Catalytic and Atom-Economic Glycosylation using Glycosyl Formates and Cheap Metal Salts

Article abstract of DOI:10.1002/cssc.202000733

Benzylated glycosyl formates have been synthesized in one step from the corresponding hemiacetal or orthoester with formic acid as the sole reagent. The glycosyl formates are used as glycosyl donors under catalytic conditions with cheap metal catalysts ba

Full text of DOI:10.1002/cssc.202000733

Chemistry–Sustainability–Energy–Materials
Accepted Article
Title: Catalytic and Atom Economic Glycosylation using Glycosyl
Formates and Cheap Metal Salts.
Authors: Liang Yang, Christian Herrstedt Hammelev, and Christian
Marcus Pedersen
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemSusChem 10.1002/cssc.202000733
Link to VoR: https://doi.org/10.1002/cssc.202000733
01/2020

10.1002/cssc.202000733
ChemSusChem
COMMUNICATION
Catalytic and Atom Economic Glycosylation using Glycosyl
Formates and Cheap Metal Salts.
Liang Yang,[a]$ Christian H. Hammelev[a]$ and Prof. Christian M. Pedersen*[a]
Dedicated to Professor Richard R. Schmidt on the occasion of his 85th birthday.
Abstract: Benzylated glycosyl formates have been synthesized in
one-step from the corresponding hemiacetal or orthoester using
formic acid as the sole reagent. The glycosyl formates were used as
glycosyl donors under catalytic conditions using cheap metal catalysts
based on iron or bismuth. A 13C-NMR method was developed and
evaluated for screening reactions conditions, giving precise
information on selectivity, yield and by-products formed. The major
side reaction observed was the transesterification giving the
formylated acceptor and returning the hemi-acetal. Using this
approach catalyst loadings and solvents were optimized and the
scope of the glycosylation was evaluated using a variety of glycosyl
donors and acceptors. Proof of concept for a traceless glycosylation
utilizing a dual purpose iron catalyst that both catalyze glycosylation
and dehydrogenation of formic acid, has furthermore been provided.
esters have recently been demonstrated using several donor
types and activation by metal triflates, albeit still with a limited
scope.[4,10–13] The simplest imaginable donor, with an ester leaving
group, would be a glycosyl formate. As it comes to our knowledge,
surprisingly, formates have not been studied as glycosyl donors.
Strikingly few reports on their synthesis have appeared and none
in the context of glycosylation chemistry. The first synthesis of
glycosyl formate came from Hough and Lewis, who isolated 1-O-
formoyl--D-glucopyranose tetraacetate as an undesired
byproduct.[14] Formates have later been synthesized as a
precursor for the enol ether obtained via a Wittig-type reaction by
Vodcadlo and co-workers.[15] Frauenrath and coworkers prepared
1-O-vinyl glycosides from the corresponding formates using
Tebbe’s reagent.[16] Here, we present our approach to develop a
simple glycosylation method, which uses cheap reagents, is atom
economic, environmental benign and catalytic.
Introduction
Results and Discussion
Catalytic glycosylation has received increasing interest in the last
decades. Simple glycosides can readily be synthesized using
Fischer’s classic method, but only when the nucleophile can be
used in a huge excess, i.e. often as the solvent. An early example
of a catalytic glycosylation in oligosaccharide synthesis came
from the Kochetkov group, who introduced orthoesters as
glycosyl donors in the early Sixties.[1] For glycosylations that are
more complicated glycosyl fluorides[2] and trichloroacetimidates
(TCAs)[3] became the method of choice in the early Eighties.
Today, TCAs are still one of the most popular glycosyl donors for
complex glycosylations, such as oligo- and glycoconjugate
synthesis.[4] The closely related N-Phenyl trifluoroacetimidates
have also become an important donor type being more stable
than the TCAs.[5] In recent years gold catalyzed activation of
alkyne derivatives have received a lot of attention and shown their
worth in syntheses of natural products.[6,7] Despite these well-
developed and popular methods, there is still a need for
improvements. Especially the cost of specialty reagents, stability
and donor preparation are limitations for several of the methods
mentioned above. There are also environmental concerns using
halogenated reagents, heavy metal promoters and it is
undesirable in terms of atom economy to use large leaving groups
in glycosylation reactions on larger scales. As none of the
commonly used donor system, used in catalytic glycosylation,
fulfill the demands for a scalable, cheap and environmentally
friendly glycosylation protocol, a fundamentally new approach is
required. A simplification could be to use glycosyl esters, but
these are generally considered to be poor donors, which requires
relative harsh activation conditions, thereby limiting their use in
oligosaccharide synthesis.[8],[4] Methods using a remote activation
and hence a more complicated aglycon, have been developed to
cope with the low reactivity, but these methods are generally not
catalytic and required expensive reagents in stoichiometric
amounts, such as Tf2O, for the activation.[9] Catalytic activation of
The previous syntheses of glycosyl formates used glycosyl
bromides,[14,16] glycosyl chlorides,[15] thioglycosides[16] or coupling
reactions (EDC) with the hemiacetal.[16] All methods, which are
out of our scope, i.e. to make a simple and cheap glycosylation
method.
