFBA. Consequently, they reduced the acids to 0.05% TFA and 0.4 mM HFBA. In a study of peptides and proteins, Walcher and colleagues (8) noted the absolute MS signal intensities were more than three times higher with HFBA versus TFA, but a significant increase in the noise level was obvious with HFBA. Wang and colleagues (22) in a study of intact protein separations reported concentrations greater than 1.5 mM HFBA (versus TFA) dramatically reduced protein ion intensities. 

Piraud and colleagues (28) analyzed underivatized amino acids with gradients containing 4 mM NFPA, 0.5 mM and 1.5 mM TDFHA, or 0.5 mM PDFOA, and did not observe significant variation of the MS signal. Midttun and colleagues (38) used a ternary gradient with an effective constant concentration of 2.5 mM HFBA and noted that ionization suppression was most pronounced for pyridoxine 5'-phosphate and pyridoxamine 5'-phosphate.  

Gao and colleagues (34), in the determination of methadone in human plasma, added 6.3 mM HFBA, NFBA, TDFHA, and PDFOA to the samples, but adding 93 mM HFBA to the sample reconstitution solution gave a 10.7-fold higher response. Naldi and colleagues (7) found a C4 column and 3.1 mM HFBA in the analysis of histone proteins gave better selectivity than 0.4% formic acid, but the ionization suppression was higher with HFBA. Lecaroz and colleagues (32) used 20 mM PFPA and samples prepared in mobile phase containing PFPA. By using a step gradient with a decreasing concentration of PFPA, gentamicin was eluted into the mass spectrometer at nearly zero PFPA concentration, solving the problem of ionization suppression. Rinne and colleagues (39) tested acetic acid, TFA, PFPA, and HFBA for the analysis of catecholamines on a porous graphitic carbon column and observed severe signal suppression with TFA, PFPA, and HFBA, compared to acetic acid. Jin and colleagues (21), in the analysis of diallyldimethylammonium chloride, chose 5 mM HFBA because 20 mM suppressed the signal. 

Hakkinen and colleagues (19) analyzed underivatized polyamines and found 0.1% HFBA gave less signal suppression than 0.1% PFPA. Ding and colleagues (20) compared PFPA and HFBA and selected 9.6 mM HFBA because it gave a better separation of synephrine, but they observed ionization suppression. 

Cherlet and colleagues (14) found that 20 mM PFPA reduced the signals of dihydrostreptomycin and streptomycin to 50% and 56%, respectively, compared to direct infusion without PFPA. Gu and colleagues (12) analyzed amino acids using 1 mM TDFHA as solvent A and acetonitrile as solvent B. High variation in intra-and interday precision of peak area ratios was attributed to poor ionization of glycine, likely due to interferences from solvent clusters. Orioli and colleagues (9) analyzed histidine and peptide adducts in urine samples, comparing HFBA and formic acid to 5 mM NFPA, which significantly suppressed the extracted ion current, but less than HFBA (a 15–40% increase with NFPA versus HFBA). Shen and colleagues (40) tested five different perfluorocarboxylic acids, and as the chain length increased, the observed atmospheric pressure chemical ionization (APCI) — not ESI — ion current increased.

De Person and colleagues (11) compared the performance of two MS instruments that differed in their atmospheric pressure sources, using amino acids and several perfluorocarboxylic acids. Previous authors observed a higher signal decrease with instruments of linear geometry in comparison with mass spectrometers of orthogonal or even Z-spray geometry. Whatever the spectrometer, the large differences in the response factors for all amino acids infused under the same conditions indicated that ion desolvation was highly dependent on the nature of the amino acid residue. Differences between the interfaces suggested that ion transmission was strongly dependent upon the ion source geometry. The MS signal increased up to NFPA compared with shorter chain homologs but decreased dramatically with TDFHA and PDFOA, with the Z-spray interface. However, with the curtain gas interface, the highest relative intensity was obtained with TDFHA and PDFOA, yielding significantly increased amino acid MS response over their shorter chain homologs. Signal suppression was evaluated with the curtain gas interface using only NFPA, and the optimum concentration was between 10 and 20 mM. However, with the Z-spray interface, the greater the NFPA concentration, the higher the amino acid signal, and therefore it was compatible with very high surfactant concentrations. They concluded that the Z-spray is more sensitive than the curtain gas interface to the concentration of ion-pair reagent and consequently is a less robust instrument for this type of analysis. However, these authors have not accounted for the partitioning effects of the ESI probe, where some amino acids move to the fringe of the spray plume while some remain in the core. The momentum effect coupled to electrostatics causes dispersion in the mass-to-charges within the plume. The Z-spray requires a 90° turn while the curtain gas interface requires in excess of a 90° turn to get into the sample orifice, which can cause differences in sampling and significantly bias the intensity. This is akin to comparing lenses with different focal points. If a "common" ESI probe was pointed at the orifice, then one could isolate whether it is the Z-spray or curtain gas geometry that plays a role. Without this experiment, the results are inconclusive regarding the ESI geometry.  

