Fast pyrolysis of 13c-labeled cellobioses: Gaining insights into the mechanisms of fast pyrolysis of carbohydrates

Article abstract of DOI:10.1021/jo5025255

A fast-pyrolysis probe/tandem mass spectrometer combination was utilized to determine the initial fast-pyrolysis products for four different selectively ¹³C-labeled cellobiose molecules. Several products are shown to result entirely from fragmentation of the reducing end of cellobiose, leaving the nonreducing end intact in these products. These findings are in disagreement with mechanisms proposed previously. Quantum chemical calculations were used to identify feasible low-energy pathways for several products. These results provide insights into the mechanisms of fast pyrolysis of cellulose.

Full text of DOI:10.1021/jo5025255

Note
pubs.acs.org/joc
13
Fast Pyrolysis of C‑Labeled Cellobioses: Gaining Insights into the
Mechanisms of Fast Pyrolysis of Carbohydrates
John C. Degenstein,† Priya Murria,‡ Mckay Easton,† Huaming Sheng,‡ Matt Hurt,‡ Alex R. Dow,‡
Jinshan Gao,‡,§ John J. Nash,‡ Rakesh Agrawal,† W. Nicholas Delgass,† Fabio H. Ribeiro,†
and Hilkka I. Kenttamaa*,‡
̈
‡
†School of Chemical Engineering and Department of Chemistry, Purdue University, West Lafayette, Indiana 47907, United States
S
* Supporting Information
ABSTRACT: A fast-pyrolysis probe/tandem mass spectrometer combination was utilized to determine the initial fast-pyrolysis
products for four different selectively 13C-labeled cellobiose molecules. Several products are shown to result entirely from
fragmentation of the reducing end of cellobiose, leaving the nonreducing end intact in these products. These findings are in
disagreement with mechanisms proposed previously. Quantum chemical calculations were used to identify feasible low-energy
pathways for several products. These results provide insights into the mechanisms of fast pyrolysis of cellulose.
ast pyrolysis is an efficient method for converting biomass
the initial fast-pyrolysis products of cellobiose and four
Fto a low energy-density liquid (bio-oil) that can be further selectively 13C-labeled cellobioses: commercially available
upgraded for use as fuel.1 Optimization of fast pyrolysis to
maximize the yields of compounds containing six or more
carbons, which represent some of the most valuable potential
end products, requires detailed knowledge of the dominant
pyrolysis reactions.2 However, the pathways and mechanisms
(e.g., stepwise radical or ionic or concerted) of fast pyrolysis
reactions are still largely unknown even for cellulose, the
simplest component of biomass, as well as for analogous di- and
oligosaccharides.2,3 The goal of this study was to explore these
mechanisms by using selective 13C isotope labeling since this
has not been performed previously.
The inherent capability of mass spectrometers to separate
ions that differ in their mass-to-charge ratios (such as those
with and without a 13C label) makes the combination of this
technique with fast pyrolysis of selectively labeled carbohy-
drates a powerful approach for mechanistic studies. We report
here the results obtained upon examination of the initial fast-
pyrolysis products of unlabeled cellobiose (a glucose dimer
with the same linkage as in cellulose; see Figure 1) and four
selectively 13C-labeled cellobioses by tandem mass spectrom-
etry. It should be noted that “initial” products are defined here
as the products that first leave the heated surface of the fast-
pyrolysis probe. The experimental studies were complemented
by extensive quantum chemical calculations.
[1-13C]glucopyranosylglucose and glucopyranosyl[1-13C]-
glucose (Omicron Biochemicals, South Bend, IN) as well as
glucopyranosyl[3-13C]glucose and glucopyranosyl[5-13C]-
glucose (synthesized according to literature procedures5,6).
The pyrolysis probe uses a resistively heated flat platinum
ribbon (2.1 mm × 35 mm × 0.1 mm), which was placed inside
the atmospheric-pressure chemical ionization (APCI) source of
the LQIT. The ribbon was heated to 600 °C at a rate of 1000
°C s−1. The initial products of pyrolysis evaporated from the
heated surface into a nitrogen atmosphere at 100 °C (at
atmospheric pressure) in the ion source and were quenched (it
should be noted that this approach is very different from that
used in laboratory pyrolysis reactors7). A soft ionization
method, APCI with ammonium hydroxide dopant, was utilized
to ionize the products since this method typically generates4
+
only one ion (either the NH4 adduct or the protonated
molecule) without fragmentation for each product molecule of
the type of interest here. The reactor and analysis suites were
designed to prevent fragmentation of the pyrolysis products.
