Reactivity of D-fructose and D-xylose in acidic media in homogeneousphases
Maxime B. Fusaro, Vincent Chagnault*, Denis PostelLaboratoire de Glycochimie, des Antimicrobiens et des Agroressources (LG2A), 33 rue Saint-Leu, F-80039 Amiens Cedex 1, France
a r t i c l e i n f o
Article history:Received 20 January 2015Received in revised form20 March 2015Accepted 23 March 2015Available online 31 March 2015
Keywords:CarbohydratesAcidic mediaHMFDehydration reactionConversion of monosaccharides
Chemistry development of renewable resources is a real challenge. Carbohydrates from biomass arecomplex and their use as substitutes for fossil materials remains difficult (European involvement on theincorporation of 20% raw material of plant origin in 2020). Most of the time, the transformation of thesepolyhydroxylated structures are carried out in acidic conditions. Recent reviews on this subject describehomogeneous catalytic transformations of pentoses, specifically toward furfural, and also the trans-formation of biomass-derived sugars in heterogeneous conditions. To complete these informations, theobjective of this review is to give an overview of the structural variety described during the treatment oftwo monosaccharides (D-Fructose and D-xylose) in acidic conditions in homogeneous phases. The reac-tion mechanisms being not always determined with certainty, we will also provide a brief state of the artregarding this.
To our knowledge, research about the reaction of carbohydratesin acidicmediadatesback to themid19thcenturyespeciallywith thereaction of cane sugar in dilute sulfuric or hydrochloric acid andheat.1 The first publications on this subject report the production ofheterocyclic aldehydes suchas furfural and5-hydroxymethylfurfural(5-HMF) by thermal treatment of carbohydrates (starting fromstarch or pure monosaccharides) under acidic conditions.2,3 Thesederivatives have various valorization areas covering applicationssuch as solvents, additives for fuels or as precursors for ‘buildingblocks’ in chemical synthesis.4,5 5-HMF is unfortunately not pro-ducedon industrial scale in contrast to furfural.However, furans suchas these are already within the top 10 bio-based building blocks.5
The important development of plant biorefinery (valorization ofthe whole plant) has among its main objectives, the rapid substi-tution of a part of the fossil molecules by bio-based chemicals(European involvement on the incorporation of 20% rawmaterial ofplant origin in 20206). Therefore, the behavior of carbohydrates inacidic media is still an important subject of research and many
. Chagnault).
studies aim to clarify mechanisms. Publications about access tolevulinic acid, furfural and 5-HMF but also to other small molecules,under various conditions, have seen their number grow exponen-tially (Fig. 1).
These studies focus mainly on the aspects of selectivity andconversion of monosaccharides to the desired product formation.8
The variety of products that is provided by these transformations ishowever much broader and a multitude of molecules have beenidentified.
The data concerning the formation of other structures by thetransformation of carbohydrates in acidic conditions are spreadover more than 170 years. Furthermore, these reports have oftenbeen the subject of verifications and modifications. Marcotullioet al. and Fatehi et al. have recently published reviews focused onthe specific production of furfural or hydroxymethylfurfural.8,9 Fortheir part, Zhang et al. have given an overview of the trans-formation of biomass-derived sugars using heterogeneous condi-tions.10 The objective of this review is to supplement theseinformations with an overview of the structural variety obtainedduring the treatment of D-fructose (hexose) and D-xylose (pentose)
Fig. 1. Number of publications dealing with A) Levulinic acid, B) Furfural, C) 5-HMF.7
Fig. 2. Example of structures of humic segments described in the literature.27,28
M.B. Fusaro et al. / Carbohydrate Research 409 (2015) 9e1910
in acidic conditions in homogeneous phases. The reaction mecha-nisms being not always determined with certainty, we will alsoprovide a brief state of the art regarding this.
2. Reaction of carbohydrates in acidic media
Whatever the medium used, characterization of the productsformed by acidic pathways remains difficult. Only few methodsallow monitoring their appearance and their in situ characteriza-tionwithout prior treatment to decrease or neutralize the acidity ofthe reaction mixture. This is even more true when the literature isold, analytical techniques then being limited. Over the years,studies have confirmed or disproved molecules already describedand helped to identify new ones. Dilute acids are most commonlyused to study the reactivity of carbohydrates in acidic media. Takinginto account researches described in the literature, experimentalconditions are rarely similar and it is not easy to compare the re-sults obtained, even for the same substrate. In fact, each study hasdifferent conditions of temperature, sugar concentration, solvent,reaction time, pressure or acidity. The first compounds resultingfrom the structural modification of a monosaccharide in an acidicmedium were described in 1840 by Stenhouse,2 Mulder1 and in1895 by Düll.11 These are the most stable and easily isolable com-pounds: furfural, levulinic acid and 5-HMF.