Scheme 1 Optimized synthesis of glycosyl formates using neat HCOOH.
We therefore decided to investigate other methods for the
formylation, with the aim to go directly from the hemiacetal using
a simple formylation reagent. First approach was to use a mixed
anhydride, trimethylacetic formic anhydride,[17] which is readily
synthesized from the trimethylacetyl chloride and sodium formate,
but unfortunately this reagent is not shelf stable. As a model
substrate we chose 2,3,4,6-tetra-O-benzyl glucose, which was
formylated to give the desired formate as mainly the -anomer
( = 22:78) in 34% yield. As the yield was modest and the
reagents not practical to use, alternative methods were therefore
sought. The ideal reagents in terms of atom economy, would be
to use formic acid it selves. Inspired by the work of Rostami[18] on
simple alcohols, neat formic acid was used together with the
substrate to give a satisfying 52% yield (74% based on recovered
starting material) as mainly the -product ( = 9:1). Full
conversion of the hemiacetal was not achievable and attempts to
drag the equilibrium by adding drying reagents did not change the
outcome of the reaction. Addition of iodine as a catalyst, was
found to be needless and hence only formic acid and the
hemiacetal had to be mixed (Scheme 1). With access to the
This article is protected by copyright. All rights reserved.

10.1002/cssc.202000733
ChemSusChem
COMMUNICATION
glycosyl formate D1 its glycosylation properties could be
evaluated.
In order to quantify the products an inert internal reference
compound with a NOE enhancement factor similar to the products
was needed. Mesitylene was found to be a good candidate with a
simple signal pattern (138, 127 and 21 ppm respectively), which
was not overlapping with the product signals in the crude reaction
mixture. The NMR-quantification was first optimized and
evaluated by recording the spectra at different delay times on a
model sample containing the glycoside 1. The accuracy of the
calculated amount of 1, based on integrals, were then obtained
at different delay times. From this it was clear that the mesitylene
signal at 127ppm gave the most precise and consistent
measurements and this even after a D1 of 5s (See SI for details).
The method was evaluated by adding a known amount of another
sugar based compound, 1 or peracetylated -D-glucopyranose,
showing satisfying accuracy between the NMR determined and
weighted amount. With the NMR parameters in place the
screening of glycosylation conditions could begin by using
standard conditions varying one parameter at the time. As a
model substrate D1 was used as the donor and methoxyethanol
as the acceptor, since it resembles the stereoelectronics of a
sugar alcohol. First different catalysts (25 mol%) and promoters
were screened in CH2Cl2 at 25°C with a reaction time of 20h (See
Table S1 in SI for detailed information).
Scheme 2 Glycosylation using Mukaiyama’s catalytic system.
As transesterification is a known problem when using glycosyl
esters, the initial glycosylation conditions were adapted from
Mukaiyama, who used TMS protected acceptors to lower the
nucleophilicity and thereby limit the transesterification.[19]
Therefore, a combination of SnCl4 and AgClO4 as the catalytic
system, together with TMS protected 2-methoxy ethanol as the
acceptor, was used in the first experiments (Scheme 2).
Pleasingly, the donor was fully converted at 0°C, but the outcome
was a mixture of the desired glycosides (46%), trehaloses and
hemiacetals.
To account for this product limiting parameter in the study, a GT-
selectivity was introduced, as a measurement (GT: Glycosylation
Transesterfication). This parameter simply gives the selectivity
towards the glycosylation product – higher percentage equals
more glycosylation product and hence less transesterification.