Mass balances cannot be obtained using this new method-
ology;4 however, relative abundances4 of the products were
determined, and much larger pyrolysis products were detected
than reported previously.8,9 Each reported mass spectrum is a
result of a weighted average (based on total ion current) of the
individual mass spectra measured during pyrolysis. No
significant changes were detected in the product distribution
The pyrolysis/tandem mass spectrometry technique utilized
here has been described in detail in the literature.4 It combines
a pyrolysis probe that can be heated very fast (up to 20 000 °C
s−1) with a Thermo Scientific linear quadrupole ion trap
(LQIT) mass spectrometer. This technique was used to study
Received: November 4, 2014
Published: January 6, 2015
© 2015 American Chemical Society
1909
DOI: 10.1021/jo5025255
J. Org. Chem. 2015, 80, 1909−1914

The Journal of Organic Chemistry
Note
Figure 1. Positive ion-mode mass spectra showing the initial fast-pyrolysis products (either as ammonium adducts or protonated molecules) of (a)
glucopyranosyl[1-13C]glucose and (b) unlabeled cellobiose ionized by APCI with ammonium hydroxide dopant. The structures of the intact
molecules are shown at the far right in each spectrum. The products that are 13C-labeled in the top spectrum are connected with red dotted lines to
the corresponding unlabeled products in the lower spectrum.
during cellobiose pyrolysis; in other words, the relative
abundances were constant during pyrolysis.
MS methodologies because of their thermal instability and low
volatility, which may at least partially explain the incomplete
mass balance of previous oligosaccharide pyrolysis studies.8,9
The results presented here are in agreement with recent
mechanistic studies10−12 suggesting that glycolaldehyde may
form directly from cellobiose (or a longer glucosaccharide)
rather than through a glucose intermediate.
The structures of most of the ions were examined by using
MS2 experiments (i.e., by isolating them and subjecting them to
collision-activated dissociation (CAD)). When necessary, the
structures of the fragment ions were examined by isolating
them and subjecting them to further CAD (MS3 experiment).
High-resolution data needed to determine the elemental
compositions of the ions were collected using an LQIT/
Fourier-transform ion cyclotron resonance mass spectrometer
coupled with the pyrolysis probe as described above.
The ionized initial fast-pyrolysis products of glucopyranosyl-
[1-13C]glucose (with the label at the reducing end or the end
containing a free hydroxyl group on the carbon adjacent to the
ring oxygen) are shown in the mass spectrum in Figure 1a.
Comparison of these products to those produced from
unlabeled cellobiose (Figure 1b) revealed that only one
The initial fast pyrolysis products of unlabeled cellobiose are
shown in Figure 1b, and they are in agreement with those
reported previously.4 The identified products include hydrox-
ymethylfurfural (protonated molecule of m/z 127), levogluco-
san (NH4+ adduct of m/z 180 and a protonated molecule of m/
product is partially 13C-labeled (levoglucosan, NH4 adducts
+
of m/z 180 and 181), while all of the others are either
completely labeled or do not contain the label at all.
+
+
z 163), glucose (NH4 adduct of m/z 198), and β-D-
Specifically, cellobiosan (NH4 adduct of m/z 343) and the
+
+
glucopyranosylglycolaldehyde (NH4 adduct of m/z 240). β-
product formed via loss of water from cellobiosan (NH4
D-Glucopyranosylglycolaldehyde is likely formed upon the loss
of two glycolaldehyde (or isomeric) molecules from cellobiose
upon fast pyrolysis. On the other hand, the pyrolysis product
adduct of m/z 325) are completely 13C-labeled, as expected,
while all of the other products are unlabeled (Figure 1a). For
example, glucose, β-D-glucopyranosylglycolaldehyde and β-D-
glucopyranosylerythrose contain intact rings that originate
exclusively from the nonreducing end since they have no label.
The formation of glucose from the nonreducing end of
cellobiose must occur via scission of the aglyconic bond (i.e.,
the other C−O bond of the oxygen atom involved in the
glycosidic linkage) rather than the glycosidic bond as previously
proposed.3,13 In the literature, the only mechanisms that explain
formation of glucose involve glycosidic bond cleavage and
thermohydrolysis, which form glucose either from the reducing
end only or from the reducing and nonreducing ends in equal
amounts, respectively.3,14,15 On the contrary, the present results
indicate that the formation of glucose occurs exclusively from
the nonreducing end of cellobiose (within the detection limits
of our pyrolysis/MS experiment). This observation suggests
that there is at least one additional glucose formation pathway
that has not been proposed in the literature. Work is underway
+
generating ions of m/z 300 (NH4 adduct) is formed via the
loss of a single glycolaldehyde (or isomeric) molecule from
+
cellobiose, and the product generating ions of m/z 282 (NH4
adduct) is formed via the loss of glycolaldehyde (or isomeric)
molecule and water from cellobiose. Evidence in support of
these products being initial products instead of secondary
products was obtained from the previously determined4
structure of the β-D-glucopyranosylglycolaldehyde product.
