Synthesis of Aryl C-Glycosides via Iron-Catalyzed Cross Coupling of Halosugars: Stereoselective Anomeric Arylation of Glycosyl Radicals

Article abstract of DOI:10.1021/jacs.7b03867

We have developed a novel diastereoselective iron-catalyzed cross-coupling reaction of various glycosyl halides with aryl metal reagents for the efficient synthesis of aryl C-glycosides, which are of significant pharmaceutical interest due to their biological activities and resistance toward metabolic degradation. A variety of aryl, heteroaryl, and vinyl metal reagents can be cross-coupled with glycosyl halides in high yields in the presence of a well-defined iron complex, composed of iron(II) chloride and a bulky bisphosphine ligand, TMS-SciOPP. The chemoselective nature of the reaction allows the use of synthetically versatile acetyl-protected glycosyl donors and the incorporation of various functional groups on the aryl moieties, producing a diverse array of aryl C-glycosides, including Canagliflozin, an inhibitor of sodium-glucose cotransporter 2 (SGLT2), and a prevailing diabetes drug. The cross-coupling reaction proceeds via generation and stereoselective trapping of glycosyl radical intermediates, representing a rare example of highly stereoselective carbon-carbon bond formation based on iron catalysis. Radical probe experiments using 3,4,6-tri-O-acetyl-2-O-allyl-α-d-glucopyranosyl bromide (8) and 6-bromo-1-hexene (10) confirm the generation and intermediacy of the corresponding glycosyl radicals. Density functional theory (DFT) calculations reveal that the observed anomeric diastereoselectivity is attributable to the relative stability of the conformers of glycosyl radical intermediates. The present cross-coupling reaction demonstrates the potential of iron-catalyzed stereo- and chemoselective carbon-carbon bond formation in the synthesis of bioactive compounds of certain structural complexity.

Full text of DOI:10.1021/jacs.7b03867

Article
pubs.acs.org/JACS
Synthesis of Aryl C‑Glycosides via Iron-Catalyzed Cross Coupling of
Halosugars: Stereoselective Anomeric Arylation of Glycosyl Radicals
Laksmikanta Adak,†,‡ Shintaro Kawamura,†,‡ Gabriel Toma,†,‡ Toshio Takenaka,†,‡ Katsuhiro Isozaki,†,‡
Hikaru Takaya,†,‡ Akihiro Orita,∥ Ho C. Li,§ Tony K. M. Shing,§ and Masaharu Nakamura*,†,‡
†International Research Center for Elements Science, Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan
‡Department of Energy and Hydrocarbon Chemistry, Graduate School of Engineering, Kyoto University, Kyoto, 615-8510, Japan
§Department of Chemistry and Center of Novel Functional Molecules, The Chinese University of Hong Kong, Shatin, New
Territories, Hong Kong SAR, China
∥Department of Applied Chemistry, Okayama University of Science, Ridai-cho, Okayama 700-0005, Japan
S
* Supporting Information
ABSTRACT: We have developed a novel diastereoselective iron-
catalyzed cross-coupling reaction of various glycosyl halides with
aryl metal reagents for the efficient synthesis of aryl C-glycosides,
which are of significant pharmaceutical interest due to their
biological activities and resistance toward metabolic degradation.
A variety of aryl, heteroaryl, and vinyl metal reagents can be cross-
coupled with glycosyl halides in high yields in the presence of a
well-defined iron complex, composed of iron(II) chloride and a
bulky bisphosphine ligand, TMS-SciOPP. The chemoselective
nature of the reaction allows the use of synthetically versatile
acetyl-protected glycosyl donors and the incorporation of various
functional groups on the aryl moieties, producing a diverse array
of aryl C-glycosides, including Canagliflozin, an inhibitor of sodium-glucose cotransporter 2 (SGLT2), and a prevailing diabetes
drug. The cross-coupling reaction proceeds via generation and stereoselective trapping of glycosyl radical intermediates,
representing a rare example of highly stereoselective carbon−carbon bond formation based on iron catalysis. Radical probe
experiments using 3,4,6-tri-O-acetyl-2-O-allyl-α-D-glucopyranosyl bromide (8) and 6-bromo-1-hexene (10) confirm the
generation and intermediacy of the corresponding glycosyl radicals. Density functional theory (DFT) calculations reveal that the
observed anomeric diastereoselectivity is attributable to the relative stability of the conformers of glycosyl radical intermediates.
The present cross-coupling reaction demonstrates the potential of iron-catalyzed stereo- and chemoselective carbon−carbon
bond formation in the synthesis of bioactive compounds of certain structural complexity.
varied functional and protecting groups are tolerated and
undesired elimination and/or epimerization reactions are
suppressed.5 However, aryl C-glycoside synthesis using radical
C-glycosidation has been a challenge and is virtually
undeveloped methodology because of the paucity of suitable
arylating agents for the glycosyl radical intermediates. This is
primarily due to the inertness of aromatic groups or compounds
toward glycosyl radicals.6
INTRODUCTION
■
C-glycosides represent an important class of natural and/or
synthetic bioactive compounds of high medicinal significance
and therefore have received considerable synthetic attention.1
Specifically, aryl C-glycosides are of marked pharmaceutical
interest because of their biological activities and resistance
toward metabolic degradation.2 As a notable example, aryl C-
glycoside-based inhibitors of sodium-glucose cotransporter 2
(SGLT2) have been utilized in the treatment of type 2 diabetes.
Although numerous synthetic methods to prepare aryl C-
glycosides are available,3,4 the development of efficient,
stereoselective, cost-effective, and nontoxic methods is critical
for the advancement of biologically active and pharmaceutically
important carbohydrate analogues.