Since that time, the development of analytical techniques andinterest in the valorization of bio-based compounds have bothexpanded the range of identified molecules and also proposedmechanisms of transformation that may involve different in-termediates and pathways reaction. However, an exhaustive censusremains difficult.12 We will focus herein on the description of thestructures obtained in the most used acidic conditions, startingfrom a hexose and a pentose (D-fructose and D-xylose). The mech-anisms proposed to explain their formation are from four types ofreactions: isomerizations, dehydrations, fragmentations andcondensations.
Table 1Products obtained from the reaction of D-fructose in dilute acidic media according to the
2.1.1. Structures obtainedD-Fructose remains the predominant hexose for which reac-
tivity in acidic media has been studied and It can be obtainedindustrially in large quantity by enzymic conversion of glucosefrom corn starch. It proves to be a prime substrate to access 5-HMF. However, many studies have shown that this compound isfar from being the only molecule formed in such conditions. Moststructures which can be obtained appear in references in the workof Antal et al.,13 Dumesic et al.14 and more recently De Jong,
Scheme 1. Set of products resulting from the reaction of D-fructose in dilute acidic media and for which a reaction mechanism is described (A: Horv�ath,12 B: Mednick,31 C:Bobleter,21 D: Antal,32 E: Antal,13 F: C€ammerer,33 G: Anet,34 H: Yoshida,35 I: Brown,26 J: Moreau,16 K: Feather,36 L: Ferrier,37 M: Van Bekkum,38 N: Civelekoglu,39 O: Kiermayer,40 P:Lund,30 Q: Van Bekkum,19 R: Tatum,17 S: Hodge,25 T: Tatum18).
M.B. Fusaro et al. / Carbohydrate Research 409 (2015) 9e19 11
Scheme 2. Formation of 5-HMF via acyclic intermediates, according to Moreau et al.16
Scheme 3. Products formed from oxocarbenium ion
M.B. Fusaro et al. / Carbohydrate Research 409 (2015) 9e1912
Heeres, De Vries et al.15 According to the authors, they can beclassified into categories related by the reactions involved in theirformation (Table 1).
Humins comprise a mixture of highly colored insoluble mole-cules, with an oligomeric or polymeric structure.27,29 The analysis oftheir structures is still the subject of research28,30 as well as theintermediates involved in their formation (Fig. 2).
Scheme 1 includes a compilation of all the compounds obtainedduring the reaction of D-fructose in an acidic medium and for whicha mechanism has been proposed. It consists of four distinct partsaccording to the molecules resulting from 2,3-enediol, 1,2-enediol,oxocarbenium ions and retroaldolisation reactions.
2.1.2. Mechanisms of formationThe formation of 5-HMF has been the subject of many studies.
Two mechanisms are described, via intermediates of the linearenediol type or via a cyclic oxocarbenium ion.
The formation of this compound has been reported by Anet,34
Tatum et al.17 et Moreau et al.16 who described 1,2- and 2,3-enediol as reaction intermediates (Scheme 2). The enediol systemis obtained by the Lobry De BruyneAlberda Van Ekenstein trans-formation, usually conducted under basic conditions, which helpsto explain the enolization of aldoses to ketoses and conversely.41
However, these enediols can also be formed in an acidic mediumas reported by the works of Speck.42 Regardless of the enediolformed, it is followed by a dehydration involving the carbon atomC-4, and the formation of an additional double bond. A last
s using D-fructose (A: Horv�ath,12 B: Mednick31).
Fig. 3. Structures of the difructose dianhydrides (DFAs). a Old notation system. b D-fructose-D-glucose mixed dianhydrides.
M.B. Fusaro et al. / Carbohydrate Research 409 (2015) 9e19 13
cyclization step involves the site either 3,6 (leading to 2-(2-hydroxyacetyl)furan) or 2,5 (leading to 5-HMF). These enediol in-termediates can also lead to fragmentations by retroaldolisationand enable access to furfural.17
2.1.2.1. Mechanisms involving intermediates of oxocarbenium type.Various studies (Feather,36 Antal et al.,13 Horv�ath et al.12) report the
involvement of oxocarbenium ions in the formation of compoundsincluding furfural derivatives, anhydrides or humins. These authorsemphasize the contribution of cyclic intermediates rather than asequence implying acyclic structures by relying on studies ofdehydration of D-fructose in D2O that highlight less than 1% ofdeuterium integration in 5-HMF at the C-3 position. This resultcasts doubt on mechanisms involving linear intermediates. As the
Fig. 4. Different trisulfated compounds obtained by the reaction of concentrated sul-furic acid on free carbohydrates at �5 �C during 2.5 h.
