Growth of Mass Spec From 1960s to Present: Interview With Professor O. David Sparkman

Professor O. David Sparkman of University of the Pacific in Stockton, CA, is a recognized authority on mass spectrometry. His career spans the exponential growth phase of MS starting in the late 1960s, which encompasses the first 50 years of American Laboratory.

RLS: I recall hearing that MS was first reduced to practice about a hundred years ago. Can you describe the early years?

ODS: Mass spectrometry was developed as an outgrowth of J.J. Thomson’s discovery of the electron ca. 1898. Just prior to WWI, physicist F.W. Aston, a protégé of Thomson, interested in measuring the stable isotopes of the elements, developed the actual first working instrument. Aston used MS in identifying 212 of the 281 naturally occurring nuclides known at the time. It provided accurate mass and isotopic abundance that helped lead to more accurate atomic weights and a new standard for mass, carbon-12.

RLS: MS was used in petroleum firms such as Shell during WWII. Was this for petroleum characterization or something else?

ODS: Characterization of petroleum was indeed a major application due to the fact that petroleum is a complex mixture, often with nitrogen, oxygen, and sulfur heteroatoms. The composition also varied by sample location. However, the major success during WWII was development of synthetic rubbers. Recall that Japan had invaded Malaysia, which was the Allies’ source of natural rubber.

The instruments used ~70 eV electrons emitted from a hot filament to ionize the analytes in gas phase. The mass analyzers were strong electromagnets. The high vacuum required was provided by oil and mercury diffusion pumps, which were slow and messy. Instruments required lots of maintenance.

Another major role played by mass spectrometry involved producing the pure radioactive material used in the development of the first atomic weapons.

RLS: What about commercial development of MS?

ODS: The first developments were in the U.S. by Consolidated Engineering Corporation (CEC) and Westinghouse. These instruments were used by several different petroleum companies during WWII. After WWII, commercial mass spectrometers came from Germany (MAT), the U.K. (AEI and VG), Japan (JEOL and Hitachi), and France (Thomson-CHF).

Mass spectrometry was well-involved with organic chemistry by the mid- to late 1950s, when gas chromatography first came on the scene. In the late 1950s, chemists R.S. Gohke and F.W. McLafferty at Dow Chemical interfaced the newly developed gas chromatograph with a Bendix time-of-flight (TOF) mass analyzer. Both the GC and the MS used gas phase analytes. The mass spectrometer gave the best results when the analyte was a pure compound and the GC separated mixtures into pure compounds. The gas phase analytes inside the MS are ideally suited for electron ionization (EI). The MS was the ultimate detector for gas chromatography because the data could be both qualitative and quantitative.

Chemists J.C. Holmes and F.A. Morell at Philip Morris were challenged by characterizing tobacco products, including combustion products. The goal was to show that inhalation of tobacco combustion products was benign. They favored the magnetic sector mass analyzer from CEC and are also credited with the beginning of GC-MS.

RLS: When I was in grad school from 1963 to 1966, organic chemists were just starting to use MS. However, MS was not widely used. What changed?

ODS: Those mass spectrometers used either velocity focusing of ions with magnetic or a combination velocity and energy (with electric fields) focusing for separation of ions based on their mass-to-charge ratio (m/z) values. This is the mass of an ion divided by the number of charge states. Most of the ions in organic mass spectrometry were single charge; therefore, the m/z value and the mass were considered to be the same. Magnet and vacuum technology were in their infancy. Getting data out of these instruments was far from easy. The instruments had to be mounted on stable floors, which put them in the basement of most buildings. They also required dimly lit rooms to prevent photobleaching of the photosensitive recording paper.

The concept of GC-MS was demonstrated by both the Dow and Philip Morris groups, but the technology was still problematic. The major challenge was that gas chromatography required above-atmospheric pressure gas in contrast to the MS, which required a very high vacuum. There were several devices, such as the Llewellyn membrane separator, the Watson-Biemann efusion separator, and the jet separator from Finland, which removed most of the GC carrier gas from the analytes before the analytes entered the MS. These helped, but the MS instruments at the time required very high vacuum.