Approximately 60% of all the reviewed publications used C18 columns of varying dimensions (from semipreparative to nanospray) packed with particle sizes of 3–5 μm, and only one used a column packed with sub-2-μm particles (40). About 20% used C8 columns, and with only two exceptions (32,41), all C8 columns were used with HFBA. Six (8,6,17,42–44) used polystyrene–divinylbenzene (PSDVB) columns, five (17,39,45–47) used columns packed with porous graphitic carbon, two (36,48) used C30, and one each used C12 (9), C4 (7), cyano (49), and chiral (16) packing in columns. Because of the very acidic pH of the unbuffered perfluorocarboxylic acids, earlier publications reported some silica stationary phase degradation over time, but later publications used newer silica stationary phases that are resistant to acid and heat. PSDVB and porous graphitic carbon are more resistant to acid hydrolysis than silica. However, Rinne and colleagues (39) reported that 0.1% PFPA in the mobile phase was involved in on-column oxidation of catecholamines, by slow oxidation of the porous graphitic carbon material (to ionic graphite salts). 

One third of all the reviewed publications used isocratic elution and two thirds used gradient elution. In some instances of gradient elution, the concentration of the ion-pair reagent was kept constant, but in most cases, it decreased. Although nearly 20 reviewed publications report system peaks, column equilibration issues, and peak drift, about 80% do not report these phenomena, in spite of the disruption of the dynamic equilibrium with gradient elution. This might be due to the increasing popularity of "fast" liquid chromatography (LC)–MS-MS, where "interferences" can be "tuned out" and system peaks are not as obvious, although peak drift and equilibration issues could persist. 

Only 25% of the reviewed publications reported a pH value for the mobile phase or for any portion of the mobile phase used in the analyses. The majority of the reported pH values for mobile phases are "s/w pH" (older definition was "apparent pH") values and half are buffered mobile phases (pH 3.0–4.0, with a few exceptions); 30% are unbuffered mobile phases (pH 1.5–3.0); and 20% are mixtures of perfluorocarboxylic acids and TFA, for example (pH 1.7–2.8). For a discussion of pH measurements in aqueous buffer–organic mobile phases, there are several recent reviews (50–55). 

We have measured the pH of aqueous PDFOA at various concentrations in our laboratory as: 0.01%, pH 4.4; 0.02%, pH 3.1; 0.06%, pH 2.9; and 0.08%, pH 2.8. For comparison, Kotrebai and colleagues (56) report aqueous 0.1% TFA, pH 2.05; 0.1% PFPA, pH 2.21; and 0.1% HFBA, pH 2.55. Garcia and colleagues (6) report a pH of 2.5 for 0.2% formic acid and a pH of 2.5 for 0.025% TFA. Shen and colleagues (40) state that "TFA is the strongest acid with a p

 of 0.3, while PDFOA is the weakest of the series, with a p

 of 2.5." These investigators prepared two mobile phases with the same concentration of NFPA, but adjusted the s/w pH value to 6.8 and 2 and tested them in the LC separation of diastereomers. Very little shift (5%) in retention time was observed. However, Snoble and colleagues (57) did detect slight retention shifts for a nine-peptide test mixture with two-day-old, room-temperature mobile phases containing dilute amounts of formic acid and methanol.  

Goss (58) reported that the literature has used a p

 of 2.8 for PDFOA, based upon a measurement of 50:50 (v/v) ethanol–water. For existing compound data, a p

 value measured in a 50:50 alcohol–water mixture is typically about 1 log unit higher than what is measured in water. An artifact that leads to overestimation of the p

 value of PDFOA, sorption to interfaces such as the water surface or the walls of a glass vessel can occur to an extent that is unknown for ordinary molecules. The titrations will be shifted significantly toward higher pH values. The calculated p

 of PDFOA is –0.5, while the calculated p

Only half of all the reviewed publications report a controlled column temperature, which is surprising because it is especially important with ion pairing in reversed-phase chromatography, as Dolan (59) notes. A total of 16 publications report column temperatures above 30 °C; 13 between 35–55 °C; and one each at 60 °C (42), 70 °C (37), and 100 °C (40). Among the potential advantages of elevated column temperatures are faster separations with reduced backpressure, sharper peaks, and potentially reduced ionization suppression in ESI–MS detection. Selectivity differences also have been reported with different column temperatures. For example, Castro and colleagues (60) reported that a change in a C8 column temperature from ambient to 50 °C using 15 mM HFBA gradient elution "completely changed the elution order of several quaternary ammonium herbicides." 

Several recent publications (61–63) review elevated temperature chromatography and its potential advantages. In our review, we have not found any obvious patterns in the relatively small number of articles using the perfluorocarboxylic acids at elevated column temperatures, especially above 55 °C. However, because most analytes are eluted faster with higher column temperature, it could be useful to reexamine equilibration times (that is, mobile phase volumes), selectivity, and mobile phase ion-pair concentrations with TDFHA and PDFOA on the newer acid-and heat-resistant C18 stationary phases. 

Five reviewed publications discuss system peaks in detail, while six report equilibration issues, and eight mention peak drift with perfluorocarboxylic acids. System peaks are not unique to ion-pairing chromatography and can have many causes. Several recent reviews (64–66) discuss this phenomenon in detail, so we will limit our discussion of "ghost peaks, pseudo peaks, eigenpeaks, vacancy peaks, induced peaks, artifact peaks, dip peaks, perturbation peaks, or spurious peaks," to reports of this phenomenon specific to these reagents. Interestingly, system peaks are not reported after 2002 in the reviewed literature, although equilibration issues and peak drift are reported as recently as 2007 and 2008, respectively.  