Scheme 4. General scheme proposed by García Fern�andez et al. explaining the DFAsformation.46
M.B. Fusaro et al. / Carbohydrate Research 409 (2015) 9e1914
existing equilibrium between 3-deoxyglucos-2-ene and the 3-deoxyglucosone (Scheme 2)43 requires, by tautomerism, theincorporation of a deuterium atom in position C-3. Similarly, thepassage through the 1,2-enediol should lead to the incorporation ofdeuterium at C-1. The low percentage of deuterium incorporatedtends to favor the involvement of cyclic intermediates (Scheme 3).These results have been corroborated by in situ NMR analyzes.12
In summary, two different pathways may be advanced whichinvolve oxocarbenium intermediates based on a scaffold eitherpyranose (17) or furanose (18). These can explain the formation ofthe major products identified (5-HMF; humins; 2,6-anhydro-b-D-fructofuranose 19; difructose dianhydrides).
Even if these new results12 tend to confirm the pathwayimplying cyclic intermediates, no mechanism has been definitivelyrefuted. To date, both are accepted by the scientific community.15
Difructose dianhydrides (DFAs) are spiroketals resulting fromthe cyclodimerization of two molecules of D-fructose. They wereinitially observed (at the beginning of the 1920s) by reaction of D-fructose44 or inulin45 (polysaccharide mixture mainly consisting offructose units) in concentrated hydrochloric acid. However, theyields obtained in these conditions remain modest due toconcomitant degradation reactions. In addition, the purificationeven partial of the mixture took several years and involved severalresearch groups. Even if a slight improvement, in term of yields,was achieved using trifluoroacetic acid,46 purification of thesecompounds remained extremely complex. This difficulty makes
Chart 1. DFAs relative abundance in a crude reaction mixture resulting from the re-action of Lewatit® S2328 (Hþ) on the D-fructose, at 90 �C.50
problematic their identification, quantification and evaluation oftheir biological properties.
Some of these compounds (16 known in the literature, Fig. 3)were identified few years later as natural molecules produced byplants47,48 (like for DFAs 1,10 et 15) or bymicroorganisms49 (DFAs 1,10, 15 et 16). They are found in artichokes and Jerusalem arti-chokes,48 chicory and roasted coffee beans, tequila,50 and especiallyin commercial caramels where they represent up to 18% byweight.51
One of the most effective synthetic methods is the use of Lew-atit® S2328 (Hþ), a strongly acidic ion exchange resin having sul-fonic acid functional groups;50 García Fern�andez's coworkers havebeen able to get a more accurate analysis of the resultant mixtureusing GC/FID (Chart 1).
Moreover, it is the first description (which still constitutes areference) of a precise relative distribution of DFAs obtainedstarting from D-fructose. However, a different relative distributionhas been observed during thermal treatment of inulin.52 Defaye,García Fern�andez et al. have proposed a general scheme for theDFAs formation (Scheme 4).
This mechanism goes through the activation of D-fructoseresulting in the formation of an oxocarbenium ion (step A), whichwill lead to fructodisaccharides after dimerization (step B). DFAsare then obtained by intramolecular glycosylation (step C). Whilethe mechanism may seem simple, many factors can influence themixture composition.50
2.1.2.2. Retroaldolization products. Whatever the favored mecha-nism for the reactivity of oses in acidic conditions, they can also
M.B. Fusaro et al. / Carbohydrate Research 409 (2015) 9e19 15
lead to low molecular weight products resulting from retro-aldolization reactions. The retroaldolization products obtainedstarting from D-fructose are described in Scheme 5. These com-pounds can be detected at the end of the reaction as final products,or, as reported by C€ammerer et al., be involved in the formation of5-HMF by aldolization.33
2.2. Example of a pentose: D-xylose
2.2.1. Structures obtainedWhile D-fructose remains the hexose for which the reactivity in
acidic media has been the most studied, it is the reactivity of D-xylose which has been widely explored in the pentose series. Infact, this allows an easy access to furfural. However, this is not theonly product to be formed under these acidic conditions and almostall of the compounds obtained were identified by Antal et al.53 Aswith D-fructose, these products can be divided into several cate-gories depending on reactions allowing their formation (Table 2).