(
)
퐺푇 − 푠푒푙푒푐푡푖푣푖푡푦 = 푛 퐺푙푦푐표푠푦푙푎푡푖표푛 푝푟표푑푢푐푡 /((푛(퐺푙푦푐표푠푦푙푎푡푖표푛 푝푟표푑푢푐푡) +
푛(푇푟푎푛푠푒푠푡푒푟푖푓푖푐푎푡푖표푛 푝푟표푑푢푐푡)) 푥 100%
The GT-selectivity in the initial glycosylation was 75%, which is
reasonable, but the formation of trehaloses, suggested that the
silylated acceptor is competing with the liberated hemiacetal as
the nucleophile (glycosyl acceptor). Because of these problems
together with the price and handling of the catalysts, the approach
using a silylated acceptor was discontinued. Activation using
unprotected alcohols as acceptors and Lewis acid catalysts was
therefore studied next (see SI for reaction schemes). BF3·OEt2 (3
equiv.) was used at rt and gave a high GT-selectivity of 91%, a
high selectivity and a yield of 67%. Using FeCl3 (3 equiv.)
maintained the high GT-selectivity of 92% and increased the yield
to 82%, albeit with a lower -selectivity. Using a strong Brønsted
acid, Tf2NH (20 mol%), resulted in low GT-selectivity and yield
and hence seems not to be a viable approach for this catalytic
glycosylation. From this initial catalyst screening it was clear that
finding the optimal conditions for the activation of glycosyl
formates required a fast and precise method to quantify the
various products of the complex reactions. Proton NMR did not
provide good enough resolution to distinguish and quantify the
Scheme 3 Conditions for NMR screening of catalysts.
The strong Brønsted acid, TfOH, gave full conversion and a yield
of 51%, but a moderate GT-selectivity of 60%. Weak Lewis acids
did not activate the donor (LiOTf, AgOTf or KPF6). Boron based
Lewis acids (BF3•OEt2 or B(C6F5)3) did yield the glycoside, albeit
not with full conversion of the donor and only a modest GT-
selectivity, when used in catalytic amounts. From the screening,
it was found that Fe3+ and Bi3+ salts were the best catalysts in
terms of GT-selectivity, yield and anomeric selectivity. The anions
used did also influence the reactions and generally the triflates
were performing better than the corresponding halide salts (FeCl3
vs Fe(OTf)3 and BiBr3 vs Bi(OTf)3. The role of the triflate ions
could be to stabilize a transient intermediate and thereby limiting
the transesterification pathway.[21] Addition of KPF6 as a non-
coordinating anion did influence both the Fe(OTf)3 and Bi(OTf)3
catalyzed reactions, but in different ways. Using Fe(OTf)3 together
with KPF6 diminished the -selectivity, whereas the selectivity
crude reactions, which consisted of the glycosyl donor, glycosides, increased when using Bi(OTf)3 and KPF6, which is somewhat
hemiacetals and trehaloses - all of them as a mixture of anomers.
13C-NMR on the other hand gave well-resolved anomeric peaks
located in the region between 90 and 105 ppm. There are
however problems associated with using 13C-NMR for
quantification: firstly the low abundance of 13C and therefore low
signal strength and secondly its long spin-lattice relaxation time
(T1) and decoupling of attached protons, which results in nuclear
Overhauser effect distorting the integrals. The signal strength
could be overcome by using a NMR spectrometer with a cryo-
probe, high sample concentration and more scans. The problem
with signal inconsistency related to long relaxation times was to
be solved by increasing the delay time to 4 times T1.
Carbohydrates in CDCl3 have T1’s in the range of 0.6s,[20] i.e. ca.
2.4 s between scans should be sufficient for complete relaxation
of the carbohydrate.
puzzling. KPF6 alone did not activate the donor. Adding an acid
scavenger, 2,4,6-tri-tert-butylpyrimidine (TTBP), quenched the
reactions catalyzed with either Fe(OTf)3 or Bi(OTf)3. This either by
complexing the LA or trapping a small amount of TfOH or formic
acid formed, which would then suggest a Brønsted acid to be the
actual catalyst. From the screening, Fe(OTf)3 and Bi(OTf)3/KPF6
were found to be the best catalysts in terms of GT-selectivity, -
selectivity and yield. To study the influence of a coordinating
solvent, the reactions with the best performing catalysts, from the
screening, were repeated in THF. The reactions were generally
slower and only the best catalyst gave full conversion. The GT-
selectivity was generally as good or better than the reactions in
CH2Cl2 and yields of up to 91% could be achieved using
Bi(OTf)3/KPF6 and 79% with Fe(OTf)3 as the catalyst, but this is
without anomeric selectivity. Zn(OTf)2, BF3·OEt2 or I2 resulted in
This article is protected by copyright. All rights reserved.

