Commercialization of LC–MS: The Second Decade

In 1997, in an article that appeared in what then was LCGC International (now LCGC Europe), I departed from my usual technical writing, which up to then had been primarily for peer review. The LCGC editor had requested help recognizing the publication's 10th anniversary. Mass spectrometry (MS) and liquid chromatography–mass spectrometry (LC–MS) in particular, given the rapidly evolving commercial landscape, I thought made investigating the phenomenon worthwhile. Little did I realize how much effort a writer must devote to developing and presenting ideas in finished form. I must have forgot that lesson when, five years later, I agreed to write the "MS — The Practical Art" column every other month.

Michael P. Balogh

Nearly 35 bimonthly columns later, I happened across that initial article, "Commercialization of LC–MS 1987–1997: A Successful Decade in Review" (LCGC International, 10 (11) 728–737 [1997]). It made me curious about just how many technological advances the field of MS had witnessed during the ensuing decade and what had not changed.

Of course, the technological basis, or history, remains unchanged for what was then called the "hyphenated" online practice of LC–MS. What has changed, though, is this: as an analytical method, LC–MS has become so ubiquitous in industry and research that more practitioners now approach it as a single, integrated analytical process. Based upon the rapid growth of the American Society of Mass Spectrometry (ASMS) indicated by the annual conference attendance in recent years, the cross-section of who constitutes today's practitioners has broadened as well. We have abandoned the sense of technological duality that the hyphenated usage implied. My 1997 article attempted to outline evolution of the mass spectrometer as it was becoming a true mass detector "mixing significant technological events with significant commercialization initiatives over [those] 10 years." It fleetingly acknowledged important events in the development of LC–MS to that point, such as Blakely and Vestal's 1983 publication on thermospray which, though no longer used, was the first LC–MS enabling technology (1). John Fenn (along with Koichi Tanaka and Kurt Wüthrich) received the Nobel Prize in Chemistry, in 2002, for his work on the most significant aspect of LC–MS practice — electrospray ionization (ESI) — based upon work published in the early 1980s (2,3). Among those receiving credit for commercializing the concept were Bruins and Covey who, along with Henion, published commercially promising ESI designs in 1987 (4).

A harbinger of things to come, the first truly hyphenated system appeared in 1982. Hewlett Packard's mass spectrometry detector (MSD) used a gas chromatograph to introduce analytes into the mass spectrometer's ion source. Of course, the gas chromatograph limited the MSD to analyzing only compounds of sufficient vapor pressure. However, no such restriction attended coupling a high performance liquid chromatography (HPLC) system to the mass spectrometer. LC–MS had its own problems: vacuum system and flow-rate requirements evolved large, costly, often unreliable instruments.

In 1993, Waters (Milford, Massachusetts) introduced a benchtop instrument for LC–MS. Designed to operate using particle beam technology, which in the era of ESI exhibited certain performance issues, it nonetheless deserves footnote-level respect for giving us true electron ionization (EI) spectra for an LC separation. That said, the first LC–MS designed for widespread adoption is probably Vestec's thermospray-equipped, modified HP MSD, which the company offered at the 1986 Pittsburgh Conference and Exposition (Pittcon) for the almost magical price of $100,000. Thermospray of course faded quickly, and so did the dream of vastly capable but inexpensive LC–MS.

As is so often the case, the optimism of commercial interests proved unwarranted. In 1988, a year after Extrel Corporation had introduced particle beam technology at Pittcon; Hewlett Packard (now Agilent Technologies, Santa Clara, California) introduced a version derived from the original particle beam instrument. The HP product manager claimed the interfacing of LC and MS brought to the LC–MS market the reproducibility of results and ease of use characteristic of gas chromatography (GC)–MS. In those days, industry watchers estimated market sales to be $15 to $20 million, with thermospray instruments accounting for 60–80% of the business (5). Always a point of industry contention, and one whose public discussion typically is avoided for reasons of competition, Finnigan (now Thermo Fisher Scientific, Waltham, Massachusetts) management succinctly stated at Pittcon 1993 that estimates of market size were almost "guaranteed" to be incorrect and were "probably rather conservative." Indeed, the total market for 1993 was predicted at $640 million. In the years immediately following, it became obvious that both sides of the argument were incorrect.