Scheme 6 summarizes all the molecules obtained by the reac-tion of D-xylose in diluted acidic media and for which a mechanism
Table 2Products obtained from the reaction of D-xylose in diluted acidic media, according to An
has been suggested. It consists of three distinct parts, from acyclicintermediates, cyclic intermediates and retroaldolization reactions.
2.2.2. Mechanisms of formationThe formation of furfural has been the subject of many discus-
sions. Two mechanisms are proposed: from open chain (enedioltype) or cyclic intermediates.
2.2.2.1. Mechanisms involving open chain intermediates (enedioltype). The production of furfural involving an 1,2-enediol inter-mediate has been proposed in several works, including those ofIsbell56 and Anet34 (Scheme 7). Once again, formation of this in-termediate is initiated by the transformation of Lobry DeBruyneAlberda Van Ekenstein. This is followed by a double dehy-dration involving sites 3 and 4, adding two additional conjugatedunsaturations, and leading to intermediate 20. Finally, a cycload-dition stepwill allow the formation of furfural (involvement of sites2,5) or of 2,3-dihydroxycyclopent-2-ene-1-one (involvement ofsites 1,5).
tal's review
Fragmentation Condensation
- Formaldehyde55
- Formic acid- Acetaldehyde55
- Crotonaldehyde55
- Lactic acid53
- Dihydroxyacetone53
- Glyceraldehyde53
- Pyruvaldehyde53
- ‘Humins’
Scheme 6. Set of products resulting from the reaction of D-xylose in dilute acidic media and for which a reaction mechanism is described (A: Isbell,56 B: Antal,53 C: Zeitsch,57 D:Bobleter,21 E: Antal32).
M.B. Fusaro et al. / Carbohydrate Research 409 (2015) 9e1916
2.2.2.2. Mechanisms implying cyclic intermediates. Several studiesreported in the literature (Antal et al.,53 Zeitsch57) have examinedthe involvement of cyclic intermediates in the formation offurfural (Scheme 8). According to these works, two differentpathways involving cyclic intermediates can explain the formationof furfural. The first pathway (B) is also divided into two parts, oneof which goes through a cyclic oxocarbenium ion, both leading tothe formation of 2,5-anhydroxylose (compound 21) via an intra-molecular rearrangement. Furfural is then obtained after a doubledehydration. The second pathway (C), suggested in 2000 byZeitsch, goes through a successive formation of a pyranosidicoxocarbenium ion, a ring opening, a serie of protonations/de-hydrations and a final cyclization leading to furfural. However,according to a study carried out on D-xylose-1-14C, the labeledcarbon is found almost exclusively at the carbonyl of furfural,which is in contradiction with pathway C (for which the carbonylis in position C-5).58
2.2.2.3. Retroaldolization products. As for D-fructose, D-xylose indiluted acidic media can lead to low molecular weight productsresulting from retroaldolization reactions (Scheme 9).
As we have seen, both with D-fructose and D-xylose, thestructure and the mechanisms of formation of certain com-pounds are still not understood. For some other products, their
reaction mechanisms have been suggested before the productshave been observed experimentally. Thus, if Antal et al.13 assumethe formation of a tetrose, coproduct of the D-fructose retro-aldolization leading to glycolaldehyde, the confirmation of thishypothesis and the identification of D-erythrose, were donethanks to the works of Yoshida et al.,35 sixteen years later.Similarly, while epimerization of D-xylose to D-lyxose is sug-gested in the Anet mechanism,34 it has been detected onlytwenty-five years later by Antal.53 Its ketopentose form (D-xylulose) has not yet been identified.
