Condensation and dehydration reactions of l-sorbose in ecofriendly melt systems

Article abstract of DOI:10.1080/00397911.2012.711882

L-Sorbose 1 was converted to 5-hydroxymethylfurfural 8 in choline chloride melts in moderate yields. A simplified synthesis of sorbosylurea tetraacetate 4 in two steps using urea melts was performed, achieving good yields. Copyright

Full text of DOI:10.1080/00397911.2012.711882

Synthetic Communications1, 43: 2452–2456, 2013
Copyright # Taylor & Francis Group, LLC
ISSN: 0039-7911 print=1532-2432 online
DOI: 10.1080/00397911.2012.711882
CONDENSATION AND DEHYDRATION REACTIONS
OF L-SORBOSE IN ECOFRIENDLY MELT SYSTEMS
Carolin Ruß,1 Alireza Haji Begli,2 and Burkhard Koenig1
1Institute of Organic Chemistry, University of Regensburg,
Regensburg, Germany
2
¨
SUDZUCKER AG Mannheim=Ochsenfurt, Central Department Research,
Entwicklung, Services (CRDS), Obrigheim, Germany
GRAPHICAL ABSTRACT
Abstract L-Sorbose 1 was converted to 5-hydroxymethylfurfural 8 in choline chloride melts
in moderate yields. A simplified synthesis of sorbosylurea tetraacetate 4 in two steps using
urea melts was performed, achieving good yields.
Supplemental materials are available for this article. Go to the publisher’s online edition
of Synthetic Communications1 to view the free supplemental file.
Keywords 5-Hydroxymethylfurfural; low melting mixtures; L-sorbose; sorbosylurea
INTRODUCTION
In 2011, a review by Zebiri et al. criticized the fact that the use of L-sorbose in
organic chemistry is exceedingly poor, and they encouraged the development of new
synthetic methods.[1] For this reason, reactions of L-sorbose in the unconventional
carbohydrate melts developed in our working group were examined.[2–4] Low-
melting mixtures allow the conversion of biomass at high concentration into
valuable intermediates.
L-Sorbose is one of the rare natural L-sugars and is mainly used as a starting
material in the synthesis of vitamin C. Therefore, it is the most readily available
L-sugar on a large scale.[1,5] As 5-epimer of D-fructose, its structural similarity would
suggest a comparable reactivity. Both fructose and sorbose form a mixture of two
pyranoid 1 and two furanoid stereoisomers 3 (plus negligible amounts of acyclic
form 2) (Scheme 1). The composition is not only strongly dependent on the solvent
employed but also on the temperature.[6,7] Simple derivatization reactions with fruc-
tose often yield product mixtures.[5] In contrast to fructose, sorbose highly favors the
Received June 7, 2012.
Address correspondence to Burkhard Koenig, Institute of Organic Chemistry, University of
Regensburg, D-93040 Regensburg, Germany. E-mail: burkhard.koenig@chemie.uni-regensburg.de
2452

SORBOSE REACTION IN MELT
2453
Scheme 1. Pyranose 1 and furanose 3 form and acyclic form of L-sorbose 2.
a-pyranoid form in solution: In water, the a-pyranoid form amounts to 87–98% in a
temperature range of 25–85 ꢀC.[6] The benefit of sorbose is that reactions to
pyranosid derivatives proceed in a more uniform manner and allow better yields
than in the case of fructose.
We therefore investigated the reaction behavior of L-sorbose in four
carbohydrate melts: urea, dimethylurea, ethylene urea, and choline chloride melts.
Significant conversions were only observed in urea and choline chloride melts.
RESULTS AND DISCUSSION
From our former investigations we know that glycosyl ureas are formed in
carbohydrate–urea melts.[8] L-Sorbose also forms stable melts with urea, giving a
melting point of 79–80 ꢀC.
Under acidic conditions, approximately 30% of the L-sorbose were mainly con-
verted to sorbosyl urea 4 after 1.5 h reaction time at 80 ꢀC (estimated by 13C NMR)
(Scheme 2). Most of the starting material was recovered, but degradation to 5-hydro-
xymethyl furfural (5-HMF) and to brown humins was also observed, especially for
longer reaction times and at higher temperatures. The isolated N-(1,3,4,5-tetra-O-
acetyl-L-sorbopyranosyl)urea 4 has (2R)-configuration (a-anomer), which was
ascertained by nuclear Over hauser spectrascopy (NOESY) experiments. A second
fraction (approx. 67 wt% of the total yield) was an inseparable mixture presumably
composed of sorbosyl acetate, the b-anomer of sorbosyl urea 4, and the sorbosyl-
biuret, inferred from 13C NMR spectra and electrospray ionization–mass spec-
trometry (ESI-MS) measurements.
The synthesis of the peracetylated derivative 4 was reported by Tokuyama and
Katsuhara in 1966 (Scheme 3).[9] They obtained the peracetylated sorbosyl urea 4
(a -anomer and b-anomer in 15% yield each) by immediate ammonolysis of the
isocyanate 6 starting from 1,3,4,5-tetra-O-acetyl-L-sorbopyranosyl chloride 5.
Comparing the overall yields of both methods starting from unmodified
L-sorbose, the synthesis in the melt yielded 15% of product 4 over two steps while
Scheme 2. Synthesis of N-(1,3,4,5-tetra-O-acetyl-L-sorbopyranosyl)urea 4 in sorbose–urea melts under
acidic conditions.