Chaimbault and colleagues (26) used gradient elution for basic amino acids, based upon simultaneously increasing the concentration of acetonitrile and decreasing the concentration of the ion-pair reagent. At the end of each chromatogram, a system peak was observed, in addition to the solute peaks. "System peaks are observed when an injection is made in a chromatographic system previously equilibrated with a mobile phase consisting of a solution with a constant concentration of an additive that is involved in the retention mechanism. These peaks appear after perturbation of the equilibrium, when some of the adsorbed additive molecules are desorbed and migrate along the column as an excess additive band." System peaks can mask the elution of one or more analytes with evaporative light scattering detection, but MS can detect the analytes under a system peak. PDFOA induced two system peaks: one near, but before the end of the chromatogram, and the second at the end of the chromatogram, probably due to impurities in PDFOA that accumulated during column equilibration and were desorbed at the same time as the surfactant. 

Petritis and colleagues (67) ensured that no irreversible surface modification occurred between each experiment by using extensive column washes (3 h to flush the column). Previous experiments using various perfluorocarboxylic acids with UV detection had problems with baseline drifts and UV spikes due to UV-absorbing contaminants. With evaporative light scattering detection (ELSD) and ESI–MS, however, if the contaminants are volatile, they will not be a problem. The adsorption of different perfluorocarboxylic acids on the stationary phase was determined by measuring breakthrough with a conductivity meter. Adsorbed quantities of surfactant increased with surfactant concentration in the mobile phase, as well as with the surfactant alkyl chain length on all three C18 columns tested. Whatever C18 column was used and the column dimensions, equilibration times were almost the same for a given surfactant at a given concentration, and the total adsorbed quantities of surfactant were similar on the three different columns tested. Moreover, for a given perfluorocarboxylic acid, the equilibration time was independent of the surfactant concentration in the mobile phase. These surfactants had the following approximate equilibration times: PDFOA, 3 h; TDFHA, 1.25 h; NFPA, 15 min; and HFBA, 5 min. Note that all of these experiments were conducted with unthermostated columns at room temperature. Furthermore, it is the volume of the mobile phase at a given concentration of ion-pair reagent (not time) that passes through the column that is important (50). However, because a range of flow rates are used in the literature, equilibration time is referenced here for convenient comparison of different ion-pair reagents. 

Chaimbault and colleagues (45) conducted further experiments with this "breakthrough system" using a porous graphitic carbon column at 10 °C. HFBA and NFPA had short equilibration times whatever the surfactant concentration range (0.5–15 mM). However, TDFHA required 35 min and PDFOA required 105 min to be equilibrated. On this column with NFPA, there were no system peaks during gradient elution, perhaps because of the very low quantities of NFPA adsorbed on the porous graphitic carbon surface, in contrast with that of PDFOA. 

Petritis and colleagues (24,68,69) continued work using the "breakthrough system" with underivatized amino acids on a C18 column at 25 °C and indirect conductivity detection. System peaks can appear in a chromatographic system containing more than one eluent component after injection of a solution with a composition different from the bulk eluent. For example, an injection of 10 μL of water into the equilibrated system caused a small positive peak in the void volume and an intense negative peak later (main system peak). For TDFHA, a small system peak appeared after about one hour and for PDFOA after about 3 h. There was a linear relationship between the mobile phase concentration of reagent and the peak area of the system peak. Moreover, the system peak always was eluted at the same time as the breakthrough of the adsorption isotherms. Even with ELSD, system peaks can be observed after a violent gradient, or after violent desorption from a chromatographic support of high concentrations of volatile compounds. The detector signal given by these peaks is attributed to sudden differences in mobile phase volatility. However, small differences in mobile phase composition occurring after perturbation of an equilibrated chromatographic system are not detectable with ELSD. Later analysis of underivatized small peptides using ESI–MS and TDFHA with an acetonitrile gradient elution produced a system peak that increased in size with increasing TDFHA concentration at the end of the chromatogram and masked the elution of some peptides. Reequilibration time was about 80 min with TDFHA, but only 15 min with NFPA. In the analysis of underivatized amino acids, the elution order was always the same whatever the concentration of PDFOA. However, with some reagents such as TDFHA, inversions in the elution order of some amino acids occurred with increasing TDFHA concentration. 

Zhu and colleagues (70) tested HFBA, NFPA, and PDFOA in the analysis of acetylcholine, and all gave good resolution. However, because HFBA had a shorter equilibration time, it was chosen. Kwon and colleagues (31) used 6.5 mM NFPA in the analysis of underivatized amino acids with gradient elution and a 20-min equilibration time. The retention time reproducibility for the 20 amino acids tested was better than 5%. In the analysis of 76 amino acids, Piraud and colleagues (28) used a 15-min equilibration time with 0.5 mM TDFHA before the first gradient, as well as between two gradients, which gave good repeatability of the retention times, even after 20 consecutive gradients. However, the first gradient had to be discarded. Gao and colleagues (34) added HFBA, NFPA, TDFHA, and PDFOA to the sample and found the uptake of the reagents on the C18 column mainly determined the retention of analytes. With HFBA or NFPA in the reconstitution solution, gradient LC did not require extensive equilibration times and carryover was reduced. However, it was difficult for systems using TDFHA and PDFOA to equilibrate. This method also might potentially address the carryover problem for other polar compounds (for example, aminoglycosides) that form ion pairs. Spacil and colleagues (71) used 3.2 mM NFPA and found in cases of gradient elution, a 15-min interval was required between runs for column equilibration. 

