Automated Chemical Ionization in GC-MS

Gas chromatography with mass spectrometric detection using chemical ionization (CI) has demonstrated great utility in a wide variety of application areas ranging from forensics to environmental to clinical and even into the life sciences, traditionally thought of as liquid chromatography applications. The tunable selectivity of CI has lent specificity to many analyses and has resulted in enhanced detection in complicated matrices. However, GC-MS methods employing CI have typically been more difficult to develop and maintain. Typical difficulties in CI methods are:

  • Choice of CI modes: Some analytes respond better in positive CI (PCI) and others in negative CI (NCI). Beyond response, the type of spectral information desired (e.g., molecular weight or composition) can be a determining factor
  • Optimization of CI method parameters: A variety of CI reagent gases are possible and their flows determine the source pressures and thereby the concentrations of CI species involved in the ionization. Also, source temperature plays a role more often from the chromatographic and practical aspects of robust operation. Often CI methods are built on the more commonly understood electron impact (EI) ionization methods but a common complication is sometimes just recognizing which compound is which under CI conditions, because of the drastic differences between EI and CI spectra
  • Replicating the CI method: After developing a method, reproducing it requires capturing some subtle aspects of the CI parameters that are not often easily discerned.

The advent of software control of all critical aspects of the CI process on the Agilent 5975 MSD series instruments (Agilent Technologies, Inc., Santa Clara, CA) dramatically simplifies the CI method development process to the degree that CI is as easy as EI. This article presents a few of the features that make this possible and gives a case example to illustrate this new approach to CI.

Automated CI control

The Agilent 5975 MSD series provides control of the key parameters necessary for high-quality CI analysis. In brief, these features are:

  • Automated PCI and NCI tuning using a dedicated CI tuning agent (perfluorodimethyltrioxadodecane) in PCI with methane and NCI with both methane and ammonia. Completely electronic control of reagent gases allows software to optimize reagent gas flow
  • Reagent ion tuning: Reagent ions for PCI optimization are provided for methane, isobutane, and ammonia, which are the common gases, but can be customized for other agents such as the common liquid agents (methanol, acetonitrile, pentane, etc.). Automatic reagent gas flow “surveys” can be generated to provide method optimization parameters. The CI reagent gas manifold supports plumbing for two reagent gases
  • Automated “EI” tuning of the CI source: The Agilent MSD CI source is designed for optimal CI performance, but the CI source can be automatically tuned to provide EI spectra to assist method development
  • Figure 1 - Example of an automated sequence that acquires a standard or sample in EI and CI modes. Note the advantage of automated acquisition of CI in modes employing both methane and ammonia reagent gases.

  • Automated EI, PCI, and NCI acquisitions in sequence: A sample or series of samples can automatically be analyzed by GC-MS in EI, PCI, and NCI modes in sequence. Further, the CI modes can also be acquired using two different reagent gases, e.g., PCI with methane or ammonia and NCI with methane or ammonia. Such a sequence is illustrated in Figure 1.

An illustration of use of a few of these features may be more instructive than general statements.

Figure 2 - Electron impact ionization mass spectrum of dioctylphthalate. Note the very low abundance of the molecular ion and relatively sparse fragment information.

The EI spectra of phthalates are well known to all mass spectroscopists. These compounds are everywhere and are common to the background of laboratories in their atmosphere and frequently in laboratory solvents. With one exception, all the common phthalates exhibit 149 m/z as the base peak in their spectra, and more defining fragment information is less than 20% in abundance. Figure 2 shows the spectra for dioctylphthalate acquired under EI conditions. Applying the sequence of Figure 1 to a dioctylphthalate standard using PCI and NCI with both methane and ammonia reagent gases gave the spectra shown in Figure 3 and with the total reconstructed ion current intensities.

Figure 3 - Spectra of dioctylphthalate in all CI modes: PCI with ammonia (upper left), PCI with methane (lower left), NCI with methane (lower right), and NCI with ammonia (upper right). Note the differences in spectra and total response (shown at center).

The PCI spectra differ significantly between methane and ammonia reagent gases. As expected, the “gentler” ionization processes at work with ammonia are reflected in the spectrum as a base peak formed by protonation of the molecule. In contrast, the more aggressive methane reveals much more fragmentation (which is useful for revealing structural features of molecules) but also a high abundance of the protonated molecule and a base peak at the familiar 149 m/z. The total reconstructed ion current response is much higher for ammonia PCI than methane.

The NCI results differ more subtly. The phthalate spectrum generated in methane NCI reflects results of primarily electron capture negative ion CI (ECNICI). Here, the methane gas acts as a buffer gas that thermalizes electrons emitted by the filament so that the electrons can be captured by the phthalate. ECNICI offers extremely sensitive detection for molecules that, by nature or by chemical art (derivatization), are electrophilic. Examples include the polychlorinated biphenyls (PCBs), polyaromatics, organochlorine pesticides, and the fluorinated derivatives of various drugs or biologicals. The selective nature of this electron capture ionization mechanism discriminates against the typical background interferences, which have very little tendency to capture electrons and therefore do not respond. The result is very high signal-to-noise and very low detection limits in matrices that typically are a problem for EI analysis.

Ammonia also offers an approach to ECNICI and actually provides higher thermalizing power than methane; thus it can be used at lower source pressures than methane. But, as the differences in the NCI methane and ammonia spectra indicate, ammonia provides an additional mechanism due to formation in the source of NH2 ion. This anion produces actual NCI and this additional process usually gives some structural information beyond methane ECNICI, here in the form of the fragment at 277 m/z.

This example shows how automated surveys in CI can offer users a full array of choices about how to conduct their analysis depending on their intentions. For example, ammonia PCI gives simple phthalate spectra and with relatively intense response. The richer spectra provided by PCI with methane yield more structural information that may be useful to the analyst. In this example, side-chain compositions are more easily determined. NCI shows a similar situation but reversed in that more molecular anion is produced in methane than in ammonia, but here ammonia yields more structural information. In both PCI and NCI, ammonia demonstrates the potential for superior detection due to great total response. Based on these data, the analyst could examine a sample, and the role of matrix interferences would decide which approach was more appropriate. As a general comment, this emphasizes the usefulness of ammonia as a reagent gas in CI.

Conclusion

CI can serve as a valuable tool and provide information such as the molecular weight of a compound, some insight into a compound’s structure, or a means for sensitive and selective detection of a compound in complicated matrices. Automation of CI allows standards or samples to be easily surveyed with minimal user effort in all MS modes. This gives the analyst the opportunity to rapidly develop an optimal method. Integral to this capability are the automated tuning and control of the reagent gases, which is critical to method replication. But, since all key CI parameters are automated, CI is as easy and simple as EI in tuning, method development, and use.

Additional reading

www.chem.agilent.com. Select the category Literature.

5975 inert MSD: Benefits of the Enhancements in Chemical Ionization Operation (5989-4347EN). On phthalate analysis, refer to A new approach to phthalate analysis by GC-MS (5988-2244EN).

The authors are with Agilent Technologies, Inc., 5301 Stevens Creek Blvd., Santa Clara, CA 95051, U.S.A.; tel.: 408-553-7191; fax: 408-553-6500; e-mail: [email protected].

Comments