Recently transition-metal-catalyzed cross-coupling reactions
have emerged as a useful method for the synthesis of aryl C-
glycosides. In particular, nickel and cobalt catalysts have enabled
the cross-coupling reactions between glycosyl bromides and
arylzinc or arylmagnesium reagents, providing the saturated aryl
C-glycosides, which has not been achieved via conventional
palladium catalysis.7,8 Gagne
́
and co-workers reported a nickel-
Among the various existing approaches for the synthesis of C-
glycosides,1 C-glycosidation via anomeric glycosyl radical
intermediates has been widely used because of its advantages
over other methods based on the corresponding glycosyl anions
and cations: the reaction conditions are generally mild so that
catalyzed procedure for the synthesis of aryl C-glycosides (eq
Received: April 18, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.7b03867
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
1)9 and suggested that the reaction might proceed through the
generation of glycosyl radical intermediates. Cossy and co-
versatile acetate-protected glycosyl donors as well as the
inclusion of various functional groups on the aryl moiety. In
addition, the practical advantages of iron catalysts derived from
the toxicologically benign nature and cost-effectiveness of iron
makes the present reaction a synthetically and industrially
feasible organic transformation.
RESULTS AND DISCUSSION
■
1. Arylation of Glycosyl Chlorides and Bromides:
Screening of Catalysts, Ligands, and Organometallic
Nucleophiles. We began our study by exploring the effects of
catalysts and ligands by using a challenging substrate, 2,3,5-tri-
O-acetyl-β-D-ribofuranosyl chloride 1, which is not viable in
nickel- and cobalt-based aryl C-glycosidation reactions.9a,10a
The reactions of 1 with phenylzinc reagent prepared from
PhMgBr and ZnCl2(tmeda), Ph2Zn(tmeda)·2MgBrCl were
thus studied (Table 1 and Chart 1).15,16 The yield of the
desired product 1a depended on the catalysts, and all the
reactions proceeded with excellent β-selectivity of product 1a
(entries 1−9, Table 1). The coupling reaction proceeded
without addition of a ligand, albeit in a low yield (30%, entry 1,
Table 1).16 In the absence of the catalyst, the coupling reaction
did not progress (entry 2). Several bisphosphine ligands such as
Table 1. Screening of Catalyst Precursors and Ligands in the
a
Reaction of 1 and 2 with Phenylzinc Reagent
workers reported the cobalt-catalyzed cross-coupling reaction of
glycosyl bromides and aryl Grignard reagents in the synthesis of
aryl C-glycosides, and illustrated the formation of an anomeric
radical intermediate (eq 2).10 These reports revealed the
potential utility of iron-group metals in transition-metal-
catalyzed aryl C-glycosidation reactions via radical intermedi-
ates. There has been, nonetheless, no successful iron catalysis in
this synthetically attractive organic transformation.
Despite the recent surge in the use of iron catalysts in cross-
coupling reaction technologies,11 the iron-catalyzed cross-
coupling has thus remained virtually unexplored in the synthesis
of aryl C-glycosides: Cossy reported a single example where the
cross-coupling of pyranosyl bromides with phenylmagnesium
bromide proceeded in the presence of an iron salt and Xantphos
to give the desired phenyl C-glycoside as a minor product (17−
42% yield), due to competing side reactions (e.g., the undesired
glucal formation as a major product) (eq 3).10a,12 Our recent
studies regarding the precise control of iron catalysis in the
cross-coupling reaction of alkyl halides13,14 revealed that an
iron-chiral bisphosphine ligand (BenzP*) enabled the stereo-
selective trapping of alkyl radicals;14a therefore, we envisaged
that the aryliron intermediate derived from well-defined iron
phosphine complexes [FeCl2(SciOPP)] would be a suitable
arylating agent for glycosyl radical intermediates to achieve a
high cross-coupling selectivity in the synthesis of aryl C-
glycosides.
% yield of
b
RSM
bc
,
entry substrate
iron catalyst (mol %)
FeCl3 (3)
product
(%)
1
2
3
4
5
6
7
1
1
1
1
1
1
1
30
0
0
none
0
0
0
0
0
0
FeCl3 (3) + DPPE (3)
FeCl3 (3) + DPPP (3)
FeCl3 (3) + DPPBz (3)
FeCl3 (3) + SciOPP (3)
48
50
52
52
55
FeCl3 (3) +
TMS-SciOPP (3)
8
1
1
2
2
2
2
2
2
2
FeCl2(TMS-SciOPP) (3)
FeCl2(TMS-SciOPP) (5)
FeCl3 (3)
72
0
0
d
9
81
10
11
12
13
14
15
16
11
0
67
85
58
77
56
55
29
none
FeCl3(3) + DPPE(3)
FeCl3(3) + DPPP(3)
FeCl3(3) + DPPBz(3)
FeCl3(3) + SciOPP(3)
7
21
17
32
50
FeCl3 (3) +
TMS-SciOPP (3)
17
18
19
2
2
2
FeCl2(TMS-SciOPP) (3)
FeCl2(TMS-SciOPP) (5)
FeCl2(TMS-SciOPP) (5)
77
78
20
18
12
Herein, we present the efficient synthesis of various aryl C-
glycosides, for the first time, using the iron-catalyzed cross-
coupling of glycosyl chlorides and bromides with a variety of
aryl-, heteroaryl-, and vinylmetal reagents (eq 4). The reactions
are high yielding and diastereoselective and tolerate many
functional groups, which allows for the use of synthetically
de
,
84
a
b
Reactions were conducted on 0.25 mmol scale. Yields were
determined by 1H NMR using 1,1,2,2-tetrachloroethane as an internal
c
standard unless otherwise noted. Recovery of starting material.
d
e
Isolated yield. Reaction was run at rt.
