Applications of Gctofms

1. Analysis of Complex Mixtures

While there are limitless opportunities to use TOFMS in conventional GC-MS analyses, there are also a variety of applications developed using GC-TOFMS that take advantage of the potential to provide more information in less time. Figure 14 illustrates the 3-minute analysis of a complex citrus standard mixture, using time-compressed chromatography, which normally requires 40-60 minutes using conventional GC-MS. Compression of the time axis occurs without any loss in analytical resolution. This time-compression is potentially useful for a variety of high-throughput screening applications; and thus, although only 53 peaks are visible in the RTIC, deconvolution detected 198 analytes. Indeed, it is this rapid and comprehensive qualitative analysis capacity that documents one of the most powerful aspects of TOMS with TAD.

2. Fast Chromatography

Another class of applications involves the use of fast chromatography, where the chromatographic technology applies sample compression techniques to generate narrow peak widths in addition to compressing the analysis time, as described earlier. Figure 15 illustrates the analysis of volatile contaminants (17). A cryofo-cusing gas injector system (Chromatofast, Inc.) is combined with a HP6890 gas chromatograph and the LECO Pegasus II TOF mass spectrometer to generate

time (sec)

Figure 14 Analysis of a complex citrus standard mixture. Although only 53 peaks are evident in the RTIC, deconvolution revealed the presence of 198 compounds. (Data courtesy of Dr. Nancy Myers, LECO Corporation.)

time (sec)

Figure 14 Analysis of a complex citrus standard mixture. Although only 53 peaks are evident in the RTIC, deconvolution revealed the presence of 198 compounds. (Data courtesy of Dr. Nancy Myers, LECO Corporation.)

peak widths of only a few hundred milliseconds, for analytes also separated by only a few hundred milliseconds. Deconvolution generates individual mass spectra for each component. The use of fast chromatography in this application reduces the analysis time and also lowers the achievable detection limit by generating narrow, and thus taller, peaks. This particular application is illustrative of throughput gains that can be achieved with analyses that require rapid feedback or nearly constant monitoring of dynamic systems, and applications to flavor and fragrance analyses. Thus, significant reductions in analysis time must also be accompanied by relatively short data processing times. The quality of the mass spectral data generated using TOFMS combined with efficient deconvolution algorithms, result in data processing times of only a few seconds for the above application.

3. Trace Analyses

Many quality control analyses require the detection of minor contaminants in a sample product. These types of analyses can be inherently difficult because, by definition, the contaminants being searched for will occur in low amounts, and the number and type of contaminants present in the sample may be unknown. The analysis is further complicated if the contaminant coelutes with a more dominant

Figure 15 Fast chromatography of volatile contaminants. Data were acquired using fast chromatography and TOFMS at 50 spectra/sec. (From Ref. 32).
Figure 16 Trace contaminant analysis. Deconvolution and subsequent analysis of contaminant standards confirmed the identity of each analyte. (Data courtesy of Dr. Nancy Myers, LECO Corporation.)

product analyte. Figure 16 illustrates an extreme example of this application type, where several contaminants elute under the solvent peak of a flavor sample. In this case, the contaminants were 400,000 times less abundant than the propylene glycol solvent, and yet subsequent deconvolution produced library-matchable spectra for each one.

4. Comprehensive 2D-GC/TOFMS

As an alternative method for the analysis of very complex mixtures, comprehensive 2D-GC combines two chromatographic retention characteristics to distin guish between similar analytes. This technique has been successfully applied in the analysis of petroleum samples where many hundreds of analyte species are present. Yet, while the benefit of analyte separation is realized in some cases, identification of many analyte peaks has thus far been precluded. Figure 17 illustrates the benefit of multichannel detection in the analysis of herbicides (19). The comprehensive 2-D field was generated using a 3 m, 0.25 mm I.D., 0.25 |m film thickness DB-5 first column and a 1.5 m, 0.1 mm I.D. 0.1 |m film thickness OV-1701 second column. The custom-designed modulator operated at a period of 4 seconds. The LECO Pegasus II TOF mass spectrometer was used as the detector, operating at 100 spectra/sec. The topographical chromatogram generated in this application uses two time axes and exhibits a complex background from which the analytes must be discriminated. The combination of 2D-GC with MS creates a fourth dimension of information to aid in definitive identification. As shown in the deconvoluted chromatogram of the first peak cluster, the modulation process generates very narrow peak widths of approximately 125 msec with each compound analytically resolved. These narrow peak widths mandate highspeed detection that, combined with the need for analyte identification, generally precludes all detection methods other than TOFMS as the method of choice.

5. Deconvolution by Differential Fragmentation

To date, deconvolution routines have realized their differentiating parameters using intervals of time, because rarely do components elute having both identical ion fragmentation patterns and identical chromatographic temporal behavior. Time-based deconvolution routines struggle, however, when compounds generating the same fragmentation pattern partially coelute. In these cases, other more sophisticated routines must be employed to determine the individual elution manifolds, and they completely fail when compounds having the same fragmentation pattern exactly coelute in time (congruent coelution). In these situations, the parameters of the chromatographic separation (column type and resolving power, carrier gas, temperature program, etc.) are changed in attempts to achieve even marginal temporal separation. Unfortunately, extensive method development is time consuming and costly, and some chemical isomers are unresolvable even under optimal conditions.

An unforeseen advantage of TOFMS with TAD has been manifested in the reproducibility of the ion fragmentation patterns. This presents the m/z axis itself to be used in deconvolution. To date, information on this axis has been used only to calculate confidence factors for the presence of components detected by the time-based routines and not to deconvolute coeluting species. Because of the unskewed quality of the TOFMS data, coeluting species, usually isomers, generating all ions of the same m/z but in differing relative intensities, can be deconvo-luted by assigning appropriate amounts of the measured m/z intensities to each

Figure 17 2D-GC-TOFMS of a herbicide mixture.

species. This involves a search for differential fragmentation between each species that can be used to determine the relative amounts of each using simple linear equations.

Figure 18 illustrates the method of deconvolution by differential fragmentation applied to 2,5- and 2,6-dimethylpyrazine, a pair of isomers found in peanut butter that coelute under common chromatographic conditions (31). In this case, the relative abundance of the ions at m/z 39 and m/z 42 is different for each isomer. Because of the reproducibility of the fragmentation pattern, this rather small differentiation can be used to calibrate a deconvolution scale that allows changes in the relative abundance of these two m/z's to be used for quantitation. Once the ratio of these m/z's versus relative concentration is calibrated, a single mass spectrum measured at the apex of the chromatographic peak will yield the relative abundance of each coeluent. Quantitation is finalized by integration of

Figure 18 Deconvolution by differential fragmentation of geometric isomers. Analysis of a standard mixture of 80% 2,5- and 20% 2,6-dimethylpyrazine, using the working curve, extrapolated to 75% 2,5-dimethylpyrazine.

the coeluting peak along the time axis with subsequent apportionment of this sum into the two components on the basis of this measured ratio.

In an interesting manner, the technique of differential fragmentation closes the loop on the application of deconvolution. Where small temporal differences exist between elution peaks, success has been shown in situations in which each analyte has a unique ion; in which only one coeluting analyte has a unique ion; and in which there are no unique ions. This technique applies where there is complete coelution, no temporal differentiation, and all the ions are shared. It can't get any more complex than that.

0 0

Post a comment