Safe and Scalable Continuous Flow Azidophenylselenylation of Galactal to Prepare Galactosamine Building Blocks

Article abstract of DOI:10.1021/acs.oprd.9b00456

Differentially protected galactosamine building blocks are key components for the synthesis of human and bacterial oligosaccharides. The azidophenylselenylation of 3,4,6-tri-O-acetyl-d-galactal provides straightforward access to the corresponding 2-nitrog

Full text of DOI:10.1021/acs.oprd.9b00456

This is an open access article published under a Creative Commons Attribution (CC-BY)
License, which permits unrestricted use, distribution and reproduction in any medium,
provided the author and source are cited.
Communication
pubs.acs.org/OPRD
Cite This: Org. Process Res. Dev. XXXX, XXX, XXX−XXX
Safe and Scalable Continuous Flow Azidophenylselenylation of
Galactal to Prepare Galactosamine Building Blocks
†,‡
†
Monica Guberman, Bartholomaus Pieber, and Peter H. Seeberger*,†,‡
́
̈
†Department of Biomolecular Systems, Max Planck Institute of Colloids and Interfaces, Am Mu
̈
hlenberg 1, 14476 Potsdam,
Germany
‡
̈
Department of Chemistry and Biochemistry, Freie Universitat Berlin, Arnimalle 22, 14195 Berlin, Germany
S
* Supporting Information
The highly exothermic APS reaction is commonly carried
out using TMSN3 as an azido source,14 due to lower explosion
hazard as well as higher solubility in organic solvents when
compared to NaN3.12,16,17 While relatively good yields were
reported,14 this transformation is difficult to reproduce reliably.
Yields of 23 experiments carried out in our department over
the past 10 years range from 10% to 65% with an average yield
of 35% (see Supporting Information for details). The poor
selectivity and irreproducibility are likely a consequence of
deficiencies in mixing and temperature control during batch
reactions. The potentially explosive and toxic nature of azido
compoundsin particular the possibility of hydrazoic acid
formationcauses severe safety risks and limits reaction scale-
up in conventional batch approaches.18 We hoped to address
these issues by performing the APS reaction under continuous
flow conditions and to benefit from reproducible and efficient
mixing and heat transfer.19−21 The concentration of potentially
hazardous side products generated at any time is minimal
during the flow regime and facilitates reaction scale-up.19−21
ABSTRACT: Differentially protected galactosamine
building blocks are key components for the synthesis of
human and bacterial oligosaccharides. The azidophenyl-
selenylation of 3,4,6-tri-O-acetyl-D-galactal provides
straightforward access to the corresponding 2-nitro-
genated glycoside. Poor reproducibility and the use of
azides that lead to the formation of potentially explosive
and toxic species limit the scalability of this reaction and
render it a bottleneck for carbohydrate synthesis. Here, we
present a method for the safe, efficient, and reliable
azidophenylselenylation of 3,4,6-tri-O-acetyl-D-galactal at
room temperature, using continuous flow chemistry.
Careful analysis of the transformation resulted in reaction
conditions that produce minimal side products while the
reaction time was reduced drastically when compared to
batch reactions. The flow setup is readily scalable to
process 5 mmol of galactal in 3 h, producing 1.2 mmol/h
of product.
KEYWORDS: azides, continuous flow, galactosamine,
azidophenylselenylation, carbohydrate chemistry
RESULTS AND DISCUSSION
■
Common APS reaction protocols in batch14 call for the
addition of TMSN3 to a solution of 3,4,6-tri-O-acetyl-D-galactal
(1) and Ph2Se2 in anhydrous DCM at −30 °C. After the
addition of bisacetoxy iodobenzene (BAIB), the reaction
mixture is warmed to −10 °C. Reaction times are variable and
typically range from 4 to 16 h.14,22 Low temperatures during
mixing minimize explosion hazards and avoid high concen-
trations of azido radicals due to the fast reaction between
TMSN3 and BAIB23 that would lead to unproductive
formation of N2 and undesired bisazido side products.
INTRODUCTION
■
Galactosamine (GalN) is ubiquitous in living organisms, and
the N-acetyl galactosamine (GalNAc) is one of the nine
monosaccharide building blocks that give rise to all
mammalian glycans.1 Many bacteria present GalNAc-contain-
ing glycans on their cell surface.2 Synthetic oligosaccharides are
valuable tools to investigate glycan function and are used as
diagnostics,3 vaccines,4 and drugs.5 Suitably protected
monosaccharide building blocks are needed to assemble the
desired oligosaccharides.6 While commercially available glucos-
amine (GlcN) can be used as starting material for N-acetyl
glucosamine (GlcNAc) building blocks, the high price of
galactosamine hampers large scale synthetic applications.
Therefore, several strategies to prepare 2-nitrogenated glyco-
side analogues from inexpensive starting materials have been
developed.7−13 Azidophenylselenylation (APS) is a commonly
used nitrogen transfer reaction to prepare galactosamine
building blocks from the corresponding galactals.14−16 APS
introduces two functional groups in a single step, the resulting
selenoglycosides are compatible with a wide range of
protecting group manipulations and can be activated to
prepare 1,2-cis and 1,2-trans glycosides (Scheme 1).
Initial experiments showed that mixing 3,4,6-tri-O-acetyl-D-
galactal (1), Ph2Se2, and TMSN3 did not result in any reaction
over a 2 h period, and the reaction only started after BAIB
addition (see Supporting Information). Consequently, a flow
setup consisting of two reagent feeds (Feed A: 3,4,6-tri-O-
acetyl-D-galactal (1), Ph2Se2 and TMSN3 in anhydrous DCM;
Feed B: BAIB in anhydrous DCM) was assembled (Figure 1).