B
DOI: 10.1021/jacs.7b03867
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Chart 1. Structures of Ligands
Table 2. Iron-Catalyzed C-Glycosidation of Various
Phenylmetal Reagents
a
b
glycosyl
halide
% yield of
RSM
bcd
, ,
entry
1
Ph−M
C-glycoside (α/β)
(%)
1
Ph2Zn(tmeda)·
81 (1/>99)
0
2MgBrCl
DPPE, DPPP, DPPBz, and SciOPP (Chart 1) showed
comparable reactivity and provided the desired cross-coupling
products in 48−50% yields (entries 3−6). TMS-SciOPP
afforded the corresponding cross-coupling product in 55%
yield (entry 7). A well-defined iron complex, FeCl2(TMS-
SciOPP), emerged as the most effective catalyst, affording 1a in
72% yield (entry 8). The highest yield (81%) was obtained with
a 5 mol % of the catalyst (entry 9).
The effects of catalysts and ligands were also investigated
using a more challenging substrate, 2,3,4-tri-O-acetyl-β-D-
ribopyranosyl chloride 2 (entries 10−19, Table 1). All reactions
showed excellent β-selectivity of product 2a. Several bi-
sphosphine ligands showed different reactivities (entries 1−7
vs 10−16). Among the bisphosphine ligands, TMS-SciOPP
provided the cross-coupling product in 50% yield (entry 16).
FeCl2(TMS-SciOPP) was found to be the most effective
catalyst, and it provided 2a in 77% yield (entry 17). The highest
yield (84%) was obtained with 5 mol % of the catalyst, and the
reaction was performed at rt (entries 18 and 19).17
e
2
1
1
1
3
Ph3Al−3MgCl2
PhB(pin) + t-BuLi
PhMgBr
30 (1/>99)
22 (1/>99)
0
0
0
0
0
f
3
g
4
5
Ph2Zn(tmeda)·
2MgBrCl
96 (73/27)
6
3
3
3
3
Ph2Zn·2MgBrCl
Ph3Al·3MgCl2
PhB(pin) + t-BuLi
PhMgBr
18 (70/30)
80 (50/50)
70 (76/24)
51
e
7
10
12
94
f
8
g
9
0
a
b
Reactions were conducted at 0 °C for 12 h. Yields were determined
by H NMR using 1,1,2,2-tetrachloroethane as an internal standard.
Recovery of the starting material 1 and 3. Tri-O-acetyl-D-glucal was
obtained as the byproduct in entries 5−9. 1.7 equiv of TMEDA was
added in entries 2 and 7. 20 mol % of MgBr2 was added as a
cocatalyst, and the reaction was run at rt. 5 equiv of TMEDA were
1
c
d
e
f
g
added in entries 4 and 9, and PhMgBr was added at 0 °C over 3 h
using a syringe pump.
cross-coupling protocol (entry 1, Table 2) was tested with
various arylzinc reagents; Table 3 illustrates the scope of the
reaction with 1 for the synthesis of diverse aryl C-ribofurano-
sides.4 All reactions showed excellent β-D-selectivities (α/β = 1/
>99). The reaction of 1 with a phenylzinc reagent and other
After screening the catalyst precursors and ligands, we
commenced a search for suitable organometallic nucleophiles
for the phenylation of 2,3,5-tri-O-acetyl-β-ribofuranosyl chloride
(1) and 2,3,4,6-tetra-O-acetyl-α-D-glucosyl bromide (3) in the
presence of a catalytic amount of FeCl2(TMS-SciOPP) (Table
2). Phenylzinc reagent Ph2Zn(tmeda)·2MgBrCl proved to be
the most effective nucleophile in the model reaction (entries 1
and 5). Phenylaluminum14k as well as phenylboron14m also
showed promising reactivities and provided 3a in 80% and 70%
yield (entries 7 and 8). Notably, this is the first example of the
use of organoaluminum and boron reagents in the cross-
coupling reaction of glycosyl halides.9a,10a
Table 3. Iron-Catalyzed Aryl C-Glycosidation of
a
Ribofuranosyl Chloride 1
The reaction of 1 gave the desired C-glycoside 1a with nearly
perfect β-selectivity (81%, α/β = 1/>99). The phenylation of
2,3,4,6-tetra-O-acetyl-α-D-glucosyl bromide (3) gave the corre-
sponding C-glucopyranoside 3a as a mixture of anomeric
isomers (α/β = 50/50−76/24) (entries 5−8). The reaction
with a phenyl aluminum reagent resulted in lower diaster-
eoselectivity (50:50 in entry 7). Although details are unclear, we
postulate tentatively the strong Lewis acidity of aluminum
species influences the fluctuating conformation of an
intermediate glucosyl radical (vide infra): the diastereoselectiv-
ity is slightly diverse when using other phenyl nucleophiles.
Coexisting TMEDA was critical to prevent the undesired
elimination reaction leading to glucal formation; when the
phenylzinc reagent was prepared in the absence of TMEDA, the
catalytic cross-coupling reaction did not proceed (entry 6).18
PhMgBr did not give the desired product, and 3 was recovered
almost quantitatively (entry 9),10a suggesting that the Grignard
reagent may allow over-reduction of the catalytic iron(II)
species (see the discussion regarding the mechanism of the
reaction described below).
a
Reactions were performed on a 0.25−0.50 mmol scale. See the
2. β-Selective Arylation of Ribofuranosyl and Ribopyr-
anosyl Chlorides with Arylzinc Reagents. The standardized
Supporting Information for details regarding the reaction conditions
for each case.
C
DOI: 10.1021/jacs.7b03867
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
substituted arylzinc reagents containing methoxy, fluoro,
phenyl, and N,N-dimethylamino groups in the para-position
gave the corresponding products 1a−1e in 75−84% yields.19 3-
Methoxy phenylzinc also provided a good yield (78%) of 1f. A
terminal olefin moiety, which often undergoes isomerization to
an internal olefin under transition metal catalysis,20 remained
intact under the present conditions and provided 1g in 71%
yield. 2-Naphthyl and 5- N-methylindolylzinc also reacted
smoothly to give 1h and 1i in 73% and 64% yields, respectively.
