Stereoselective acetylation of hemicellulosic C5-sugars

Article abstract of DOI:10.1016/j.carres.2017.03.008

The stereoselective peracetylation of α-D-xylose (1) and α-L-arabinose (4) using a combination of triethylamine and acetic anhydride in the presence or absence of a catalytic amount of dimethylaminopyridine (DMAP) is described. The peracetylated D-xylose and L-arabinose alpha pyranose anomers 2α and 5α are obtained in 97% and 56% yields respectively. The peracetylated D-xylose beta pyranose anomer 2β is obtained in 71% yield through simple modification of the reaction conditions. Details regarding synthesis and isolation optimization studies under different conditions are presented below. The stereoselective peracetylation reactions disclosed here have been used to separate mixtures of D-xylose and L-arabinose as their peracetylated derivatives 2β and 5α in 47% and 42% yields and can provide pure pentoses after deacetylation.

Full text of DOI:10.1016/j.carres.2017.03.008

Accepted Manuscript
Stereoselective acetylation of hemicellulosic C5-sugars
Zachary D. Herde, Prathap John, Dania Alvarez-Fonseca, Jagannadh Satyavolu,
Christopher T. Burns
PII:
S0008-6215(17)30007-1
10.1016/j.carres.2017.03.008
CAR 7337
DOI:
Reference:
To appear in: Carbohydrate Research
Received Date: 3 January 2017
Revised Date: 9 March 2017
Accepted Date: 9 March 2017
Please cite this article as: Z.D. Herde, P. John, D. Alvarez-Fonseca, J. Satyavolu, C.T. Burns,
Stereoselective acetylation of hemicellulosic C5-sugars, Carbohydrate Research (2017), doi: 10.1016/
j.carres.2017.03.008.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Graphical Abstract

ACCEPTED MANUSCRIPT
Stereoselective Acetylation of Hemicellulosic C5-Sugars
Zachary D. Herde,b Prathap John,b Dania Alvarez-Fonseca,b Jagannadh Satyavolu,b
Christopher T. Burns*,a,b
aDepartment of Chemistry, University of Louisville, 2320 South Brook Street, Louisville, Kentucky, 40292
bConn Center for Renewable Energy Research, University of Louisville, Louisville, Kentucky, 40292
Abstract
The stereoselective peracetylation of
triethylamine and acetic anhydride in the presence or absence of a catalytic amount of
dimethylaminopyridine (DMAP) is described. The peracetylated -xylose and -arabinose alpha
pyranose anomers 2α and 5α are obtained in 97 % and 56 % yields respectively. The peracetylated
α-D-xylose (1) and α-L-arabinose (4) using a combination of
D
L
D
-
xylose beta pyranose anomer 2β is obtained in 71 % yield through simple modification of the reaction
conditions. Details regarding synthesis and isolation optimization studies under different conditions are
presented below. The stereoselective peracetylation reactions disclosed here have been used to separate
mixtures of D-xylose and L-arabinose as their peracetylated derivatives 2β and 5α in 47 % and 42 %
yields and can provide pure pentoses after deacetylation.
Keywords
acetylation; stereoselective; carbohydrate; pentoses; xylose; arabinose
1

ACCEPTED MANUSCRIPT
Introduction
The growing importance of carbohydrates in the synthesis of value-added chemicals,1,2,3
conversion to biofuels,4 and the understanding of their importance to a multitude of biological processes5
has led to an increase in research efforts for new, practical, and stereoselective methods for the synthesis
of carbohydrate derivatives. While much attention has focused on the use of C6-sugars (hexoses) derived
from cellulose6, less attention has been paid to derivatization and use of C5-sugars (pentoses) obtained
from hemicellulose rich biomass.7 Acetylation is one of the most common reactions in carbohydrate
chemistry since it provides an efficient method for the conversion of hydrophilic hydroxyl groups to
hydrophobic acetate functionalities that are more soluble in organic solvents. A common route used for
the peracetylation of sugars with acetic anhydride requires pyridine as both the solvent and base for the
acetylation reaction despite its unpleasant odor and known toxicity.8 While the yields are acceptable, the
presence of a mixture of regioisomers (pyranose and furanose) as well as anomers (α- and β-) makes this
reaction less than desirable when a single well-defined stereochemical product is preferred. Addition of
4-(N,N)-dimethylaminopyridine (DMAP), as a co-catalyst increases the rate of the acetylation, reaction
but does not lead to a change in the stereochemical outcome of the reaction.9 The solvent free
peracetylation of various mono- and disaccharides using a catalytic amount of iodine in neat acetic
anhydride was disclosed in 2004.10 The method is operationally simple, leads to exclusive formation of
the peracetylated sugars as their pyranosyl esters in high yields (> 90%), and results in favorable α:β
anomer ratios (10:1 or higher) in the case of several C6 mono- and disaccharides. For the C5-sugars,
xylose and -arabinose, while the peracetylated product yields are high the α:β anomer ratios are lower
(4:1 for -xylose) as well as opposite for the two C5 sugars (1:3 α:β ratio for -arabinose). The use of
D-
L
D
L
anhydrous sodium acetate in refluxing acetic anhydride,11 which forms the β-glycosyl acetate in good
yield, is the only known stereoselective acetylation reaction. Unfortunately, this method is not
synthetically useful if the α-glycosyl acetate is the desired product. Numerous other methods for the
peracetylation of carbohydrates exist in the literature,12-13 each with its own advantages and
2

ACCEPTED MANUSCRIPT
disadvantages, but none of these methods lead to the stereoselective formation of the β-glycosyl acetate
versus α-glycosyl acetate through simple modification of the reaction conditions. Therefore, there still is
a need for a general procedure for the steroselective peracetylation of carbohydrates.
Hemicellulose based sugars (e.g., D-xylose and L-arabinose) can provide a platform for synthesis
of a variety of industrially important chemicals that are currently derived from petroleum. At the
beginning of the bio-refinery process, hemicellulose rich biomass is selectively extracted to produce a C-5
sugar rich hydrolyzate.14 Crystalline
D-xylose and L-arabinose are isolated, as their pyranose
regioisomers, from the pentose rich hydrolysate using a boronic acid mediated methodology.15 This two-
step process provides a mixture of the two, pure, monosaccharides required for subsequent synthetic
transformations with the first step often being the peracetylation of the isolated C-5 sugar. The classic
pyridine based peracetylation protocol8 was initially used, but there were several drawbacks. The long
reaction time (12 hrs), low yield of the α-pyranose peracetylated C5-sugar which we desired for further
synthetic manipulations, and inability to isolate pure product made the use of this methodology far from
desirable for our integrated bio-refinery. Optimization of this step, in terms of both stereoselectivity and
our desire to use greener reaction conditions, was critical to make the overall process cost effective and
sustainable within our integrated bio-refinery. We describe herein the stereoselective peracetylation of
D-
xylose and -arabinose using a combination of triethylamine and acetic anhydride in the presence or
L
absence of a catalytic amount of DMAP. The desired alpha or beta pyranosyl esters are obtained in high
yield and purity through simple modification of the reaction conditions. Synthesis and isolation
optimization studies follow.
Results and Discussions
Peracetylation using Neat Pyridine. We sought to develop a methodology whereby C5-sugars,
primarily D-xylose and L-arabinose, could be peracetylated in a stereoselective fashion where simple
modification of the reaction conditions led to the formation of the α-pyranoid or β-pyranoid hemiacetal
anomers in high yield and purity. We began by examining the product distribution from the
3

ACCEPTED MANUSCRIPT
peracetylation of both C5-sugars using the known pyridine/acetic anhydride methodology. Peracetylation
1
of
α
-D
-xylopyranose (1), whose full H and 13C NMR characterization can be found in the general
experimental section, was carried out using the method of Shindo (Scheme 1).16 To a white suspension of
1 in neat pyridine, acetic anhydride
was added over the course of several minutes at 25 ˚C and then the reaction was allowed to stir under N2.
Within 15 minutes of stirring, the reaction mixture had turned clear and golden in appearance. After
stirring for 12 hours at 25 ˚C, the clear golden solution was subjected to an aqueous workup (see
1
experimental section) and the resulting isolated yellow oil characterized by H NMR. The results are
shown in Table 1 (entries 1 and 2). The overall yield of the peracetylation reaction was near quantitative,
1
but multiple products were observed. H NMR characterization of the isolated reaction mixture in CDCl3
showed the yellow oil was comprised of two regioisomers, tetraacetylated
D-xylopyranose (TXP) 2 and
tetraacetylated -xylofuranose (TXF) 3, in a ratio of 94:6. For both regioisomers 2 and 3 there also
D
existed α- and β-anomers. The ratio of TXP anomers was 2α:2β = 6:4 and the TXF anomers was 3α:3β
= 1:1. Chemical shift assignments for the four TXP and TXF anomers were made via comparison with
literature values for 2α,17 2β,18 and 3α/3β.19 Relative ratios of the four products were determined by
comparing C(1)-H integral values of each anomer located at δ 6.42 ppm (doublet, TXF 3α), δ 6.21 ppm
(doublet, TXP 2α), δ 6.11 ppm (singlet, TXF 3β), and δ 5.71 ppm (doublet, TXP 2β) in the low field
1
region of the H NMR . Observation of four compounds in the isolated product mixture came as a
surprise since we could find no mention of the presence of furanose isomers 3α/3β in the published
literature (Table 1, entries 1 and 2).16, 18
4