The synergy of the GC and MS improved over the 1960s, leading to commercial GC-MS systems from many of the MS vendors referenced above. Perkin-Elmer was particularly successful by combining its GC with the Hitachi RMU mass spectrometer. During this period, in-line coupling separations, including sample prep to other techniques such as various spectrometry techniques, was a fad called “hyphenation” as in GC-MS or LC-IR spectroscopy. GC-MS was the most successful example of hyphenation in the 1970s.

RLS: Ah, the 1970s—wasn’t that the time that computers started to control instruments, gather and process data, and do semiautomatic preparation of reports?

ODS: Yes, the computer revolution with rack mounts of PDP-8s, etc., was important because the data from a GC-MS analysis was so voluminous that the analyst could easily be overwhelmed. Finnegan Instruments introduced the quadrupole (known today as the transmission quadrupole) mass analyzer that had a large dynamic range. Because it was smaller and required less extreme vacuum, this MS was much less expensive to purchase and easier to operate. However, its most outstanding feature was the minicomputer data system (developed at Stanford University) that allowed for both instrument control and data analysis. This was the first GC-MS sold with a computer-implemented control and data analysis system. The nascent U.S. EPA was so impressed that they purchased 10 GC-MS systems. These were placed in various locations around the U.S. to monitor drinking water for halogenated organic compounds. This single event propelled Finnegan to supernova status in the constellation of analytical instruments.

RLS: In the late 1960s, Varian and Hewlett-Packard were both located on adjacent reservations in Palo Alto, CA. Varian had a position in NMR, which was internally developed. H-P also saw opportunities in analytical instruments, particularly GC. Varian reached across San Francisco Bay to acquire Wilkins Aerograph in Walnut Creek. H-P acquired F&M Scientific in Maryland. The latter made rack-mounted GCs that resembled refrigerators in size. Aerograph’s were the size of today’s countertop microwave ovens. How did this play out?

ODS: Finnegan was located in San Jose, CA, a few miles from Palo Alto and 30 miles from Aerograph in Walnut Creek. T.Z. Chu, the charismatic leader of Varian-Aerograph, had helped Robert Finnegan start up Finnegan Instruments. Eventually, T.Z. joined Finnegan as president. This led to a huge brain drain from Varian-Aerograph to Finnegan. When asked about this, T.Z. explained that he could teach almost anyone about MS, but chromatography is a much more complex art. He said that most of Finnegan’s service problems in GC-MS were with separations, not MS. Another interesting data point: In about 80% of Finnegan’s service calls, the technician did not even use a screwdriver (the problem was a training or experience issue and not a malfunction that required removing a panel, module, or printed circuit board).

RLS: What will the 1980s be remembered for?

ODS: A great deal happened in the field of mass spectrometry in the late 1970s through the 1980s. This was the era of advances in “ion optics” (optical spectroscopists explained advances in optical spectrometers with displays of the photon path, so mass spectrometrists described the flow of ions using analogous terms, with “ions” replacing “photons”).

The first notable event was the development of the tandem quadrupole analyzer. This employs a pair of transmission quadrupoles separated by a collision cell. The first quadrupole is used to select ions of specific m/z values. These ions are passed to the collision cell, which induces fragmentation of the ions. The second quadrupole is then used for the m/z analyses of the fragment ions. The collisional fragmentation process has advantages in GC-MS; however, it is most advantageous in LC-MS, where the initial ionization process is less energetic (“softer”), which reduces fragmentation, yielding ions that are representative of the intact molecule. MS-MS facilitates structural elucidation of large molecules. Other mass analyzer types are used in MS-MS, (e.g., tandem quadrupole-TOF). MS-MS always means tandem m/z analyzers (usually tandem quadrupoles or tandem quadrupole/TOF instruments).

Another development of this era was the 3-D quadrupole ion trap. The quadrupole technology for manipulating ions was developed by Wolfgang Paul in Germany in the mid-1950s. Paul later shared a Physics Nobel Prize for the quadrupole ion trap part of this technology. However, the ion trap was not commercialized during the life of Paul’s patents. George Stafford of Finnegan designed a series of mass spectrometers utilizing 3-D quadrupole ion traps (QIT) for mass analyzers.