Peak drift can occur in LC using these reagents, even with standard solutions after several injections over time, or be exacerbated by sample matrices. With one exception (72), all the publications mentioned in the following reported controlled column temperatures ranging from 20 °C to 40 °C. Cherlet and colleagues (72) observed important matrix effects on retention time stability and peak shape for the analysis of gentamicin in tissue samples using isocratic elution with 20 mM PFPA. A shift of retention times and peak tailing were observed when comparing the sample and standard chromatograms. Gammelgaard and colleagues (41), using isocratic elution with 3 mM NFPA, noted the retention times and sensitivity changed slowly during the lifetime of the columns. Wybraniec and colleagues (73) observed with 0–5 mM PFPA and HFBA, there was some shifting of retention time for betacyanins, using gradient elution. Qu and colleagues (74) used 1 mM PDFOA and gradient elution for the analysis of glutamine and other amino acids and attributed peak drift to the accumulation of PDFOA in the column, which slowly modified the surface of the C18 phase. To avoid this, they flushed the column after every six chromatographic runs. In a later article (75), they used even more extensive flushing when 0.8 mM TDFHA was used with gradient elution. Piraud and colleagues (76), using 0.5 mM TDFHA with gradient elution, observed a slight drift of retention time after analyzing many biological samples for amino acids. However, this was limited by using acetonitrile column washing every 14–16 samples. Heller and colleagues (77) used column temperature regulation (30 °C) and isocratic elution with 55 mM TFA to eliminate occasional retention time drift, which contributed to peak-shape variations. Heavily used columns produced ghost peaks in the analysis of control samples of gentamicin. Liu and colleagues (78), using 5 mM NFPA with gradient elution, reported that long-term injection of amino acids lead to retention time drifts. A thorough column cleaning with methanol was necessary when the retention time of amino acids was seen to decrease after weeks of sample injections. The retention time shift in this study is opposite to what was reported by Qu and Piraud. Eckstein and colleagues (79) reported that the chromatographic durability with gradient elution using HFBA versus PDFOA was improved, as thousands of cerebrospinal fluid samples were analyzed for glutamine and related compounds, with no change in retention or durability. Andrieux and colleagues (33), using isocratic elution with 10.9 mM HFBA, reported that column washing after every 50 samples preserved column life for more than 1000 injections with no shift in retention time or degradation of peak shape. 

The volatile perfluorocarboxylic acids offer a range of choices for selectivity and are valuable and versatile ion-pair reagents. They also provide good MS-compatible (and a host of other detection methods) LC gradient and isocratic separations for an increasingly diverse selection of polar analytes (without requiring derivatization), as well as proteins and peptides. In recent years, nearly half of all the reviewed publications covered analytes other than aminoglycosides, amino acids, proteins, and peptides. 

While some significant ionization suppression has been reported in about 20% of the reviewed publications, mostly with HFBA, this disadvantage has been manageable in most cases. The longer alkyl chain homologs, used at lower concentrations, offer the potential for reduced ionization suppression. Furthermore, knowing whether the MS instrument ESI geometry truly has an effect on the ionization suppression of specific analyte–ion pair reagent combinations might help to minimize this suppression. 

Good column temperature control (including mobile phase prewarming for elevated temperatures) and proper mobile-phase pH adjustment and measurement are essential for reproducibility in all LC methods, and ion-pairing LC is no exception. Beginning with the nature of the sample, one should ask: Where did it come from? Is it stable when dissolved in acidic mobile phase? Will the ion-pair reagent in the sample reduce carryover (for example, gentamicin), eliminate double peaks, show "matrix effects," or affect the column equilibrium? Also consider LC column stationary phase compatibility with very acidic mobile phases, potentially at elevated temperatures; column equilibration; gradients used; flushing; storage; and dedication of the column to a specific application. 

Faster LC separations and acid-and heat-resistant C18 stationary phases offer new opportunities to explore the performance of PDFOA and TDFHA at column temperatures up to 90 °C. With a fast separation, the thermal stability of various polar analytes might not limit such an approach (64), and column equilibration times with these reagents might be reduced dramatically. Indeed, it can be advantageous to delay the too-rapid elution of some very polar analytes under such conditions with these reagents. Furthermore, PDFOA has shown greater selectivity than HFBA and NFPA for amino acids, for example, and is an effective reagent at lower concentrations than NFPA or HFBA (for example, 0.5 mM vs. 5 mM, or 4–8 mM, respectively). Also, PDFOA often is associated with less ionization suppression than HFBA and NFPA (11), potentially increasing sensitivity in ESI–MS detection. Daily preparation of fresh mobile phase containing methanol or the use of acetonitrile as an organic modifier could prevent the formation of esters in the mobile phase (40,57). 

According to Desmet (80), the advantage of sub-2-μm particles diminishes if columns are run at elevated temperatures. Nevertheless, Shen and colleagues (40) achieved fast baseline resolution of diastereomers with a column temperature of 100 °C using sub-2-μ particles and NFPA. The purity of PDFOA and all ion-pair reagents, including the potential effects of branched isomers, should be considered carefully in future investigations. Although LC methods using hydrophilic interaction chromatography and other techniques are becoming increasingly popular, we believe the perfluorocarboxylic acids have much to offer and their use, especially in LC–ESI-MS, will continue to increase.  