Not only ribofuranosyl chloride but also ribopyranosyl
chloride 2 participated in the aryl C-glycosidation; the cross-
coupling reactions of 2 with a variety of aryl and heteroarylzinc
reagents furnished the corresponding products in good to
excellent yields (65−90%) with excellent β-D-selectivities (α/β
= 1/>99) (Table 4).19 The trimethylsilylmethyl (TMSM)
Table 5. Iron-Catalyzed Aryl C-Glycosidation of O-Acetyl-α-
bromo-D-glucose 3
a
Table 4. Iron-Catalyzed Aryl C-Glycosidation of
a
Ribopyranosyl Chloride 2
a
b
Reactions were performed on a 0.25−1.25 mmol scale. 10 mol % of
c
FeCl2(TMS-SciOPP) was used. O-Acetyl-α-D-glucosyl chloride (3′)
d
was used, and the reaction was run at rt for 36 h. Arylzinc reagent was
e
prepared from 2-pyridylZnBr, Me3SiCH2MgCl, and TMEDA. 5 mol
% of FeCl2(TMS-SciOPP) was used. Arylzinc reagent was prepared
f
from arylmagnesium chloride and ZnBr2 (tmeda). See the Supporting
g
Information for detailed reaction conditions for each case. Minor
anomer, β-3h, is shown.
Table 6. Iron-Catalyzed Aryl C-Glycosidation of Various
a
Bromosugars 4−6
a
b
Reactions were conducted on a 0.25−0.50 mmol scale. Reaction was
c
performed on the gram scale. Arylzinc reagent was prepared from 2-
pyridylZnBr, Me3SiCH2MgCl, and TMEDA. See the Supporting
Information for detailed reaction conditions for each case.
group, which can act as a nontransferable dummy ligand for
pyridyl zinc reagents, participated in the reaction21 and
provided the corresponding product 2l in 68% yield. The
isomerization of the terminal olefin was not observed (2j). The
high reactivity and stereoselectivity of glycosyl chlorides using
the iron catalyst demonstrate its advantages over nickel- and
cobalt-based glycosidation reactions, which cannot be applied
for the diastereoselective arylation of glycosyl chlorides.9a,10a
3. Diastereoselective Arylation of Glycosyl Bromides
with Arylzinc Reagents. To our delight, the iron catalyst also
showed excellent reactivity and selectivity toward various
glycosyl bromides (Tables 5 and 6). The cross-coupling
reactions of 3 with different aryl and heteroarylzinc reagents
afforded the corresponding products (3a−3g) in 61−96% yields
and with varied anomeric selectivities (α/β = 75/25−3/97),
favoring the α-D or β-D anomer depending on the aryl
a
b
Reactions were conducted on a 0.25−0.50 mmol scale. Reaction was
c
performed on the gram scale. 10 mol % FeCl2(TMS-SciOPP) was
d
used. O-Acetyl-α-D-mannosyl chloride (4′) was used and the reaction
was run at rt for 36 h. See the Supporting Information for details.
nucleophile (Table 5). The reaction of O-acetyl-α-D-glucosyl
chloride 3′ with phenylzinc gave 3a in 51% yield with anomeric
selectivity of α/β = 68/32.22 The reactions with N-
methylindolylzinc proceeded smoothly to afford the corre-
D
DOI: 10.1021/jacs.7b03867
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
sponding coupling product 3b in 95% yield. Electron-deficient
aromatic groups were introduced using an appropriate arylzinc
reagent to give the corresponding aryl C-glycosides (3d and 3e)
in 90% and 76% yield, respectively. The reactions with O-
tolylzinc proceeded smoothly and provided 3f in 89% yield. A
notable stereochemical switch was observed when the 2-tolyl
group was replaced with a 2-anisyl group to produce the β-D-
anomer 3g with high selectivity (α/β = 3/97). This selectivity
may result from electronic (or coordinating) effects of the
methoxy group, as judged by the 45/55 selectivity of the 2-
tolylzinc reagent. These results contrast with the corresponding
nickel-9a and cobalt-catalyzed10a arylation reactions, which are
not suitable for introducing 2-substituted aromatic and pyridyl
rings. The cross-coupling reaction of 3 with an arylzinc reagent
bearing a heteroaromatic ring was also applicable for the
synthesis of pharmaceutically important molecule 3h in 61%
yield.
Table 7. Iron-Catalyzed Cross Coupling of Glycosyl Halides
with Alkenylzinc Reagents
a
Next, our cross-coupling protocol was applied to other
glycosyl halides (Table 6). The cross-coupling reactions of O-
acetyl-α-D-mannosyl bromide and chloride (4 and 4′) were
highly diastereoselective and produced only the α-D-anomers
(α/β = > 99/1). Electron-neutral, electron-rich, and electron-
deficient arylzinc reagents participated in the high-yielding
arylation (90−98%, 4a−4d). The phenylation of O-acetyl-α-
bromo-D-galactose 5 proceeded diastereoselectively to give only
the α-D-anomer 5a in 89% yield. A disaccharide, hepta-O-acetyl-
α-D-maltosyl bromide 6, also participated and afforded the
desired aryl C-glycoside, 6a in 88% yield.