10.1002/cssc.202000733
ChemSusChem
COMMUNICATION
slow reactions (see SI for details). To study the effect of other
solvents, Fe(OTf)3 was chosen as the catalyst and the
glycosylation performed in eight common organic solvents.
Except when using diglyme, full conversion of the donor was
observed in all solvents. Interestingly, dioxane gave as high -
selectivity as CH2Cl2, whereas THF did not (Table 1; entry 1, 4
and 2). MeCN gave a small excess of the -anomer, whereas the
non-polar solvents toluene and trichloroethylene both gave low
selectivity and yield (entry 7 and 8). The GT selectivity remained
high for all solvents except diglyme. Using HFIP did not provide
any glycosylation product (Table 1; entry 6).
6
7
Bi(OTf)3 (5)
Bi(OTf)3 (25)
KPF6(25%)
Bi(OTf)3 (10)
KPF6(25%)
Bi(OTf)3 (5)
KPF6(120)
100
100
57
79
47:53
48:52
57
74
8
100
40
82
21
95
50
26
47:53
44:56
47:53
44:56
50:50
81
7
9
10
11
12
Bi(OTf)3 (25)
KPF6(120%)
Bi(OTf)3 (10)
KPF6 (120%)
Bi(OTf)3 (5)
KPF6 (120)
100
47
91
19
4
16
Table 1 Solvent screening.
Entry
Solvent
Conv.%
*
GT-sel.
(%)*
:*
Yield
(%)*
Reaction conditions: 180 mol D1, 1.5 eq. 2-methoxyethanol and catalyst
in 2 mL THF at rt for 20h. *Data obtained from crude 13C-NMR.
1
2
3
4
5
6
7
8
CH2Cl2
THF
100
100
100
100
28
76
80
81
77
38
-
76:24
48:52
45:55
75:25
60:40
-
55
79
64
70
5
With the optimized reaction conditions established, the acceptor
scope was studied using three other alcohols as nucleophiles.
Cyclohexanol was glycosylated using D1 giving the glucoside as
an  mixture in 69% yield (entry 2, table 3). The more hindered
acceptor adamantanol A1 gave a similar yield (62%, but with
some -selectivity (entry 3, table 3). Glucosylating the less
reactive sugar alcohol A2 gave a satisfactory isolated yield of 71%,
but low anomeric selectivity. The much less reactive 4-OH
glycosyl acceptor A3 gave 32% yield and no selectivity (entry 3-
5, table 3). From these initial results with the common benzylated
glycosyl donor a clear tendency to diminished yields, when
glycosylating more hindered acceptors, was observed. In order to
study the influence of the donor properties on the glycosylation
outcome it was decided to synthesize three other perbenzylated
glycosyl formates with different stereochemistry and reactivity
profiles. The galactosyl formate was synthesized via the
trichloroethyl galactoside obtained from the peracetylated sugar
using BF3O•Et2O as the Lewis acid. Benzylation followed by zinc
mediated anomeric deprotection gave the hemiacetal in modest
yields. The corresponding perbenzylated manno hemiacetal was
synthesized from the perbenzylated mannose by acid treatment
in 63% over 2 steps. Finally, the perbenzylated xylo hemiacetal
was synthesized from the methyl xylopyranoside, by
perbenzylation and anomeric deprotection by aqueous acid.
Introduction of the formyl group was performed in neat formic acid.
MeCN
1,4-Dioxane
Diglyme
HFIP
100
100
100
0
PhMe
68
76
49:51
56:44
46
40
CHClCCl2
Reaction conditions: 180 mol D1, 1.5 equiv. 2-methoxyethanol and 25 mol%
Fe(OTf)3 in 2 mL solvent at rt for 3h. *Data obtained from crude 13C-NMR.
Next, the dependence of the catalyst and additive loading was
studied in THF. Fe(OTf)3 still performed satisfactory with 10%
loading, but both GT selectivity and conversion took a hit, when
lowering the loading to 5% (Table 2, entry 1-3). This was not the
case when using Bi(OTf)3, where the results remained unaffected
by lowering the loading (Table 2, entry 4-6). Addition of the KPF6
(20 %), did however change this: at higher loadings the yield and
GT-selectivities were improved, but at 5% only 40% conversion
and a low GT-selectivity could be achieved. Adding an excess of
KPF6 (120%) did not improve the outcome of the reaction, but
lowered both the conversion and the GT selectivity further,
although without affecting the anomeric selectivity (Table 2, entry
10-12). The optimal conditions were found to be Bi(OTf)3 (25%) in
combination with the additive KPF6 (120%) or alternatively
Fe(OTf)3 down to 10% loading.