The glycosidic bond (i.e., the C−O bond at the oxygen atom
on the carbon adjacent to the ring oxygen) in this pyrolysis
product is still in the same position and has the same
stereochemistry (β-1) as in cellobiose. If β-D-glucopyranosyl-
glycolaldehyde were not an initial product but instead the result
of recombination of smaller initial products, one would expect a
mixture of linkage positions and stereochemistry. These larger
products (>162 Da) are typically not seen with pyrolysis/GC−
1910
DOI: 10.1021/jo5025255
J. Org. Chem. 2015, 80, 1909−1914

The Journal of Organic Chemistry
Note
Figure 2. Positive ion-mode mass spectra showing the initial fast-pyrolysis products (either as ammonium adducts or protonated molecules) of (a)
[1-13C]glucopyranosylglucose and (b) glucopyranosyl[3-13C]glucose ionized by APCI with ammonium hydroxide dopant. The structures of the
intact molecules are shown at the far right in each spectrum. The only product that is labeled in the top spectrum but unlabeled in the bottom
spectrum is glucose, as indicated by a red dotted line. The ion labeled with * corresponds to an unknown impurity.
+
to explore alternate reaction pathways that explain formation of
glucose from the nonreducing end.
reducing end of cellobiose (to yield the NH4 adduct of m/z
240 for the unlabeled cellobiose), a third 13C-labeled cellobiose,
glucopyranosyl[3-13C]glucose, was synthesized. The ionized
initial fast-pyrolysis products of this compound are shown in
the mass spectrum in Figure 2b. The presence of a major ion of
m/z 301 (containing 13C) supports the above finding that the
first glycolaldehyde (or isomeric) molecule eliminated from
cellobiose upon pyrolysis contains C-1 and C-2 of the reducing
end (i.e., to form the ion of m/z 301). Furthermore, since
elimination of the second glycolaldehyde (or isomeric)
molecule yields an ion of m/z 241 (containing 13C), this
glycolaldehyde unit must contain C-5 and C-6 of the reducing
end. Examination of a fourth 13C-labeled compound,
glucopyranosyl[5-13C]glucose, confirmed all of the above
conclusions. In particular, the elimination of C-5 of the
reducing end in the second eliminated glycolaldehyde (or
isomeric) molecule was verified by the observation of β-D-
glucopyranosylglycolaldehyde with no 13C.
The results also demonstrate that the first glycolaldehyde (or
isomeric) molecule eliminated from cellobiose upon pyrolysis
contains the 13C label (NH4 adduct of m/z 300); this process
+
must involve the loss of 13C-1 and most likely C-2 of the
reducing end of cellobiose. Identification of the origin of the
initially eliminated glycolaldehyde on the basis of carbon
labeling indicates that certain mechanisms may be more feasible
than others. For example, the retro-aldol mechanism
considered here and elsewhere results in the elimination of
glycolaldehyde from the C-1 and C-2 positions.14,15 On the
other hand, 1,2-dehydration followed by retro-Diels−Alder
reaction would result in elimination of glycolaldehyde from the
C-5 and C-6 positions.
Levoglucosan must be formed via at least two pathways since
+
two different ions (NH4 adducts of m/z 180 and 181,
corresponding to ammoniated molecules with and without the
13C label) are produced. Hence, levoglucosan is likely formed
from both the reducing and nonreducing ends of cellobiose.
The ionized initial fast-pyrolysis products of another labeled
cellobiose, [1-13C]glucopyranosylglucose, with a 13C label in
the nonreducing end, are shown in the mass spectrum in Figure
2a. Comparison of these products to those obtained for
unlabeled cellobiose revealed that all of the ionized products
with m/z values greater than 180 contain the 13C label. These
results support the above conclusion that glucose is produced
exclusively from the nonreducing end of cellobiose. Further,
both labeled and unlabeled levoglucosan were observed, in
agreement with the above proposal that more than one
mechanism must lead to levoglucosan and that it is likely to be
formed from both the reducing and nonreducing ends. Finally,
the results show that the first two glycolaldehyde (or isomeric)
molecules eliminated from cellobiose come from the reducing
end.