2.2.2.4. A special case: nitric acid. Finally, it is noted that nitric acid,one of the commonmineral acids, is not studied in the case of sugardecomposition in an acidic medium. Indeed, the acid property ofthis one is in competition with its oxidizing power. Thus, in 1888,Sohst and Tollens showed that it is possible to form D-mannaric acidfrom D-mannose using nitric acid as an oxidizer (Scheme 10).59
Fischer did the same with D-glucaric acid from D-glucose.60
Their work has been taken up by Kiely et al. to improve theprotocols used, and make them more attractive from a commercialpoint of view.61,62
2.2.2.5. Another special case: sulfuric acid. In addition to self-condensation products, different authors have reported sulfation
M.B. Fusaro et al. / Carbohydrate Research 409 (2015) 9e19 17
reactions during the treatment of D-glucose by cold sulfuric acid.Turvey in 196563 and Takiura et al. in 1970,64 report the productionof variously substituted compounds. The first one mentioned thepossibility of obtaining a complex mixture of monosulfated (at C-6)and polysulfated D-glucose by dissolving the monosaccharide inconcentrated sulfuric acid at 0 �C. Takiura et al. describe the pro-duction of mono-, di- and trisulfated D-glucose by diluting D-glucose in concentrated sulfuric acid at �5 �C. The ratio betweenthe various final products depends on the reaction time. Theexclusive formation of the D-glucose 1,3,6-trisulfate (compound 22,Fig. 4) is reached after 2.5 h. The D-mannose 1,3,6-trisulfate (com-pound 23), D-galactose 1,3,6-trisulfate (compound 24) and the D-fructose 1,2,4-trisulfate (compound 25) have been obtained in thesame conditions.
Other authors, in the example of Nagasawa et al., also reportedthe formation of sulfated oligosaccharides during polymerization ofmonosaccharides in concentrated sulfuric acid.65 Nevertheless,
they have not been able to determine an exact structure for any ofthese polymers.
3. Conclusion
Carbohydrate chemistry is a research field that is still of in-terest. Environmental challenges are such important that it isnecessary to find alternatives to the scarcity of fossil substances.Only a few of molecules resulting from the conversion of sugars inacidic media can be produced on an industrial scale. However,these reactions enable access to a very important structural vari-ety. This allows envisaging the production of a whole range ofreagents that could substitute or complete those obtained fromnon-renewable resources. However further studies are neededbefore exploiting the structures obtained from biomass. Thesescientific advances must result in a better understanding of the
Scheme 10. Oxidation of D-mannose and D-glucose with nitric acid.
M.B. Fusaro et al. / Carbohydrate Research 409 (2015) 9e1918
mechanisms involved in the conversion of carbohydrates intoacidic environments.
Author contributions
All authors have given approval to the final version of themanuscript.
M.B. Fusaro et al. / Carbohydrate Research 409 (2015) 9e19 19
Acknowledgments
We thank CNRS (France) for financial support and the RegionPicardie and Europe (FEDER) for a grant (EROMS Project, to M. B.Fusaro).
References
1. Mulder GJ. J Prakt Chem 1840;21:203e40.2. Stenhouse J. Justus Liebigs Ann Chem 1840;35:301e4.3. Maillard LC. Gen�ese des mati�eres prot�eiques et des mati�eres humiques: action de la
glyc�erine et des sucres sur les acides -amin�es. Masson et Cie: Paris; 1913.4. Van Putten R-J, Dias AS, de Jong E. In: Catalytic process development for
5. Bozell JJ, Petersen GR. Green Chem 2010;12:539e54.6. European Commission 2020. A strategy for smart, sustainable and inclusive
growth. Brussels COM(2010)7. Graphic from Web of Knowledge in 2013.8. Danon B, Marcotullio G, de Jong W. Green Chem 2014;16:39e54.9. Dashtban M, Gilbert A, Fatehi P. RSC Adv 2014;4:2037e50.
10. Guo X, Yan Y, Zhang Y, Tang Y. Prog Chem 2013;25:1915e27.11. Düll G. Chem Ztg 1895;19:216.12. Akien GR, Qi L, Horv�ath IT. Chem Commun 2012;48:5850e2 (Cambridge, U. K.).13. Antal Jr MJ, Mok WSL, Richards GN. Carbohydr Res 1990;199:91e109.14. Rom�an-Leshkov Y, Chheda JN, Dumesic JA. Science 2006;312:1933e7.15. van Putten R-J, van der Waal JC, de Jong E, Rasrendra CB, Heeres HJ, de Vries JG.
