Multidimensional Gas Chromatography for Precise, High-Resolution

Apr 26, 2012 - Centre for Green Chemistry, School of Chemistry, Monash University, Wellington Road, Clayton 3800, Australia. ABSTRACT: A novel hybrid ...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/ac

Novel Hybrid Comprehensive 2D − Multidimensional Gas Chromatography for Precise, High-Resolution Characterization of Multicomponent Samples Blagoj Mitrevski and Philip J. Marriott* Centre for Green Chemistry, School of Chemistry, Monash University, Wellington Road, Clayton 3800, Australia ABSTRACT: A novel hybrid (sequential) comprehensive 2D − multidimensional gas chromatography (GC × GC − MDGC) method for complex sample manipulation and separation is described. It incorporates a separation step that approximates slow modulation GC × GC, prior to microfluidic Deans switch heart-cutting of a targeted region(s) into a third analytical column. It allows discrete single or multiple components, bands or regions, or any combination of these to be selected and excised from within the 2D GC × GC separation space. The excised individual components can be further collected and studied. Alternatively, any unresolved or poorly resolved components, or regions that require further separation, can be transferred to an additional (third) column separation step. The method is applied to separation and quantitative analysis of oxygenates in a thermally stressed algae-derived biofuel oil by using flame ionization detection (FID), without any prefractionation. This permits oxygenated compounds to be fully resolved from saturated (matrix) compounds, which are completely excluded from the third column. Improved separation was obtained between target classes (aldehydes, 2-ketones, alcohols, acids). Excellent calibration linearity, and retention time and peak area reproducibility were obtained for 14 oxy-compounds present in trace amount in the complex biofuel matrix. Accuracy of microfluidic transfer to the third column, and the profile reproducibility before and after heart-cut operations, was demonstrated by extracting single components from a complex coffee volatile sample.

T

of 2D columns equal to about the peak capacity of the 1D column. Subsequent development of a fast modulation approach by Phillips7 permitted use of a single 2D column, to implement comprehensive 2D GC (GC × GC), which samples the whole effluent from 1D to 2D. This requires fast repetitive analysis, with modulation period PM < 1wb (modulation ratio MR > 1)8 by using a short 2D column (0.5−2 m), which has limited capacity. This results in high total peak capacity, with narrow peaks (50−200 ms) produced by the fast analysis 2D column, and consequently requires relatively high acquisition frequency detection,9 with time-of-flight MS the instrument of choice for many GC × GC-MS applications.10 Innovations on the theme of MDGC and GC × GC separations11 include for example 1D GC-FID in parallel with fast cryotrapping MDGC.12 A switching system has been described to select between a GC × GC and a MDGC method.13 A loop-type configuration for single component isolation in a cryotrap following Deans switching allows multiple injection/target collection with NMR analysis.14 A method that switches from GC × GC-FID mode to fast sequential heart-cut MDGC-qMS to permit identification of oxygenates in algae-derived fuel oil.15 The latter provides bulk-

here is continuing interest in development of a gas chromatography (GC) method that is fast, sensitive, selective, affordable, and simple. It should also allow specific measurement of both target trace and bulk components in the presence of highly complex matrices and abundant interfering components. This may be for diverse applications such as trace pesticides in food or environmental samples, chemical class speciation in petrochemical samples, or metabolite profiles in biological systems. Because a single GC dimension has inadequate peak capacity to resolve all components in complex mixtures,1 a move toward multidimensional (MD) separation systems is understandable and inevitable. Proposed by Simmons and Snyder in 19582 and refined by use of a pneumatic Deans switching system,3 in MDGC regions of the first dimension column (1D) effluent are diverted to a second column (2D), of different selectivity, to expand peak capacity. MDGC uses low frequency discrete sampling of 1D, and a relatively long 2D column, although variants on MDGC technology have been proposed.4,5 Because few heart-cuts (H/C) are applied within one run, this limits applicability of the technique for global sample characterization. However, slower acquisition rate detectors, for example a quadrupole mass spectrometer (qMS), are suitable with this method. Giddings6 proposed that a coupled-column method equivalent to a 2D planar experiment would require a total number © 2012 American Chemical Society

Received: February 13, 2012 Accepted: April 26, 2012 Published: April 26, 2012 4837

dx.doi.org/10.1021/ac300429y | Anal. Chem. 2012, 84, 4837−4843

Analytical Chemistry

Article

Figure 1. (a) Modulated flow from the first two columns is diverted to the monitor detector FID1 when DS is off; (b) Modulated flow is diverted to the 3DL column when the DS is activated, with DS activation time < PM. Remaining effluent from 2DM passes to FID1.



matrix-interference free identification of oxygenated compounds, which was impossible in the classical 1D GC-MS method. However, 10 repeat runs for the whole sample were required for high-resolution separation and precise qMS identification of all target compounds. A 3D system (GC × GC × GC) was described by Siegler et al.,16 which progressed in successive stages from slow, through medium, to fast modulation. Progressively faster modulation operation was required at each stage. Complex MDGC methods are available for various applications in the petroleum industry, for example PIONA analysis; GC × GC methods were proposed to simplify this determination.17 Harynuk and Gorecki described a stop-flow method to expand the capacity on the 2D column in GC × GC to allow greater resolution.18 These demonstrate the lengths that researchers go to achieve improved resolution of complex samples. The method proposed here examines algal-derived biofuel and coffee volatiles. Biofuel requires assessment of fuel thermal and storage stability. Oxidative degradation, in the absence of sufficient antioxidant, promotes oxidative instability and leads to formation of alkylperoxyl radicals and secondary oxidation products.19 Such degradation products must be measured at trace level with accurate quantification. Low-level oxidation products (