Figure 1 Glycosyl formates synthesized and model acceptors used in this study.
As with the gluco hemiacetal the conversion was modest, but
clean and the desired glycosyl formates could easily be prepared
on a gram scale with recovery of the remaining hemiacetal.
Isolated yields based on recovered starting material were 85% for
the galactoside, 82% for the xyloside and 64% for the mannoside.
Glycosylations were performed using the conditions optimized for
the glucosyl donor, i.e. Bi(OTf)3 (25%) in combination with the
additive KPF6 (120%). The excess of acceptor was kept to 20
mol% in order to compare the performance of the donor with
different acceptors. Galactosylation using donor D2 with
adamantanol A1 gave high -selectivity and a yield of 58% (entry
6, table 3). Using the 6-OH acceptor A2 the yield and selectivity
Table 2 Catalytic glycosylation with Fe(OTf)3 or Bi(OTf)3 in THF with different
catalyst loading.
Entry
Cat.
Conv.
(%)*
100
100
49
GT-Sel
(%)*
80
:
Yield
(%)*
79
(mol%)
1
2
3
4
5
Fe(OTf)3 (25)
Fe(OTf)3 (10)
Fe(OTf)3 (5)
Bi(OTf)3 (25)
Bi(OTf)3 (10)
48:52
49:51
48:52
45:55
48:52
71
69
40
19
100
100
57
49
57
57
This article is protected by copyright. All rights reserved.

10.1002/cssc.202000733
ChemSusChem
COMMUNICATION
were similar (entry 7, table 3). Galactosylation of the more
hindered 4-OH acceptor A3 resulted in 29% yield and poor
selectivity (entry 8, table 3). Mannosylations with D3 gave good
yields with A1 and A2, and -selectivity (entry 9 and 10, table 3).
Using the hindered acceptor A3 again reduced the yield (entry 11,
table 3). More or less the same trends in yields were observed
with the xylosyl donor D4 (entry 12-14, table 3). Interestingly, -
selectivity was observed when glycosylating acceptor A2 and A3
(entry 13-14, table 3).
but no gas development could be detected and this catalytic
system was therefore also dismissed. Interestingly, Beller and
coworkers, recently demonstrated that Fe(III)PP3 could catalyze
the dehydrogenation of formic acid without the addition of base.[28]
The best results were however at higher temperatures (60°C) and
in propylene carbonate, which is unattractive in glycosylations
due to its high boiling point. Changing the conditions to using THF
at room temperature was, satisfyingly, still able to dehydrogenate
formic acid albeit with a slower pace. Adding the model acceptor
methoxyethanol or the donor D1 to the reaction did not quench
the dehydrogenation, but no glycosylation took place with both
present. To circumvent this a slight excess, relative to the ligand,
of Fe(OTf)3, was added. Using 20 mol% ligand tris[(2-
diphenylphosphino)ethyl]phosphine [P(CH2CH2PPh2)3, PP3] and
25 mol% Fe(OTf)3 for the activation of the glycosyl formate at
70°C gave the glycoside in 61 % with a GT-selectivity of 68%
demonstrating the proof of concept for a traceless catalytic
glycosylation (entry 18, Table 3).
Scheme 4 Synthesis of donor D5 from the orthoester 2.
In order to improve the anomeric selectivity and to study whether
anchimeric assistance would be able to decrease the reaction
times[22,23] and improve the GT-selectivity, 2-O-benzoyl-3,4,6-tri-
O-benzyl--glucopyranosyl formate D5 was synthesized from the
1,2-orthoester 2, using formic acid in CH2Cl2, in one step with an
89% isolated yield. Using neat formic acid resulted in a more
complex reaction mixture due to partly deprotection of the 2-O-
benzoyl group. Due to the relative long reaction times with armed
glycosyl donors, disarmed glycosyl donors carrying more acyl
groups were not investigated.[24]
Scheme 5 Traceless glycosylation using Fe(OTf)3 and PP3 in THF.