The liquid streams were pumped using HPLC pumps24 and
mixed in a T-piece before entering a residence time unit that
was immersed in a thermostatic bath. Sample loops were
connected to the liquid stream via six-way valves to introduce
the reagent solution to the flow stream. The reaction mixture
Received: October 22, 2019
Published: December 4, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.oprd.9b00456
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
a
Scheme 1. Galactosamine Building Blocks Are Obtained via Nitrogen Transfer to Glycals
a
PG: Protecting group.
Figure 1. Continuous flow setup for the azidophenylselenylation of 3,4,6-tri-O-acetyl-D-galactal (1).
finally passed a backpressure regulating unit (17 bar) and was
quenched offline. Pressurizing the system was essential to avoid
the formation of a biphasic gas/liquid flow pattern resulting
from the formation of N2 to guarantee reproducible flow
processes. To ensure immediate quenching, the reaction
mixture was collected in a vigorously stirred mixture of
dichloromethane and an aqueous solution containing Na2S2O3
(to reduce unreacted BAIB) and NaHCO3 to prevent the
formation of hydrazoic acid (for details the see Supporting
Information).
electrophilic selenium species, but the observation of 1,2-cis
glycoside 5 argues against the involvement of a cyclic selenium
cation that was previously suggested.23 Isolation and character-
1
ization of all compounds allowed us to identify a distinct H
NMR signal for all monosaccharide products 1−8 and enabled
1
us to assess the reaction selectivity using crude H NMR
analysis (Figure 2; see the Supporting Information for details).
Initial continuous flow reactions revealed a rather invariable
selectivity between desired product 2 and its unwanted isomers
3−5 (ratios 2/3/4/5 greater than 10:1:1:1), as well as high
fluctuations of the amounts of bisazido and 1-O-acetyl side
products 6−8 (Figure 2A−C; see the Supporting Information
for details). Occasionally, reaction outcomes changed drasti-
cally as acetyl glycoside 8 was produced mainly, as well as an
increased amounts of bisazido side products 6−7 and the
absence of selenoglycosides 2−3 (Figure 2D). Comparison of
1H NMR spectra of the organic layer before and after solvent
evaporation revealed that 6−8 were formed during solvent
evaporation rather than during the continuous flow process.
Systematic investigations showed that filtration through a bed
of silica of the quenched reaction mixture prior to solvent
evaporation avoids this reproducibility issue (see Supporting
Information). It should be noted that removal of selenide
impurities through silica filtration is required for an adequate
crystallization of 2. The use of silica filtration prior to
evaporation therefore does not introduce an extra step to the
preparative protocol, but simply exchanges the order between
silica filtration and solvent evaporation. Presumably, seleno-
glycosides 2−3 are converted into 6−8 via radical-mediated
homolytic cleavage of the C−Se anomeric bond or β-
elimination to afford galactal 1 as an intermediate.23,27,28
Alternatively, oxidation of the phenylselenyl group to the
corresponding selenone would afford a good leaving group,
rendering C1 susceptible to nucleophilic attack.29 Involvement
of BAIB on the formation of 6−8 under reduced pressure
In order to optimize the reaction conditions, the crude
reaction mixtures were carefully analyzed by 1H NMR.
Preliminary experiments showed the presence of the desired
azidophenylselenylation product 2 and up to six different side
products (3−8, Figure 2), including the talo stereoisomer 3,
the regioisomers 4 and 5, the bisazido monosaccharides 6 and
7, and 1-O-acetyl glycoside 8 (Figure 2). Remarkable
selectivity toward α-glycosides (2−6, 8) was noted, with the
β-bisazido species (7) only observed occasionally. A
mechanism where an initial ligand exchange on the BAIB
generates the azido radicals was previously proposed.23
Subsequent addition of the azido radical to the glycal double
bond to form an anomeric (pyranos-1-yl) radical accounts for
the regioselectivity toward 2-azido products (2, 3, 6−8). The
pseudoaxial disposition of the C4-substituent hindering the β-
face accounts for the stereoselectivity for equatorial (galacto
products 2, 5−8) over axial (talo products 3−4) substituents at
C2.25 α-Selectivity is commonly observed in galactal reactions,
and it is attributed to the anomeric effect and the pseudoaxial
disposition of the C4-OAc.11 BAIB is the source for the
acetoxy moiety in acetyl glycoside 8. Side products bearing
acetoxy groups were observed in the presence of BAIB in the
azidophenylselenylation of olefins, where acetoxyphenylsenelyl
compounds were identified.26 Formation of both 2-phenyl-
selenyl products (4, 5) is likely due to the production of
B
DOI: 10.1021/acs.oprd.9b00456
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
Figure 2. 1H NMR spectroscopic analysis of crude reaction mixtures of the APS reaction of galactal 1. 1H NMR spectra were recorded after solvent
1
evaporation. Highlighted signals correspond to the H chemical shifts used to assess the presence of 1−8 in the reaction crude (see Supporting
Information). Representative examples: (A) excellent outcome, only regio- and stereoisomers of 2 (3−5) observed as side products; (B) acceptable
outcome, where side products 3−5 and minimum amounts of 8 are observed; (C) complex case, with incomplete conversion and products 2−7;
(D) extreme case with 8 as main product, absence of glycosyl selenides 2 and 3, and presence of side products 4−7.