ACCEPTED MANUSCRIPT
The pyridine/acetic anhydride peracetylation method was also examined using α-L-arabinose (4),
1
which was also characterized by H and 13C NMR (see general experimental section), since 4 can be
obtained from the hemicellulose derived, pentose rich hydrolyzate.15 Peracetylation of 4 was carried out
using the method of Jakeman (Scheme 2). To a white suspension of 4 in neat pyridine, acetic anhydride
was added over the course of several minutes at 25 ˚C and then the suspension was allowed to stir under
N2. After stirring for 12 hours at 25 ˚C, the clear golden solution was subjected to an aqueous workup
(see experimental section) and the resulting isolated yellow oil characterized by 1H NMR. The results are
shown in Table 1 (entries 3 and 4). The overall yield of the peracetylation 4 was near quantitative but
multiple products were identified as was observed for peracetylated
characterization of the isolated reaction
α-D
-xylopyranose. 1H NMR
mixture in CDCl3 showed the yellow oil was comprised of two regioisomers, tetraacetylated
L-
arabinopyranose (TAP) 5 and tetraacetylated -arabinofuranose (TAF) 6, in a ratio of 75:25. For both of
L
the regioisomers, 5 and 6, there also existed α- and β-anomers. The ratio of TAP anomers was 5α:5β =
6:4 and the TAF anomers was 6α:6β = 4:6. Chemical shift assignments for the four TAP and TAF
anomers were made via comparison with literature values for 5α/5β,20 and 6α/6β.21 Relative ratios of the
four products were determined by comparing C(1)-H integral values of each anomer located at δ 6.37
ppm (doublet, TAF 6α), δ 6.34 ppm (doublet, TAP 5α), δ 6.19 ppm (singlet, TAF 6β), and δ 5.64 ppm
1
(doublet, TAP 5β) in the low field region of the H NMR . As with the peracetylation of
α-D-
xylopyranose, the observation of four compounds in the isolated product mixture was unexpected,
5

ACCEPTED MANUSCRIPT
especially with such a large proportion of 6α/6β present, since only the presence of the pyranose products
5α and 5β are described in the literature.9, 20
Table 1. Peracetylation of C5 sugars in neat pyridine
peracetylated C5 sugarsb
hemiacetal ratioc
anomer ratiod
entry
C5 sugar
α-D-Xyl (1)
α-D-Xyl (1)
α-L-Ara (4)
α-L-Ara (4)
acetylated product yield (%)a pyranose furanose
αp βp αf
βf
1
2
3
4
2 / 3
2 / 3
5 / 6
5 / 6
100
100
100
100
94
95
73
75
6
5
60 40 50 50
68 32 57 43
60 40 40 60
60 40 40 60
27
25
b
abased on mass of all peracetylated C5 sugars present in isolated reaction mixture. pyranose/furanose and anomer
c
ratios determined by 1H NMR of the reaction mixture after workup. pyranose = αp + βp; furanose = αf + βf. dp =
pyranose; f = furanose.
Peracetylation in Presence of DMAP/Amine. The product regiochemistry and lack of
stereochemical control observed in our initial peracetylation reactions led us to explore alternative
conditions that would allow for the isolation of an anomerically pure pyranose product in high yield.
Our idea was to optimize the peracetylation reaction by making the reaction more atom economic,
decreasing the reaction time, and using greener reagents where possible. The use and amount of pyridine
required in these reactions was not acceptable as pyridine is a harmful, noxious chemical that is
environmentally detrimental and harmful to human health. We decided to evaluate the use of
triethylamine (NEt3) as a greener alternative to pyridine in these reactions.22 Use of NEt3 for the
peracetylation of glucose, xylose, and rhamnose was disclosed in 200023 as a footnote in a table but no
indication of the reaction’s stereochemical outcome was discussed. Furthermore, no actual experimental
details were given in the text of the paper nor the accompanying supporting information. Based on this
initial literature lead, we explored the peracetylation of 1 to form 2α (Scheme 3). Acetic anhydride was
added to a CH2Cl2
6

ACCEPTED MANUSCRIPT
suspension that contained 1 (10.0 g, 66.6 mmol), eight equivalents of NEt3, and a catalytic amount of
DMAP cooled to 0 ˚C. After one hour of stirring at 0 ˚C, the reaction mixture was clear and orange in
color. The reaction was stirred at 25 ˚C for 19 hours, subjected to an aqueous workup (see experimental
section), a viscous orange oil was isolated and assessed using 1H NMR. The results are shown in Table 2
(entry 9). As observed previously with neat pyridine, the peracetylation yield was quantitative.
Gratifyingly, our acetylation reaction conditions yielded excellent regio- and stereochemical results as
well. The isolated orange oil consisted of tetraacetylated D-xylopyranoses 2α, characterized by a doublet
at δ 6.21 ppm (J = 4.0, equatorial C(1)-H), and 2β, characterized by the doublet at δ 5.71 ppm (J = 6.8,
axial C(1)-H), in a 94:6 ratio of the two anomers. The furanose steroisomers 3α and 3β were not present
in the isolated oil. This result was in stark contrast to the peracetylation of 1 in neat pyridine where
formation of tetraacetylated xylopyranoses 2α and 2β was favored but a 60:40 mixture of anomers
prevented the isolation of either. With this very encouraging result in hand, we sought to optimize the
reaction conditions further towards the exclusive formation of 2α.
7

ACCEPTED MANUSCRIPT
Table 2. Peracetylation of α-
D
-xylose (1) in the presence of DMAP/aminea
peracetylated
hemiacetal ratiod
D
-xylosec
anomer ratioe
entry amine
0 ˚C, t1 (h) 25 ˚C, t2 (h) yield (%)b
2α/β
100
100
100
100
100
100
93
3α/β
2α 2β 3α 3β
1
2
3
4
5
6
7
8
9f
NEt3
NEt3
1
1
4
4
99
81
-
-
-
-
-
-
7
-
-
98
98
98
98
98
98
2
2
2
2
2
2
-
-
-
-
-
-
-
-
-
-
-
-
NEt3
0.5
0.5
0.5
0.5
-
2
94
NEt3
0.5
0.5
-
100
99
NEt3
NEt3
99
NEt3
2.5
2
87
72 28 50 50
pyridine
NEt3
1
97
100
100
99
94
1
6
-
-
-
-
1
19
100
aConditions: α-
D-xylose (2 g, 13 mmol, 1 equiv), DMAP (10 mol %), amine (8 equiv), and Ac2O (7 equiv) reacted
b
in CH2Cl2 (10 mL) at indicated temperatures for indicated time. based on mass of all peracetylated C5 sugars
1
present in isolated reaction mixture. cpyranose/furanose and anomer ratios determined by H NMR of the reaction
f
mixture after workup. dpyranose = αp + βp; furanose = αf + βf. ep = pyranose; f = furanose. large scale
peracetylation: α-D-xylose (10.0 g, 66.6 mmol, 1 equiv), DMAP (10 mol %), amine (8 equiv), and Ac2O (7 equiv)
reacted in CH2Cl2 (50 mL) at indicated temperatures for indicated time.
In an effort to streamline the reaction conditions, we first explored the length of time required for
the reaction to stir at 25 ˚C after the addition of acetic anhydride at 0 ˚C. Our hope was to maintain the
high product selectivity for the formation of 2α observed in our first experiment and considerably shorten
the time required, thus making our reactions more efficient. The effect of time at 25 ˚C on yield as well
as regiochemistry/stereochemistry was evaluated by performing the peracetylation of 1 (2 g, 13 mmol) at
0 ˚C then allowing the reaction to stir for different lengths of time (4, 2, 0.5 hours) at 25 ˚C followed by
aqueous work-up and isolation (Table 2, entries 1-5). Assessment of each isolated reaction mixture by 1H
NMR showed that regardless of the time the reaction was allowed to stir at 25 ˚C the peracetylation yield
of 1 was high with only TXP anomers 2α and 2β present. Next, we investigated shortening the time the
reaction was stirred at 0 ˚C. We had observed that the white CH2Cl2 suspension dissolved completely
within 30 minutes at 0 ˚C to yield a clear, pale yellow solution after the addition of acetic anhydride was
complete. When we allowed the reaction to only stir for 30 minutes at 0 ˚C, complete dissolution of the
8