Glycosylations with this donor indeed improved the -selectivity,
but the yield remained similar to the reactions with a 2-OBn group
(entry 15-17, table 3).
The gas development was difficult to follow on this scale and at
this temperature. To confirm whether the catalytic system
produced gas under the glycosylation conditions the reaction was
followed by Head-space MS, where the gas development could
be monitored. Detection of CO2(g) supported dehydrogenation of
formic acid hence a “traceless” glycosylation. In order to verify that
these conditions have a broader scope. The same conditions, i.e.
PP3 and Fe(OTF)3 in THF, were used for the glycosylation with
D5 and methoxyethanol. This gave 49% of the -product (entry
19, Table 3). This aspect could be of interest for larger scale
synthesis or in setups where recycling of reagents is of
importance.
With proof of concept of using glycosyl formates as glycosyl
donors the next step was to study whether the formed formic acid
could be removed in situ by a catalytic transformation to hydrogen
gas and carbon dioxide. This would make the leaving group
traceless and ensure complete reaction via Le Châtelier’s
principle. The formic acid dehydrogenation is a common and
industrially important reaction, albeit unpresented in glycosylation
chemistry. Common catalysts for this dehydrogenation are
transition metal catalysts. Pd/C has been used and due to its
availability, this was the first catalyst to study. O’Neil has shown
that formic acid in water dehydrogenates to CO2 and H2, results
that were easily reproduced and demonstrated by collecting the
gas produced (see SI for details).[25] Transferring this approach to
CH2Cl2, which was one of the preferred glycosylation solvents, did
not result in any gas development. Adding Et3N did not change
this; neither did the addition of cyclohexene as a H2 scavenger.
Beller and coworkers have recently demonstrated that ruthenium
based catalysts are effective in dehydrating formic acid.[26,27] To
study this, the readily available catalyst, RuCl2(PPh3)3 was
studied using THF as the solvent and Et3N as an additive. Under
these conditions gas development readily appeared. As Et3N is
incompatible with a Lewis acid catalyzed glycosylation, more
sterically hindered bases were also investigated. TTBT has often
been used as an acid scavenger in glycosylation reactions,
especially when performing pre-activation of the glycosyl donor.
Despite, that we earlier observed inhibition of the glycosylation
with this acid scavenger it was tried out with the rhenium catalyst,
Table 3 Glycosylations
Entry
Donor
D1
Acceptor
Yield (:)[a]
79% (1:1)
69% (1:1)
62% (3:2)
71% (1:1)
32% (1:1)
58% (1:0)
1
2
3
4
5
6
D1
cHexOH
A1
D1
D1
A2
D1
A3
D2
A1
This article is protected by copyright. All rights reserved.

10.1002/cssc.202000733
ChemSusChem
COMMUNICATION
[1]
N. K. Kochetkov, A. J. Khorlin, A. F. Bochkov, Tetrahedron Lett.
1964, 5, 289–293.
7
D2
D2
D3
D3
D3
D4
D4
D4
D5
D5
D5
D1
D5
A2
A3
A1
A2
A3
A1
A2
A3
A1
A2
A3
62% (0:1)
29% (2:1)
68% (1:0)
58% (1:0)
35% (1:0)
51% (1:1)
44% (0:1)
36% (1:3)
67% (1:6)
72% (0:1)
81% (0:1)
61% (1:1)
49% (0:1)
[2]
[3]
T. Mukaiyama, Y. Murai, S. Shoda, Chem. Lett. 1981, 10, 431–432.
R. R. Schmidt, J. Michel, Angew. Chemie Int. Ed. English 1980, 19,
731–732.
8
9
[4]
[5]
[6]
[7]
[8]
M. M. Nielsen, C. M. Pedersen, Chem. Rev. 2018, 118, 8285–8358.
B. Yu, H. Tao, Tetrahedron Lett. 2001, 42, 2405–2407.
S. Hotha, S. Kashyap, J. Am. Chem. Soc. 2006, 128, 9620–9621.
Y. Li, Y. Yang, B. Yu, Tetrahedron Lett. 2008, 49, 3604–3608.
B. Helferich, E. Schmitz-Hillebrecht, Berichte der Dtsch. Chem.
Gesellschaft (A B Ser. 1933, 66, 378–383.
10
11
12
13
14
15
16
17
18
19
[9]
K. S. Kim, Y. J. Lee, H. Y. Kim, S. S. Kang, S. Y. Kwon, Org.
Biomol. Chem. 2004, 2, 2408–2410.