2454
C. RUß, A. H. BEGLI, AND B. KOENIG
Scheme 3. Synthesis of N-(1,3,4,5-tetra-O-acetyl-L-sorbopyranosyl)urea 4 according to Tokuyama and
Katsuhara.[9]
Scheme 4. Synthesis of HMF 8 from L-sorbose 3 in a choline chloride melt.
the four-step procedure of Tokuyama et al. provided an overall yield of 6%. Accord-
ing to that publication, the transformation to the chloride 5 was performed in 32%
yield and the direct reaction via the isocyanate 6 afforded both anomers in 30% yield.
As no yield was given for the acetylation of L-sorbose, it was assumed that the
reaction proceeded with a maximum yield of 65%.[10]
Our previous investigations showed that choline chloride melts promote the
production of 5-HMF 8 from carbohydrates.[11] The production of 5-HMF 8 from
L-sorbose has already been patented in 1948 by Haworth (yield: 27%)[12] and was
later investigated in organic solvents catalyzed by lanthanide(III) ions (61.4%),[13]
in DMA-LiCl and dimethylsulfoxide (DMSO) by Binder et al. (37–60%),[14] and sub-
critical water by Asghari et al. (50%).[15] Other reaction conditions to make 5-HMF 8
from different sugars were recently reviewed.[16,17]
As expected, the only product from a choline chloride–sorbose melts was 5-HMF
8 (Scheme 4). After 1.5 h at 80 ꢀC under acidic conditions using Amberlyst 15 as catalyst,
the starting material was completely consumed and 24% of 5-HMF was obtained.
Under the same reaction conditions used by Ilgen et al. (100 ꢀC, 0.5 h, 10%
p-TsOH),[11] L-sorbose yielded only 11% of 5-HMF while 67% were obtained starting
from D-fructose. The lower 5-HMF yields starting from L-sorbose were previously
observed by Seri et al.[13] and Asghari et al.[15] They suggested that the different con-
figurations of the hydroxyl groups at the C-3 and the C-4 atom in L-sorbose impair
the dehydration reaction, leading to lower yields.
CONCLUSION
In summary, new reactions of L-sorbose were investigated using low melting
carbohydrate mixtures as an ecofriendly reaction medium. We showed that the
important intermediate 5-HMF can be produced from L-sorbose in choline chloride
melts under acidic conditions.
Furthermore, a new, time-saving synthesis of sorbosylurea tetraacetate 4 is
reported here. Using urea melts, we successfully simplified the literature four-step
synthetic route to a two-step procedure starting from unprotected L-sorbose, without
the need for expensive reagents and inconvenient reaction conditions. For the first