Table 1. Screening of Reaction Conditions for the Flow APS Reaction of Galactal 1
b
yield (%)
a
entry
tres/min
T (°C)
Ph2Se2 (equiv)
TMSN3 (equiv)
BAIB (equiv)
Conv (%)
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
9
25
25
25
25
25
20
25
25
40
27
27
27
27
27
27
0
2
2
2
1.3
2
2
2
2
2
2
1.3
1.3
2
100
98
98
95
100
95
62
100
100
78
74
58
72
79
71
41
65
60
6
6
5
6
6
6
3
7
8
7
5
7
7
7
4
8
7
7
7
5
8
7
7
5
8
7
0
0
0
0
0
0
4
4
4
0
0
0
0
0
0
3
0
0
0
0
0
0
0
0
0
0
4
1.3
1.3
1.3
1.6
1.6
1.6
1.6
1.6
1
1.3
1.3
1.3
1.3
1.3
40
40
7
a
b
Conversion of galactal (1) determined by 1H NMR using 1,2,4,5-tetramethylbenzene as internal standard. NMR yields determined by 1H NMR
using 1,2,4,5-tetramethylbenzene as internal standard.
resulting from incomplete reduction in the aqueous quenching
was ruled out, as no BAIB signals were observed during H
NMR analysis.
Once the flow setup with a reliable quenching protocol and a
suitable method to assess the reaction outcome was
established, reaction conditions were screened aiming to
1
C
DOI: 10.1021/acs.oprd.9b00456
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
maximize the yield of selenoglycoside 2 by achieving full
conversion and minimizing side product formation (Table 1).
Best results were obtained with 1.6 equiv of Ph2Se2, 2 equiv of
TMSN3, and 1.3 equiv of BAIB at room temperature (Table 1,
entry 5) to afford target product 2 in 79% yield within 25 min.
Excess Ph2Se2 was crucial to minimize the formation of 6−8
and was necessary for complete substrate consumption (entries
1, 2, and 5). A 2-fold excess of BAIB did not lead to a
significant formation of 1-O-acetyl product 8, but a low mass
balance (full consumption of 1, sum of the yields for 2−5 =
73%) was obtained (entry 3). Lowering the reaction
temperature to 0 °C resulted in a significantly reduced
conversion, without any improvement in selectivity toward 2
(entry 7). At higher temperatures (40 °C), significantly lower
selectivity was observed due to the formation of bisazido side
products 6−7 (Table 1, entries 8−9). The selectivity between
2−5 remained roughly unaltered by temperature changes
(ratios 2/3/4/5 greater than 10:1:1:1) and did not differ from
control experiments in batch (see Supporting Information).
For the large scale production of 2, the residence time was
extended to 35 min to guarantee full conversion, and the
overall flow rate was increased from 0.24 to 0.34 mL/min for
enhanced throughput. These conditions were tested at a 1
mmol scale and resulted in a 72% NMR yield of 2.
Recrystallization yielded 281 mg (61% yield) of 2. Only a
seventh of the time originally reported in batch is required in
flow for the full conversion of 1 mmol of starting material (1).
The isolated yield from the flow approach ranks together with
the top yields observed in batch (65%), and well above the
average yield of 35% (see Supporting Information). These
results were subsequently reproduced on a gram scale APS
reaction (5 mmol of 1) that resulted in 63% (1.46 g) of
analytically pure selenogalactoside 2.
EXPERIMENTAL SECTION
■
General Methods. Reagents and solvents were obtained
from commercial sources, unless stated otherwise. Anhydrous
DCM was obtained from a Solvent Dispensing System (J.C.
Meyer), and anhydrous chloroform was prepared by adding
preactivated molecular sieves (4 Å, Roth) to HPLC grade
chloroform (Merck). Analytical thin layer chromatography
(TLC) was performed on glass precoated TLC-plates SIL G/
UV254 sheets (Macherey-Nagel) and visualized with 254 nm
light and sugar stain (3.70 mL of p-anisaldehyde in 140 mL of
a solution 3.5% H2SO4 in ethanol). NMR spectra were
obtained using Ascend 400 (Bruker) and Agilent 400 MHz
NMR Magnet (Agilent Technologies) spectrometers at 400
MHz (1H) and 100 MHz (13C). CDCl3 was used as solvent,
and chemical shifts (δ) are reported in ppm relative to the
residual solvent peaks (CDCl3: 7.26 ppm 1H, 77.16 ppm 13C).
Assignments were supported by COSY and HSQC experi-
ments. IR spectra were measured with a Spectrum 100 FT-IR
Spectrometer (PerkinElmer). Only diagnostic signals are listed.
Specific rotations were measured using a UniPol L 1000
polarimeter (Schmidt + Haensch). ESI-HRMS were performed
with a Xevo G2-XS Q-Tof (Waters). HPLCs were performed
on Agilent 1200 Series systems.
Safety note: Even under continuous flow conditions,
extreme caution in the handling of potentially hazardous
starting materials and product mixtures is required. The
possibility of formation of poisonous and explosive
hydrazoic acid must be contemplated at all times.