[10]
[11]
[12]
H. Christensen, M. S. Christiansen, J. Petersen, H. H. Jensen, Org.
Biomol. Chem. 2008, 6, 3276–3283.
M. R. Rasmussen, M. H. S. Marqvorsen, S. K. Kristensen, H. H.
Jensen, J. Org. Chem. 2014, 79, 11011–11019.
S. K. Kristensen, S. Salamone, M. R. Rasmussen, M. H. S.
Marqvorsen, H. H. Jensen, European J. Org. Chem. 2016, 2016,
5365–5376.
[13]
Y. Yokoyama, T. Hanamoto, S. Suzuki, K. Shimizu, H. Furuno, J.
Inanaga, Heterocycles 2009, 79, 967–983.
[14]
[15]
L. Hough, A. W. Lewis, Carbohydr. Res. 1975, 42, 173–174.
S. Cecioni, D. J. Vocadlo, J. Am. Chem. Soc. 2017, 139, 8392–
8395.
[16]
[17]
[18]
[19]
J. Yuan, K. Lindner, H. Frauenrath, J. Org. Chem. 2006, 71, 5457–
5467.
E. J. Vlietstra, J. W. Zwikker, R. J. M. Nolte, W. Drenth, Recl. des
Trav. Chim. des Pays-Bas 1982, 101, 460–462.
A. Rostami, A. Khazaei, H. A. Alavi-Nik, Z. Toodeh-Roosta, Chinese
J. Catal. 2011, 32, 60–64.
[a] Yields and anomeric ratios were determined after flash chromatography.
Conclusions
T. Mukaiyama, T. Takashima, M. Katsurada, H. Aizawa, Chem. Lett.
1991, 20, 533–536.
In conclusion, we have developed a method for utilizing glycosyl
formates as glycosyl donors under mild and catalytic conditions at
ambient temperatures. The glycosyl formates are synthesized
from the corresponding hemiacetal or orthoester in one step using
formic acid as the sole reagent. The new donors were found to be
storable and easy to handle. Proof of concept for a traceless
glycosylation using a catalytic system for both donor activation
and formic acid dehydrogenation has additionally been
demonstrated.
[20]
[21]
[22]
[23]
[24]
K. Bock, L. D. Hall, Carbohydr. Res. 1975, 40, C3–C5.
B. Dhakal, L. Bohé, D. Crich, J. Org. Chem. 2017, 82, 9263–9269.
L. K. Mydock, A. V Demchenko, Org. Lett. 2008, 10, 2107–2110.
D. Crich, T. Hu, F. Cai, J. Org. Chem. 2008, 73, 8942–8953.
D. R. Mootoo, P. Konradsson, U. Udodong, B. Fraser-Reid, J. Am.
Chem. Soc. 1988, 110, 5583–5584.
[25]
[26]
[27]
[28]
B. J. O’Neill, E. I. Gürbüz, J. A. Dumesic, J. Catal. 2012, 290, 193–
201.
B. Loges, A. Boddien, H. Junge, M. Beller, Angew. Chemie Int. Ed.
2008, 47, 3962–3965.
Acknowledgements
A. Boddien, B. Loges, H. Junge, M. Beller, ChemSusChem 2008, 1,
751–758.
The China Scholarship Council (file no. 201806220084 for L.Y.).
Theis Brock Nannested is acknowledged for helping with head-
space MS and Christian Tortzen for help with NMR experiments.
A. Boddien, D. Mellmann, F. Gärtner, R. Jackstell, H. Junge, P. J.
Dyson, G. Laurenczy, R. Ludwig, M. Beller, Science 2011, 333,
1733–1736.
Keywords: Glycosylations • Glycosyl formates • Catalysis •
Traceless • NMR reaction optimization
This article is protected by copyright. All rights reserved.

10.1002/cssc.202000733
ChemSusChem
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 1:
COMMUNICATION
Text for Table of Contents
Key Topic*
Author(s), Corresponding Author(s)*
Page No. – Page No.
Title
*one or two words that highlight the emphasis of the paper or the field of the study
Layout 2:
COMMUNICATION
Catalytic Glycosylation
Liang Yang, Christian H. Hammelev and
Christian M. Pedersen*[
Page No. – Page No.
Catalytic and Atom Economic
Glycosylation using Glycosyl
Formates and Cheap Metal Salts.
Text for Table of Contents
This article is protected by copyright. All rights reserved.