Advances in environmental organic mass spectrometry

---

Environmental analysis, as a rule rather than the exception, involves the analysis of mixtures of compounds present in a sample matrix. The process is analogous to finding a tree in a forest. The problem is compounded by the fact that there are many types of trees and a wide range of sizes and ages of each type of tree.

Environmental analytical procedures consist of (1) representative sampling (2) a sample preparation procedure to isolate and concentrate compounds from the matrix and (3) instrumental analysis. Instruments consist of a device to separate the components in a mixture (i.e., separating the types of trees) followed by a detector (identifying the size and age of each type of tree) that has sufficient sensitivity (how small an amount can I detect?) and selectivity (how confident am I in its identification?).

The gas chromatograph (GC) is a very efficient separation device developed in the 1950s. A series of major improvements over 50 years has produced the modern GC - capable of separating one hundred or more compounds of a complex mixture. Mass spectrometers (MS) have been available commercially since the 1940s. Within this instrument, ions and ion fragments are formed from molecules; these fragments together form a fingerprint of each specific molecule. Because of the great separating power of the GC and the ability to identify molecules based on data from the mass spectrometer, the interfacing of these technologies seemed inevitable. One problem - GCs were operated at atmospheric pressure with flow rates 20–30 mL/minute exiting the instrument, whereas mass spectrometers operate under high vacuum conditions (<10-6 torr).

Development of technologies to couple these seemingly incompatible systems began in the late 1950s. Interfaces were developed to remove the bulk of the carrier gas from the support coated open tubular (SCOT) or packed GC columns as they are more commonly known. The interfaces included synthetic membranes, the Watson-Biemann effusion separator and the jet separator. The most significant breakthrough was the development of the wall coated open tubular (WCOT) or capillary GC column with its relatively low carrier gas flow rates of the order of 1 to 2 mL/minute. The use of this type of column eliminated the need for an interface and the column could be coupled directly to the MS. Commercial GC-MS with data systems (DS) were not available until the late1960s. Electron ionization (EI) mass spectra, obtained under a high vacuum with a beam of 70 eV electrons, are, with some exceptions, fairly independent upon instrument design. Therefore, the EI mass spectra obtained with one GC-(EI)MS system can be compared with the EI mass spectra from other GC-(EI)MS systems and with the EI mass spectra in commercial databases such as National Institute of Standards and Technology (NIST) and Wiley where collections of mass spectra now exceed 200,000. This makes the GC-(EI)MS system a powerful tool for the identification of individual compounds in mixtures, e.g., environmental samples.

The GC-MS/DS became a routine instrument for environmental analysis of organic compounds in the 1970s and 1980s. At the Ontario Ministry of the Environment (MOE), GC-MS/DS technology was first acquired in 1974. The Dupont 21-491B GC-MS, with a packed column GC was used to analyze volatile organic compounds (VOCs) with an off-line purge-and-trap (PT) system. The Finnigan 4000 GC-MS with a capillary column GC, acquired in 1979 was used to analyze semi-volatile (extractable) organic compounds. These instruments were replaced in 1999 by their bench-top counterparts. This calibre of instrument can be seen on TV shows such as CSI.

GC-MS systems are capable of analyzing a broad range of organic compounds that includes volatile, non-polar and low molecular weight (<500 Da) organic compounds that are mobile in the environment. Early GC-MS systems incorporated low resolution mass spectrometers (LRMS) capable of detecting masses to the nearest integral value. However, carbon monoxide (mass = 28) cannot be separated from molecular nitrogen (mass = 28) and chlorinated dioxins (CDDs) could not be separated from polychlorinated biphenyls (PCBs).

In 1983, the MOE acquired the VG ZAB-2F, a double-focusing magnetic sector mass spectrometer, i.e., a 'high resolution' mass spectrometer (HRMS), having a mass accuracy of three to four decimal places.
With this mass spectrometer, carbon monoxide (mass = 27.9949) could be separated from molecular nitrogen (mass = 28.0062) and (chlorinated) dioxins could be differentiated from PCBs. This capability significantly increased the selectivity of the instrumentation.