General Procedure and Equipment for Azidophenyl-
selenylation of Galactal under Continuous Flow
Conditions. For screening conditions (0.49 mmol of 1):
The flow setup was assembled using 1.6 mm O.D. × 0.8 mm
I.D. FEP tubing (residence time unit volume: 6 mL) and
connected by a simple T-mixer (0.8 mm I.D.). The
temperature of the residence time unit was regulated at 27
°C using a thermostatic bath unless stated otherwise. 3,4,6-Tri-
O-acetyl-α-D-galactal 1 was coevaporated with anhydrous
toluene twice and kept under high vacuum for at least 30
min prior to every reaction. Galactal 1 was defined as the
limiting reagent, and concentrations of other reagents and flow
rates were calculated according to desired stoichiometry and
residence time. Sample loops were made out of 1.6 mm O.D. ×
0.8 mm I.D. FEP (loop A: 3 mL; loop B: 5 mL) and were
washed and filled with anhydrous DCM prior to reactions. The
sample loops were loaded with the following solutions for
continuous flow experimentens: Solution A: 0.163 M solution
of galactal 1, Ph2Se2, and TMSN3 in anhydrous DCM. Solution
A was loaded in loop A and injected using pump A set to a flow
rate fA (Figure 1), 1.5 min after injection of solution B had
started. Solution B: BAIB in anhydrous DCM. Solution B was
loaded in loop B and then injected using pump B set to a flow
rate fB = fA.
Knauer BlueShadow 40P pumps were used for pumping. A
Vapourtec R2 series was used for sample loop command and
in-line pressure monitoring. An exchangeable cartridge back
pressure regulator (BPR) loaded with a 17-bar cartridge
(Upchurch Scientific) was placed downstream, and the
reaction mixture was collected and quenched by dropping it
into a mixture of DCM (50 mL) and aq. sat. NaHCO3 (25
mL), Na2S2O3 (25 mL) that was stirred vigorously. After phase
separation, the organic layer was passed through a 6 cm silica
gel plug preloaded on a disposable sample syringe with a filter,
CONCLUSION
■
A thorough analysis revealed that the APS reaction is more
complex than previously thought, and six different side
products can be formed that significantly reduce the yield of
the desired selenogalactoside 2. An additional degree of
complexity is added when considering that the formation of
the target product is reversible. Several factors account for low
reproducibility of standard batch APS reactions. The lack of
control in the mixing of TMSN3 and BAIB and slow heat
transfer processes produce variable results, since elevated local
concentrations of azido species can lead either to an increase in
bisazido side products (6−7) or to low conversions if azido
species are consumed in N2 formation.30 Additionally,
significant differences in the reaction outcome can result
from insufficient quenching, as the desired product is prone to
form bisazido and 1-O-acetyl compounds. Performing the
reaction in flow allowed for the safe reaction of TMSN3 with
BAIB at room temperature, enabling to establish reaction
conditions that narrowed the number of side products and
reduced the reaction time significantly. Effective and
reproducible quenching conditions were key to achieve
reproducible, selective reactions. The flow setup allowed for
a straightforward scale-up, as the results observed in an
analytical scale were steadily reproduced throughout the entire
process of a gram-scale synthesis.
D
DOI: 10.1021/acs.oprd.9b00456
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
eluted with additional 50 mL of DCM (fraction 1, “f1”), and
then eluted with 100 mL of DCM/Acetone 95:5 (fraction 2,
“f2”) and 100 mL of DCM/Acetone 90:10 (fraction 3, “f3”).
The solvent of f2 was evaporated under reduced pressure and
product 2 as a white solid (281 mg, 0.60 mmol, 61% for
0.98 mmol scale; 1.46 g, 3.09 mmol, 63% for 4.90 mmol scale).
IR (film): 2958 cm−1 (w, C−H Ar) 2113 cm−1 (s, N3), 1749
1
cm−1 (s, CO), 1227 cm−1 (s, C−O ester). H NMR (400
1
was analyzed by H NMR to assess reaction outcome, and/or
MHz, CDCl3) δ 7.64−7.57 (m, 3H), 7.36−7.27 (m, 3H), 6.00
(d, J = 5.4 Hz, 1H, H-1), 5.47 (d, J = 2.5 Hz, 1H, H-4), 5.11
(dd, J = 10.8, 3.2 Hz, 1H, H-3), 4.67 (t, J = 6.4 Hz, 1H, H-5),
4.26 (dd, J = 10.8, 5.4 Hz, 1H, H-2), 4.05 (qd, J = 11.4, 6.6 Hz,
2H, H-6,6′), 2.15 (s, 3H, OC(O)CH3), 2.07 (s, 3H,
OC(O)CH3), 1.98 (s, 3H, OC(O)CH3). 13C NMR (101
MHz, CDCl3) δ 170.5, 170.1, 169.7 (3 × OC(O)CH3), 134.9,
129.3, 128.3, 127.7, 84.3 (C-1), 71.3 (C-3), 69.1 (C-5), 67.3
(C-4), 61.7 (C-6), 58.9 (C-2), 20.79, 20.77, 20.76 (3 ×
OC(O)CH3). HRMS (ESI) calcd for C18H21N3O7SeNa,
494.0437; found [M + Na]+, 494.0427.
crystallized from isopropanol to obtain target product 2.
Fractions f1 and f3 were kept until the absence of
carbohydrates (only required for screening and optimization
purposes) was confirmed by thin layer chromatography and/or
1
1H NMR. Quantifications were performed by H NMR of the
crude reaction mixture, using 1,2,4,5-tetramethylbenzene as an
internal standard (IS). Peak areas of IS δ 6.91 (s, 2H) and of
the diagnostic of compounds 1−8 (see Supporting Informa-
tion) were used for calculations.
Experimental Data for Monosaccharides 1−8. Galactal
1 was synthesized following previously reported procedures.22
Monosaccharides 2−8 were isolated from flow APS reactions.