4. Vinyl C-Glycoside Synthesis. Having established the
optimal procedure for aryl-C-glycoside synthesis, we examined
the cross-couplings between glycosyl halides and alkenylzinc
reagents because vinyl C-glycosides are also potentially useful
intermediates leading to a variety of carbasugars and C-glycosyl
compounds.23 Among them, intra- and intermolecular glycosyl
radical additions to alkynes24 are useful, because of their mild
reaction conditions, which allow for the presence of various
polar functional groups on the sugar derivatives. The synthesis
of vinyl C-glycosides was performed using glycosyl halides 1−4,
and the desired cross-coupling products were obtained in good
to excellent yields (70−96%) with high diastereoselectivities
(Table 7).
a
b
Reactions were performed on a 0.25 mmol scale. The anomeric ratio
1
was determined by H NMR analysis of the crude reaction mixture.
c
Reaction was run at rt.
5. Deprotection of Acetyl Group. An advantage of the
present reaction is the use of acetyl protection, which is easily
removed under mild conditions. Tetraacetylmannose α-4a was
synthesized on a gram scale (eq 5) and deprotected by
transesterification using Bu2SnO/MeOH25 to afford the desired
1-α-phenyl D-mannose (7a) in a 90% yield (eq 6). Because of
the small molecular weight of acetyl protecting groups, the loss
of the molecular weight during the deprotection is far less than
those of benzyl-, benzoyl-, and pivaloyl-protection, which are
commonly used in organometallic reactions of glycosyl
donors.26 The acetyl groups of isolated β-3h were also cleanly
removed using the tin oxide catalyst in refluxing methanol (eq
7), which stereoselectively provided a representative type 2
antidiabetic drug, Canagliflozin (7b, 95% yield).
Scheme 1. Cross-Coupling Reaction Using Radical Probe 8
6. Mechanistic Considerations: Radical Probe Experi-
ments and Plausible Catalytic Cycle. In order to gain
mechanistic insight into the cross-coupling reaction, we
performed the cross-coupling reaction with a radical probe 8
and phenylzinc under the reaction conditions described above
(Scheme 1). The radical probe experiment produced a
diastereomeric mixture of bicyclic compound 9 (d.r. = 60:40),
supporting the intermediacy of the corresponding glycosyl
radical intermediate.5b,e,10a We observed the exclusive formation
of the ring-closing product, 9, and no anomeric arylation
product 9′. This selectivity can be attributed to the fast 5-exo-
trig cyclization of the allyl ether moiety and subsequent cross-
coupling reaction with a phenylzinc reagent.27
E
DOI: 10.1021/jacs.7b03867
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
6-Bromo-1-hexene 10 was next adopted as a model radical
probe to investigate the dependence of the ratio of the direct
arylation and cyclization/arylation on catalyst loading.28 The
reaction with phenylzinc gave a mixture of direct arylation
(uncyclized) product 11 and cyclized 12 at different catalyst
loadings, which was consistent with the formation of an alkyl
radical intermediate as described above. This radical probe
reaction with various catalyst loadings of FeCl2(TMS-SciOPP)
resulted in the observation of an almost linear relationship
between the ratio of 11/12 and the catalyst loading (inset in
Scheme 2). This supports the possibility that once formed, the
radical intermediate to give the corresponding aryl C-glycoside
and iron(I) species D possessing one bulky TMS-SciOPP
ligand.14l Iron(I) species is reported to show high reactivity
toward haloalkanes in Negishi and other types of coupling
reactions.31 We therefore consider that D is the most likely
species responsible for the homolytic cleavage of the C−X bond
of halosugars,32 and once it forms under the reaction conditions
the radical chain reaction may proceed smoothly.30a
The mechanism of C−C bond formation of the glycosyl
radical with the intermediate A in Figure 1 has been unclear and
under intensive theoretical investigation. We are tentatively
postulating two possible pathways, e.g., a reversible formation of
aryl(glycosyl)iron(III) species B and subsequent reductive
elimination, or an ipso-substitution of the aryl group of the
intermediate A by the attack of the glycosyl radical via TS C to
furnish the cross-coupling product and the iron(I) species D.
The correlation between the conformational stability of glycosyl
radical interemediates and the observed anomeric stereo-
selectivity (vide infra) implies the second mechanism via the
direct ipso-substitution may operate but does not exclude the
possibility of the first trap/reductive elimination mechanism
(Figure 1).33
Scheme 2. Cross-Coupling Reaction Using Model Radical
Probe 10
7. Correlation of Stereoselectivity of Anomeric
Arylation to Conformational Stability of Glycosyl
Radicals. In the preceding sections, we show that the arylation
of glycosyl halides proceeds through the formation of glycosyl
radicals. Giese and Togo reported that the stability of glycosyl
radical conformations could influence the stereoselectivity of
intermolecular reaction of glycosyl radicals with olefins to
produce the corresponding alkyl C-glycosides.34,35 In our case,
the observed anomeric diastereoselectivity of the aryl C-
glycosides can also be correlated to the conformation of the
glycosyl radical intermediates. The results obtained by density
functional theory (DFT) calculations are summarized in Figure
2.17
alkyl radical intermediate escapes from the solvent cage and
undergoes cyclization/arylation or direct arylation with an aryl
iron species, which is different from the one that reacts with the
alkyl halide to generate the corresponding alkyl radical
intermediate.29,30
Figure 1 shows a plausible mechanism for the present cross-
coupling reaction based on a bimetallic (out-of-cage) mecha-
Aryl C-glycosidation of mannopyranosyl bromide 4 showed
excellent α-D-selectivity (α/β = >99/1) of 4a−4d, (Table 6). As
4
shown in Figure 2a, the C1 conformer of the mannopyranosyl
1
radical is more favorable than the C4 conformer (ΔG = 3.2
kcal/mol) where the α-face exhibits less steric hindrance than
the β-face. The recombination (C−C bond forming step)
between the mannopyranosyl radical and aryliron species A may
take place from the selective α-face and bring about the α-
selectivity of the products.