(Chlorinated) dioxins could be analyzed with certainty that they were not PCBs. The MOE used this instrument to develop a novel ‘lock mass’ technique to improve sensitivity by a factor of three over standard HRMS methods of analyzing dioxins. This was a significant advance in HRMS technology.

In 1991, the VG AutoSpec, the first in this series of high resolution mass spectrometers, was obtained. This was followed by Micromass AutoSpec Ultima’s in 1999, 2002, 2003 and 2006 and an AutoSpec-Q in 2003.

These high resolution mass spectrometers are used to analyze dioxins/furans, dioxin-like PCBs, brominated diphenyl ethers, i.e., brominated flame retardants (BFRs), and polychlorinated naphthalenes (PCNs), N-nitrosamines (including NDMA), taste and odour compounds (including geosmin and 2-methylisoborneol) and selected disinfection by-products. In addition, water-soluble organic compounds such as ethylene glycol can be analyzed directly by direct aqueous injection (DAI)/GC-HRMS at sub-ppm detection limits.

In 1987, the MOE acquired the Finnigan MAT TSQ-70, a tandem mass spectrometer (MS/MS), to analyze dioxins. A tandem mass spectrometer achieves selectivity by operating two LRMS systems in series. While this type of selectivity is different from HRMS, the confidence in the identity of the compound is comparable. With respect to sensitivity, the MOE developed an optimization procedure for tuning that achieved detection limits within a factor of two to three of the GC-HRMS system. This was another significant advance in technology. The Finnigan MAT TSQ-70 was replaced in 2005 with a bench-top system, the Varian 1200L. GC-MS, GC-HRMS and GC-MS/MS systems provided the MOE with the capabilities to detect suspected toxins at trace levels with a high degree of confidence.

However, there remained some groups of compounds that could not be analyzed by GC-MS technology. Included were thermally-labile compounds (decompose when heated), polar/ionic compounds and high molecular weight compounds. These types of compounds are more amenable to analysis (separation) by a high performance liquid chromatograph (HPLC). The major obstacle was development of technology to couple the HPLC with the MS, i.e., coupling a liquid with a high vacuum system. While development of GC-MS technology was initiated in the late 1950s, development of LC-MS technology was not initiated until the late 1970s. Early LC-MS systems included the moving-belt (MB), ThermoSpray (TSP) and particle beam (PB) interfaces but they had limited applications. The MB interface was mechanically complex but allowed the acquisition of EI spectra, i.e., same type of data obtained from a GC-MS, from an LC effluent. This allowed searching of the same databases such as NIST and Wiley.

The successful coupling of an LC effluent with the high vacuum system of a mass spectrometer involves three processes, namely nebulisation, desolvation and ionization. Nebulisation is the creation of a fine spray with the mixing of a gas such as helium or nitrogen with the liquid effluent. Desolvation is the separation of the liquid from the compounds of interest. Ionization is the creation of ions from these compounds. The order in which these three processes occur defines the LC interface and ultimately determines its areas of applications.

At the MOE, LC-(PB)MS was acquired in 1991. The VG Trio-2. With PB, nebulisation occurs first followed by desolvation with ionization being last. The difficulties in desolvation of neutral droplets and in aiming a neutral beam into the mass spectrometer resulted in variable and low transport efficiency across the interface.
This instrument was used to analyze pesticides by LC-MS. Selectivity was improved over LC-UV systems but detection limits were high and calibration curves were non-linear. However, the mass spectra obtained by LC-(PB)MS were searchable through databases such as NIST. PB remains as the only LC-MS technique in which EI spectra can be obtained from an LC effluent.

The introduction of the atmospheric pressure ionization (API) techniques, electrospray ionization (ESI) and atmospheric pressure chemical ionization (APCI) in the early 1990s, allowed for the widespread introduction of LC-MS systems. With ESI, ionization occurs first (in solution) followed by nebulisation and desolvation. With APCI, nebulisation occurs first followed by ionization (in the vapour phase) and desolvation. Because ionization occurs at the first and second stages respectively, sensitivity has been improved significantly.

One of the first applications of LC-(ESI)MS was the analysis of the ionic herbicides diquat and paraquat. ESI technology was obtained in 1995 on a 1986 vintage instrument.