ACCEPTED MANUSCRIPT
reactants was observed in each case and no change in the conversion of 1 and product distribution of the
TXP anomers 2α and 2β was observed (Table 2, entries 3-5). If the addition of acetic anhydride to a
CH2Cl2 suspension of 1 and NEt3 is conducted at 25 ˚C an exothermic event is observed and some CH2Cl2
1
is lost to evaporation. Assessment of the isolated reaction mixture by H NMR revealed that multiple
peracetylated products were present (Table 2, entry 7). The TXP and TXF regioisomers were observed in
a 97:3 ratio and all four possible anomers were present in amounts similar to those observed when the
reaction was performed in neat pyridine (Table 1, entries 1 and 2). It was clear that performing the
reaction at 0 ˚C was crucial to obtaining the high level of regio- and stereocontrol we desired for the
peracetylation of 1. Combining the time and temperature results discussed above together (Table 2, entry
6), we performed the acetylation of 1 at 0 ˚C. The CH2Cl2 suspension was stirred for 30 minutes at 0 ˚C
at which time no precipitate was observed, and then the reaction was quenched with H2O at 0 ˚C. Using
this optimized set of conditions, we were able to stereoselectively peracetylate α-D-xylopyranose to form
2α along with a trace of 2β (2α:2β = 98:2) in 99 % yield. The results for the stereoselective
peracetylation of xylose detailed above compare very favorably with the solvent free method developed
by Field,10 where a high peracetylation yield of 1 was obtained but the stereochemical outcome of the
reaction was not as high (2α:2β = 4:1).
The amount of control obtained over the product distribution for peracetylation of 1 through
simple modifications of the reaction conditions was quite surprising. It is clear that the acetylation
reaction is fast and control of the reaction temperature is crucial to limiting the number of products
isolated. While it is clear that a catalytic amount of DMAP does increase the rate of the esterification
reaction,24-25 keeping the reaction temperature at 0 ˚C is necessary to prevent the faster isomerization of
xylose between the pyranose and furanose regioisomers.26 The nature of the amine base used during the
peracetylation of 1 is not crucial with regards to obtaining a high yield of 2α. When pyridine is used in
place of NEt3 (Table 2, entry 8) the TXP regioisomer was obtained in 97 % isolated yield with anomers
9

ACCEPTED MANUSCRIPT
2α and 2β present in a 99:1 ratio. The presence of a catalytic amount of DMAP in the reaction is crucial
to obtaining a high yield of anomer 2α as will be illustrated later.
Our new method for the formation of 2α from 1 is a considerable improvement over published
literature methods12 with regards to choice of reagents, overall reaction time, and most importantly
product regio- and stereoselectivity. We next explored the peracetylation of α-L-arabinose (4) with acetic
anhydride using the conditions developed for the stereoselective peracetylation 1 (Scheme 4). Acetic
anhydride was added to a CH2Cl2 suspension that contained 4 (2.0 g, 13.0 mmol), eight equivalents of
NEt3, and a catalytic amount of DMAP cooled to 0 ˚C. Unlike the peracetylation of
Scheme 4
OH
OAc
AcO
O
0.1 DMAP / 8 amine / 7 Ac2O
O
HO
5
6 6
/
CH2Cl2, N2
1) 0 °C, t1
2) 25 °C, t2
OH
AcO
OH
OAc
4
5
1, the white suspension of 4 present at 0 ˚C did not dissolve within 30 minutes. Instead, the white
suspension was stirred for two hours at 0 ˚C before 4 completely dissolved and became a clear yellow
solution. Once all of 4 had dissolved the reaction mixture was warmed to 25 ˚C, stirred for 2 hours,
subjected to an aqueous workup (see experimental section) where a yellow crystalline solid was isolated
1
and assessed using H NMR. The results are shown in Table 3 (entry 1). As observed in the
peracetylation of 1, the overall yield of peracetylated 4 was high. The ratio of regio- and stereoisomers
was dramatically improved versus the peracetylation of 4 performed in neat pyridine but did not have the
1
high degree of stereoselectivity observed in the formation of 2α from 1. H NMR characterization of the
isolated reaction mixture in CDCl3 showed the yellow solid was comprised of TAP and TAF regioisomers
5 and 6 in a 95:5 ratio. For regioisomers 5 and 6 there also existed α- and β-anomers. The ratio of TAP
anomers was 5α:5β = 99:1 and the TAF anomers was 6α:6β = 1:1. This result differs from the solvent
free peracetylation of 4 disclosed by the research group of Field,10 where the presence of furanosyl esters
10

ACCEPTED MANUSCRIPT
6α and 6β was not observed and the ratio of pyranosyl esters 5α and 5β was 1:3. While the two
furanosyl esters 6α:6β could not be completely eliminated from the isolated reaction mixture, the minor
amounts of both present along with the significantly diminished quantity of 5β allowed for the isolation of
5α as a white solid in 56 % yield by recrystallization of the reaction mixture using absolute ethanol
(Table 3, entry 2).
Table 3. Peracetylation of α-
L
-arabinose (4) in the presence of DMAP/aminea
peracetylated
hemiacetal ratiod
L
-arabinosec
anomer ratioe
entry amine
0 ˚C, t1 (h) 25 ˚C, t2 (h) yield (%)b
5α/β
6α/β
5α
99
98
85
98
5β 6α 6β
1
2
3
4
NEt3
NEt3
2
2.5
-
2
2
88
91(56)f
98
95
5
1
2
50 50
60 40
95
5
NEt3
2.5
2
77
23
6
15 53 47
67 33
pyridine
4.5
100
94
2
aConditions: α-
L-arabinose (2 g, 13 mmol, 1 equiv), DMAP (10 mol %), amine (8 equiv), and Ac2O (7 equiv)
b
reacted in CH2Cl2 (10 mL) at indicated temperatures for indicated time. based on mass of all peracetylated C5
1
sugars present in isolated reaction mixture. cpyranose/furanose and anomer ratios determined by H NMR of the
reaction mixture after workup. dpyranose = αp + βp; furanose = αf + βf. ep = pyranose; f = furanose. fisolated yield
of 5α.
The melting point of the isolated white crystalline solid was determined to be 96-97 ˚C. 1H NMR
analysis in CDCl3 confirmed that the isolated solid was indeed 5α as a doublet at δ 6.34 ppm (J = 3.2 Hz)
was observed for the equatorially disposed hydrogen attached to C1 in 5α.20 The published melting point
for 5α was incorrectly assigned to 5β in the literature.27 1H NMR analysis of filtrate residue (1.12 g)
showed that it was still enriched in 5α versus 5β/6α/6β with an observed ratio of 81:19. Attempts to
improve the yield of 5α through recrystallization of the filtrate residue using smaller amounts of absolute
ethanol met with no success.
As was observed in the peracetylation of 1 using DMAP/NEt3, the temperature of the reaction
mixture after the addition of acetic anhydride played an important role. When the reaction was not stirred
at 0 ˚C, allowing all solids to completely dissolve, an exothermic event was observed and some CH2Cl2
1
was lost to evaporation. Assessment of the isolated yellow syrup by H NMR revealed that multiple
11

ACCEPTED MANUSCRIPT
peracetylated products were present (Table 3, entry 3). The TAP and TAF regioisomers were observed in
a 77:23 ratio and all four anomers were present in amounts similar to those observed when the reaction
was run in neat pyridine (Table 1, entries 3 and 4). The nature of the amine base used during the
peracetylation of 4 is not crucial with regards to obtaining good selectivity for the formation of 5α. When
pyridine is used in place of NEt3 (Table 3, entry 4) the overall conversion of 4 into peracetylated products
was quantitative, although a longer time stirring at 0 ˚C was required for all solids to completely dissolve.
1H NMR characterization (CDCl3) of the isolated reaction mixture showed the colorless crystalline solid
was comprised of TAP and TAF regioisomers 5 and 6 in a 94:6 ratio, with anomer 5α being highly
enriched (93:7), very similar to what was observed when NEt3 was employed in this reaction. This is the
first reported example of the synthesis and isolation of pure 5α in good yields. It was reported recently
that heating 4 at 130 ˚C in the presence of KOAc and Ac2O produced a mixture of 5α and 5β that was
1
enriched in 5α (5.4:1) although no H NMR data was provided and no attempt to isolate 5α was
discussed.28 The method disclosed here provides access to pure 5α that is straightforward and scalable.
Peracetylation in the Absence of DMAP. The synthesis of 2α and 5α discussed above
illustrates that temperature plays an important role in determining the regio- and stereochemical outcome
of the peracetylation reactions. Our results indicate the identity of the amine base is not crucial in these
peracetylation reactions. All of the reactions discussed above were performed with a catalytic amount (10
mol %) of DMAP present. It is clear that DMAP does increase the rate of esterification reactions,24-25 but
it is unknown what affect the absence of DMAP will have on the peracetylation reaction’s product
distribution. In an effort to further streamline the reaction conditions, both in terms of overall cost and
atom economy, we were interested in exploring the product distribution of the peracetylation reaction
when it was performed in the absence of DMAP.
Using the reaction conditions outlined for the stereoselective synthesis of 2α but omitting DMAP
(Scheme 5), acetic anhydride was added to a CH2Cl2 suspension that contained 1 (2.0 g, 13.0 mmol) and
NEt3 (8 equiv.) cooled to 0 ˚C. After two hours of stirring at 0 ˚C, the reaction mixture was still a white
12