Aryl C-glycosidation of ribopyranosyl chloride 2 demon-
strated excellent β-D-selectivity (α/β = 1/>99) of 2a−2l (Table
1
4). In the case of ribopyranosyl radicals, the C4 conformer is
more stable and exhibits strong steric hindrance at the α-face
with the aryliron species and would result in β-D-selectivity
(Figure 2b).
Aryl C-glycosidation of glucopyranosyl bromide 3 showed
varied α/β selectivities, ranging from 75/25 to 3/97 (3a−3g as
in Table 5). This was probably due to the fluctuating
conformational changes of the glucosyl radical intermediates
(Figure 2c), which makes the diastereoselectivity sensitive to
the structure of the aryl nucleophiles.34
Figure 1. Plausible reaction mechanism (P P, X, and R denote TMS-
SciOPP, halogen, and H or CH2OH, respectively.).
From our DFT calculations, we found that the B2,5 boat
4
1
conformer (average structure of C1 and C4 conformers) is
slightly more stable than the other conformers (ΔG1 = 0.25
kcal/mol and ΔG2 = 0.57 kcal/mol). The radical at C-1 of the
boat conformation has p-character, providing a planar carbon
center (i.e., it is a π radical) with potential to form both anomers
through α- and β-attacks.34 As shown in Table 5, the α/β
nism, which we recently proposed in the iron-catalyzed
enantioselective cross-coupling reaction.29,30 The catalytic
cycle starts from the formation of reactive divalent organoiron
species A, which is generated from FeCl2(TMS-SciOPP) and an
aryl zinc reagent. The organoiron A reacts with a glycosyl
F
DOI: 10.1021/jacs.7b03867
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 2. Relative DFT stabilities of conformers of glycosyl radical intermediates.
selectivity depended on the aryl nucleophiles. Although the
reason for the unique high selectivity in the case of 2-anisyl is
unclear, the interaction or coordination of the methoxy group to
the iron center or to the accompanying Lewis acidic metal salts
may alter the coordination environment of the iron center.
Our DFT study on the ribofuranosyl radical also revealed that
it displays only one stable conformer, possessing an envelope
structure with a trigonal planar carbon radical (Figure 2d). The
two-acetoxy groups sterically crowded the α-face of the
furanosyl radical, leaving the β-face open for further
reactions.34a,36 Based on this structure, anomeric C−C bond
formation with an aryliron species can be expected to proceed
selectively from the β-face and provided the corresponding aryl
C-ribofuranosides (1a−1i) with exclusive β-D-selectivity (α/β =
1/>99). Although there remains room for discussion on the
detailed mechanism of the C−C bond forming step, the
qualitative understanding and prediction of the anomeric
selectivity is shown to be possible by using DFT analysis of
conformational stability of glycosyl radical intermediates.
stereo- and chemoselective C−C bond formation and spurs
their application to the synthesis of complex molecules of
biological importance.
EXPERIMENTAL SECTION
■
Representative Procedure for the Cross Coupling of O-
Acetyl-α-D-mannosyl Bromide 4 with Phenylzinc Reagent
(Gram Scale Reaction). A THF solution of phenylmagnesium
bromide (12.2 mL, 1.11 M in THF, 13.6 mmol) was added to a
suspension of ZnCl2 (tmeda) (1.72 g, 6.8 mmol) in 6 mL of THF at
ambient temperature; the reaction mixture was stirred at 0 °C for 45−
60 min. 2,3,4,6-Tetra-O-acetyl-α-D-mannopyranosyl bromide (1.64 g,
4.0 mmol) was added to the reaction mixture at 0 °C followed by a
THF solution of FeCl2 (TMS-SciOPP) (1.6 mL, 0.1 M in THF
solution, 0.16 mmol). The coupling reaction was carried out for 18 h at
0 °C. After dilution with 50 mL of CH2Cl2 at 0 °C, the solution was
filtered using a pad of Florisil and concentrated in vacuo. The cross-
coupling product 4a (1.6 g, 98% yield) was obtained as a white solid
after silica-gel column chromatography (hexane/EtOAc = 80/20).
Representative Procedure for Deprotection of Acetyl Group
from Sugar Derivative. The β-D-anomeric product 3h (141 mg, 0.23
mmol), dibutyltin oxide (25 mg, 0.092 mmol), and MeOH (1.7 mL)
were placed in a Schlenk tube, and the mixture was stirred at the reflux
for 24 h. After cooling to ambient temperature, 5.0 mL of water were
added and the mixture was stirred for 10 min at rt. The entire reaction
mixture was transferred into a 100 mL round-bottom flask, and water
was evaporated by azeotropic distillation with additional MeOH. The
residue was dried in vacuo and purified by short silica-gel column
chromatography (MeOH/EtOAc = 10/90) to give the desired product
Canagliflozin 7b as an off-white solid (97 mg, 95% yield).
CONCLUSION
■
In summary, we have developed the efficient synthesis of aryl C-
glycosides and vinyl C-glycosides through the iron-catalyzed
stereoselective cross-coupling reaction of a variety of glycosyl
chlorides and bromides with diverse aryl, heteroaryl, and vinyl
metal reagents. This is the first example of the highly selective
cross-coupling of rather unreactive glycosyl chlorides. The
reaction proceeded via a glycosyl radical intermediate, and the
anomeric diastereoselectivity of the products was explained by
the relative stability of glycosyl radical intermediate and thus
was predictable to some extent. We hope that this simple,
efficient, and practical synthesis of C-glycosides will provide a
new entry to development of a wide array of functional sugar
derivatives of medicinal and clinical interest. We believe that the
present study demonstrates the potential of iron-catalyzed
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.7b03867.
G
DOI: 10.1021/jacs.7b03867
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(d) Beuck, C.; Singh, I.; Bhattacharya, A.; Hecker, W.; Parmar, V. S.;
Seitz, O.; Weinhold, E. Angew. Chem., Int. Ed. 2003, 42, 3958.
