High-Res Mass Spec

When it comes to high-resolution mass spectrometry, Fourier transform ion cyclotron resonance used to be the only game in town. FTICR MS still has the highest resolving power of all MS techniques, but improvements in other mass analyzers have made them suitable for some applications that previously required the big guns of FTICR.

High resolution is particularly important as a means to obtain high mass accuracy and exact—as opposed to nominal—mass measurements. Resolution becomes especially important when overlapping sample components compromise the accuracy of mass measurements. “You need enough resolution to separate mixture components to enable getting high mass accuracy,” says Richard D. Smith of Pacific Northwest National Laboratory (PNNL).

This idea of needing “enough” resolution means that improvements in mass analyzers that have traditionally not been considered high resolution are giving users more ways to obtain high-resolution, accurate mass measurements. Perhaps just as important, these newer instruments allow researchers to perform qualitative identification and quantitative analysis on the same instrument. These advances are making high-resolution MS attractive and accessible to more users than before.

In addition to FTICR, researchers can now also choose between mass analyzers such as Thermo Fisher Scientific’s Orbitrap and time-of-flight (TOF) systems from several manufacturers. Picking the right mass analyzer for the job depends on what trade-offs are acceptable between resolving power and scan speed.

High resolution and the accompanying improved mass accuracy facilitate the analysis of complex mixtures in applications such as proteomics, metabolite identification, and petroleum research. In proteo­mics, the better mass accuracy made possible by high resolution allows researchers to be more confident about their protein identifications. In small-molecule applications, better resolution and improved mass accuracy can peg an identification to a single empirical formula instead of a range of possible formulas.

Resolution (Δm) is the mass difference needed to distinguish two peaks. As resolution improves, more components can be separated. However, when comparing the performance of different mass analyzers, people are more likely to talk in terms of resolving power, which is defined as m/Δm, where m is the mass of an ion and Δm is the resolution. Both resolution and resolving power change as a function of the mass-to-charge ratio, and the extent of the change depends on the mass analyzer. Among mass analyzers now considered to have high resolution, resolving power ranges from tens of thousands for TOF to millions for FTICR.

Resolving power is especially important when dealing with unknowns, says Alan G. Marshall, a chemistry professor at Florida State University and director of the ion cyclotron resonance (ICR) program at the National High Magnetic Field Laboratory. “You only know there’s more than one peak if you have enough resolution to tell them apart,” he says. For simplicity, most discussions of resolving power focus on distinguishing two peaks of equal height. If the peaks are not of equal height, then more resolving power is necessary, Marshall says.

Depending on the strength of the magnetic field used, FTICR, which uses superconducting magnets, can achieve resolving power of more than 1 million. But such levels can often be overkill. Thermo Fisher Scientific’s Orbitrap, another type of Fourier transform mass analyzer, can achieve resolving power of 100,000, and TOF mass analyzers from a number of vendors can attain resolving powers of 40,000-plus. This lower—but still considered high—resolving power is sufficient for many applications.

TOF works by accelerating ions to the same kinetic energy and then measuring the time they take to travel through a defined path length. From this information, the mass of each ion can be calculated.

The two FT techniques differ in how they achieve their resolving power. ICR measures the motion of ions rotating around a magnetic field, whereas the Orbitrap measures the oscillation of ions back and forth within an electrostatic field. In both cases, the frequency of motion depends on the mass and charge of the ion and the strength of the magnetic or electrostatic field. A Fourier transform is a mathematical operation used to convert those frequencies to a mass spectrum.

FTICR instruments rely on big magnets to obtain high resolution. The magnetic field strengths of commercial systems start at 7 tesla and currently reach as high as 18 tesla. Bruker, with its solariX line of instruments, is the main FTICR vendor. Thermo Fisher Scientific also sells some FTICR instruments, but the company steers most users to the Orbitrap family of mass spectrometers, which tend to be less expensive because they don’t use large superconducting magnets. Varian also sold FTICR spectrometers, but Agilent Technologies, which closed its acquisition of Varian late last month, has not yet decided the fate of the product line, according to Gustavo Salem, vice president and general manager of Agilent’s Biological Systems Division.