The early 1990s fell far short of sales expectations, the consequence of a general economic slowdown. (Remember, these were the days when costly analytical purchases were driven largely by a rapidly building pharmaceutical industry.) A 1996 American Chemical Society report cited 1991 sales exceeding $450 million, but with the caveat that "growth in these areas is difficult to predict."

ASMS as Litmus Test

Between 1985 and 2004, it was my good fortune to co-organize the LC–MS interest workshop at the annual conference of ASMS (www.asms.org). During that time, the success of LC–MS transformed ASMS. Once an annual event for thought leaders and innovators in MS research, the conference also came to represent an educational resource for nonspecialists. In 1987, 2200 members attended. By 1997, that number had doubled. In keeping with the growing interest, attendance at the LC–MS interest workshop grew, by some estimates to more than 1200 from a few hundred.

I began taking note of changes in abstract topics matched to overall attendance during a stint on the program committee in 2003–2005, (and in the years since — the conference administrator, Brent Watson, kindly obliges with statistics each year). The ASMS conference has grown in the early years following 2000 to nearly 6000 attendees. Once the LC–MS interest wave crested, and LC–MS became the ubiquitous tool it is today, the conference organization reflected that interest. In these much larger conferences, LC–MS is no longer the esoteric focus of a special interest group. Indeed, the technology now appears as a topic or reference in virtually every discussion. As for the LC–MS interest group, it has reverted to a more reasonable size, returning to a more utilitarian, narrow charter.

In the early years, presentations rarely associated LC columns and separations with LC–MS practice. LC served merely as an "inlet" for compounds dissolved in the condensed phase, and it was the mass spectrometer that bore the brunt of separations work. A survey in 1985 noted that of 400 chemists interviewed, less than one third were "comfortable with prospective applications and operation of LC–MS" (6).

Figure 1

A figure I excerpted from a 1993 article (7) showing comparative analyte ranges for the major LC–MS interfaces appears quite dated today: particle beam and thermospray, two of the four noted along with atmospheric pressure chemical ionization (APCI) and ESI, pretty much vanished years ago. On the other hand, I noted in my article emerging technologies just gaining ground:

"Significant recent introductions, with much shorter commercialization histories (as of this date), such as the reengineered ion-trap (commercialized initially by Finnigan, now Thermo Fisher Scientific) and time-of-flight (TOF) instruments also are worth a passing note for their potential role in shaping the future direction for LC–MS detection."

It was obvious by the time the article appeared that a few years earlier, "The interface with the greatest potential impact was . . . previewed as a commercial possibility in 1983 (8), when preformed ions in solution were described . . . . The advantage of electrospray, readily embraced by protein chemists, was the ability to charge multiple sites on . . . relatively large molecular weights . . . well above the operating range of the quadrupole."

Looking back over the 16 years since the article was published, it is obvious that quadrupoles still serve as the analytical backbone of LC–MS practice. Evolution of the technology brought ion traps to prominence from their attendant problems of space-charge effects (too many ions in the trap) and quantitation issues to much-improved, linear traps coupled with ion cyclotron devices such as the Orbitrap (Thermo Fisher Scientific) or as an adjunct to tandem designs (MDS Analytical Technologies Sciex, Toronto, Ontario, Canada).

A Few Statistics

Consider that in recent years, the ASMS conference attracts about 6000 attendees who submit almost 3000 abstracts for talks and posters (low of 2730, in 2009, and recent high of 2907, in 2008). Granted, we must bear in mind the content in recent years has become heavily weighted away from environmental and industrial applications toward life science (searching for individual instances where an abstract contains the term "peptide" for instance returns 1023 hits out of 2730 abstracts), but the consistency in attendance numbers year-to-year can reveal some shifts in technology adoption trends. For instance, "ion trap" and "quadrupole" are too common (500 or more instances for each across numerous applications). Moreover, the ASMS program committee rules out trade names as part of the acceptance process. Yet some such terms are unavoidable and unique. "Orbitrap," though a trade name, is an example of commercialized novel technology. Since 2006, Orbitrap cites per abstract rose from 39 at its introduction to ASMS to 429 in 2009. Similarly, we can judge unique entities such as SYNAPT (Waters Corporation,) and the ability to combine ion mobility as an orthogonal discriminator (in essence multiplying the resolution or peak capacity of a mass spectrometer), which shows a tripling of its unique instances since 2007.

The ASMS conference is, of course, an MS venue. With that in mind cites for "UPLC" (Waters' ultra high performance LC system) more than tripled, to 142, since 2006, which is impressive because LC is simply an "inlet" device to a mass spectrometer even today (by comparison, "HPLC," long-established as the inlet technology, returns 540 instances).