(1) R. King, R. Bonfiglio, C. Fernandez-Metzler, C. Miller-Stein, and T. Olah, 

(5)  S.A. Gustavsson, J. Samskog, K.E. Markides, and B. Langstrom, 

(6) M.C. Garcia, A.C. Hogenboom, H. Zappey, and H. Irth, 

(7) M. Naldi, V. Andrisano, J. Fiori, N. Calonghi, E. Pagnotta, C. Parolin, G. Pieraccini, and L. Masotti, 

(8) W. Walcher, H. Toll, A. Ingendoh, and C.G. Huber, J. 

(9) M. Orioli, G. Aldini, M.C. Benfatto, R.M. Facino, and M. Carini, 

(10) K. Petritis, V. Dourtoglou, C. Elfakir, and M. Dreux, 

 (11) M. de Person, P. Chaimbault, and C. Elfakir, 

 (13) B.G. Keevil, S.J. Lockhart, and D.P. Cooper, 

Article

In Part II, we review MS ionization suppression; column, pH, and temperature selection; and system peaks and column equilibration issues associated with the use of five volatile perfluorocarboxylic acid reagents in LC.

Review of Volatile Perfluorocarboxylic Acids as Ion Pair Reagents in LC: Part II

The description of a chromatographic procedure in a 

) monograph contains necessary information that enables the procedure to be reproduced in a laboratory and typically includes system suitability requirements. System suitability is based upon the concept that "the equipment, electronics, analytical operations, and samples to be analyzed constitute an integral system that can be evaluated as such" (1), and it is used to demonstrate the system is stable and suitable for its intended use. In some instances, the system fails to meet those requirements, resulting in the need for corrective actions. The deviations can be temporary (short-term excursions caused by a particular factor or combinations of factors) or more permanent. The causes for such deviations might be assignable or might be of uncertain origin attributable to normal system variability. 

Because this situation can arise, the System Suitability section of 

 acknowledges that adjustments to operating conditions might be necessary to meet system suitability requirements and includes the maximum variation allowed for each chromatographic parameter. Users are allowed to implement these adjustments to bring the system into suitable performance without a full validation of the analytical procedure. The influence of the changes should be assessed by verification of the procedure under the adjusted chromatographic conditions. Chromatographers must demonstrate that the changes introduced will not affect the performance of the procedure adversely. If the allowed adjustments to the system are not successful and assistance from USP staff has not resolved the problem, the procedure might no longer be suitable for use with the article being tested. This conclusion could be a laboratory-specific issue (for example, a bad column) or it could be serious enough to warrant revision of the existing monograph. This article discusses the flexibility and limitations of the System Suitability section in the current version of 

 and provides some specific examples to help analysts understand how to take advantage of the flexibility that is allowed in this chapter.

Adjustments to the Chromatographic System

 allows a number of adjustments to the chromatographic system in response to issues with a procedure's system suitability. These are the maximum allowable variations unless the individual monograph directs otherwise.

 Although the replacement of any solvent in the mobile phase constitutes a change and not an adjustment, the amounts of the minor components (specified as 50% or less) in the mobile phase can be adjusted up to ±30% relative to the particular component. However, the change in any component cannot exceed ±10% absolute (that is, in relation to the total mobile phase). Adjustments can be made to one minor component in a ternary mixture.

 The pH of the aqueous buffer used in the preparation of the mobile phase can be adjusted to within ±0.2 units of the value or range specified.  

The column temperature can be adjusted by as much as ±10 °C. 

 The concentration of the salts used in the preparation of the aqueous buffer for the mobile phase can be adjusted up to ±10% relative to the particular component, provided the permitted pH variation (see above) is met. 

Deviations from the wavelengths specified in the method are not permitted.  

Column length: The column length can be adjusted by as much as ±70%.

Column inner diameter: The column inner diameter can be adjusted, provided that the linear velocity is kept constant (see Flow Rate). 

 When column dimensions have been modified, the flow rate can be adjusted using the following formula: 

 is the flow rate indicated in the monograph, in milliliters per minute; 

 is the adjusted flow rate, in milliliters per minute;  

 is the length of the column indicated in the monograph;  

 is the column inner diameter indicated in the monograph; and 

 is the internal diameter of the column used.

				Additionally, the flow rate can be adjusted by ±50% (2).			

The injection volume can be reduced as far as is consistent with accepted precision and detection limits, but no increase is permitted. 

Particle size can be reduced by as much as 50% but cannot be increased.

The degree of verification needed to implement these adjustments varies and depends upon the purpose of the procedure, the extent of the changes, and the performance characteristics that could be affected by the changes — see USP General Information Chapter Verification of Compendial Procedures 

 (3). Meeting system suitability requirements might not be sufficient to demonstrate that the procedure has been verified successfully. Especially in older monographs, the requirements for the system suitability test might be weak, and other experiments might be needed to demonstrate acceptable system performance.

The following examples demonstrate the flexibility and limitations offered by the current version of 

Mobile phase composition is one of the primary causes for the inability to meet the system suitability criteria provided in an official monograph or in a monograph proposal published in 

). An example is the olanzapine proposal that appeared in 

 (4). The proposal included the following chromatographic conditions and system suitability criteria: 

Mobile phase: Buffer–acetonitrile (50:50). 