The availability of less expensive high-resolution instruments such as the Orbi­trap XL and TOF spectrometers has reduced the demand for FTICR, which was already a niche market. “A lot of people who were on the lower fringes of the FTICR market went to the Orbitrap,” says Darwin Asa, marketing manager at Bruker Daltonics. “That’s skewed our FTICR business more to the higher end.” Bruker offers 7- or 9-tesla magnets in its low-end FTICRs and 12-, 15-, or 18-tesla magnets in its high-end instruments, Asa says.

Despite the narrowing market for FTICR, Bruker continues to improve its systems. At the American Society for Mass Spectrometry meeting, which was held late last month in Salt Lake City, Bruker announced the addition of a MALDI ionization source to its solariX FTICR system. The matrix-assisted laser desorption/ionization source is aimed at imaging small molecules in tissue, particularly for pharmaceutical applications. “It’s the only system that can give you small-molecule imaging at therapeutic doses,” Asa says.

A drawback of both FT techniques—ICR and Orbitrap—is that high resolving power comes at the expense of scan speed. For example, the Orbitrap achieves a resolving power of 100,000 at one scan per second. Such scan speeds can’t keep up with the increasingly narrow peaks generated by today’s fast separations techniques. Doubling the scan speed to two scans per second cuts the resolving power in half.

For many users, as long as the resolution is above a certain threshold, the scan speed takes priority. Doubling or tripling the scan speed while maintaining current resolving power would be more beneficial to users than would an equivalent improvement in resolving power, says Ian D. Jardine, vice president of global R&D at Thermo Fisher Scientific.

The Orbitrap is making its mark in biological applications, Jardine says. He used to focus presentations on the Orbitrap technology itself, but he now emphasizes the research employing the Orbitrap that is published in journals such as Science, Nature, and Cell. “Mass spectrometry is moving cellular biology forward incredibly rapidly,” Jardine says.

One application in which high resolution is particularly beneficial is “top down” proteomics, in which intact proteins are fragmented in the mass spectrometer. For such experiments, “you want high resolution, if you can get it,” says Neil L. Kelleher of Northwestern University. “If you want the maximal level of information, 50,000 resolving power is just barely enough to resolve isotopes at 50 kilodaltons,” he says. He uses the term “precision proteomics” to describe proteomics experiments with high levels of resolving power and mass accuracy that provide “hyperconfident” identifications.

For top-down proteomics, Kelleher currently uses high-resolution tandem MS systems, with ICR or ion traps for the first stage and either ICR or Orbitrap to record fragment ion spectra. But for top-down proteomics to catch on more broadly, it will have to be possible to carry it out in a “plug and play” manner on less expensive systems, he says. Kelleher says he would like to try to use cheaper “benchtop” Orbitrap systems to simplify and lower the cost of top-down proteomics experiments.

Kelleher also advocates precision proteomics for “bottom up” proteomics, a more common type of experiment in which proteins have been digested first. Given the relatively small masses of peptides produced by protein digestion, high precision can be achieved with less resolving power than is required for top-down proteomics. “For bottom-up, 100,000 is overkill,” he says. With 50,000 resolving power, users can confidently identify unmodified peptides, he says.

John R. Yates III, who studies proteo­mics at Scripps Research Institute in La Jolla, Calif., finds that the accuracy of peptide identification in bottom-up proteomics is not dramatically improved by obtaining high-resolution data. The mass assignments do become more accurate with higher resolution, but “there’s probably a limit to how much mass accuracy is usable,” he says. Yates and coworkers run their proteomics experiments on an Orbitrap backed off to 60,000 resolving power. If that were insufficient for his analyses, “we would probably bite the bullet and scan slower,” he says.