ACCEPTED MANUSCRIPT
suspension. The suspension was warmed to 25 ˚C with stirring and the white solid slowly dissolved.
After stirring the clear yellow solution for two hours at 25 ˚C, it was subjected to an aqueous workup (see
experimental section), and a white crystalline solid was isolated. The results of this reaction are shown in
Table 4 (entry 1). As observed in our earlier reactions, the overall yield of peracetylated xylose was
almost quantitative (95 %). The regiochemical outcome of the peracetylation reaction in the absence of
DMAP again favored the formation of the pyranose isomers 2 over the furanose isomers 3 in a 94:6 ratio.
The stereochemical outcome of the reaction was unexpected in both physical appearance and molecular
composition. The isolated white crystalline solid consisted of tetraacetylated D-xyloses 2α and 2β with
an anomer ratio of the 19:81 for 2α (doublet at δ 6.21 ppm (J = 4.0, equatorial C(1)-H)) and 2β (doublet
at δ 5.71 ppm (J = 6.8, axial C(1)-H)) which was opposite the results obtained when the peracetylation
reactions were performed in the presence of DMAP.
Scheme 5
OAc
AcO
OH
HO
O
O
8 amine / 7 Ac2O
OAc
2
3 /3
CH2Cl2, N2
1) 0 °C, t1
2) 25 °C, t2
OAc
OH
OH
2
1
While the two furanose anomers 3α:3β could not be completely eliminated from the isolated reaction
mixture, the minor amounts of both as well as the significantly diminished quantity of 2α allowed for the
isolation of 2β as a white solid in 71 % yield by recrystallization of the reaction mixture using absolute 1-
propanol (Table 4, entry 8). The melting point of the isolated white crystalline solid 2β was determined
to be 124-125 ˚C. The observed melting point is similar to the value published in the literature.29 1H
NMR analysis in CDCl3 confirmed that the isolated solid was indeed 2β as a doublet at δ 5.71 ppm (J =
6.8 Hz) was observed for the axially disposed hydrogen attached to C1 in 2β.30 1H NMR analysis of
filtrate residue (4.76 g) showed that it was depleted in 2β versus 2α/3α/3β with an observed ratio of
12:88.
13

ACCEPTED MANUSCRIPT
Table 4. Peracetylation of α-
D
-xylose (1) in the presence of aminea
peracetylated
hemiacetal ratiod
D
-xylosec
anomer ratioe
entry amine
0 ˚C, t1 (h) 25 ˚C, t2 (h) yield (%)b
2α/β
94
94
95
93
93
95
93
92
95
3α/β
2α 2β 3α 3β
1
NEt3
NEt3
2
2
4.5
-
2
2
2
2
2
2
2
2
2
95(48)f
88(51)f
87(40)f
83(49)f
82(44)f
90(48)g
92(63)g
93(71)g
84
6
6
5
7
7
5
7
8
5
19 81 25 75
14 86 29 71
20 80 17 83
11 89 25 75
12 88 22 78
2
3
NEt3
4
NEt3
5
NEt3
-
6
NEt3
-
9
91 33 67
12 88 33 67
91 20 80
68 32 57 43
7h
NEt3
-
8i
NEt3
-
9
9
pyridine
-
aconditions: α-
D-xylose (2 g, 13 mmol, 1 equiv), amine (8 equiv), and Ac2O (7 equiv) reacted in CH2Cl2 (10 mL) at
indicated temperatures for indicated time. bbased on mass of all peracetylated C5 sugars present in isolated reaction
c
mixture; value in parenthesis is isolated yield of 2β. pyranose/furanose and anomer ratios determined by 1H NMR
d
e
of the reaction mixture after workup. pyranose = αp + βp; furanose = αf + βf. p = pyranose; f = furanose. fethanol
used as crystallization solvent. 1-propanol used as crystallization solvent. hconditions: α-
D-xylose (5 g, 33.3 mmol,
g
1 equiv), amine (8 equiv), and Ac2O (7 equiv) reacted in CH2Cl2 (25 mL) at indicated temperatures for indicated
time. iconditions: α-
D
-xylose (10 g, 66.6 mmol, 1 equiv), amine (8 equiv), and Ac2O (7 equiv) reacted in CH2Cl2 (50
mL) at indicated temperatures for indicated time.
It was determined that the peracetylation of 1 in the absence of DMAP does not have to be cooled to 0 °C
and the reaction can be performed at 25 °C without an adverse effect on the isolated yield of 2β (Table 4,
entries 4 through 8). This makes the synthesis of 2β from 1 in the presence of NEt3 and acetic anhydride
operationally straightforward and simple.
Unlike our earlier results for the stereoselective peracetylation of 2α, where the nature of the
amine base used was not crucial to obtaining high stereoselectivity, a change in the amine used for the
peracetylation of 1 in the absence of DMAP led to very different results. When pyridine is used in place
of NEt3 (Table 4, entry 9), the overall conversion of 1 to peracetylated products was high but multiple
1
products were observed. H NMR characterization of the isolated reaction mixture in CDCl3 showed the
colorless oil was comprised of regioisomers 2 and 3 in a ratio of 95:5. For both 2 and 3 there also existed
α- and β-anomers. The ratio of TXP anomers was 2α:2β = 68:32 and the TXF anomers was 3α:3β =
15

ACCEPTED MANUSCRIPT
57:43. These isomer ratios are identical to those obtained when 1 was peracetylated in neat pyridine
(Table 1, entry 2) and are not desirable. It is clear that to obtain high isolated yields of 2β the choice of
amine base is important. The success of this reaction using NEt3 creates an ideal situation where the
outcome of the reaction can be predetermined by the presence or absence of DMAP. This shows the
potential for using this reaction early in a biorefinery chemical process, forming two products that could
be used in different applications; 2α for continued processing into fuels and materials,31 and 2β for
surfactants and medical applications.32
Our new method for the formation of 2β from 1 is an improvement over published literature
methods29 with regards to choice of reagents and overall reaction time. We next explored the
peracetylation of α-L-arabinose (4) with acetic anhydride using the conditions developed for the
peracetylation in the absence of DMAP (Scheme 6). Acetic anhydride was added to a CH2Cl2 suspension
that contained 4 (2.0 g, 13.0 mmol) and eight equivalents of NEt3 cooled to 0 ˚C. Unlike the
peracetylation of 4 in the presence of DMAP, the white suspension of 4 did not dissolve even after 4
hours at 0 ˚C. The white suspension was warmed to 25 ˚C, stirred for 2 hours, the resulting clear yellow
solution was subjected to an aqueous workup (see experimental section), and a yellow glass was isolated.
The results are shown in Table 5 (entry 1). As observed in previous peracetylation reactions, the overall
yield of peracetylated 4 was high. The regiochemical outcome of the peracetylation reaction of 4 in the
absence of DMAP favored the formation of the pyranosyl ester 5 over the furanozyl ester 6. 1H NMR
characterization of the isolated reaction mixture in CDCl3 showed the yellow glass was comprised of TAP
and TAF regioisomers 5 and 6 in a 67:33 ratio. The ratio of TAP anomers was 5α:5β = 9:91 and the TAF
16

ACCEPTED MANUSCRIPT
anomers was 6α:6β = 44:56. Attempts to isolate 5β from the reaction mixture via crystallization were
not successful.
Table 5. Peracetylation of α-
L
-arabinose (4) in the presence of aminea
peracetylated
hemiacetal ratiod
L
-arabinosec
anomer ratioe
entry amine
0 ˚C, t1 (h) 25 ˚C, t2 (h) yield (%)b
5α/β
6α/β
5α 5β 6α 6β
91 44 56
11 89 46 54
93 43 57
1
2
3
NEt3
NEt3
NEt3
4
4
-
2
2
4
78
82
88
67
33
9
66
34
70
30
7
aConditions: α-
L-arabinose (2 g, 13 mmol, 1 equiv), amine (8 equiv), and Ac2O (7 equiv) reacted in CH2Cl2 (10 mL)
b
at indicated temperatures for indicated time. based on mass of all peracetylated C5 sugars present in isolated
reaction mixture. cpyranose/furanose and anomer ratios determined by H NMR of the reaction mixture after
1
d
e
workup. pyranose = αp + βp; furanose = αf + βf. p = pyranose; f = furanose.
Peracetylation as a Separation Method. As mentioned above, we are pursuing the use of D-
xylose and L-arabinose as starting materials for synthesis of a variety of industrially important chemicals
traditionally obtained from petroleum. We have disclosed a boronic acid mediated isolation protocol
-arabinose in a 83:17 ratio of the two pentoses.15 This process does provide
which delivers D-xylose and L
a mixture of pure monosaccharides but separation and isolation of each pentose is required for subsequent
synthetic transformations to be accomplished. Based on our ability to isolate pure 2β and 5α using our
peracetylation methodologies followed by selective crystallization, we felt this could be a route to isolate
pure
D
-xylose and
L-arabinose, albeit as their acetylated derivatives (Scheme 7 and 8).
Using the reaction conditions outlined for the stereoselective synthesis of 2β, acetic anhydride
was added to a CH2Cl2 suspension that contained a 85:15 mixture of 1 and 4 (2.0 g, 13.0 mmol) and NEt3
(8 equiv.) cooled to 0 ˚C (Scheme 7). After two hours of stirring at 25 ˚C, the reaction mixture was a
clear yellow solution. The solution was subjected to an aqueous workup (see experimental section), and a
white crystalline solid was isolated. The results of this reaction are shown in Table 6 (entry 1).
17