Selenoglycoside 2 was isolated through recrystallization (see
below). Fractions enriched in monosaccharides 3−8 were
isolated exclusively for characterization purposes. Monosac-
charides 2−5 and 6−7 showed similar chromatographic
behavior, such that it was not possible to achieve full peak
resolution using preparative normal-phase HPLC. The
isopropanol filtrate of a 0.49 mmol flow APS synthesis was
purified using preparative HPLC purification, from which a
fraction enriched in 3 was isolated and repurified through
semipreparative HPLC to afford pure selenoglycoside 3.
Monosaccharides 4−8 were isolated from flow APS syntheses
on 0.49 mmol scale without filtration through silica prior to
evaporation (see Supporting Information). Purification via
silica gel chromatography (toluene/acetone 0−10% as eluent)
and subsequent preparative normal-phase HPLC chromatog-
raphy (hexanes/AcOEt 2−20% as eluent) allowed isolating
fractions enriched in 4−5, 6−7, and 8. For normal-phase (NP)
HPLC, YMC-Diol-300-NP columns were used (analytical: 150
mm × 4.60 mm I.D.; semipreparative: 150 mm × 10.0 mm
I.D.; preparative: 150 mm × 20.0 mm I.D.), with hexanes/
EtOAc as eluent (flow rates: 1.0, 5.0, and 15.0 mL/min for
analytical, semipreparative, and preparative chromatography,
respectively). The following gradient was used: (1) isocratic
2% EtOAc in hexanes (5 min); (2) linear gradient 2 to 20%
EtOAc in hexanes (30 min); (3) linear gradient to 100%
EtOAc (10 min).
Phenyl 3,4,6-Tri-O-acetyl-2-azido-2-deoxy-1-seleno-α-D-
talopyranoside (3). IR (film) 2950 cm−1 (w, C−H Ar),
2109 cm−1 (s, N3), 1750 cm−1 (s, CO), 1260 cm−1 (s, C−O
1
ester). H NMR (400 MHz, CDCl3) δ 7.59 (dd, J = 7.7, 1.8
Hz, 2H), 7.35−7.28 (m, 3H), 5.89 (s, 1H, H-1), 5.45−5.42
(m, 1H, H-4), 5.31 (t, J = 3.8 Hz, 1H, H-3), 4.70 (ddd, J = 7.3,
5.7, 1.7 Hz, 1H, H-5), 4.23−4.11 (m, 2H, H-6,6′), 4.08−4.06
(m, 1H, H-2), 2.20 (s, 3H, OC(O)CH3), 2.10 (s, 3H,
OC(O)CH3), 2.02 (s, 3H, OC(O)CH3). 13C NMR (101
MHz, CDCl3) δ 170.6 (OC(O)CH3), 170.5 (OC(O)CH3),
169.6 (OC(O)CH3), 134.34, 129.57, 128.63, 128.03, 83.1 (C-
1, JC−H = 174 Hz), 69.7 (C-5), 68.4 (C-3), 65.8 (C-4), 61.9
(C-6), 59.8 (C-2), 20.8, 20.7 (3 × OC(O)CH3). HRMS (ESI)
calcd for C18H21N3O7SeNa, 494.0437; found [M + Na]+,
494.0439.
3,4,6-Tri-O-acetyl-2-deoxy-2-selenophenyl-1-azido-α-D-
talopyranoside (4)14 and 3,4,6-Tri-O-acetyl-2-deoxy-2-
selenophenyl-1-azido-α-D-galactopyranoside (5). Mixture
of glycosides 4 and 5 (1.7:1). IR (film) 2965 cm−1 (w, C−H
Ar), 2113 cm−1 (s, N3), 1746 cm−1 (s, CO), 1214 cm−1 (s,
1
C−O ester). H NMR (400 MHz, CDCl3) δ 7.60−7.52 (m,
2H), 7.35−7.27 (m, 3H), 5.71 (d, J = 1.6 Hz, 0.6H, H-1 4),
5.67 (d, J = 4.0 Hz, 0.4H, H-1 5), 5.39−5.35 (m, 1H), 5.33−
5.30 (m, 0.6H, C−H 4), 5.23 (dd, J = 11.8, 3.1 Hz, 0.4H, C−H
5), 4.40 (ta, J = 6.5 Hz, 1H, H-5 4 and 5), 4.15 (m, 2H, H-6,6’
4 and 5), 3.57 (dd, J = 11.8, 4.0 Hz, 0.4H, H-2 5), 3.42 (ddd, J
= 5.1, 1.4, 0.8 Hz, 0.6H, H-2 4), 2.22 (s, 1.8H, C(O)CH3 4),
2.11 (s, 1.8H, C(O)CH3 4), 2.10 (s, 1.2H, C(O)CH3 5), 2.07
(s, 1.8H, C(O)CH3 4), 2.06 (s, C(O)CH3. 1.2H, 5), 1.89 (s,
1.2H, C(O)CH3 5). 13C NMR (101 MHz, CDCl3) δ 170.70
(C(O)CH3), 170.66 (C(O)CH3), 170.1 (C(O)CH3), 170.00
(C(O)CH3), 169.98 (C(O)CH3), 169.8 (C(O)CH3), 134.5,
134.4, 130.3, 129.7, 129.4, 128.9, 128.5, 128.2, 91.6 (C-1 4,
3,4,6-Tri-O-acetyl-α-D-galactal (1).31 1H NMR (400 MHz,
CDCl3) δ 6.47 (d, J = 6.1, 1H, H-1), 5.56 (s, 1H, H-3), 5.43 (s,
1H, H-4), 4.73 (d, J = 6.1, 1H, H-2), 4.38−4.28 (m, 1H),
4.28−4.16 (m, 2H), 2.13 (s, 3H), 2.09 (s, 3H), 2.03 (s, 3H).