Even longtime FTICR users are shifting to other high-resolution mass analyzers for some applications. In fact, Florida State’s Marshall—perhaps the most vocal advocate for FTICR—has found applications for which he needed to use other techniques. For example, when he needed to analyze singly charged ions with a mass-to-charge ratio too high for FTICR, Marshall turned to TOF, which is faster, more sensitive, and has a wider mass range than either of the FT methods, he acknowledges.

Smith, another longtime FTICR user, has recently been relying increasingly on TOF. His lab has been pairing it with fast multiplexed ion-mobility separations, a combination he sees becoming his mass analyzer of choice for most applications. As his research has shifted more toward biological questions, he has realized that he doesn’t always need the highest mass-resolving power. For bottom-up proteomics analysis of peptides with chromatographic and ion-mobility separations in front of the mass spectrometer, “the number of cases where you’d have mixtures of peptides that you can’t resolve with 50,000 resolving power is extremely small,” Smith says. Although his lab still uses FTICR for some experiments, he expects the new high-resolution platforms to become dominant for broad applications.

The real advantage of TOF relative to the FT methods is speed, especially with the rise of fast chromatographic techniques. “People don’t buy mass spectrometers and just inject samples,” says Lester Taylor, LC/MS platforms and programs manager at Agilent. “They put something in front of it.” Fast chromatography requires that the mass spectrometer scan quickly to collect enough points to define each chromatographic peak. “I’ve been in situations where you have to relax the chromatography constraints to make your peak wide enough to sample sufficiently” by MS, Taylor says.

A major application area for high-resolution TOF MS is metabolite identification. “A lot of drug metabolism people are heavily committed to UHPLC,” Asa says, referring to ultra-high-pressure liquid chromatography. The speed of TOF lets researchers get high throughput from their UHPLC systems, and the resolution and mass accuracy enable them to confidently identify unknown metabolites.

In recent years, mass spec vendors have greatly improved the resolution of TOF mass analyzers. Bruker’s maXis achieves a resolving power of 50,000, and instruments from AB Sciex, Waters, and Agilent achieve values of more than 40,000. JEOL’s new SpiralTOF, which has an unusual ion trajectory, obtains 60,000 resolving power. In each case, the level of resolving power has been reached by a combination of system geometry and electronics.

“There’s no one magic modification that makes it happen,” Asa says. “It has a lot to do with focusing the ions, cooling the ions, and making sure that we’re not losing the signal, resolution, and accuracy as those ions are going through the system and making it to the detector.”

Vendors have also focused on developing instruments that can allow users to perform high-resolution qualitative identification and sensitive quantification on the same instrument. “Until now, customers have done their high-resolution, accurate-mass work on one dedicated platform and then moved to a triple-quadrupole platform for quantitation,” says Dominic Gostick, director of biomarker MS, pharmaceutical, and proteomics businesses at AB Sciex.

At the American Society for Mass Spectrometry meeting, AB Sciex launched its new TripleTOF 5600, which achieves more than 40,000 resolving power across the mass range without loss of sensitivity or scan speed, the company says. “Our customers told us they needed 30,000 resolving power for most of their applications,” Gostick says. “We wanted to develop an instrument that maintained high resolution and mass accuracy while delivering the speed—up to 100 scans per second—and sensitivity of a high-performance triple quad for robust quantitative analysis,” he says.

Waters’ entry into the quantitative, high-resolution TOF market, the Synapt G2, was introduced last year. The high resolution—routinely accessible at UHPLC-compatible speeds—is achieved by a combination of flight-tube geometry and detector electronics, according to Ronan O’Malley, group manager for TOF product management at Waters.

But the Synapt adds another type of resolution with the inclusion of an ion-mobility cell, in which ions can be separated on the basis of size and shape, as well as mass-to-charge ratio. “The second-generation Synapt’s detector electronics provide enhanced dynamic range and allow for exact mass measurement of components separated by the high-resolution ion-mobility cell, enabling quantitative analysis of previously unresolvable isobaric species,” O’Malley says. Isobaric species are chemically distinct entities that have the same nominal mass.

