Carbohydrate pyrolysis mechanisms from isotopic labeling: Part 3. The Pyrolysis of d-glucose: Formation of C3 and C4 carbonyl compounds and a cyclopentenedione isomer by electrocyclic fragmentation mechanisms

https://doi.org/10.1016/j.jaap.2007.12.005Get rights and content

Abstract

The flash pyrolysis of d-glucose was investigated by the use of 13C labeling, in conjunction with GC/MS. Co-pyrolysis of uniformly labeled and unlabeled d-glucose established the extent of unimolecular formation of each of the pyrolysis products. A complete set of singly labeled d-glucose isotopologs was used to determine the origin of specific carbons within each of the pyrolysis products. The results were compared with the expected labeling patterns that arise when the cyclic Grob 1,3-diol fragmentation and the tandem alkaline pinacol rearrangement/retro-aldol fragmentation (TAPRRAF) discovered from the pyrolysis of glycerin are used to initiate breakage of the six-carbon chain of d-glucose. The most promising rationalizations provided by this exercise are presented herein, for the formation of six C3 and eight C4 acyclic carbonyl-containing pyrolysis products, and for 3-cyclopentene-1,2-dione.

Introduction

Carbohydrate pyrolysis is of central importance in the thermochemical conversion of biomass to sources of energy [1], [2], [3], the formation of flavors during the cooking of food [4], and in the formation of cigarette smoke constituents generated from carbohydrates in tobacco by the act of smoking [5], [6]. Yet despite this importance, much of the detailed chemistry remains poorly understood, and the pathways of product formation of numerous pyrolysis products remain debatable [7]. We have elected to model the carbohydrate pyrolysis chemistry by the flash pyrolysis for the first time of a complete set of appropriately 13C-labeled d-glucose (1, Fig. 1) isotopologs. The resulting products were examined for identity and isotopic content by means of gas chromatography/mass spectroscopy (GC/MS). This method allows a large number of products to be examined simultaneously using only a limited number of pyrolysis experiments. Flash pyrolysis is particularly appropriate for modeling the formation of tobacco smoke constituents, given the steep temperature gradients and short contact times inherent to a cigarette being puffed [8].

In the first paper of this series [9], isotopic labeling with 13C was applied to gain a clear picture of the detailed mechanisms by which glycerin (2, Fig. 1) fragments by carbon–carbon bond breakage during pyrolysis to form acetaldehyde. In part 2 [10], the mechanisms discovered from glycerin pyrolysis were conceptually applied to the pyrolysis of d-glucose (1) as a typical monosaccharide, as a means of providing general methods of initial fragmentation by carbon–carbon bond breakage, and were further discussed specifically with reference to the formation of C1 and C2 carbonyl-containing products. Fructose (3, Fig. 1) was specifically invoked as an intermediate for some of the products; further migration of the carbonyl group was also proposed to provide intermediary “3-ketohexose” (4, Fig. 1). Here we extend the same concepts to an explanation of the formation of the carbonyl-containing products of three or four carbons (Fig. 2). Since five-carbon fragments containing multiple oxygen functionality generally have the wherewithal to cyclize forming furans, furanoids, or cyclopentanoids, such rarely provide linear acyclic products that survive long enough to be observed. Furan formation will be discussed in Part 4 of this series [11].

With the sole exception of 3-cyclopentene-1,2-dione, cyclopentanoids will not be addressed in detail in this initial series of papers. Although the overall incorporation of labeled atoms is readily determined for such molecules as 2-cyclopentenone or 2-hydroxy-2-cyclopentenone (and favors the last five carbons of d-glucose), the fragmentation modes are too ambiguous to allow ready location of the label within the molecule, as compared to acyclic species. Internal aldol cyclization of species such as 4-oxopentanal or 5-hydroxy-4-oxopentanal may well be involved, but the isotopes are unable to provide confirmation.

Section snippets

Experimental

The experimental details were as provided in Part 2, inasmuch as the products were obtained during the same pyrolysis experiments as described therein [10]. A gas-separating column was used to provide some of the data relating to methacrolein (19): (Restek Rt-QPLOT, 30 m and 0.32 mm ID). Dimeric dl-glyceraldehyde and dimeric 1,3-dihydroxyacetone were obtained from Aldrich and used as received.

Unlabeled (naturalomer [9]) substrates afforded products which were identified based on their retention

Establishing the maintenance of carbon–carbon bond connectivity during the pyrolysis of d-glucose

The experiment involving the pyrolysis of a homogeneous blend of naturalomer (98.9%12C6) and 13carbo-ubiquilog (99+%13C6) d-glucose (mixed in aqueous solution) to establish the extent of reaction unimolecularity was described previously [9], [10]. The extent of unimolecularity determined thereby are presented in Table 1 (C3 products) and Table 2, Table 3 (C4 products). The net numbers reported were not statistically adjusted for fortuitous recombination of isotopically homogeneous fragments,

Conclusion

By the use of isotopic labeling with 13C, we have established explicitly the dominance of unimolecular mechanisms for the formation of a wide range of low-molecular weight carbonyl compounds that are formed from the pyrolysis of d-glucose. The two novel electrocylic fragmentation mechanisms originally discovered during glycerin pyrolysis and used to rationalize the dominant observed labeling patterns for C1 and C2 carbonylic products [10] here continued to provide rationalizations for larger

Acknowledgements

The support of management and staff of Philip Morris, USA, in pursuing this work is gratefully acknowledged. This work was presented in part at the meeting of CORESTA, Xian, China, September 3, 2001.

References (34)

  • P. Tomasik et al.

    Adv. Carbohydr. Chem. Biochem.

    (1989)
  • R. Talhout et al.

    Food Chem. Toxicol.

    (2006)
  • J.B. Paine et al.

    J. Anal. Appl. Pyrol.

    (2007)
  • G.R. Ponder et al.

    Carbohydr. Res.

    (1993)
  • G.R. Ponder et al.

    Carbohydr. Res.

    (1993)
  • G.L. Lookhart et al.

    Carbohydr. Res.

    (1978)
  • M.M. Mulder et al.

    J. Anal. Appl. Pyrol.

    (1991)
  • A. Van Der Kaaden et al.

    J. Anal. Appl. Pyrol.

    (1983)
  • M.S. Feather et al.

    Adv. Carbohydr. Chem.

    (1973)
  • M. Kaltschmitt et al.

    Biomass Gasification and Pyrolysis: State of the Art and Future Prospects

    (1997)
  • E.J. Soltes et al.
  • J.L. Schauer et al.

    Environ. Sci. Technol.

    (2001)
  • H.R. Burton et al.

    Beitrag. Zur Tabakforsch.

    (1975)
  • S.C. Moldoveanu

    Analytical Pyrolysis of Polymeric Carbohydrates

    (1998)
  • R.R. Baker

    High Temp. Sci.

    (1975)
  • J.B. Paine III, Y. Pithawalla, J.D. Naworal, Part 2. Glucose C1 & C2 fragments, J. Anal. Appl. Pyrol., preceding...
  • J.B. Paine III, Y. Pithawalla, J.D. Naworal, Part 4. Furans, J. Anal. Appl. Pyrol., in...
  • Cited by (0)

    View full text