Current State of the Art

From the highest-level view, major corporate acquisitions and mergers seem to have slowed in recent years, and the rapid emergence of revolutionary technologies has morphed, with a few exceptions, into slower, evolutionary change. Agilent Technologies, with quadrupole and TOF technology, has joined the top tier and continues to promote chip-introduction schemes. Waters Corporation has followed suit with the advent of its TRIZAIC UPLC system. Waters' addition of ion mobility to its MS products provides the means to directly control the sample introduced into the mass spectrometer in an orthogonal fashion, layering a variety of ion manipulation techniques with differentiation of molecules by conformational cross-section. Thermo Scientific has steadily improved the speed of acquisition and resolution of its matched devices, the ion trap and the Orbitrap. The company's recently introduced dual-chamber ion-trap system, the LTQ Velos, begins to address some of the impediments inherent in ion-trap technology like space-charge and quantitation limitations.

Applied Biosystems today no longer exists as a corporation, surviving as a brand in the market as part of Life Technologies, a biotechnology manufacturer, after merging with Invitrogen (San Diego, California). The company maintains its joint venture with MDS Analytical Technologies Sciex, which continues to improve the top-end, tandem quadrupole series.

Based upon the same slanted attendance profile, we can with some reservation look at some year-to-year trends:

• A data base query for ion suppression returns 69 instances representing unique abstracts, which seems typical (76 in 2008; 64 in 2007; and 87 in 2006). A disproportionately large amount of our analytical effort is based upon condensed phase practice, which is amenable to ESI. But ESI is prone to ion suppression and constantly requires better ways of remediation.

• Metabolomics at 139 hits shows a significant increase. In 2008, 80 was an increase from 46 in 2007 (39 in 2006, 54 in 2005).

• Biomarkers remains fairly constant with 355 instances versus 400 in 2008 and 350 in 2007.

• Informatics as a general term returns only 10 hits, while bioinformatics returns 154, down from 239 in 2008.

Predicting the Future

Making predictions is a tempting but risky exercise when you are committing to print. Yet in the 1997 article, I did venture to state that the "halcyon days of interface development were over," which appears to be true. A recent study report by Strategic Directions International, Inc., a market research firm that tracks instrument business trends indicates that, as of 2008, four manufacturers (Agilent, Applied Biosystems, Thermo Fisher Scientific, and Waters Corporation) comprised 70% of the mass spectrometer market, with between 13% and 25% shares each.

According to the report:

"Despite being well over a $2 billion market, demand for MS is expected to grow significantly, if not stronger than in recent years. This is despite current weakness in major economies the world over, as well as some pull back in the pharmaceutical industry. The fact that MS is such an important tool in high-end basic research in so many leading-edge industries will help to increase demand at a 9% annual rate through 2012. The market will be led for the foreseeable future by the more advanced methods, including Fourier transform (FT)-MS, tandem LC–MS, and quadrupole time-of-flight QTOF LC–MS (9)."

The same report sees "typical" prospects for GC–MS settling for 5% growth, with half the business resulting from replacement sales of an installed base said to be reaching 30,000 units. Improvements in GC-TOF and, to a lesser degree, GC-GC tandem account for increased sales.

Often overlooked, even in these instances, is the central role single quadrupole instruments still play. A relatively large, installed base (7000 is estimated in this report), has Waters and Agilent competing closely, with equal shares dominating 90% of this market segment.

QTOF instruments are reported to be at 4000 installed, proving, as I predicted in a 2004 column, that complexity coupled with improvements made this hybrid technology accessible for a variety of quantitative and qualitative applications (10).

Alan Millar (Senior Manager for QTOF Product Development, Waters) recently compiled an historical overview of the efforts of only a single company during the past decade. Eight nominal mass single-quadrupole instruments have been brought to market in the past 16 years, beginning with the Platform in 1993. A total of 11 triple- or tandem-quadrupole instruments appeared between 1989 and today, beginning with the Quattro model. Hybrid-TOF instruments account for eight unique product developments in 13 years, beginning with the original QTOF with 5000 resolution (FWHM) in 1996 to today's recent G2 introduction capable of 50,000 resolution. Adding matrix-assisted laser desorption ionization (MALDI)- and TOF-only designs would easily more than double that number in a similar time span.