SORBOSE REACTION IN MELT
2455
time, the structure of the selectively formed a-pyranose form was proven by
two-dimensional NMR techniques. We introduced two new solvent-free systems
here based on nontoxic and renewable resources for the conversion of L-sorbose.
EXPERIMENTAL
L-Sorbose (0.6 g, 3.3 mmol), urea (1.4 g, 26.7 mmol), and NH4Cl (0.2 g,
3.7 mmol) were heated in a 10-mL reaction flask at 80 ꢀC until a clear melt was
formed. Amberlyst 15 (150 mg) was added, and the reaction was stirred for 1.5 h
at that temperature. Methanol was added, solids were filtered off, and the solvent
evaporated. Then, 6 mL (5.8 g, 74.3 mmol) of pyridine and 3.5 mL of Ac2O (3.8 g,
37.0 mmol) were added to the mixture, and the mixture was first stirred for 1 h in
the ultrasounic bath and then overnight at room temperature without sonication.
The reaction was stopped by pouring the solution into saturated NaHCO3 solution.
The aqueous phase was extracted three times with CH2Cl2. The combined organic
layers were washed twice with 2N HCl and twice with water, dried over MgSO4,
and evaporated. The pure product was crystallized from ethyl acetate, affording a
colorless powder (0.215 g, 0.55 mmol, 16%).
1H NMR (300 MHz, CDCl3): d (ppm) ¼ 2.02, 2.04, 2.08, 2.10 (s, 12 H, OAc), 3.75
3
2
(dd, 1 H, H-6), 4.01 (dd, 1H, J(H-5, H-60) ¼ 5.91 Hz, H-60), 4.03 (dd, 1H, J(H-1,
H-10) ¼ 12.36 Hz, 166.24 Hz, H-1), 4.58 (dd, J1 ¼ 12.33 Hz, J2 ¼ 166.24 Hz, 1H,
H-10), 5.00–5.09 (m, 1H, H-5), 5.24–5.40 (m, 5H, H-3, H-4, urea); 13C NMR (75 Hz,
CDCl3): d (ppm) ¼ 20.5, 20.6, 20.5, 20.7 (CH3-Ac), 59.9 (C-5), 62.0 (C-1), 68.3 (C-4),
=
69.0 (C-5), 70.4 (C-3), 85.3 (C-2), 156.6 (urea), 168.6, 169.8, 170.1, 170.2 (C O, Ac).
1H NMR (600 MHz, CDCl3): d (ppm) ¼ 2.01, 2.04, 2.07, 2.10 (4 ꢁ s, 4 ꢁ 3H,
OAc), 3.74 [‘‘dd’’, 2J(H-6ax, H-6eq) ¼ 10.98 Hz, 3J(H-6ax, H-5) ¼ 10.98 Hz, 1 H,
H-6ax], 4.00 [dd, J(H-5, H-60) ¼ 5.91 Hz, J(H-6ax, H-6eq) ¼ 11.07 Hz, 1H, H-6eq],
3
2
2
2
4.04 [d, 1H, J(H-1, H-10) ¼ 12.37 Hz, H-1], 4.58 [d, J(H-1, H-10) ¼ 12.37 Hz, 1H,
3
H-10], 5.02–5.06 (m, 1H, H-5), 5.26 [d, J(H-3, H-4) ¼ 9.87 Hz, 1H, H-3], 5.35–5.38
(m, 2H, H-4, urea); 13CNMR (75 Hz, CDCl3): d (ppm) ¼ 20.5, 20.6, 20.6, 20.7 (CH3-
Ac), 59.8 (C-5), 62.0 (C-1), 68.3 (C-4), 69.0 (C-5), 70.4 (C-3), 85.3 (C-2), 156.7 (urea),
=
168.6, 169.8, 170.1, 170.2 (C O, Ac); FT-IR (ATR): v (cmꢂ1) ¼ 3487, 3458, 3341,
1753, 1744, 1697, 1665, 1620, 1614, 1591, 1564, 1547, 1535, 1441, 1433, 1377, 1371,
1275, 1229, 1211, 1138, 1103, 1061, 1032, 908; mp: 192 ꢀC. ESI-MS: m=z (%) ¼ 407.9
20
(MNH4þ, 100), 431.9 (MHþ þ MeCN, 37), 390.9 (MHþ, 15).½aꢃD ¼ ꢂ37.395 (c 1.000,
23:5
chloroform) (a-anomer) (literature[9]: ½aꢃD ¼ ꢂ34.4 (c 0.973, chloroform).
SUPPORTING INFORMATION
General experimental information, NOE spectrum of compound 7, and syn-
thesis of HMF from sorbose can be found via the Supplementary Content section
of this article’s web page.
ACKNOWLEDGMENTS
We thank the Suedzucker AG and the Fachagentur fuer Nachwachsende
Rohstoffe for financially supporting this project.

2456
C. RUß, A. H. BEGLI, AND B. KOENIG
REFERENCES
1. Zebiri, I.; Balieu, S.; Guilleret, A.; Reynaud, R.; Haudrechy, A. Eur. J. Org. Chem. 2011,
2905–2910.
2. Imperato, G.; Eibler, E.; Niedermaier, J.; Koenig, B. Chem. Commun. 2005, 1170–1172.
3. Imperato, G.; Hoger, S.; Lenoir, D.; Koenig, B. Green Chem. 2006, 8, 1051–1055.
4. Imperato, G.; Vasold, R.; Koenig, B. Adv. Synth. Catal. 2006, 348, 2243–2247.
5. Lichtenthaler, F. W. Carbohydr. Res. 1998, 313, 69–89.
6. Angyal, S. J. Adv. Carbohydr. Chem. Biochem. 1984, 42, 15–68.
7. Schneider, B.; Lichtenthaler, F. W.; Steinle, G.; Schiweck, H. Liebigs Ann. Chem. 1985,
1985, 2443–2453
8. Russ, C.; Ilgen, F.; Reil, C.; Luff, C.; Haji Begli, A.; Koenig, B. Green Chem. 2011, 13,
156–161.
9. Tokuyama, K.; Katsuhara, M. Bull. Chem. Soc. Jpn. 1966, 39, 2728–2733.
10. Schlubach, H. H.; Graefe, G. Justus Liebigs Ann. Chem. 1937, 532, 211–227.
11. Ilgen, F.; Ott, D.; Kralisch, D.; Reil, C.; Palmberger, A.; Koenig, B. Green Chem. 2009,
11, 1948–1954.
12. Haworth, W. N.; Wiggins, L. F. Patent GB600871, 1948.
13. Seri, K.; Inoue, Y.; Ishida, H. Chem. Lett. 2000, 22–23.
14. Binder, J. B.; Cefali, A. V.; Blank, J. J.; Raines, R. T. Energy Environ. Sci. 2010, 3,
765–771.
15. Salak Asghari, F.; Yoshida, H. Ind. Eng. Chem. Res. 2006, 45, 2163–2173.
16. Rosatella, A. A.; Simeonov, S. P.; Frade, R. F. M.; Afonso, C. A. M. Green Chem. 2011,
13, 754–793.
17. Tong, X.; Ma, Y.; Li, Y. Appl. Catal. A-Gen. 2010, 385, 1–13.

Copyright of Synthetic Communications is the property of Taylor & Francis Ltd and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.