13C NMR (101 MHz, CDCl3) δ 170.7, 170.4, 170.3
(OC(O)CH3), 145.5 (C-1), 98.9 (C-2), 72.9 (C-5), 64.0
(C-3), 63.8 (C-4), 62.0 (C-6), 20.98, 20.93, 20.8 (OC(O)
CH3).
Phenyl 3,4,6-Tri-O-acetyl-2-azido-2-deoxy-1-seleno-α-D-
galactopyranoside (2).12 Flow APS syntheses were performed
as described above, with stoichiometry and temperature as in
Table 1, entry 5. The following modifications on the reaction
setup were performed for the scale-up: The flow reactor was
built using 1.6 mm O.D. × 0.8 mm I.D. FEP tubing (reactor
volume: 12 mL). Sample loops were built with 3.2 mm O.D. ×
1.6 mm I.D. FEP or PFA tubing (Loop A: 6 mL and Loop B:
10 mL for 0.98 mmol scale; Loop A: 30 mL and Loop B: 34
mL for 4.9 mmol scale). Flow rates were set at fA = f B = 0.17
mL/min. Loop A was cooled to 0 °C for 4.9 mmol scale
reaction. Recrystallization from isopropanol afforded the
JC−H = 177 Hz), 90.9 (C-1 5, JC−H = 173 Hz), 69.8, 69.3, 69.0,
67.6, 66.7, 66.2, 62.0, 61.9, 46.0 (C-2 4), 43.4 (C-2 5), 21.07
(C(O)CH3), 21.03 (C(O)CH3), 20.86 (2 × C(O)CH3), 20.81
(C(O)CH3), 20.7 (C(O)CH3). HRMS (ESI) calcd for
C18H21N3O7SeNa, 494.0437; found [M + Na]+, 494.0438.
3,4,6-Tri-O-acetyl-2-azido-2-deoxy-1-azido-α,β-D-galacto-
pyranoside (6, 7).32 Mixture of glycosides 6 and 7 (7:1). IR
(film) 2113 cm−1 (s, N3), 1745 cm−1 (s, CO), 1224 cm−1 (s,
C−O ester). 1H NMR (400 MHz, CDCl3) δ 5.46 (dd, J = 3.3,
1.4 Hz, 0.87H, H-4 6), 5.43 (d, J = 3.4 Hz, 0.87H, H-1 6), 5.40
(dd, J = 11.0, 3.2 Hz, 0.87H, H-3 6), 5.35 (d, J = 3.3 Hz,
0.13H, H-4 7), 4.82 (dd, J = 10.9, 3.3 Hz, 0.13H, H-3 7), 4.71
(d, J = 7.9 Hz, 0.13H, H-1 7), 4.46 (t, J = 7.0 Hz, 0.87H, H-5
6), 4.16−4.03 (m, 2H, H-6,6′ 6, 7), 3.91 (td, J = 6.6, 1.2 Hz,
0.13H, H-5 7), 3.76 (dd, J = 11.0, 3.4 Hz, 0.87H, H-2 6), 3.67
E
DOI: 10.1021/acs.oprd.9b00456
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
Etzler, M. E., Eds.; Cold Spring Harbor Laboratory Press: Cold Spring
Harbor, 2009.
(dd, J = 10.9, 8.0 Hz, 0.13H, H-2 7), 2.16 (s, 0.39H,
OC(O)CH3 7), 2.15 (s, 2.61H, OC(O)CH3 6), 2.06 (m, 6H, 2
× OC(O)CH3 6, 7). 13C NMR (101 MHz, CDCl3) δ 170.72
(C(O)CH3), 170.68 (C(O)CH3), 170.3 (C(O)CH3), 170.1
(2) Knirel, Y. A. O-Specific Polysaccharides of Gram-Negative
Bacteria. In Microbial Glycobiology; Moran, A. P., Ed.; Elsevier: 2010;
pp 57−73.
(3) Wang, S. K.; Cheng, C. M. Glycan-Based Diagnostic Devices:
Current Progress, Challenges and Perspectives. Chem. Commun. 2015,
51, 16750−16762.
(4) Wilson, R. M.; Danishefsky, S. J. A Vision for Vaccines Built
from Fully Synthetic Tumor-Associated Antigens: From the
Laboratory to the Clinic. J. Am. Chem. Soc. 2013, 135, 14462−14472.
(5) Ernst, B.; Magnani, J. L. From Carbohydrate Leads to
Glycomimetic Drugs. Nat. Rev. Drug Discovery 2009, 8, 661−677.
(6) Werz, D. B.; Ranzinger, R.; Herget, S.; Adibekian, A.; Von der
Lieth, C. W.; Seeberger, P. H. Exploring the Structural Diversity of
Mammalian Carbohydrates (“glycospace”) by Statistical Databank
Analysis. ACS Chem. Biol. 2007, 2, 685−691.
(7) Winterfeld, G. a.; Schmidt, R. R. Nitroglycal Concatenation: A
Broadly Applicable and Efficient Approach to the Synthesis of
Complex O-Glycans. Angew. Chem., Int. Ed. 2001, 40, 2654−2657.
(8) Leblanc, Y.; Fitzsimmons, B. J.; Springer, J. P.; Rokach, J. [4 + 2]
Cycloaddition Reaction of Dibenzyl Azodicarboxylate and Glycals. J.