Prices continue to be influenced by the cost of vacuum systems and control capabilities. Except for GC–MS, do not expect to see a high-end MS system for less than $100,000. A counter-argument can be made that by any standard, single-quadrupole LC–MS instruments, which do sell typically for much less than $100,000, offer state-of-the-art quadrupole technology today, lacking only the extensive additional capabilities of the more complex mass spectrometer. Median, or average, prices often appear in descriptions, except in the case of high-volume instruments (single-quadrupole and ion-trap instruments) and even those are often bundled in special arrangements. But without stating detailed specifications, you will find that the prices of higher-performing instruments overlap.

Reiterative Nature of Technology

As we have seen, technology can have a very brief half-life. Some make a great appearance only to fail to thrive after three or five years. Some, on the other hand, need enabling technologies to bolster their utility — TOF waited 60 years before fast electronics, software, and much improved engineering design made them the practical analytical tool they are today. Software became more capable and hardware clearly improved in recent years, and an improved understanding of fundamentals often allows us to build on such improvements.

A couple of recent introductions illustrate how sound fundamentals can reappear to our advantage. Both Bruker (Billerica, Massachusetts) and Waters have recently commercialized an atmospheric pressure GC (APGC) adaptation for their LC–ESI-MS systems. The Waters license derives from DuPont and the work of Charles McEwen (currently Houghton professor of Chemistry and Biochemistry, University of the Sciences, Philadelphia, Pennsylvania). McEwen is perhaps better known these days for his work developing yet another classic fundamental, the atmospheric solids analysis probe (ASAP). The ASAP technology allows an operator to simply place a solid or liquid sample on the end of a tube, insert the tube into a stream of heated nitrogen, where the sample is volatilized, and look at the resulting ions from the mixture. Direct links to Horning's work in the 1970s are readily apparent (11,12).

Debate over the exact nature and mechanics of ionization might still persist as it has with ESI while we made good use of it. But in the meantime, extremely fast chromatography demanded faster MS acquisition speeds with good fidelity. Improved software allows us to take advantage of our understanding of isotope characterization and our predictive capability. Better engineered, robust MS platforms displaying our improved understanding of gas dynamics yield sources capable of greater diversity.

For instance, an aspect of the APGC Xevo design (Waters Corporation) puts a "mini-chamber" in place between the ESI inlet in the MS source and the GC transfer line as an integral part of the changeover from LC to GC. The entire procedure takes only 5 min. When the idea of LC and GC in the same instrument was proposed not too many years ago as a viable return-on-investment argument, it failed because substandard GC performance at any cost is unacceptable. Today we see low-femtogram sensitivity in what is ostensibly an LC source design. Add recent developments in atmospheric pressure photoionization (APPI), which show the potential to deliver EI-like spectra, and you realize that the same level of training and the same software and experimental design we've been accustomed to applying for LC–ESI-MS now covers a much more diverse chemical landscape.

Michael P. Balogh "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.

References

(1) C.R. Blakley and M.L. Vestal ML, Anal. Chem. 55(4), 750–754 (1983).

(2) M. Yamashita and J.B. Fenn, J. Phys. Chem. 88, 4451–4459 (1984).

(3) M. Yamashita and J.B. Fenn, J. Phys. Chem. 88, 4671–4675 (1984)

(4) A.P. Bruins, T.R. Covey, and J.D. Henion, Anal. Chem. 59, 2642–2646 (1987).

(5) Analytical Instrument Industry Report, (East Grinstead, UK) 5 (7) July 1988.

(6) SDI Business Outlook, Strategic Developments, Inc. (Los Angeles, California) 4(13) October 1995.

(7) C.R. Creaser and J.W. Stygall, Analyst 118, 1467–1480 (1993).

(8) J.V. Iribarne, P.J. Dziedzic, and B.A. Thomson, Int. J. Mass Spec. Ion Phys. 50, 331 (1983).

(9) Strategic Directions International, Inc., Global Assessment Report, 10th edition. Market Forecast 2007–2012 published September 2008.

(10) M.P. Balogh, LCGC 22(2), 118–130 (2004).

(11) E.C. Horning, M.G. Horning, D.I. Carroll, I. Dzidic, and R.N. Stillwell, Anal Chem. 45, 936-943 (1973).

(12) D.I. Carroll, I. Dzidic, R.N. Stillwell, K.D. Haegele, and E.C. Horning, Anal. Chem. 47, 2369–2373 (1975).