Quantum chemical calculations16 at 600 °C (see the
Supporting Information for calculations performed at 25 °C)
at the M06-2X/6-311++G(d,p)//M06-2X/6-311++G(d,p)
level of theory17,18 revealed a low-energy pathway for the
consecutive elimination of one glycolaldehyde and two isomeric
ethenediol molecules19 from cellobiose, eventually producing
levoglucosan (Figure 3). This pathway is consistent with the
13C-labeling results described above. Work to identify addi-
tional pathways leading to levoglucosan and the other observed
products is in progress.
In conclusion, identification of the initial fast-pyrolysis
products of selectively 13C-labeled cellobioses using a
previously reported4 experimental method has been demon-
strated to be a powerful approach for exploring the details of
the initial reactions in the fast pyrolysis of cellobiose. Several
products that are likely to be produced by consecutive
glycolaldehyde (or isomer) eliminations from the reducing
end of cellobiose, including levoglucosan, were observed.
Literature mechanisms proposed3 for the fast pyrolysis of
In order to explore the mechanism for the elimination of the
second glycolaldehyde (or isomeric) molecule from the
1911
DOI: 10.1021/jo5025255
J. Org. Chem. 2015, 80, 1909−1914

The Journal of Organic Chemistry
Note
Figure 3. Calculated free energies (kcal mol−1) of intermediates and transition states (square brackets) for the formation of levoglucosan from
cellobiose via consecutive losses of one glycolaldehyde (GA) and two ethenediol (EDL) molecules (which are likely to eventually tautomerize to
glycolaldehyde) at 600 °C obtained at the M06-2X/6-311++G(d,p)//M06-2X/6-311++G(d,p) level of theory. The location of a 13C label at C-1 in
the reducing end is indicated by a red circle, at C-3 by a blue circle, and at C-5 by a green circle. The mass-to-charge (m/z) ratios are for unlabeled
cellobiose.
multiplet. The elemental compositions were obtained using high-
resolution mass spectrometry and ESI in positive-ion mode, and the
measured m/z values are reported. The previously synthesized
compound 2,3,4,6-tetra-O-benzyl-D-glucopyranosyl trichloroacetimi-
date (S7) was prepared from commercially available 2,3,4,6-tetra-O-
benzyl-D-glucopyranose by following a literature procedure.20 The
synthesis of labeled cellobioses glucopyranosyl[3-13C]glucose (S9) and
glucopyranosyl[5-13C]glucose (S17) are shown in Schemes 1 and 2,
respectively.
cellobiose are not consistent with the results reported here.
Quantum chemical calculations revealed a feasible low-energy
pathway for some of the products observed. Since many of the
initial products have molecular weights greater than those of
the final pyrolysis products reported for cellobiose in the
literature,8,9 these initial products may be intermediates in the
formation of the final products, which include light oxygenated
hydrocarbons.
Minimization of the production of low-molecular-weight
oxygenated hydrocarbons is an important goal for fast pyrolysis
of cellulose in order to maximize the production of fuel and
high-value chemical products. Knowledge of the fragmentation
pathways occurring during fast pyrolysis of smaller carbohy-
drates will contribute to the knowledge of fast pyrolysis of
cellulose and ultimately fast pyrolysis of whole biomass,
possibly enabling the tailoring of the product distribution
obtained upon fast pyrolysis of whole biomass.
1,2,3,4,6-Penta-O-acetyl-3-13C-β-D-glucopyranoside (S1).21
To a solution of acetic anhydride (6 mL) was added sodium acetate
trihydrate (1.13 g, 8.3 mmol, 3 equiv). The mixture was refluxed at 90
°C for 20 min, and then the labeled D-glucose (0.5 g, 2.8 mmol) was
added. The resulting mixture was stirred for 4 h and then
concentrated, dissolved in methanol, and recrystallized with cold
water. A white solid was then filtered and dried to afford S2 (1.00 g,
1
92%). H NMR (400 Hz, CDCl3, δ): 2.02, 2.03, 2.04, 2.09, and 2.12
(15H, s, CH3), 3.84−3.87 (1H, m, CH), 4.10−4.13 (1H, m, CH2),
4.28−4.32 (1H, dd, J1 = 8, J2 = 4, CH), 5.04−5.09 (d, 1H, J = 8, CH),
5.13−5.17 (m, 1H, CH), 5.42−5.47 (m, 1H, CH), 5.72−5.74 (d, 1H, J
= 8, CH). 13C NMR (100 Hz, CDCl3, δ): 170.6, 170.1, 169.4, 169.2,
168.9, 91.7, 89.1, 74.0, 69.8, 67.8, 67.5, 61.5, 20.7. HRMS (ESI): calcd
for 13CC15H22O11Na [M + Na]+, 414.1093; measured, 414.1086.