Resolution requirement: Not less than (NLT) 2.0 between olanzapine-related compound A and olanzapine. 

 proposal, some users observed that the resolution requirement could not be met easily. Figure 1 shows the chromatogram obtained under the conditions in the 

The observed resolution was 1.7. According to 

 for the specified ratio of 50:50, 30% percent of 50 is 15% absolute, but this exceeds the maximum permitted change of ±10% absolute in either component. Therefore, the mobile phase ratio of buffer–acetonitrile can be adjusted only within the range of 40:60 to 60:40. Decreasing the organic component relative to the buffer is potentially a way to obtain the required resolution. In this case, when the organic content of the mobile phase composition was changed to 47% (buffer–acetonitrile 53:47), the resolution improved to 7.0 (Figure 2). 

USP received a comment from another user stating that the drug substance has a specified process impurity that elutes very close to the drug substance. Lowering the mobile phase organic component is likely to improve the separation between the drug substance and the process impurity as well. Presented with this information, the USP Expert Committee responsible for the approval of this monograph decided to revise the mobile phase composition to buffer–acetonitrile (53:47) in the official monograph. Changing only the mobile composition according to 

 might not lead to successful results in all cases. Thus, the chapter recommends that users contact USP staff for assistance. The next example demonstrates the outcome of an interaction between the user and USP staff.

The paroxetine hydrochloride monograph in 

 (5) had the following chromatographic conditions: 					

Mobile phase composition: acetate buffer–acetonitrile–triethylamine (60:40:1).

System suitability requirements: resolution between paroxetine-related compound B and paroxetine NLT 2.0.

Tailing factor requirement for paroxetine peak: Not more than (NMT) 1.6.

Users reported that after one week of use with a new column, neither the tailing factor nor the resolution requirement could be met with Zorbax TMS, the column used in the validation studies. Although the mobile phase includes three components, lowering the acetonitrile content is likely to result in higher resolution. A systematic evaluation showed that 70% aqueous buffer (30% acetonitrile) was necessary to achieve the required resolution. 

The new chromatographic conditions included the following:  					

Mobile Phase: acetate buffer–acetonitrile–triethylamine (70:30:1). 

The new mobile phase conditions resulted in a tailing factor of more than 1.6. Tailing factors obtained in a robustness study with perturbation of column temperature and mobile composition were all greater than 1.6 (Tables I and II, respectively). 

Table I: Tailing factor and resolution data as a function of column temperature			

Because the tailing factor did not exceed 2.0 under any of the chromatographic conditions used in the robustness study, it was concluded that a tailing factor requirement of NMT 2.0 would be acceptable. The monograph was revised as follows (6): 

Table II: Tailing factor and resolution data as a function of mobile-phase composition			

Mobile phase: acetate buffer, acetonitrile, and triethylamine (70:30:1). (Note: The actetate buffer–acetonitrile–triethylamine ratio can be varied between 70:40:1 and 75:25:1 to meet the system suitability requirements.) 

Column Watch

To meet system suitability requirements, chromatographers might need to make adjustments to the high performance liquid chromatography (HPLC) operating conditions.

Adjustment to the Chromatographic System in U.S. Pharmacopeia Monographs: Flexibility and Limitations

I recently received an e-mail inquiry from a reader, along with the two chromatograms shown in Figure 1. Although not explicitly stated in the e-mail, it was clear that a debate was raging about how to best integrate this group of peaks. Proper integration procedures is a topic that comes up with surprising regularity, so I would like to look at some aspects of integration in this month's "LC Troubleshooting."  

In Figure 1, it is possible to distinguish three peaks. Peak 1 is just a shoulder on the front of peak 2, whereas peaks 2 and 3 are distinct peaks. So the question is how to best integrate this set of three peaks to get results that are the most accurate — that is, most closely reflect the true area under the peaks. In Figure 1a, a valley-to-valley integration method is used. On the one hand, it may look like this is a good approach, but it misses peak 1 altogether. And, although the integrated area (above the drawn baseline) clearly belongs to peak 2 or peak 3, there is a gross under-integration of the two peaks. That is because all of the area beneath the integration line is ignored.  

The correct way to integrate a group of peaks like this is to draw a perpendicular line from the valley between the peaks to the baseline extended between the normal baseline before and after the group of peaks, as seen in Figure 1b. For peak 1, it takes a bit of imagination to pick the correct point to drop the valley, and as we'll see in a minute, this is probably not appropriate anyway. For peaks 2 and 3, the process is simple. First draw a baseline connecting the real baseline before and after the peak group. Then draw a perpendicular line from the valley between each peak pair to the baseline. 

The errors involved in the perpendicular drop method are as follows: If the peaks are approximately the same size, and tailing or fronting is ignored, the amount of the peak tail from the first peak (peak 2 in the present case) hiding under the second peak (peak 3) should be about the same as the amount of peak front from the second peak hiding under the first. If this is the case, the errors should cancel and peak areas should be fairly accurate. If the second peak fronts significantly or if the first peak has a strong tail, the weighting will be distorted, with corresponding errors. If the peak ratio is large — for example, 20:1 — the larger peak will be little affected by the minor contribution of the smaller peak, but the smaller peak will have excess area contributed by the major peak. In this case, the accuracy for the larger peak should be much better than for the minor peak. When the resolution between the peaks is so small that a clear valley is not present, as in the case for peak 1 in this example, the perpendicular drop will grossly over-integrate the peak. A peak skim is more appropriate for integration of peak 1; see further discussion of Figure 2b below.  