Another important factor for resolution and sensitivity is the capability to shepherd ions produced by an ionization source so they reach the mass analyzer and detector. For example, Agilent uses ion-beam compression technology in its 6540 and 6538 Q-TOF systems. The technology shapes the ion beam, “making sure that it’s spatially compact and energetically homogeneous, so you haven’t got an energy spread and spatial spread that degrade resolution,” Taylor says.

JEOL’s new SpiralTOF, which has a staggered figure-eight design, takes a new multipass approach to TOF. A danger with multipass systems, in which ions retrace the same path multiple times, is that the heavy ions can eventually catch up with the light ions and become indistinguishable. With the SpiralTOF, each trip around the figure eight is slightly offset, so the ions don’t catch up with each other, says Robert B. Cody, product manager at JEOL. “That gives us a 17-meter flight path in a 1.3-meter package,” he says.

Efforts are also under way to push the resolving power of TOF beyond its current limits. Virgin Instruments, based in Sudbury, Mass., is developing an instrument with 200,000 resolving power, says Marvin L. Vestal, chief executive officer. The instrument is about 20 feet high and gets a two-story installation, with the ion source on the first floor, the ion mirror on the second floor, and a hole in the floor between. “Our major limitation is the mechanical vibration of the second floor,” Vestal says. “If you walked across the room, you could see it on the mass spectrometer.” Such unforeseen complications have kept them from hitting the predicted 200,000 resolving power, Vestal says.

One reason for constructing such an instrument is “just to establish that we know what all the limitations really are,” Vestal says. Over the past several years, he has developed a detailed theory of what is feasible with current TOF technology. He finds that resolving power should increase in direct proportion to the flight path. “If a 20-foot tube gives us 200,000 resolving power, then a 100-foot one should give us a million,” he says. Vestal doesn’t know whether Virgin will ever actually build such an instrument—even the more modest 20-foot one—commercially.

Even as TOFs continue to improve, there will always be a place for instruments with the highest levels of resolving power. Thus, advances continue in FTICR as well.

For example, the National High Magnetic Field Laboratory and the Environmental Molecular Sciences Laboratory at PNNL are collaborating on the development of new 21-tesla FTICR mass spectrometers. Projects and funding exist at both laboratories, and exploratory discussions are ongoing with potential magnet manufacturers. Such magnets would require two to three years to build, given the long lead times required for obtaining the necessary specialized superconducting wire materials.

This type of instrument is “not going to be common,” Smith says. There’s going to be only a couple of them in the U.S., he adds. “I don’t think it will ever be broadly marketed and used.” Instead, he says, the appropriate place for such instruments is a national laboratory, where users can bring samples when they need extreme resolving power, or in places where a critical mass of applications can justify the cost.

Smith already knows how he hopes to use the instrument in proteomics studies. “The very-high-end platforms will be used for initial identification, particularly in top-down proteomics applications,” he says. “Once we’ve identified the parent protein, we don’t need to dissociate it and identify it every time we make a measurement. We can establish what we call an accurate mass and time tag, and then use that information with an ion-mobility TOF platform to make routine, high-throughput measurements. With the information provided by the high-end FTICR, we can then do this much faster, with greater sensitivity, and on a much less expensive platform.”

The main focus for the new instrument at the National High Magnetic Field Laboratory will be identification of protein modifications by top-down proteomics. “You never can figure out what’s there unless you look at the whole protein to start with and then chop it up,” Marshall says. “The difficulty is that proteins are big, and it’s hard to do that.” The average mass of a human protein is about 50 kilodaltons, Marshall notes. With current instruments, Marshall can capture only about half of the proteome in a single spectrum. “At 21 tesla, we should be able to get most of the rest of the way,” he says.

Marshall also plans to apply the instrument to questions in petroleum research. His team can identify the components of petroleum up to about 800 or 900 Da, but the spectrum actually extends out to about 1,300 or 1,400 Da. Again, for this application, he expects the 21-tesla instrument to enable him to resolve all of the components.

The current crop of instruments gives users a variety of options for high-resolution MS. What is certain is that in the future mass spec will offer scientists even more choices.