Benzyl 2,3,4,6-Tetra-O-acetyl-3-13C-β-D-glucopyranoside
(S2).6 To a solution of penta-O-acetylglucose S1 (1.00 g, 2.56
mmol) and benzyl alcohol (0.61 mL, 5.60 mmol) in anhydrous
CH2Cl2 (10 mL) was added BF3·Et2O (0.41 mL, 3.33 mmol). The
reaction mixture was stirred at room temperature (RT) for 24 h and
diluted with 5% aqueous NaHCO3 (10 mL). The organic layer was
separated, washed sequentially with aqueous NaHCO3 (10 mL) and
water (10 mL), dried over Na2SO4, and concentrated. The crude
EXPERIMENTAL SECTION
■
Synthesis of Labeled Cellobioses. 3-13C-β-D-Glucose and
[1-13C]glucopyranosylglucose were purchased from Omicron Bio-
chemicals. The synthesized compounds (for the syntheses, see below)
were purified by column chromatography on a Teledyne-ISCO
CombiFlash system with a silica gel column. These compounds were
1
characterized by H and 13C NMR spectroscopy and high-resolution
mass spectrometry. The reactions were performed under an argon
atmosphere if needed. Solvents were purified and/or predried as
necessary. Analytical thin-layer chromatography (TLC) was used to
monitor the reactions. Visualization was accomplished with UV light
(254 nm) and by ethanol/H2SO4 TLC stain. CDCl3, CD3OD, and
D2O were used as NMR solvents. 1H NMR spectra were acquired on a
400 MHz instrument, and the chemical shifts (δ) are reported relative
to the residual solvent peak. 13C NMR spectra were acquired on a 100
MHz instrument, and the chemical shifts are reported in parts per
million relative to the residual solvent peak. When reporting spectral
data, the format chemical shift (integration, multiplicity, J value(s) in
Hz, identification) was used with the following multiplicity
abbreviations: s = singlet, d = doublet, t = triplet, q = quartet, m =
1
product was recrystallized from EtOH to give S2 (0.55 g, 49%). H
NMR (400 Hz, CDCl3, δ): 7.39−7.27 (5H, m, Ar−H), 5.13−4.99
(3H, m, CH, H2, H3, and H4), 4.88 (1H, d, J = 12, CH2), 4.61 (1H, d,
J = 12, CH2), 4.60 (1H, d, J = 8, CH, H1), 4.30 (1H, dd, J = 12, 4, CH,
H6), 4.19 (1H, dd, J = 12, 2, CH, H6), 3.67−3.64 (1H, m, CH, H5),
2.09 (3H, s, CH3), 2.00 (3H, s, CH3), 1.99 (3H, s, CH3), 1.98 (3H, s,
CH3). 13C NMR (100 Hz, CDCl3, δ): 170.6, 170.2, 169.3, 169.2,
136.6, 128.4, 127.9, 127.8, 127.7, 99.2, 77.0, 76.7, 72.7, 71.7, 71.4, 70.6,
70.1, 69.7, 68.5, 68.1, 20.5. HRMS (ESI): calcd for 13CC20H26O10Na
[M + Na]+, 462.1424; measured, 462.1410.
1912
DOI: 10.1021/jo5025255
J. Org. Chem. 2015, 80, 1909−1914

The Journal of Organic Chemistry
Note
Scheme 1. Synthesis of Glucopyranosyl[3-13C]glucose (S9)
Scheme 2. Synthesis of Glucopyranosyl[5-13C]glucose (S17)
Benzyl 3-13C-β-D-Glucopyranoside (S3).6 A mixture of benzyl
2,3,4,6-tetra-O-acetyl-β-D-glucopyranoside (0.896 g, 2.50 mmol),
MeOH (16 mL), triethylamine (2 mL), and H2O (2 mL) was stirred
for 5 h. The reaction was concentrated in vacuo, and the resulting
residue was purified by chromatography (5:1 CH2Cl2/CH3OH) to
CDCl3, δ): 137.0, 136.8, 129.2, 128.5, 128.1, 128.0, 126.3, 102.2,
101.8, 80.6, 80.2, 77.4, 77.0, 76.71, 74.6, 74.2, 73.0, 72.8, 71.2, 68.5,
66.2. HRMS (ESI): calcd for 13CC20H22O6Na [M + Na]+, 382.1314;
measured, 382.1310.