Not only could the 3-D QIT be used as a normal EI mass spectrometer, it allowed for another new concept called tandem-in-time mass spectrometry. Ions were either injected into or formed in the trap. The trap was tuned electronically to select and hold ions with a particular m/z value. The selected ions were next fragmented via collisional activation. The m/z values of the resulting fragments were then recorded.

The reflectron TOF mass analyzer was another significant development of this decade. Prior to reflectron technology, TOF mass analyzers used a single path from the ion source to the detector. Resolution was limited by the energy dispersion of the ions in the source. Higher-energy ions of a particular m/z value traveled faster and reached the detector before those with lower energy; however, if the detector is replaced with a strong reflecting electric field,the fast and slow ions can be refocused by the reflectron field and onto the detector located near the original ion source. The refocusing produces a very sharp signal compared to a unidirectional TOF. The speed and resolution are phenomenal, comparable to the best of the double-focusing (much more expensive) spectrometers. Detection limits were reduced because, unlike instruments that obtain a spectrum by scanning a range of m/z values, the TOF detects all ions of all m/z values formed during each cycle.

As important as the advances in ion optics were, introductions in ionization technology were much, much more significant. Prior to the mid-1980s, mass spectrometers had been limited to analytes that could be ionized in the gas phase. If the analyte was thermally labile and/or not volatile, mass spectrometry studies required formation of a nonthermally labile and volatile derivative. This was a frustrating restriction.

John Fenn developed a technique called electrospray ionization (ESI), for which he shared in the 2002 Nobel Prize in Chemistry. ESI allowed nonvolatile thermally labile compounds to be made ionic in the condensed phase and then desorbed into the gas phase for m/z analysis as they eluted from an LC column. Not only did ESI expand the range of compounds compatible with quadrupole and magnetic sector mass spectrometers, but the limits had been set by the maximum strength possible for the magnet in sector-based instruments and the maximum amplitude that could be achieved by the fixed frequency RF in quadrupoles. Provided that sufficient sites could be charged on an analyte, multiple charge analytes could be analyzed. It is important to remember that the mass-to-charge ratio of an ion is measured in mass spectrometry, not the mass of the ion. This means that a protein that has a mass of 50 kDa and 60 protonatable sites would be observed as an ion of ~m/z 833 (50,000/60) when analyzed by ESI mass spectrometry.

The other significant ionization technique of this era was MALDI (matrix-assisted laser desorption/ionization), developed by Franz Hillenkamp and Michael Karas. This desorption ionization allowed a high-mass compound like a protein or synthetic polymer to be mixed with an energy-absorbing matrix for ionization and desorption by a laser. The result was single-charge ions from molecules that had masses of kDa. Prior to this, the mass of proteins could only be measured to within a range of ~10%. With MALDI, the mass could be determined to ±2 to 3 Da. MALDI contributed to the resurgence in TOF mass spectrometry (MALDI-TOF) because these were the only instruments capable of measuring high m/z values. This resurgence led to the development of the reflectron TOF described above.

RLS: I recall that the 1990s were the decade for rapid growth of HPLC-MS.

ODS: Yes, the invention of ESI in the mid-1980s led to explosive growth in LC-MS applications in the ’90s. In addition to this lower-energy desorption ionization method, other techniques found their way into use with LC: atmospheric pressure chemical ionization (APCI) and atmospheric pressure photoionization (APPI). APCI had been developed in the mid-1970s by Evan Horning’s group at Baylor College of Medicine as an LC-MS technique. There was not much interest because the results were mainly ions representing the intact molecule with little or no fragmentation to provide structural information. Now, with the advances in tandem mass spectrometry facilitating fragmentation, APCI once again began to appear in the literature. The fact that ESI mass spectrometers could be easily switched to APCI also increased the popularity of both modes.

Electrospray made MS and MS-MS a nearly universal technology, even for smaller molecules. Ion suppression by matrix and mobile phase components were the major experimental difficulty.

During the 1990s, technical advances in rapid electronics, high vacuum, and ion optics gave TOF instruments a performance advantage in speed, detection limits, and m/z range. Perhaps you recall the maXis from Bruker (Figure 1). It featured a three-meter-high vertical fl ight tube that was also suitable as a flagpole. Advanced turbo pumps provided high vacuum, which is more important for TOF than other types of mass analyzers. Add reflectron ion optics (above), and one had a truly powerful high-throughput mass spectrometer. Later firms developed folded TOF instruments that had less characteristic flight paths and structures.