ACCEPTED MANUSCRIPT
The overall yield of peracetylated pentoses was high (76 %). The ratio of peracetylated
D-xylose
compared to peracetylated
L-arabinose (81:19) was very close to the ratio of the two starting pentoses.
The ratio of 2β versus 2α/3/5/6 in the reaction residue was determined to be 66:34. Recystallization of
the reaction mixture using ethanol allowed for the isolation of the
D-xylose derivative 2β in 45 % yield.
1
1
Melting point and H NMR analysis verified the identity of the crystalline white solid as 2β. H NMR
analysis of crystallization filtrate showed that it was depleted in 2β versus 2α/3/5/6with an observed ratio
of 18:82. Doubling the peracetylation reaction time to 4 hours at 25 °C, led to an increase in the overall
yield of peracetylated products as shown in entry 2 of Table 6. The ratio of peracetylated
D-xylose
compared to peracetylated -arabinose (81:19) was identical to the previous reaction with only a minor
L
increase in the isolated yield of 2β obtained. The pentose mixtures used in entries 2 and 3 of Table 6
were obtained from DDG derived hydrolyzate.15
Table 6. Isolation of 2β from a mixture of pentoses 1/4 using NEt3/Ac2Oa
peracetylated pentosese
acetylated C5 ratio
2/3 : 5/6
rxn mixture filtrate residue
entry ratio 1:4 25 ˚C, t (h) yield (%)d
2β : 2α/3/5/6
2β : 2α/3/5/6
1
2
3
85:15b
83:17c
83:17c
2
4
2
76(45)f
97(47)f
87(40)f
81:19
81:19
78:22
66:34
18:82
68:32
24:76
63:37
20:80
aConditions: pentoses 1/4 (2 g, 13 mmol, 1 equiv), amine (8 equiv), and Ac2O (7 equiv) reacted in CH2Cl2 (10 mL)
at 25 ˚C for indicated time. bmixture (m/m) made using commercially purchased
D-xylose/L-arabinose, pentose
1
1
ratio determined using H NMR. cisolated sugar mixture from ref 15, pentose ratio determined using H NMR.
dbased on mass of all peracetylated C5 sugars present in isolated reaction mixture; value in parenthesis is isolated
1
yield of 2β. epentose, pyranose/furanose and anomer ratios determined by H NMR of the reaction mixture after
f
workup. ethanol used as crystallization solvent.
18

ACCEPTED MANUSCRIPT
Isolation of pure
L-arabinose derivative 5α can be achieved starting with an arabinose enriched
(4.4:1) mixture of 4 and 1 using a catalytic amount of DMAP in the presence of NEt3 at 0 °C (Scheme 8).
Acetic anhydride was added to a CH2Cl2 suspension that contained a mixture of 4 and 1 (4.4:1)
(2.0 g, 13.0 mmol), eight equivalents of NEt3, and a catalytic amount of DMAP cooled to 0 ˚C. The
white suspension was stirred for two hours at 0 ˚C before all solids completely dissolved and became a
clear yellow solution. Once all of the solids had dissolved the reaction mixture was warmed to 25 ˚C,
stirred for 2 hours, subjected to an aqueous workup (see experimental section), and a yellow crystalline
solid was isolated. The results are shown in Table 7. The overall yield of peracetylated pentoses was
almost quantitative (98 %). The ratio of peracetylated
(81:19)
L-arabinose compared to peracetylated D-xylose
Table 7. Isolation of 5α from a mixture of pentoses 4/1 using DMAP/NEt3/Ac2Oa
peracetylated pentosesd
acetylated C5 ratio rxn mixture filtrate residue
entry ratio 4:1b yield (%)c
5/6 : 2
81:19
81:19
5α : 2/5/6
5α : 2/5/6
1
2
4.4:1
4.4:1
98(42)e
99(40)e
74:26
53:47
75:25
55:45
aConditions: pentoses 4/1 (2 g, 13 mmol, 1 equiv), DMAP (10 mol %), amine (8 equiv), and Ac2O
b
(7 equiv) reacted in CH2Cl2 (10 mL) at 0 ˚C for 2 hours and at 25 ˚C for 2 hours. mixture (m/m) made
using commercially purchased D-xylose/L
-arabinose, pentose ratio determined using 1H NMR.
cbased on mass of all peracetylated C5 sugars present in isolated reaction mixture; value in parenthesis
d
is isolated yield of 5α. pentose, pyranose/furanose and anomer ratios determined by 1H NMR
e
of the reaction mixture after workup. 1-propanol used as crystallization solvent.
was very close to the ratio of the two starting pentoses. The ratio of 5α versus 2/5/6 in the reaction
residue was determined to be 74:26. Recystallization of the reaction mixture using 1-propanol allowed
1
for the isolation of the
L
-arabinose derivative 5α in 42 % yield. Melting point and H NMR analysis
19

ACCEPTED MANUSCRIPT
verified the identity of the crystalline white solid as 5α. 1H NMR analysis of crystallization filtrate
showed that it was depleted in 2β versus 2/5/6with an observed ratio of 53:47.
Peracetylation is a very common protection strategy that is widely implemented in carbohydrate
synthesis. In summary, we developed a straightforward, time-conservative that allows for stereo-selective
peracetylation of the C5 sugars, D-xylose and L-arabinose, to form either
α
or
β
pyranosyl esters in high
-xylose and L-
yields. The stereoselective peracetylation reactions were used to separate mixtures of
D
arabinose obtained from hydrolyzate of distillers dried grains (DDG) to provide pure pentoses as their
peracetylated derivatives in good yields.
Experimental Section
General Experimental Details. All reactions were carried out in oven-dried glassware under N2 unless
otherwise specified. N2 was purified by passage through columns containing activated molecular
sieves and Q-5 oxygen scavenger. Dichloromethane was purified by passage through columns of
activated 4 Å molecular sieves. All reagents were purchased from commercial sources and used
without further purification unless noted. 1H and 13C NMR spectra were recorded in capped tubes on
1
13
Varian 400 and 500 spectrometers at ambient probe temperature unless otherwise indicated. H and C
1
13
chemical shifts are reported versus SiMe4 and were determined by reference to the residual H and C
solvent peaks. Chemical shift, regiochemical and stereochemical assignments for the starting D-xylose
and L-arabinose were made by preparing samples of each in DMSO-d6 and obtaining NMR spectra for
1
13
1
both solutions. For
D
-xylose, H, C33, and multidimensional NMR (1H-1H COSY and H-13C one bond
correlation) were obtained and comparison to literature chemical shift values34 allowed for the assignment
of hydroxyl protons, the C1 (anomeric) proton and only
α-D-xylose was observed. For L-arabinose in
DMSO-d6, very little information exists in the literature except for hydroxyl proton chemical shift
1
13
1
values35 so H, C33, and multidimensional NMR (1H-1H COSY and H-13C one bond correlation) were
obtained to assign the anomeric proton and determine its stereochemcial orientation. Interpretation of the
data from these NMR experiments allowed for the assignment of the starting L-arabinose as the α anomer.
20