4,6-O-Benzylidene-1,2,3-tri-O-benzyl-3-13C-β-D-glucopyra-
noside (S5).22 The partially protected S4 (0.21 g, 0.58 mmol) was
added to anhydrous DMF (2 mL), and the mixture was stirred at 0 °C
for 30 min. NaH in mineral oil (60%) (94 mg, 2.32 mmol) was then
added under argon in small portions over 30 min, with the
temperature being maintained at 0 °C. Tetrabutylammonium iodide
(54 mg, 0.14 mmol) was then added, and the mixture was stirred for a
further 2 h at 0 °C. Benzyl bromide (0.21 mL, 1.74 mmol) was then
added slowly. The reaction mixture was stirred at 0 °C for a further 30
min, allowed to warm to room temperature, and stirred for 24 h.
Methanol was added slowly to destroy the excess NaH, and the
solvents were then removed in vacuo. The residue was subjected to
column chromatography with 3:2 hexane/ethyl acetate to furnish S5 as
1
give S3 (0.44 g, 81%). H NMR (400 Hz, CD3OD, δ): 7.43−7.25
(5H, m, Ar−H), 4.78 (2H, ABq, J = 16, PhCH2), 4.36 (1H, d, J = 8
Hz, CH, H1), 3.89 (1H, dd, J = 12, 2, CH, H6), 3.70 (1H, dd, J = 12,
6, CH, H6), 3.35−3.15 (4H, m, CH, H2, H3, H4, and H5). 13C NMR
(100 Hz, CDCl3, δ): 139.1, 129.3, 129.2, 128.7, 103.3, 78.1, 75.1, 74.9,
74.1, 71.8, 71.4, 62.8. HRMS (ESI): calcd for 13CC12H18O6Na [M +
Na]+, 294.1030; measured, 294.0988.
Benzyl 4,6-O-Benzylidene-3-13C-β-D-glucopyranoside (S4).6
To a mixture of benzyl β-D-glucopyranoside (0.21 g, 0.78 mmol) and
benzaldehyde dimethyl acetal (0.14 mL, 0.94 mmol) in dimethylfor-
mamide (DMF) (2 mL) at RT was added p-toluenesulfonic acid
(TsOH·H2O) (37 mg, 0.195 mmol). The reaction mixture was stirred
for 5 min, heated to 80 °C, and stirred for 2.5 h. The mixture was
cooled to RT and subsequently concentrated at reduced pressure. The
resulting residue was partitioned between CH2Cl2 (20 mL) and
saturated Na2CO3 (20 mL). The layers were separated, and the
aqueous layer was extracted with CH2Cl2 (3 × 10 mL). The combined
organic layers were washed with water (2 × 10 mL) and brine (10
mL), dried over MgSO4, and concentrated. Purification of the resulting
residue by flash chromatography (1:1 EtOAc/hexanes) provided the
1
a pale-yellow oil (0.3 g, 81%). H NMR (400 Hz, CDCl3, δ): 7.48−
7.34 (20H, m, Ar−H), 5.65 (1H, s, CH, H7), 5.04−4.71 (8H, m, CH,
CH2), 4.48−4.44 (1H, dd, J = 8, 4, CH, H4), 3.92−3.76 (2H, m, CH,
H6, H3), 3.66−3.60 (2H, m, CH, H6, H2), 3.53−3.48 (1H, m, CH,
H5). 13C NMR (100 Hz, CDCl3, δ): 138.5, 137.1, 129.0, 128.3, 127.6,
126.0, 103.2, 101.1, 80.9, 80.7, 77.4, 76.8, 75.4, 75.1, 71.6, 68.8, 66.1.
HRMS (ESI): calcd for 13CC33H34O6Na [M + Na]+, 562.2253;
measured, 562.2227.
1,2,3,6-Tetra-O-benzyl-3-13C-β-D-glucopyranoside (S6).22
Triethylsilane (0.27 mL, 1.68 mM) and trifluoroacetic acid (0.13
mL, 1.68 mmol) were added to a solution of S5 (0.3 g, 0.56 mmol) in
anhydrous CH2Cl2 (10 mL) at 0 °C. The solution was stirred at RT
for 6 h. The reaction mixture was diluted with EtOAc (approximately
20 mL), neutralized with saturated aqueous NaHCO3 and brine, dried
1
desired product (0.21 g, 75%) as a white solid. H NMR (400 Hz,
CDCl3, δ): 7.50−7.48 (2H, m, Ar−H), 7.37−7.29 (8H, m, Ar−H),
5.46 (1H, s, benzylidene CH), 4.88 (1H, d, J = 12, CH2), 4.59 (1H, d,
J = 12, CH2), 4.53 (1H, d, J = 8, CH, H1), 4.31 (1H, dd, J = 8, 4, CH,
H4), 3.59−3.76 (2H, m, CH, C6, H3), 3.63−3.58 (4H, m, CH, H6,
H2, and 2OH), 3.35−3.30 (1H, m, CH, H5). 13C NMR (100 Hz,
1913
DOI: 10.1021/jo5025255
J. Org. Chem. 2015, 80, 1909−1914

The Journal of Organic Chemistry
Note
over Na2SO4, and evaporated to dryness. The residue was purified by
flash column chromatography (3:10 EtOAc/hexane) to get S6 (0.2 g,
60%) as a colorless oil. H NMR (400 Hz, CDCl3, δ): 7.45−7.29
(20H, m, Ar−H), 5.02−4.64 (8H, m, CH2), 4.58 (1H, d, J = 8, CH,
H1), 3.86−3.48 (6H, m, CH, CH2, ring CH), 2.60 (1H, s, OH). 13C
NMR (100 Hz, CDCl3, δ): 138.6, 138.3, 137.5, 128.1, 127.8, 102.5,
84.3, 84.1, 83.8, 82.5, 81.7, 81.4, 77.3, 77.0, 76.7, 75.2, 74.7, 74.1, 73.6,
717, 71.3, 71.1, 70.2. HRMS (ESI): calcd for 13CC33H36O6Na [M +
Na]+, 564.2433; measured, 564.2465.