(5) Selected books and papers: (a) Skrydstrup, T.; Beau, J.-M. In
Advances in Free Radical Chemistry; Zard, S. Z., Ed.; Elsevier:
Greenwich, CT, 1999; Vol. 2, p 89. (b) Praly, J.-P. Advances in
Carbohydrate Chemistry and Biochemistry; Wiley: France, 2001; Vol. 56,
p 65. (c) Yamago, S.; Yoshida, J.-i. In CRC Handbook of Organic
Photochemistry and Photobiology, 2nd ed.; Horspool, W., Lenci, F., Eds.;
CRC Press: New York, 2004; p 2/1. (d) Studer, A.; Curran, D. P.
Angew. Chem., Int. Ed. 2016, 55, 58. (e) Giese, B.; Zeitz, H.-G. In
Preparative Carbohydrate Chemistry; Hanessian, S., Ed.; Marcel Dekker,
Inc.: 1997; p 507. (f) Togo, H.; He, W.; Waki, Y.; Yokoyama, M.
Synlett 1998, 1998, 700.
(6) Radical substitution to only heteroaromatic base that introduces
furanose and hexose (Minisci reaction); see: (a) Duncton, M. A. J.
MedChemComm 2011, 2, 1135. (b) See ref 5f. (c) He, W.; Togo, H.;
Yokoyama, M. Tetrahedron Lett. 1997, 38, 5541.
(7) Selected reviews and papers on Pd-catalyzed synthesis of
unsaturated aryl C-glycosides from glycals: Reviews: (a) Merino, P.;
Experimental details, procedures, compound character-
ization data, NMR spectra, and X-ray crystallographic
data (PDF)
Crystallographic data (CIF, CIF)
(CIF)
AUTHOR INFORMATION
Corresponding Author
*masaharu@scl.kyoto-u.ac.jp
ORCID
Katsuhiro Isozaki: 0000-0002-0990-1708
Akihiro Orita: 0000-0001-8684-2951
Masaharu Nakamura: 0000-0002-1419-2117
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
́
Tejero, T.; Marca, E.; Gomollon-Bel, F.; Delso, I.; Matute, R.
■
Heterocycles 2012, 86, 791. (b) Wellington, K. W.; Benner, S. A.
Nucleosides, Nucleotides Nucleic Acids 2006, 25, 1309. Selected papers
on Heck coupling: (c) Bai, Y.; Kim, L. M. H.; Liao, H.; Liu, X.-W. J.
Org. Chem. 2013, 78, 8821. (d) Ramnauth, J.; Poulin, O.; Rakhit, S.;
Maddaford, S. P. Org. Lett. 2001, 3, 2013. On Suzuki coupling:
This work was funded in part by a grant from the JSPS through
the “Funding Program for Next Generation World-Leading
Researchers (NEXT Program),” initiated by the Council for
Science and Technology Policy, and also by the JST through
Core Research for Evolutional Science and Technology
(CREST 1102545). JSPS KAKENHI Grant Number
26620085 also supported this work. L.A. is grateful for a
research fellowship from JSPS and the MEXT project
“Integrated Research on Chemical Synthesis”. S.K. and H.C.L.
thank the ICR-KU International Short-term Exchange Program.
A.O. thanks JSPS KAKENHI Grant Numbers JP16H01164 in
Middle Molecular Strategy and JP15K05440 and Grant for
Promotion of OUS Research Projects. This work was supported
in part by the Collavorative Research Program of Institute for
Chemical Research, Kyoto University (Grant Nos. 2011014,
2012-17, 2013-16, 2014-18) and by JSPS Core-to-Core
Program “Elements Function for Transformative Catalysis and
Materials”. This article is dedicated to Prof. Teruaki Mukaiyama
in celebration of his 90th birthday (Sotsuju).
́
(e) Cobo, I.; Matheu, M. I.; Castillon, S.; Boutureira, O.; Davis, B. G.
Org. Lett. 2012, 14, 1728. (f) Lehmann, U.; Awasthi, S.; Minehan, T.
Org. Lett. 2003, 5, 2405. On Hiyama coupling: (g) Denmark, S. E.;
Regens, C. S.; Kobayashi, T. J. Am. Chem. Soc. 2007, 129, 2774.
(8) Only one example of the palladium-catalyzed synthesis of
saturated aryl C-glycosides was reported recently; it was achieved by
the cross-coupling reaction of anomeric stannanes with aromatic
halides in the presence of a catalytic palladium salt and phosphine
ligand with a large excess of copper(I) (CuCl, 300 mol %) and a
fluoride source (KF, 200 mol %) at a high reaction temperature. See:
Zhu, F.; Rourke, M. J.; Yang, T.; Rodriguez, J.; Walczak, M. A. J. Am.
Chem. Soc. 2016, 138, 12049.
(9) (a) Gong, H.; Gagne,
Gagne also reported the transition-metal-catalyzed C−C bond forming
reactions of halosugars: (b) Andrews, R. S.; Becker, J. J.; Gagne, M. R.
Angew. Chem., Int. Ed. 2010, 49, 7274. (c) Gong, H.; Andrews, R. S.;
Zuccarello, J. L.; Lee, S. J.; Gagne, M. R. Org. Lett. 2009, 11, 879.
(d) Gong, H.; Sinisi, R.; Gagne, M. R. J. Am. Chem. Soc. 2007, 129,
1908.
́
M. R. J. Am. Chem. Soc. 2008, 130, 12177.
́
́
́
REFERENCES
́
■
(1) (a) Hultin, P. G. Curr. Top. Med. Chem. 2005, 5, 1299. (b) Du, Y.;
Linhardt, R. J. Tetrahedron 1998, 54, 9913. (c) Postema, M. H. D.