ACCEPTED MANUSCRIPT
α
-
D
-xylopyranose (1): 1H NMR (DMSO-d6):
δ 6.12 (d, J = 4.8, 1H, 1-OH), 4.84 (t, J = 8.4, 1H, H-1),
4.80 (d, J = 5.2, 1H, 4-OH), 4.68 (d, J = 4.4, 1H, 3-OH), 4.45 (d, J = 6.4, 1H, 2-OH), 3.47 (m, 1H, H-5),
3.35 (m, 2H, H-5’, H-3), 3.22 (m, 1H, H-4), 3.11 (m, 1H, H-2). 13C33 NMR (DMSO-d6):
δ 92.5 (C1), 73.2
(C3), 72.4 (C2), 70.2 (C4), 61.6 (C5).
α
-
L
-arabinopyranose (4): 1H NMR (DMSO-d6):
δ 6.00 (d, J = 5.6, 1H, 1-OH), 4.85 (dd, J = 5.2, 2.4,
1H, H-1), 4.49 (d, J = 5.2, 1H, 3-OH), 4.39 (s, 1H, 4-OH), 4.38 (d, J = 2.4, 1H, 2-OH), 3.67 (m, 2H, H-4,
H-5), 3.59 (m, 1H, H-3), 3.46 (m, 1H, H-2), 3.37 (dd, J = 12.4, 4.8, 1H, H-5’). 13C33 NMR (DMSO-d6):
δ
92.7 (C1), 69.5 (C3), 69.3 (C2), 67.5 (C4), 62.6 (C5).
Tetra-O-acetyl-α-D-xylopyranose (2α δ 6.21 (d, J = 4.0, H-1 of 2α), 5.42 (t, J =
): 1H NMR (CDCl3):
10.0, H-3 of 2α), 4.98 (m, H-2, H-4), 3.89 (dd, J = 10.8, 6.0 H-5 of 2α), 3.66 (t, J = 10.8, H-5’ of 2α),
2.14 (s, 3H, OAc CH3), 2.02 (s, 6H, OAc CH3), 1.99 (s, 3H, OAc CH3).
Tetra-O-acetyl-β-D-xylopyranose (2β δ 5.71 (d, J = 6.8, 1H, H-1), 5.20 (t, J = 8.0,
): 1H NMR (CDCl3):
1H, H-3), 5.01 (m, 2H, H-2, H-4), 4.14 (dd, J = 12.0, 4.8, 1H, H-5), 3.52 (dd, J = 12.0, 8.8, 1H, H-5’),
2.10 (s, 3H, OAc CH3), 2.06 (s, 3H, OAc CH3), 2.05 (s, 6H, OAc CH3).
Tetra-O-acetyl-α-L-arabinopyranose (5α δ 6.34 (d, J = 3.2, 1H, H-1), 5.35 (m, 3H,
): 1H NMR (CDCl3):
H-2), 4.06 (d, J = 13.2 1H, H-4), 3.82 (dd, J = 13.2, 2.0, 1H, H-3), 2.15 (s, 6H, OAc CH3), 2.02 (s, 6H,
OAc CH3).
Tetra-O-acetyl-
H-2, H-4), 5.05 (dd, J = 8.8, 3.6, 1H, H-3), 3.96 (dd, J = 13.2, 4.0, 1H, H-5).
Peracetylation of xylose in neat pyridine (Table 1, entry 1).16
-xylose (1) (8.00 g, 53.3 mmol) was
β-L-arabinopyranose (5β δ 5.60 (d, J = 7.2, 1H, H-1), 5.23 (m, 3H,
): 1H NMR (CDCl3):
α-D
added to a flask under N2. Pyridine (34.4 mL, 426.4 mmol) was added to the flask and the white
suspension was stirred at 25 ˚C. Acetic anhydride (40.3 mL, 426.4 mmol) was added slowly to the flask.
The suspension was stirred at 25 ˚C overnight. The solution turned clear after 15 minutes at 25 ˚C. After
21

ACCEPTED MANUSCRIPT
stirring overnight, the solution flask was quenched with chilled DI water (100 mL). The solution was
transferred to a 500 mL separatory funnel, washed with CH2Cl2 (3 x 100 mL) and the organic layer from
each wash was collected. The combined organic layer (CH2Cl2) was then washed with 1M HCl (3 x 100
mL), chilled DI water (4 x 100 mL), saturated NaHCO3 (2 x 100 mL), and chilled DI water (2 x 100 mL).
The CH2Cl2 layer was dried over Na2SO4, filtered through celite, and the solvent was removed under
vacuum to yield a yellow oil (20.14 g, 100%). 1H NMR characterization of the isolated reaction mixture
in CDCl3 showed it contained both tetraacetylated D-xylopyranoses 2α/β and tetraacetylated D-
xylofuranoses 3α/β in a ratio of 94:6. The ratio of stereoisomers/anomers was 2α:2β = 6:4 and 3α:3β =
1:1. 1H NMR (CDCl3): key resonances of four products
δ: 6.38 (d, J = 4.4, H-1 of 3α), 6.21 (d, J = 4.0,
H-1 of 2α), 6.07 (s, H-1 of 3
β
), 5.67 (d, J = 6.8, H-1 of 2
β
), 5.42 (t, J = 10.0, H-3 of 2α), 5.16 (t, J = 8.0,
), 3.89 (dd, J = 10.8,
H-3 of 2 ), 4.98 (m, H-2 of 2α/2
β
β
, H-4 of 2α/2 ), 4.10 (dd, J = 12.0, 4.8, H-5 of 2β
β
6.0, H-5 of 2α), 3.66 (t, J = 10.8, H-5’ of 2α), 3.48 (dd, J = 12.0, 8.8, H-5’ of 2
β).
Peracetylation of arabinose in neat pyridine (Table 1, entry 3).20
α-L-arabinose (4) (4.50 g, 30.0
mmol) was added to a flask under N2. Pyridine (19.3 mL, 240.0 mmol) was added to the flask and the
white suspension was stirred at 25 ˚C. Acetic anhydride (22.6 mL, 240.0 mmol) was added slowly to the
flask. The suspension was stirred at 25 ˚C overnight. After stirring overnight, the solution flask was
quenched with chilled DI water (75 mL). The solution was transferred to a 500 mL separatory funnel,
washed with CH2Cl2 (3 x 75 mL) and the organic layer from each wash was collected. The combined
organic layer (CH2Cl2) was then washed with 1M HCl (3 x 75 mL), chilled DI water (2 x 75 mL),
saturated NaHCO3 (1 x 75 mL), and chilled DI water (1 x 75 mL). The CH2Cl2 layer was dried over
Na2SO4, filtered through celite, and the solvent was removed under vacuum to yield a yellow oil (11.11 g,
1
100%). H NMR characterization of the isolated reaction mixture in CDCl3 showed it contained both
tetraacetylated
L-arabinopyranoses 5α/β and tetraacetylated
L-arabinofuranoses 6α/β in a ratio of 73:27.
1
The ratio of stereoisomers/anomers was 5α:5β = 6:4 and 6α:6β = 4:6. H NMR (CDCl3): key resonances
of four products
δ
6.38 (m, H-1 of 6α), 6.34 (d, J = 3.2, H-1 of 5α), 6.19 (s, H-1 of 6β), 5.65 (d, J = 7.2,
22

ACCEPTED MANUSCRIPT
H-1 of 5
), 4.37 (m, H-4 of 5
DMAP/NEt3 mediated peracetylation of xylose (Table 2, entry 1).
β
), 5.36 (m, H-2 of 5
α
), 5.32 (m, H-4 of 5
β
), 5.30 (m, H-2 of 5
β), 5.11 (dd, J = 8.8, 3.6, H-3 of
5
β
α
), 4.22 (m, H-3 of 5
α), 3.78 (m, H-5 of 5
α
and 5 ).
β
α-D-xylose (1) (2.03 g, 13.32
mmol) and DMAP (0.17 g, 1.33 mmol) were added to a flask under N2. CH2Cl2 (10 mL) and NEt3 (15
mL, 106.56 mmol) were added to the flask and the white suspension was stirred at 25 ˚C. The flask was
cooled to 0 ˚C and stirred for 10 minutes. Acetic anhydride (9 mL, 93.25 mmol) was added dropwise to
the flask. The white suspension was stirred at 0 ˚C for 1 hour. After 1 hour, the clear, yellow solution was
removed from the ice bath and warmed to 25 ˚C. The solution was stirred at 25 ˚C for 4 hours. After 4
hours, the solution flask was placed in an ice bath and quenched with chilled DI water (20 mL). The ice
bath was removed after 20 minutes and the bi-phasic, yellow solution was stirred for 12 hours. The
solution was transferred to a 250 mL separatory funnel, washed with CH2Cl2 (3 x 20 mL) and the organic
layer from each wash was collected. The combined organic layer (CH2Cl2) was then washed with 1M
HCl (4 x 20 mL), chilled DI water (3 x 20 mL), saturated NaHCO3 (3 x 20 mL), and chilled DI water (2 x
20 mL). The CH2Cl2 layer was dried over MgSO4, filtered through celite, and the solvent was removed
1
under vacuum to yield a yellow glass (4.20 g, 99%). H NMR characterization of the isolated reaction
mixture in CDCl3 showed it contained 2α as the major product (98%) and a trace amount of 2β (2%).
DMAP/NEt3 mediated peracetylation of xylose (Table 2, entry 3). α-D-xylose (1) (2.00 g, 13.32
mmol) and DMAP (0.17 g, 1.33 mmol) were added to a flask under N2. CH2Cl2 (10 mL) and NEt3 (15
mL, 106.56 mmol) were added to the flask and the white suspension was stirred at 25 ˚C. The flask was
cooled to 0 ˚C and stirred for 10 minutes. Acetic anhydride (9 mL, 93.25 mmol) was added dropwise to
the flask. The white suspension was stirred at 0 ˚C for 30 minutes. After 30 minutes, the clear, yellow
solution was removed from the ice bath and warmed to 25 ˚C. The solution was stirred at 25 ˚C for 2
hours. After 2 hours, the solution flask was placed in an ice bath and quenched with chilled DI water (20
mL). The ice bath was removed after 20 minutes. The bi-phasic, yellow solution was then transferred to a
250 mL separatory funnel, washed with CH2Cl2 (3 x 20 mL) and the organic layer from each wash was
collected. The combined organic layer (CH2Cl2) was then washed with 1M HCl (4 x 20 mL), chilled DI
23