Having demonstrated the utility of LC-(ESI)MS technology, work was initiated to analyze algal toxins, specifically microcystins. This work was started on the VG ZAB-EQ but this instrument would prove to be inadequate to analyze singly- and doubly-charged microcystins in the same run.

In 2001, a fully automated triple quadrupole tandem mass spectrometer (LC-MS/MS) system was acquired to meet the needs of the microcystin analyses. The Micromass Quattro Ultima and the Applied Biosystems MDS Sciex 4000 Q Trap.
A method was developed on the Quattro Ultima to analyze total (extracellular + intracellular) microcystin-LR, -RR, -YR, -LA and anatoxin-A in the same run. At the present time, the MOE has the only method accredited and licensed for analyzing microcystins in drinking water. This method was put to the test recently when a late algal bloom in Lake Rosalind resulted in a request for analyses on an emergency basis. The method is labour-intensive because of the need to lyse the algal cells and extract the intracellular microcystins. This is necessary because the Ontario Drinking-Water Quality Standard (ODWQS) for microcystin-LR is 1.5 µg/L for the total of the extracellular (free) and the intracellular microcystin-LR. Analysis of a batch of 20 samples plus quality control (QC) samples normally takes three to five days. For this emergency, the analysis of five samples plus QCs was completed in 24 hours. Only traces of microcystins were present well below the ODWQS, and the ministry's timely analyses enabled quick action.

Identification of unknowns is analogous to identifying a unique individual in the world’s population. A GC-(LR)MS can provide the equivalent of a photograph and this can be compared to photographs in a database. In order to identify unknowns that are not in any commercially available databases, a tandem hybrid mass spectrometer was acquired in 2003. A tandem hybrid mass spectrometer can provide a photograph of the individual as well as the individual’s fingerprints. The Micromass AutoSpec-Q is shown in Figure 1.
This instrument is a combination of a double-focusing magnetic sector mass spectrometer and a quadrupole tandem mass spectrometer. It has capabilities comparable to the AutoSpec Ultima. In addition, it has advanced scanning features and gas phase ion chemistry techniques that can be used to identify unknowns. Work is in progress using this instrument to identify disinfection by-products in treated drinking water.

In 2006, a new state-of-the-art instrument was acquired by the MOE. This instrument is known as a Fourier Transform (Ion Cyclotron Resonance) Mass Spectrometer [FT(ICR)MS] or FTMS. This technology was developed in 1974 at the University of British Columbia by Drs. Melvin Comisarow and Alan Marshall. Dr. Alan Marshall is currently at the National High Magnetic Field Laboratory at Florida State University. Although this technology was demonstrated in 1974, the high pressure inlets and interfaces needed to introduce a variety of samples were not commercially available until recently.

The superconducting magnet of an IonSpec 9.4 T FTMS is shown in Figure 3. The GC and LC vacuum carts are on either side of the magnet (Figure 2 and 4). On an FTMS, the masses can be measured to 4 to 5 decimal places, i.e., carbon monoxide = 27.99492, molecular nitrogen = 28.00615. The mass accuracy is such that the weight of the electron (0.00054 Da) must be taken into account in the calculated masses. This capability can result in the identification of a compound based on a single accurate mass measurement. This would be analogous to identifying an individual not only by the individual’s photograph and fingerprints but also by the individual's DNA, the DNA being provided by the FTMS.

In addition to ultrahigh mass accuracy, ultrahigh resolution can also be obtained on an FTMS. A double-focusing magnetic sector mass spectrometer can differentiate (chlorinated) dioxins from PCB but it cannot differentiate chlorinated / brominated dioxins from (chlorinated) dioxins. The resolution required for the former is 10,000 and for the latter is >80,000. The FTMS has the capability to make this differentiation.

The FTMS will be used to provide enhanced emergency response capabilities, chemical fingerprinting, identification of new and emerging environmental contaminants and enhanced method and technology development. In this way, the MOE can continue to be at the forefront of environmental analyses.

Dr. Vince Taguchi (vince.taguchi@
ontario.ca, (416) 235-5902) is manager of the Mass Spectrometry Section, Laboratory Services Branch, Ontario Ministry of the Environment and the current chairman of the Toronto Area Mass Spectrometry Discussion Group.