Basic Energy Sciences under Award DE-SC0000997. The work
by J.C.D. was supported in part by the National Science
Foundation EFRI (0938033-DGE).
1
REFERENCES
■
(1) Mohan, D.; Pittman, C. U.; Steele, P. H. Energy Fuels 2006, 20,
848−889.
(2) Mettler, M. S.; Vlachos, D. G.; Dauenhauer, P. J. Energy Environ.
Sci. 2012, 5, 7797−7809.
2,3,4,6-Tetra-O-benzyl-β-D-glucopyranosyl-(1−4)-1,2,3,6-
tetra-O-benzyl-3-13C-β-D-glucopyranoside (S8).22 To a solution
of S6 (0.1 g, 0.18 mmol) and S7 (0.12 g, 0.18 mmol) in anhydrous
CH2Cl2 (5 mL) was added BF3·Et2O (3 μL, 0.018 mmol) at −72 °C.
After the solution was stirred at −72 °C for 1 h, the reaction mixture
was neutralized with triethylamine. The residue was purified by flash
column chromatography (1:8 EtOAc/hexane) to give S8 as a 1:3
(3) Mayes, H. B.; Broadbelt, L. J. J. Phys. Chem. A 2012, 116, 7098−
7106.
(4) Hurt, M. R.; Degenstein, J. C.; Gawecki, P.; Borton, D. J., II;
Vinueza, N. R.; Yang, L.; Agrawal, R.; Delgass, W. N.; Ribeiro, F. H.;
Kenttamaa, H. I. Anal. Chem. 2013, 85, 10927−10934.
̈
(5) Yoneda, Y.; Kawada, T.; Rosenau, T.; Kosma, P. Carbohydr. Res.
2005, 340, 2428−2435.
1
mixture of α and β isomers (0.12 g, 55%) as a colorless syrup. H
(6) Zheng, S.; Laraia, L.; O’Connor, C. J.; Sorrell, D.; Tan, Y. S.; Xu,
Z.; Venkitaraman, A. R.; Wu, W.; Spring, D. R. Org. Biomol. Chem.
2012, 10, 2590−2593.
NMR (400 Hz, CDCl3, δ): 7.42−7.19 (40H, m, Ar−H), 5.15−4.46
(18H, m, CH, PhCH2, H1, H1′), 3.92−3.35 (12H, m, CH, CH2, ring
CH). 13C NMR (100 Hz, CDCl3, δ): 139.3, 138.6, 137.6, 128.3, 127.8,
127.5, 102.4, 84.8, 83.1, 78.0, 77.3, 76.7, 75.5, 74.9, 73.2, 70.9, 69.0,
68.2. HRMS (ESI): calcd for 13CC68H70O11Na [M + Na]+, 1086.4849;
measured, 1086.4846.
(7) Laboratory pyrolysis reactors generally utilize heated surfaces in
combination with heated gaseous environments.
(8) Mettler, M. S.; Paulsen, D.; Vlachos, D. G.; Dauenhauer, P. J.
Green Chem. 2012, 14, 1284−1288.