ACCEPTED MANUSCRIPT
water (3 x 20 mL), saturated NaHCO3 (3 x 20 mL), and chilled DI water (2 x 20 mL). The CH2Cl2 layer
was dried over MgSO4, filtered through celite, and the solvent was removed under vacuum to yield a
yellow glass (3.98 g, 94%). 1H NMR characterization of the isolated reaction mixture in CDCl3 showed it
contained 2α as the major product (98%) and a trace amount of 2β (2%).
DMAP/NEt3 mediated peracetylation of xylose (Table 2, entry 4). α-D-xylose (1) (2.00 g, 13.32
mmol) and DMAP (0.17 g, 1.33 mmol) were added to a flask under N2. CH2Cl2 (10 mL) and NEt3 (15
mL, 106.56 mmol) were added to the flask and the white suspension was stirred at 25 ˚C. The flask was
cooled to 0 ˚C and stirred for 10 minutes. Acetic anhydride (9 mL, 93.25 mmol) was added dropwise to
the flask. The white suspension was stirred at 0 ˚C for 30 minutes. After 30 minutes, the clear, yellow
solution was removed from the ice bath and warmed to 25 ˚C. The solution was stirred at 25 ˚C for 30
minutes. After 30 minutes, the solution flask was placed in an ice bath and quenched with chilled DI
water (20 mL). The ice bath was removed after 20 minutes. The bi-phasic, yellow solution was then
transferred to a 250 mL separatory funnel, washed with CH2Cl2 (3 x 20 mL) and the organic layer from
each wash was collected. The combined organic layer (CH2Cl2) was then washed with 1M HCl (4 x 20
mL), chilled DI water (3 x 20 mL), saturated NaHCO3 (3 x 20 mL), and chilled DI water (2 x 20 mL).
The CH2Cl2 layer was dried over MgSO4, filtered through celite, and the solvent was removed under
vacuum to yield a yellow glass (4.26 g, 100%). 1H NMR characterization of the isolated reaction mixture
in CDCl3 showed it contained 2α as the major product (98%) and a trace amount of 2β (2%).
DMAP/NEt3 mediated peracetylation of xylose (Table 2, entry 6).
α-D-xylose (1) (2.00 g, 13.32
mmol) and DMAP (0.17 g, 1.33 mmol) were added to a flask under N2. CH2Cl2 (10 mL) and NEt3 (15
mL, 106.56 mmol) were added to the flask and the white suspension was stirred at 25 ˚C. The flask was
cooled to 0 ˚C and stirred for 10 minutes. Acetic anhydride (9 mL, 93.25 mmol) was added dropwise to
the flask. The white suspension was stirred at 0 ˚C for 30 minutes. After 30 minutes, chilled DI water (20
mL) was added to the clear, yellow solution and the resulting biphasic solution was stirred at 0 ˚C for 20
minutes. The solution was transferred to a 250 mL separatory funnel, washed with CH2Cl2 (3 x 20 mL)
24

ACCEPTED MANUSCRIPT
and the organic layer from each wash was collected. The combined organic layer (CH2Cl2) was then
washed with 1M HCl (4 x 20 mL), chilled DI water (3 x 20 mL), saturated NaHCO3 (3 x 20 mL), and
chilled DI water (2 x 20 mL). The CH2Cl2 layer was dried over MgSO4, filtered through celite, and the
1
solvent was removed under vacuum to yield a yellow glass (4.23 g, 99%). H NMR characterization of
the isolated reaction mixture in CDCl3 showed it contained 2α as the major product (98%) and a trace
amount of 2β (2%).
DMAP/NEt3 mediated peracetylation of xylose (Table 2, entry 7). α-D-xylose (1) (2.03 g, 13.32
mmol) and DMAP (0.17 g, 1.33 mmol) were added to a flask under N2. CH2Cl2 (10 mL) and NEt3 (15
mL, 106.56 mmol) were added to the flask and the white suspension was stirred at 25 ˚C. Acetic
anhydride (9 mL, 93.25 mmol) was added dropwise to the flask, where the white suspension warmed to
the touch. The white suspension was stirred at 25 ˚C for 2 hours and 30 minutes. After 2 hours and 30
minutes, the solution flask was placed in an ice bath and quenched with chilled DI water (20 mL). The ice
bath was removed after 20 minutes and the bi-phasic, yellow solution was transferred to a 250 mL
separatory funnel. The solution was washed with CH2Cl2 (3 x 20 mL) and the organic layer from each
wash was collected. The combined organic layer (CH2Cl2) was then washed with 1M HCl (4 x 20 mL),
chilled DI water (3 x 20 mL), saturated NaHCO3 (3 x 20 mL), and chilled DI water (2 x 20 mL). The
CH2Cl2 layer was dried over MgSO4, filtered through celite, and the solvent was removed under vacuum
to yield a yellow syrup (3.67 g, 87%). 1H NMR characterization of the isolated reaction mixture in CDCl3
showed it contained both 2α/β and 3α/β in a ratio of 93:7. The ratio of stereoisomers/anomers was 2α:2β
= 72:28 and 3α:3β = 1:1.
DMAP/pyridine mediated peracetylation of xylose (Table 2, entry 8). α-D-xylose (1) (2.00 g, 13.32
mmol) and DMAP (0.17 g, 1.33 mmol) were added to a flask under N2. CH2Cl2 (10 mL) and Pyridine (9
mL, 106.56 mmol) were added to the flask and the white suspension was stirred at 25 ˚C. The flask was
cooled to 0 ˚C and stirred for 10 minutes. Acetic anhydride (9 mL, 93.25 mmol) was added dropwise to
the flask. The white suspension was stirred at 0 ˚C for 1 hour. After 1 hour, the clear, colorless solution
25

ACCEPTED MANUSCRIPT
was removed from the ice bath and warmed to 25 ˚C. The solution was stirred at 25 ˚C for 2 hours. After
2 hours, the solution flask was placed in an ice bath and quenched with chilled DI water (20 mL). The ice
bath was removed after 20 minutes and the bi-phasic, colorless solution was stirred for 12 hours. The
solution was transferred to a 250 mL separatory funnel, washed with CH2Cl2 (3 x 20 mL) and the organic
layer from each wash was collected. The combined organic layer (CH2Cl2) was then washed with 1M
HCl (4 x 20 mL), chilled DI water (3 x 20 mL), saturated NaHCO3 (3 x 20 mL), and chilled DI water (2 x
20 mL). The CH2Cl2 layer was dried over MgSO4, filtered through celite, and the solvent was removed
1
under vacuum to yield a colorless syrup (4.12 g, 97%). H NMR characterization of the isolated reaction
mixture in CDCl3 showed it contained 2α as the major product (99%) and a trace of 2β (1%).
DMAP/NEt3 mediated peracetylation of xylose (Table 2, entry 9). α-D-xylose (1) (10.0 g, 66.6 mmol)
and DMAP (0.817 g, 6.66 mmol) were added to a flask under N2. CH2Cl2 (50 mL) and NEt3 (74 mL,
532.8 mmol) were added to the flask and the white suspension was stirred at 25 ˚C. The flask was cooled
to 0 ˚C and stirred for 10 minutes. Acetic anhydride (44 mL, 466.2 mmol) was added in 15 mL portions to
the flask. The Ac2O portions were added in 10 minute intervals to prevent excessive heating of the
mixture. The white suspension was stirred at 0 ˚C for 1 hour. After 1 hour, the clear, yellow-orange
solution was removed from the ice bath and warmed to 25 ˚C. The solution was stirred at 25 ˚C for 19
hours. After 19 hours, the solution flask was placed in an ice bath and quenched with chilled DI water
(100 mL). The ice bath was removed after 20 minutes and the bi-phasic, yellow-orange solution was
stirred for 12 hours. The reaction mixture was transferred to a 1 L separatory funnel and was washed with
CH2Cl2 (3 x 100 mL) and the organic layer from each wash was collected. The combined organic layer
(CH2Cl2) was then washed with 1M HCl (4 x 100 mL), chilled DI water (3 x 100 mL), saturated NaHCO3
(3 x 100 mL), and chilled DI water (2 x 100 mL). The CH2Cl2 layer was dried over MgSO4, filtered
1
through celite, and the solvent was removed under vacuum to yield an orange glass (23.56 g, 100%). H
NMR characterization of the isolated reaction mixture in CDCl3 showed it contained 2α as the major
product (96%) and a trace amount of 2β (4%).
26