Figure 1 – First maXis TOF mass spectrometer installed at the University of Queensland, Australia. (Courtesy of Bruker.)

Another important event of the 1990s was the emergence of Shimadzu in the field of mass spectrometry. The British company AEI (Allied Electronic Industries) had entered the market shortly after WWII with the manufacture of sector-based instruments for MS. AEI was acquired in the 1970s by a multidiscipline company called Kratos. Kratos began working in the area of reflectron TOF mass spectrometry utilizing “curved field reflectron technology” technology developed by the late Robert J. Cotter at Johns Hopkins. In around 1997, Shimadzu formed an alliance with Kratos to supplement its transmission quadrupole used in GC-MS. This alliance worked well. Today, Shimadzu has a complete line of instruments for GC-MS, LC-MS, and MALDI.

In the 1990s, the software really caught up with the instruments, leading to much higher value data and easier interpretation. This was facilitated by the explosive growth of the microcomputer.

RLS: Let’s move to the beginning of the 21st century. I recall this was the era of the very high-resolving-power instruments such as the ion cyclotron resonance and Orbitrap. These instruments could provide resolution between ions that differed by 0.0001 m/z units or less, which was useful in determining the precise elemental composition of nominally isobaric elemental compositions.

ODS: Not only were these two instruments opening the door to mass accuracy that had previously been elusive, but the 21st century saw advancements in the reflectron TOF and the introduction of the two-dimensional quadrupole ion trap (the linear QIT, or 2-D QIT).

Ion cyclotron resonance mass spectrometers (magnetic ion traps) were first developed by Hans George Dement, who shared the ion-trapping portion of the 1989 Nobel Prize in Physics with Wolfgang Paul. In the mid-1970s, Allen Marshall and Melvin Comisarow applied the technique of Fourier transform to the detection of ions in a cyclotron trap to determine their abundances and m/z values. Recall that Fourier transform improves detection in FTIR and FTNMR. In MS, however, achieving high m/z values and high mass accuracy required super-conducting magnets. These are expensive to buy and maintain because they use liquid helium. Helium is expensive, and occasionally under allocation. At one time there were three manufacturers of FTICR mass spectrometers: IonSpec, Thermo Fisher, and Bruker. Today, Bruker is the surviving vendor of research-grade FTICR instruments.

Thermo Fisher then introduced the Orbitrap at the 2005 ASMS meeting in San, Antonio, TX. This instrument had a mass resolving power close to that of the FTICR, but without super-conducting magnets. The Orbitrap used electrostatic charge as a method of ion separations. Since the Orbitrap’s introduction, the resolving power spec has continually increased. This, combined with the Orbitrap’s ease of use, has made it very popular.

Thermo Fisher has made numerous acquisitions over the decades. The acquisition of Finnegan has been unusually successful. In addition to the Orbitrap, Thermo’s MS team was responsible for the commercialization of both the transmission quadrupole and 3-D quadrupole ion traps mass spectrometers, and the development of 2-D QIT. No other mass spectrometer manufacturer has been responsible for such a breadth of technological introductions.

Another significant event of the early 21st century was LECO’s surprising acquisition of Meridian Instruments, which commercialized a novel integer-m/z GC-MS based on new TOF detector technology developed at Michigan State University. Today, LECO is the leader in fast GC-MS and GC×GC-MS. LECO also has a high-resolving-power GC-TOF instrument based on proprietary multiple ion path technology.

In addition to the LECO accurate m/z GC-MS, this period has seen the growth of accurate m/z GC-MS instruments such as the Agilent GC-QTOF and the JEOL GC-AccuTOF. During this same period, TOF and tandem Q-TOF instruments resulting in accurate mass have been produced by Shimadzu, Bruker, Waters, SCIEX (Danaher), and JEOL for use in LC-MS.