A related question is when, if ever, is it appropriate to use a valley-to-valley integration technique? The simple answer is that if a known baseline disturbance is present under a set of eluted peaks, valley-to-valley may be appropriate. In the case of Figure 1a, there would have to be a large, broad peak that is roughly defined by the area under the drawn baselines. This is unlikely to occur, and I can never remember encountering such a situation in my experience in the laboratory. On the other hand, in some gradient runs, there may be a small, broad rise in the baseline of blank runs that is consistent enough to allow valley-to-valley integration. In this case, however, the valleys between peaks should reach nearly to the baseline extended from before to after the peak group. This is also a rare occurrence, so the bottom line here is that valley-to-valley integration seldom is the best approach. 

While the valley-to-valley problem mentioned previously may not be encountered often, there are other integration errors that may be more common. I've shown three of these in Figure 2. In Figure 2a, the data system chose to draw the baseline (solid line) from the bottom of the dip before the peak to the baseline after the peak. This is a result of the default integration mode of connecting the baseline (lowest point) before and after the peak. The problem, of course, is that the negative peak just before the peak of interest is falsely identified as true baseline. The correct integration is shown with the dashed line. The area above the dashed line and within the peak envelope better represents the true peak area than the original integration. 

A different example of the slightly resolved peak 1 of Figure 1 is shown in Figure 2b. Here a small but distinct peak appears on the tail of the major peak. The data system chose a perpendicular drop to the extended horizontal baseline as the integration technique. It is easy to see in this case that the tail of the first peak extends under the minor second peak. The perpendicular drop will under-report the area for the large peak and over-report the area for the smaller one. The proper integration technique is to skim the smaller peak off the tail of the larger one, as shown by the dashed line. I've heard debates about whether this skim line should be linear (a tangent skim) or curved (an exponential skim) or some other shape. This may be of academic interest, but it is of little practical interest. Whenever a skim is used, it is merely an estimate of the peak area (see the conclusions at the end of this column) — consistency of integration methodology is more important than whether the skim is a tangent or curved. As a rule of thumb, if the minor peak is <10% of the height of the major one, skimming the peak is the appropriate integration technique. If the minor peak is >10% of the height of the major one, a perpendicular drop to the baseline connecting the true baseline before and after the peak group is best. 

Another very common integration error is shown in Figure 2c. Here the baseline after the peak is falsely located before the peak descends to the true baseline. The data system identifies the start and end of the peak by determining if the signal is rising (or falling) faster than the baseline would in the absence of the peak. This means that determining the end of the peak is especially problematic when the peak tails strongly, the peak is small, and the baseline drifts. When this situation exists, not only can the peak be integrated improperly, but also, the peak endpoint may vary significantly from one run to the next. The proper integration for Figure 2c is shown by the dashed line. One way to help avoid the problem of identifying a false peak-endpoint before the peak has been fully eluted, is to force the peak end to some point after you are sure the peak will reach the baseline. For example, in Figure 2c, this point could be where the peak-end is located with the dashed integration line or even a little to the right of that. Erroneously marking the peak end too late may have little practical effect, because the additional area created usually is a very small contribution to the overall peak area and, thus, a small contribution to integration error.  

If most of your experience is with UV or fluorescence detection and with peaks that are distinct on the baseline, the problems identified in Figures 1 and 2 will likely be the most common integration problems you encounter. If you switch to tandem mass spectrometric detection (MS-MS), however, you may be surprised at the results — I know that I was when I first encountered liquid chromatography (LC)–MS-MS. Figure 3a shows a typical chromatogram of the raw signal at the lower limit of quantification (LLOQ) for a method to determine a drug in plasma. The first peak is a metabolite and the second peak comprises the analyte and co-eluted 

C internal standard. I remember my first reaction to seeing such peaks: something is wrong here! But this is not at all the case. With the MS-MS instrument, at least in theory, because the electrometer counts ions exiting the third quadrupole, there is either a signal present or not. The magnitude of the signal will vary across the peak according to the concentration of the peak and the superimposed noise. When each data point is connected to the next, ragged peaks, such as in Figure 3a, are seen. The big question here is how to integrate the peaks. The baseline is noisy, but drawing an integration baseline through the middle of the noise to connect the true baseline before and after the peaks is not too challenging. The problem arises with where to locate the top of the peak. The solution to the problem is to smooth the peak. Peak smoothing applies a signal-averaging algorithm that averages the adjacent points to give a more pleasing peak appearance that is easier to integrate. If done properly, as is the case for Figure 3b, the peaks appear much more "normal" and accurately reflect the true peak area when integrated properly. Over-smoothing can compromise the signal, so care needs to be taken to avoid this.  

Before the advent of modern data systems, there was regular discussion about whether peak height or peak area was a more accurate way to quantify a peak. With manually measured peaks, peak area was estimated by triangulating the peak between points on the baseline at the beginning and end of the peak and at its apex. This meant three measurements, plus the assumption that the peak is a triangle, when it more closely represents a Gaussian distribution. Three measurements meant more error than the two required for the perpendicular height (apex and baseline), so it was argued that height was a better choice. Those arguments are somewhat moot today, because all data systems measure the peak as area or height slices from the baseline to each point along the peak. The height is the largest of these slices and the area is the sum of the slices. Most of the time, peak area will give better results because it inherently averages error across the peak, but area is not always the best choice. 