ACCEPTED MANUSCRIPT
DMAP/NEt3 mediated peracetylation of arabinose (Table 3, entry 1).
α-L-arabinose (4) (2.00 g,
13.32 mmol) and DMAP (0.17 g, 1.33 mmol) were added to a flask under N2. CH2Cl2 (10 mL) and NEt3
(15 mL, 106.56 mmol) were added to the flask and the white suspension was stirred at 25 ˚C. The flask
was cooled to 0 ˚C and stirred for 10 minutes. Acetic anhydride (9 mL, 93.25 mmol) was added dropwise
to the flask. The white suspension was stirred at 0 ˚C for 2 hours. After 2 hours, the clear, yellow solution
was removed from the ice bath and warmed to 25 ˚C. The solution was stirred at 25 ˚C for 2 hours. After
2 hours, the solution flask was placed in an ice bath and quenched with chilled DI water (20 mL). The ice
bath was removed after 20 minutes and the bi-phasic, yellow solution was transferred to a 250 mL
separatory funnel. The solution was washed with CH2Cl2 (3 x 20 mL) and the organic layer from each
wash was collected. The combined organic layer (CH2Cl2) was then washed with 1M HCl (4 x 20 mL),
chilled DI water (3 x 20 mL), saturated NaHCO3 (3 x 20 mL), and chilled DI water (2 x 20 mL). The
CH2Cl2 layer was dried over MgSO4, filtered through celite, and the solvent was removed under vacuum
to yield a yellow crystal (3.75 g, 88%). 1H NMR characterization of the isolated reaction mixture in
CDCl3 showed it contained both 5α/β and 6α/β in a ratio of 95:5. The ratio of stereoisomers/anomers
was 5α:5β = 99:1 and 6α:6β = 1:1.
DMAP/NEt3 mediated peracetylation of arabinose (Table 3, entry 2).
α-L-arabinose (4) (2.00 g,
13.32 mmol) and DMAP (0.17 g, 1.33 mmol) were added to a flask under N2. CH2Cl2 (10 mL) and NEt3
(15 mL, 106.56 mmol) were added to the flask and the white suspension was stirred at 25 ˚C. The flask
was cooled to 0 ˚C and stirred for 10 minutes. Acetic anhydride (9 mL, 93.25 mmol) was added dropwise
to the flask. The white suspension was stirred at 0 ˚C for 2 hours and 30 minutes. After 2 hours and 30
minutes, the clear, yellow solution was removed from the ice bath and warmed to 25 ˚C. The solution was
stirred at 25 ˚C for 2 hours. After 2 hours, the solution flask was placed in an ice bath and quenched with
chilled DI water (20 mL). The ice bath was removed after 20 minutes and the bi-phasic, yellow solution
was transferred to a 250 mL separatory funnel. The solution was washed with CH2Cl2 (3 x 20 mL) and
the organic layer from each wash was collected. The combined organic layer (CH2Cl2) was then washed
27

ACCEPTED MANUSCRIPT
with 1M HCl (4 x 20 mL), chilled DI water (3 x 20 mL), saturated NaHCO3 (3 x 20 mL), and chilled DI
water (2 x 20 mL). The CH2Cl2 layer was dried over MgSO4, filtered through celite, and the solvent was
1
removed under vacuum to yield a yellow crystal (3.84 g, 91%). H NMR characterization of the isolated
reaction mixture in CDCl3 showed it contained both 5α/β and 6α/β in a ratio of 95:5 while the ratio of 5α
to 5β/6α/6β was 93:7. The ratio of stereoisomers/anomers was 5α:5β = 98:2 and 6α:6β = 3:2. Absolute
EtOH (60 ml) was used to transfer the crystals to a 100 mL beaker with stir bar. The mixture was heated
and stirred to dissolve the crystals. The solution volume was reduced to 30 mL with heating. The beaker
was then removed from the heat and allowed to cool to 25 ˚C. The solution was agitated which yielded
white crystals. The crystals were collected by filtration with a Buchner funnel and dried to afford a white,
1
crystalline solid (2.38 g, 56%), mp 96-97 ˚C. The solid was characterized by H NMR (CDCl3) and
identified as 5α. The solvent from the filtrate was removed under vacuum to yield 1.12 g of an orange oil.
1H NMR characterization of the filtrate residue in CDCl3 showed the ratio of 5α/β and 6α/β was 85:15
while the ratio of 5α to 5β/6α/6β had changed to 81:19. The ratio of stereoisomers/anomers was 5α:5β =
95:5 and 6α:6β = 63:37.
DMAP/NEt3 mediated peracetylation of arabinose (Table 3, entry 3).
α-L-arabinose (4) (2.00 g,
13.32 mmol) and DMAP (0.17 g, 1.33 mmol) were added to a flask under N2. CH2Cl2 (10 mL) and NEt3
(15 mL, 106.56 mmol) were added to the flask. Acetic anhydride (9 mL, 93.25 mmol) was added
dropwise to the flask, where the white suspension warmed to the touch. The solution was stirred at 25 ˚C
for 2 hours and 30 minutes. After 2 hours and 30 minutes, the solution flask was placed in an ice bath and
quenched with chilled DI water (20 mL). The ice bath was removed after 20 minutes and the bi-phasic,
yellow solution was allowed to stir for 12 hours. The solution was transferred to a 250 mL separatory
funnel, washed with CH2Cl2 (3 x 20 mL) and the organic layer from each wash was collected. The
combined organic layer (CH2Cl2) was then washed with 1M HCl (4 x 20 mL), chilled DI water (3 x 20
mL), saturated NaHCO3 (3 x 20 mL), and chilled DI water (2 x 20 mL). The CH2Cl2 layer was dried over
MgSO4, filtered through celite, and the solvent was removed under vacuum to yield a yellow syrup (4.17
28

ACCEPTED MANUSCRIPT
1
g, 98%). H NMR characterization of the isolated reaction mixture in CDCl3 showed it contained both
5α/β and 6α/β in a ratio of 77:23. The ratio of stereoisomers/anomers was 5α:5β = 85:15 and 6α:6β =
53:47.
DMAP/pyridine mediated peracetylation of arabinose (Table 3, entry 4).
α-L-arabinose (4) (2.00 g,
13.32 mmol) and DMAP (0.17 g, 1.33 mmol) were added to a flask under N2. CH2Cl2 (10 mL) and
pyridine (9 mL, 106.56 mmol) were added to the flask and the suspension was stirred at 25 ˚C. The white
suspension was then stirred at 0 ˚C for 10 minutes. Acetic anhydride (9 mL, 93.25 mmol) was added
dropwise to the flask. The mixture was stirred at 0 ˚C for 4 hours and 30 minutes. After 4 hours and 30
minutes, the clear, colorless solution was stirred at room temperature for 2 hours. After 2 hours, the
reaction flask was placed in an ice bath and the solution was quenched with chilled DI water (20 mL).
The ice bath was removed after 20 minutes and the bi-phasic, colorless solution was allowed to stir for 12
hours. The solution was transferred to a 250 mL separatory funnel, washed with CH2Cl2 (3 x 20 mL) and
the organic layer from each wash was collected. The combined organic layer (CH2Cl2) was then washed
with 1M HCl (4 x 20 mL), chilled DI water (3 x 20 mL), saturated NaHCO3 (3 x 20 mL), and chilled DI
water (2 x 20 mL). The CH2Cl2 layer was dried over MgSO4, filtered through celite, and the solvent was
1
removed under vacuum to yield clear crystals (4.27 g, 100%). H NMR characterization of the isolated
reaction mixture in CDCl3 showed it contained both 5α/β and 6α/β in a ratio of 94:6. The ratio of
stereoisomers/anomers was 5α:5β = 98:2 and 6α:6β = 2:1.
NEt3 mediated peracetylation of xylose (Table 4, entry 1). α-D-xylose (1) (2.00 g, 13.32 mmol) was
added to a 100 mL round bottom flask under N2. CH2Cl2 (10 mL) and NEt3 (15 mL, 106.56 mmol) were
added to the flask and the white suspension was stirred at 25 ˚C. The flask was cooled to 0 ˚C and stirred
for 10 minutes. Ac2O (9 mL, 93.25 mmol) was added dropwise to the flask. The white suspension was
stirred at 0 ˚C for 2 hours. After 2 hours, the solution was still a white suspension and the flask was
removed from the ice bath and warmed to 25 ˚C. The reaction mixture was stirred at 25 ˚C for 2 hours.
After 2 hours, the clear yellow solution was placed in an ice bath and quenched with chilled DI water (20
29