In 2004, Waters introduced the Acquity UPLC instrument. This had much narrower peak widths than prior LC separations technology. This tipped the scale from the LC being the slow part of LC-MS to being faster than the MS. In UPLC, analyte is often more concentrated compared to conventional HPLC. This was a significant advantage to ESI, which is a concentration-dependent rather a mass-dependent technology.

After the Orbitrap, the next most significant MS innovation of the early 21st century was “open-air” or ambient mass spectrometry. These operate at atmospheric pressure rather than high vacuum. Early in 2005, Direct Analysis in Real Time (DART) MS was introduced by Chip Cody at JEOL. This was preceded by a month, in late 2004, by Graham Cooks’ (Purdue University) introduction of desorption electrospray ionization (DESI).

Both of these techniques have revolutionized mass spectrometry. Until that time, mass spectrometry required that analytes ions be in the gas phase at some reduced pressure and somehow separated from the carrier matrix. Oftentimes, the analyte had to be extracted from a complex matrix, such as fragrance chemicals from a flower. With DESI and DART, the source of the analyte put in front of the mass spectrometer and mass spectral data is obtained without any effort other than the sample placement.

RLS: How would you characterize the current decade (2010–2020) in MS?

ODS: LC-MS in proteomics was the clear beneficiary of the economic stimulus spending to get the world out of the major funk of the prior decade. The basic technology has matured to the point where clinical labs can use LC-MS for monitoring pain medication, including opioids. Other diagnostic applications include Vitamin D. Contract research labs use quadrupole LC-MS to monitor drugs and metabolites. One LC-MS vendor shows a picture of about 100 instruments working 24 × 7 in a CRO. Corresponding improvements in LC led to UHPLC analytical throughput that is now several analytes per minute, with state-of-the-art leaders claiming single-digit peaks per second.

Now the labs are turning out huge quantities of information, too much for humans to comprehend. Drug and disease discovery has become an exercise in informatics. Display technology needs to be improved to convey more information per view.

The other development of note since 2010 is the rapid adoption of LCMS in the clinical diagnostics space. High-volume applications include hemoglobin A1c for diabetes, opiate drugs for pain relief, and Vitamin D.

Adoption of LC-MS was hindered for at least a decade due to the FDA’s insistence that instruments and protocols used for clinical diagnostics be registered for diagnostic use. Outside the U.S.A., these systems were working well, notably in Europe and Asia. Several American labs responded with developing and marketing analytical services using LC-MS protocols as laboratory-developed tests.

With advances in software and ease of use for accurate-mass GC-MS, GC-MS has also grown significantly. One outstanding software development during this period is Cerno Biosciences’ MassWorks. Even though instruments for accurate mass measures are much easier to use and require far less sophisticated operators than their predecessors, they are still costly ($200K + to mid-$400K). MassWorks can be used with any transmission quadrupole mass spectrometer to yield high mass accuracy data when data are acquired in the profile mode. The software can be used to resolve some mass-spectral doublets. It has proven to be quite effective in being able to obtain information that was previously only possible with high-cost MS-MS. Today MassWorks provides accurate mass structural information using quadrupoles.

Another advancement of this era has been the interface of ion mobility spectrometry with mass spectrometry. Ion mobility allows for characterization of ions based on their size and shape as well as the reduction of noise. A leader in this area of ion mobility-mass spectrometry is Waters.

RLS: This decade is almost over. What do you foresee in the 2020s and beyond?

ODS: It is hard to say what the future may bring. I expect that accurate mass data will be more prevalent aided by continued advances in TOF technology. This should expand the applications base. Open-air mass spectrometry will become more prevalent since it is easier and so fast.

It seems unlikely that major improvements in ionization technology will challenge ESI, since it works so well. Similarly, I’d be surprised if a new mass analyzer could displace the Orbitrap and reflectron; however, having said that, the ideas that led to these two game-changing advances were not anticipated before they burst onto the scene in the 1990s and 2000s.

Conclusion

Early on after Aston’s development of the mass spectrograph, J.J. Thomson is credited with having said, “There are far more problems in chemistry that can be solved by mass spectrometry than any other technique.” This was true then and I believe it to be true now and into the future.

Robert L. Stevenson, Ph.D., is Editor Emeritus, American Laboratory/Labcompare; e-mail: [email protected]