In Figure 1b, the peak height for peaks 2 and 3 is shown by the vertical dashed line. When two or more peaks overlap, as in Figure 1, peak height may be a better choice than peak area. If you were to draw the tail of peak 2 and the front of peak 3 to baseline, you would conclude that there is very little overlap at the center of the peaks where the peak height is measured. This would be expected to give less error than peak area, where there is known overlap. 

On the other hand, peak height for peaks such as those in Figure 3a would be expected to have a much greater probability of picking a false peak maximum than peak area, which averages out the noise across the peak. So peak area would be the preferred technique here. With the smoothed signal, the difference would not be expected to differ as much. 

The simplest way to determine the best integration technique is to run a set of known samples, such as during or before validation and collect data for both peak height and area. Calculate the results using both techniques and use the method that gives the most accurate and precise results. 

Another common question that I get regarding integration is the complaint that goes something like, "My boss won't let me adjust integration baselines in the chromatogram. He insists that the data system must be set to integrate all peaks properly and that the regulatory agencies will not allow manual baseline adjustments." This may be a fine argument for well-behaved methods with well-resolved peaks and negligible baseline noise, for example, content uniformity or potency tests for drug substance or drug product. However, whenever trace analysis is involved, such as impurities methods, drugs in plasma, or pesticides in the environment, peak integration can be much more challenging. As the peak gets smaller, the signal-to-noise ratio (S/N) gets smaller and it is harder for the integrator to find the true baseline. This is shown for the large peak of Figure 2c, and it is easy to imagine that the difficulty increases as the peak size decreases and baseline noise and drift increase. In the laboratory I was most recently involved with, we had a process we called "peer review," where after the analyst had finished integrating the chromatograms, they were reviewed by someone else before going to the quality group, at which point the data were "locked down" and integration changes required additional documentation. In this system, I reviewed thousands of chromatograms from hundreds of batches of samples, many of which were at or near the LLOQ. Of these, my guess is that less than a dozen did not require some manual integration. While the data systems today do a good job of integration, they are not perfect.  

To protect against frivolous reintegration or adjusting baselines so that the results meet a desired result, the FDA has a guideline commonly referred to as "21 CFR Part 11," "Part 11," or just "Electronic Signatures Rules" (1). This guideline states that if manual integration is performed, four criteria must be met: the original "raw" data must be preserved for later inspection, the person who made the change must be identified, the time and date of change must be noted, and the reason for the change must be recorded. Nearly every data system available today has an "audit trail" feature that complies with these regulations. A copy of the original data is archived. Because the user must sign on to the computer to use it, the username, date, and time are automatically stamped on each event. The software prompts for an explanation for each manual integration event. For example, the correction of the baseline in Figure 2c might be noted as "wrong baseline end." So the bottom line here is that the regulatory agencies anticipate that manual integration will be required and have set down guidelines to follow in such cases. If these guidelines are followed, you should not worry about negative regulatory action when you manually integrate a chromatogram to correct an error by the data system. 

In my experience, modern data systems do a marvelous job of integration in most cases. My recommendation is to allow the data system to integrate several typical samples and let it determine how to set the integration parameters. Then inspect the results and decide if they are satisfactory. If you encounter problems, such as those illustrated Figures 1 and 2, you may want to set the integration parameters to force particular integration techniques for certain peaks in the chromatogram. However, no matter how carefully you set up the data system, it is unlikely that it will pick all the peaks properly every time. You should visually inspect every chromatogram produced by your LC system to be sure it is integrated properly, and adjust the integration if necessary. 

Want more information on integration? Look in your data system manual or try reference 2. This book contains nearly 200 pages of information about integration — much more than most people ever want to know. However, my copy is loaded with sticky notes marking practical advice. One of my favorite statements in the book, very loosely quoted, is to the effect that chromatography always trumps integration. In other words, you can try all sorts of tricks to get a useable area from peak 1 of Figure 1 or the second peak of Figure 2b, but you will never get results as good as those obtained if you had a better LC separation. Integration of poorly resolved peaks is only an estimate of the more accurate results you would get when the peaks are baseline resolved. 

"LC Troubleshooting" Editor John W. Dolan is Vice-President of LC Resources, Walnut Creek, California; and a member of LCGC's editorial advisory board. Direct correspondence about this column to "LC Troubleshooting," LCGC, Woodbridge Corporate Plaza, 485 Route 1 South, Building F, First Floor, Iselin, NJ 08830, e-mail 

For an ongoing discussion of LC trouble-shooting with John Dolan and other chromatographers, visit the Chromatography Forum discussion group at 

(1) Guidance for Industry: Part 11, Electronic Records; Electronic Signatures – Scope and Application, USFDA-CDER, Aug. 2003, 

http://www.fda.gov/ora/compliance_ref/part11

(2) N. Dyson, Chromatographic Integration Methods, 2nd ed. (Royal Society of Chemistry, Letchworth, UK, 1998). 

LCGC North America-10-01-2009

The Third Annual LCGC Pittcon Awards: Establishing a Tradition

The Reality of Lab-on-a-Chip Technology for the Mass Spectrometry Laboratory

Review of Volatile Perfluorocarboxylic Acids as Ion Pair Reagents in LC: Part II

Adjustment to the Chromatographic System in U.S. Pharmacopeia Monographs: Flexibility and Limitations

Integration Problems