Am. Chem. Soc. 1989, 111, 2995−3000.
(9) Du Bois, J. Du; Tomooka, C. S.; Hong, J.; Carreira, E. M. Novel,
Stereoselective Synthesis of 2-Amino Saccharides. J. Am. Chem. Soc.
1997, 119, 3179−3180.
(10) Danishefsky, S. J.; Koseki, K.; Griffith, D. A.; Gervay, J.;
Peterson, J. M.; McDonald, F. E.; Oriyama, T. Azaglycosylation of
Complex Stannyl Alkoxides with Glycal-Derived Iodo Sulfonamides:
A Straightforward Synthesis of Sialyl-Lewis X Antigen and Other
Oligosaccharide Domains. J. Am. Chem. Soc. 1992, 114, 8331−8333.
(11) Lemieux, R. U.; Ratcliffe, R. M. The Azidonitration of Tri-O-
Acetyl-D-Galactal. Can. J. Chem. 1979, 57, 1244−1251.
(12) Santoyo-Gonzalez, F.; Calvo-Flores, F. G.; Garcia-Mendoza, P.;
Hernandez-Mateo, F.; Isac-Garcia, J.; Robles-Diaz, R. Synthesis of
Phenyl 2-Azido-2-Deoxy-1-Selenoglicosides from Glycals. J. Org.
Chem. 1993, 58, 6122−6125.
(13) (a) Emmadi, M.; Kulkarni, S. S. Rapid Transformation of D-
Mannose into Orthogonally Protected D-Glucosamine and D-
Galactosamine Thioglycosides. J. Org. Chem. 2011, 76, 4703−4709.
(b) Emmadi, M.; Kulkarni, S. S. Orthogonally Protected D-
Galactosamine Thioglycoside Building Blocks via Highly Regiose-
lective, Double Serial and Double Parallel Inversions of β-D-
Thiomannoside. Org. Biomol. Chem. 2013, 11, 4825−4830.
(c) Emmadi, M.; Kulkarni, S. S. Synthesis of Orthogonally Protected
Bacterial, Rare-Sugar and D-Glycosamine Building Blocks. Nat.
Protoc. 2013, 8, 1870−1889.
(14) Mironov, Y. V.; Sherman, A. A.; Nifantiev, N. E. Homogeneous
Azidophenylselenylation of Glycals Using TMSN3−Ph2Se2−PhI-
(OAc)2. Tetrahedron Lett. 2004, 45, 9107−9110.
(15) Mironov, Y. V.; Sherman, A. A.; Nifantiev, N. E. Homogeneous
Azidophenylselenylation of Glucals. Mendeleev Commun. 2008, 18,
241−243.
(C(O)CH3), 96.6 (C-1 7, JC−H = 165 Hz), 92.5 (C-1 6, JC−H
=
176 Hz), 71.3 (C-3 7), 71.1 (C-5 7), 68.5 (C-3 6), 67.8 (C-4
6), 66.8 (C-5 6), 66.5 (C-4 7), 62.1 (C-6 7), 61.9 (C-6 6),
61.7 (C-2 7), 58.1 (C-2 6), 20.89 (C(O)CH3), 20.87 (C(O)
CH3), 20.84 (C(O)CH3), 20.81 (C(O)CH3), 20.79 (C(O)
CH3), 20.78 (C(O)CH3).
1,3,4,6-Tetra-O-acetyl-2-azido-2-deoxy-α-D-galacto-
pyranoside (8). [α]D25 (c 1.00, CHCl3) + 85.9. IR (film) 2116
cm−1 (s, N3), 1751 cm−1 (s, CO), 1215 cm−1 (s, C−O
ester). 1H NMR (400 MHz, CDCl3) δ 6.32 (d, J = 3.6 Hz, 1H,
H-1), 5.49−5.46 (m, 1H, H-4), 5.31 (dd, J = 11.0, 3.2 Hz, 1H,
H-3), 4.27 (t, J = 6.8 Hz, 1H, H-5), 4.08 (dd, J = 6.7, 3.4 Hz,
2H, H-6,6′), 3.94 (dd, J = 11.0, 3.6 Hz, 1H, H-2), 2.17 (s, 3H,
C(O)CH3), 2.17 (s, 3H, C(O)CH3), 2.07 (s, 3H, C(O)CH3),
2.03 (s, 3H, C(O)CH3). 13C NMR (101 MHz, CDCl3) δ
170.5 (C(O)CH3), 170.2 (C(O)CH3), 170.0 (C(O)CH3),
168.9 (C(O)CH3), 90.5 (C-1, JC−H = 180 Hz), 68.9 (C-3),
68.8 (C-5), 66.9 (C-4), 61.2 (C-6), 56.9 (C-2), 21.1 (C(O)
CH3), 20.81 (2 × C(O)CH3), 20.78 (C(O)CH3). HRMS
(ESI) calcd for C14H19N3O9Na, 396.1014; found [M + Na]+,
396.1021.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.oprd.9b00456.
■
S
Additional experimental data and H, 13C, COSY, and
1
HSQC NMR spectra of compounds 2−8 (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: peter.seeberger@mpikg.mpg.de.
■
ORCID
̈
Bartholomaus Pieber: 0000-0001-8689-388X
Peter H. Seeberger: 0000-0003-3394-8466
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully acknowledge the Max-Planck Society for
generous financial support. We acknowledge DFG SFB/
Transregio 67 for funding to M.G. B.P. acknowledges financial
support by a Liebig Fellowship of the German Chemical
Industry Fund (Fonds der Chemischen Industrie, FCI). We
are grateful to Dr. Anna Nickel (nee Chentsova) for initial
advice in flow technology to M.G. We thank Eva Settels for
support with HPLCs and other analytical devices and Olaf
Niemeyer for support with NMR studies.