Catal A 1996;145:211e24.17. Shaw PE, Tatum JH, Berry RE. Carbohydr Res 1967;5:266e73.18. Shaw PE, Tatum JH, Berry RE. Carbohydr Res 1971;16:207e11.19. Luijkx GCA, van Rantwijk F, van Bekkum H. Recl Trav Chim Pays Bas 1991;110:
343e4.20. Dunlop AP, Peters FN;. The furans, vol. 119. New York: Reinhold Pub. Corp.;
1953.21. Bonn G, Bobleter O. J Radioanal Chem 1983;79:171e7.22. Amin ES. Carbohydr Res 1967;4:96e8.23. Rice FAH, Fishbein L. J Am Chem Soc 1956;78:3731e4.24. Manley-Harris M, Richards GN. Adv Carbohydr Chem Biochem 1997;52:207e66.25. Goodwin JC, Hodge JE, Weisleder D. Carbohydr Res 1986;146:107e12.26. Brown HT, Millar JH. J Chem Soc Trans 1899;75:286e308.27. Chuntanapum A, Matsumura Y. Ind Eng Chem Res 2009;48:9837e46.28. Hoang TMC, van Eck ERH, Bula WP, Gardeniers JGE, Lefferts L, Seshan K. Green
Chem 2015;17:959e72.
29. Herzfeld J, Rand D, Matsuki Y, Daviso E, Mak-Jurkauskas M, Mamajanov I. J PhysChem B 2011;115:5741e5.
30. Patil SKR, Heltzel J, Lund CRF. Energy Fuels 2012;26:5281e93.31. Mednick ML. J Org Chem 1962;27:398e403.32. Antal Jr MJ, Mok WSL, Richards GN. Carbohydr Res 1990;199:111e5.33. C€ammerer B, Wedzicha BL, Kroh LW. Eur Food Res Technol 1999;209:261e5.34. Anet EFLJ. Adv Carbohydr Chem 1964;19:181e218.35. Salak Asghari F, Yoshida H. Ind Eng Chem Res 2006;45:2163e73.36. Feather MS, Harris JF. Adv Carbohydr Chem Biochem 1973;28:161e224.37. Collins P, Ferrier R. Monosaccharides: their chemistry and their roles in natural
products. West Sussex, England: Wiley; 1995.38. van Dam HE, Kieboom APG, van Bekkum H. Starch - St€arke 1986;38:95e101.39. Von Helberger JH, Ulubay S, Civelekoglu H. Justus Liebigs Ann Chem 1949;561:
215e20.40. Kiermayer J. Chem Ztg 1895;19:1003e5.41. de Bruyn CAL, van Ekenstein WA. Recl Trav Chim Pays-Bas 1895;14:203e16.42. Speck Jr JC. Adv Carbohydr Chem 1958;13:63e103.43. ten Dam J, Hanefeld U. ChemSusChem 2011;4:1017e34.44. Pictet A, Chavan J. Helv Chim Acta 1926;9:809e14.45. Jackson RF, Goergen MBS. J Res 1929;117:1481.46. Defaye J, Gadelle A, Pedersen C. Carbohydr Res 1985;136:53e65.47. Uchiyama T. Agric Biol Chem 1983;47:437e9.48. Schlubach HH, Knoop H. Justus Liebigs Ann Chem 1933;504:19e30.49. Kawamura M, Uchiyama T. In: Norio S, Noureddine B, Onodera S, editors.
Recent advances in fructooligosaccharides research. Kerala, India: ResearchSignpost; 2007. p. 273e96.
51. Defaye J, García Fern�andez JM. Carbohydr Res 1994;256:C1e4.52. Christian TJ, Manley-Harris M, Field RJ, Parker BA. J Agric Food Chem 2000;48:
1823e37.53. Antal Jr MJ, Leesomboon T, Mok WS, Richards GN. Carbohydr Res 1991;217:
71e85.54. Reichstein T, Oppenauer R. Helv Chim Acta 1933;16:988e98.55. Rice FAH, Fishbein L. J Am Chem Soc 1956;78:1005e9.56. Isbell HS. J Res Nat Bur Stand 1944;33:45e61.57. Zeitsch KJ. In: Karl JZ, editor. Sugar series, vol. 13. London: Elsevier; 2000.
p. 3e7.58. Bonner WA, Roth MR. J Am Chem Soc 1959;81:5454e6.59. Sohst O, Tollens B. Justus Liebigs Ann Chem 1888;245:1e27.60. Fischer E. Ber Dtsch Chem Ges 1891;24:539e46.61. Smith TN, Hash K, Davey C-L, Mills H, Williams H, Kiely DE. Carbohydr Res
2012;350:6e13.62. Carpenter CA, Hardcastle KI, Kiely DE. Carbohydr Res 2013;376:29e36.63. Turvey JR. Adv Carbohydr Chem 1965;20:183e218.64. Takiura K, Yuki H, Honda S, Kojima Y, Chen L. Chem Pharm Bull 1970;18:
429e35.65. Nagasawa K, Tohira Y, Inoue Y, Tanoura N. Carbohydr Res 1971;18:95e102.