Glucopyranosyl[3-13C]glucose (S9).22 A solution of S8 (0.12 g,
1.15 mmol) in MeOH (10 mL) was hydrogenated in the presence of
10% Pd/C (15 mg) at atmospheric pressure at RT for 36 h. After the
catalyst was filtered off, the reaction mixture was evaporated to give S9
(α:β = 1:3, 34 mg, 87%) as a colorless solid. 1H NMR (400 Hz, D2O,
δ): 4.56 (1H, d, J = 8, CH, H1′), 4.41 (1H, d, J = 8, CH, H1), 3.83−
3.15 (12H, m, CH, CH2, ring CH). 13C NMR (100 Hz, D2O, δ):
103.1, 96.3, 92.4, 76.8, 76.6, 76.1, 75.4, 74.9, 74.6, 73.8, 72.1, 70.7,
70.0, 61.0, 60.5. HRMS (ESI): calcd for 13CC11H22O11Na [M + Na]+,
366.1060; measured, 366.1068.
(9) Patwardhan, P. R.; Satrio, J. A.; Brown, R. C.; Shanks, B. H. J.
Anal. Appl. Pyrolysis 2009, 86, 323−330.
(10) Mayes, H. B.; Nolte, M. W.; Beckham, G. T.; Shanks, B. H.;
Broadbelt, L. J. ACS Sustainable Chem. Eng. 2014, 2, 1461−1473.
(11) Seshadri, V.; Westmoreland, P. R. J. Phys. Chem. A 2012, 116,
11997−12013.
(12) Paine, J. B., III; Pithawalla, Y. B.; Naworal, J. D. J. Anal. Appl.
Pyrolysis 2008, 82, 10−41.
(13) Hosoya, T.; Nakao, Y.; Sato, H.; Kawamoto, H.; Sakaki, S. J.
Org. Chem. 2009, 74, 6891−6894.
(14) Zhou, X.; Nolte, M.; Mayes, H. B.; Shanks, B. H.; Broadbelt, L.
J. Ind. Eng. Chem. Res. 2014, 53, 13274−13289.
(15) Vinu, R.; Broadbelt, L. J. Energy Environ. Sci. 2012, 5, 9808−
9826.
(16) Frisch, M. J., et al. Gaussian 09, revision D.01; Gaussian, Inc.:
Wallingford, CT, 2009.
Glucopyranosyl[5-13C]glucose (S17).22 The same synthesis
strategy as described above for S9 was employed for the synthesis of
glucopyranosyl[5-13C]glucose (S17). A solution of S16 (0.12 g, 1.15
mmol) in MeOH (10 mL) was hydrogenated in the presence of 10%
Pd/C (15 mg) at atmospheric pressure at RT for 36 h. After the
catalyst was filtered off, the reaction mixture was evaporated to give
S17 (α:β = 1:3, 34 mg, 87%) as a colorless solid. 1H NMR (500 MHz,
D2O, δ): 4.56 (1H, d, J = 5, CH, C1′), 4.41 (1H, d, J = 5, CH, C1),
3.83−3.15 (12H, m, CH, CH2, ring CH). 13C NMR (125 MHz, D2O,
δ): 102.5, 95.6, 91.7, 78.8, 78.5, 78.3, 75.9, 75.4, 74.7, 74.2, 73.8, 71.7,
71.2, 71.1, 70.0, 69.4, 60.5, 60.1, 59.9. HRMS (ESI): calcd for
13CC11H22O11Na [M + Na]+, 366.1060; measured, 366.1066.
(17) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys.
1980, 72, 650−654.
(18) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−241.
(19) It is well-known that ethenediol converts readily into its
tautomer glycolaldehyde; therefore, β-D-glucopyranosylethenediol was
assumed to readily tautomerize to β-D-glucopyranosylglycolaldehyde.
(20) Kim, S.; Song, S.; Lee, T.; Jung, S.; Kim, D. Synthesis 2004,
847−850.
ASSOCIATED CONTENT
* Supporting Information
NMR spectra, tables of cartesian coordinates, and complete ref
16. This material is available free of charge via the Internet at
http://pubs.acs.org.
■
S
(21) Pulsipher, A.; Yousaf, M. N. Chem. Commun. 2011, 47, 523−
525.
(22) Malz, F.; Yoneda, Y.; Kawada, T.; Mereiter, K.; Kosma, P.;
Rosenau, T.; Jager, C. Carbohydr. Res. 2007, 342, 65−70.
̈
AUTHOR INFORMATION
Corresponding Author
*E-mail: hilkka@purdue.edu.
■
Present Address
§J.G.: Department of Chemistry and Biochemistry and Center
for Quantitative Obesity Research, Montclair State University,
Montclair, NJ 07043.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors gratefully acknowledge partial financial support by
the Center for Direct Catalytic Conversion of Biomass to
Biofuels (C3Bio), an Energy Frontier Research Center funded
by the U.S. Department of Energy, Office of Science, Office of
1914
DOI: 10.1021/jo5025255
J. Org. Chem. 2015, 80, 1909−1914