(16) Czernecki, S.; Ayadi, E.; Randriamandimby, D. Seleno
Glycosides. 3. Synthesis of Phenyl 2-(N-Acetylamino)-and 2-Azido-
2-Deoxy-1-Seleno-α-D-Glycopyranosides via Azido-Phenylselenyla-
tion of Diversely Protected Glycals. J. Org. Chem. 1994, 59, 8256−
8260.
(17) Czernecki, S.; Randriamandimby, D. Azido-Phenylselenylation
of Protected Glycals. Tetrahedron Lett. 1993, 34, 7915−7916.
(18) Langerman, N. Chemical Safety: Explosion hazard in synthesis
of azidotrimethylsilane http://cen.acs.org/articles/92/i43/Chemical-
Safety-Explosion-hazard-synthesis.html (accessed Oct 10, 2019).
(19) Baxendale, I. R. The Integration of Flow Reactors into
Synthetic Organic Chemistry. J. Chem. Technol. Biotechnol. 2013, 88,
519−552.
ABBREVIATIONS
■
APS, azidophenylselenylation; BAIB, bisacetoxy iodobenzene;
GalN, galactosamine; GalNAc, N-acetylgalactosamine; GlcN,
glucosamine; TLC, thin layer chromatography
REFERENCES
■
(20) Gutmann, B.; Cantillo, D.; Kappe, C. O. Continuous-Flow
Technology - A Tool for the Safe Manufacturing of Active
(1) Essentials of Glycobiology, 2nd ed.; Varki, A., Cummings, R. D.,
Esko, J. D., Freeze, H. H., Stanley, P., Bertozzi, C. R., Hart, G. W.,
F
DOI: 10.1021/acs.oprd.9b00456
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
Pharmaceutical Ingredients. Angew. Chem., Int. Ed. 2015, 54, 6688−
6728.
(21) Plutschack, M. B.; Pieber, B.; Gilmore, K.; Seeberger, P. H. The
Hitchhiker’s Guide to Flow Chemistry. Chem. Rev. 2017, 117,
11796−11893.
(22) Eller, S.; Collot, M.; Yin, J.; Hahm, H. S.; Seeberger, P. H.
Automated Solid-Phase Synthesis of Chondroitin Sulfate Glycosami-
noglycans. Angew. Chem., Int. Ed. 2013, 52, 5858−5861.
(23) Mironov, Y. V.; Grachev, A. A.; Lalov, A. V.; Sherman, A. A.;
Egorov, M. P.; Nifantiev, N. E. A Study of the Mechanism of the
Azidophenylselenylation of Glycals. Russ. Chem. Bull. 2009, 58, 284−
290.
(24) Chloroform was used as carrier solvent for improving the
performance of the pump.
(25) Nowacki, A.; Liberek, B. Comparative Conformational Studies
of 3,4,6-Tri-O-Acetyl-1,5-Anhydro-2-Deoxyhex-1-Enitols at the DFT
Level. Carbohydr. Res. 2018, 462, 13−27.
(26) Tingoli, M.; Tiecco, M.; Testaferri, L.; Temperini, A.
Iodosobenzene Diacetate and Diphenyl Diselenide: An Electrophilic
Selenenylating Agent of Double Bonds. Synth. Commun. 1998, 28,
1769−1778.
́
́
(27) Santoyo-Gonzalez, F.; Calvo-Flores, F. G.; Hernandez-Mateo,
́
F.; García-Mendoza, P.; Isac-García, J.; Perez-Alvarez, M. D. Radical
β-Elimination of Vicinal Phenylselenide and Xanthate Azides in Sugar
Derivatives. Synlett 1994, 1994, 454−456.
(28) SanMartin, R.; Tavassoli, B.; Walsh, K. E.; Walter, D. S.;
Gallagher, T. Radical-Mediated Synthesis of α-C-Glycosides Based on
N-Acyl Galactosamine. Org. Lett. 2000, 2, 4051−4054.
(29) Tiecco, M.; Tingoli, M.; Testaferri, L. The Use of Selenium
Containing Intermediates to Mediate or Catalyze New Organic
Reactions. Pure Appl. Chem. 1993, 65, 715−722.
(30) Pedersen, C. M.; Marinescu, L. G.; Bols, M. Radical
Substitution with Azide: TMSN3 − PhI(OAc)2 as a Substitute of
IN3. Org. Biomol. Chem. 2005, 3, 816−822.
(31) Shiao, T. C.; Rodrigue, J.; Renaudet, O.; Roy, R. Synthesis of
3,4,6-Tri-O-Acetyl-D-Galactal. In Carbohydrate Chemistry: Proven
́
Synthetic Methods, Vol. 2; van der Marel, G. A., Codee, J. D. C.,
Eds.; CRC Press: Boca Raton, 2014; pp 253−256.
(32) Snider, B. B.; Lin, H. An Improved Procedure for the
Conversion of Alkenes and Glycals to 1,2-Diazides Using Mn(OAc)3·
2H2O in Acetonitrile Containing Trifluoroacetic Acid. Synth.
Commun. 1998, 28, 1913−1922.
G
DOI: 10.1021/acs.oprd.9b00456
Org. Process Res. Dev. XXXX, XXX, XXX−XXX