Tetrahedron 1992, 48, 8545. (d) Levy, D. E. In The Organic Chemistry
(10) (a) Nicolas, L.; Angibaud, P.; Stansfield, I.; Bonnet, P.;
Meerpoel, L.; Reymond, S.; Cossy, J. Angew. Chem., Int. Ed. 2012,
51, 11101. Recently, Cossy reported the synthesis of C-ribofurano-
sides: (b) Nicolas, L.; Izquierdo, E.; Angibaud, P.; Stansfield, I.;
Meerpoel, L.; Reymond, S.; Cossy, J. J. Org. Chem. 2013, 78, 11807.
of Sugars; Levy, D. E.; Fugedi, P., Eds.; CRC Press LLC: Boca Raton,
̈
FL, 2006; p 269. (e) Levy, D. E.; Tang, C. The Chemistry of C-
Glycosides; Pergamon: Tarrytown, NY, 1995.
(11) (a) Furstner, A. ACS Cent. Sci. 2016, 2, 778. (b) Guer
́
inot, A.;
Cossy, J. Top. Curr. Chem. (Z) 2016, 374, 49. (c) Lv, L.; Li, Z. Top.
Curr. Chem. (J) 2016, 374, 38. (d) Legros, J.; Figadere, B. Nat. Prod.
̈
(2) Selected reviews and articles: (a) McGill, J. B. Diabetes Ther.
2014, 5, 43. (b) Fujita, Y.; Inagaki, N. J. Diabetes Invest. 2014, 5, 265.
(c) Kurosaki, E.; Ogasawara, H. Pharmacol. Ther. 2013, 139, 51.
(d) Shah, N. K.; Deeb, W. E.; Choksi, R.; Epstein, B. J.
Pharmacotherapy 2012, 32, 80. (e) Nomura, S.; Sakamaki, S.; Hongu,
M.; Kawanishi, E.; Koga, Y.; Yamamoto, Y.; Ueta, K.; Nakayama, K.;
̀
Rep. 2015, 32, 1541. (e) Jia, F.; Li, Z. Org. Chem. Front. 2014, 1, 194.
(f) Nagashima, H. Bull. Chem. Soc. Jpn. 2017, Advanced Publication.
DOI: 9076110.1246/bcsj.20170071.
́
(12) C-Mannosylation with an alkenyl Grignard reagent was reported,
whereas this protocol was not applied to aryl Grignard reagents or any
aryl nucleophiles; see: Bensoussan, C.; Rival, N.; Hanquet, G.;
Colobert, F.; Reymond, S.; Cossy, J. Tetrahedron 2013, 69, 7759.
(13) Selected reviews: (a) Nakamura, E.; Hatakeyama, T.; Ito, S.;
Ishizuka, K.; Ilies, L.; Nakamura, M. Org. React. 2013, 83, 1.
(b) Nakamura, E.; Yoshikai, N. J. Org. Chem. 2010, 75, 6061.
Tsuda-Tsukimoto, M. J. Med. Chem. 2010, 53, 6355. (f) Bokor, E.;
Kun, S.; Goyard, D.; Tot
Rev. 2017, 117, 1687.
́ ́
h, M.; Praly, J.-P.; Vidal, S; Somsak, L. Chem.
(3) Selected book and reviews on C-glycosides: (a) C-Glycoside
Synthesis; Postema, M. H. D., Ed.; CRC Press Inc: FL, 1995. (b) Lee,
D. Y. W.; He, M. Curr. Top. Med. Chem. 2005, 5, 1333. (c) Bililign, T.;
Griffith, B. R.; Thorson, J. S. Nat. Prod. Rep. 2005, 22, 742.
(d) Lemaire, S.; Houpis, I. N.; Xiao, T.; Li, J.; Digard, E.; Gozlan,
(c) Czaplik, W. M.; Mayer, M.; Cvengros,
ChemSusChem 2009, 2, 396. (d) Sherry, B. D.; Furstner, A. Acc. Chem.
Res. 2008, 41, 1500. (e) Bolm, C.; Legros, J.; Le Paih, J.; Zani, L. Chem.
̌
J.; von Wangelin, A. J.
C.; Liu, R.; Gavryushin, A.; Dien
Org. Lett. 2012, 14, 1480.
(4) Selected review and papers on C-nucleosides: (a) Stambasky, J.;
̀
e, C.; Wang, Y.; Farina, V.; Knochel, P.
̈
̌
Rev. 2004, 104, 6217.
́
(14) Selected papers on the iron-catalyzed cross-coupling reaction of
alkyl halides with Grignard reagents: (a) Jin, M.; Adak, L.; Nakamura,
M. J. Am. Chem. Soc. 2015, 137, 7128. (b) Ghorai, S. K.; Jin, M.;
Hocek, M.; Koco
Singh, I.; Hemmings, J.; Seitz, O. J. Org. Chem. 2007, 72, 8811.
(c) Novak, P.; Pohl, R.; Kotora, M.; Hocek, M. Org. Lett. 2006, 8, 2051.
̌
́
vsky, P. Chem. Rev. 2009, 109, 6729. (b) Hainke, S.;
́
H
DOI: 10.1021/jacs.7b03867
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Hatakeyama, T.; Nakamura, M. Org. Lett. 2012, 14, 1066.
(c) Hatakeyama, T.; Fujiwara, Y.; Okada, Y.; Itoh, T.; Hashimoto,
T.; Kawamura, S.; Ogata, K.; Takaya, H.; Nakamura, M. Chem. Lett.
2011, 40, 1030. (d) Noda, D.; Sunada, Y.; Hatakeyama, T.; Nakamura,
excellent yields; see: Xiang, J.; Toyoshima, S.; Orita, A.; Otera, J.
Angew. Chem., Int. Ed. 2001, 40, 3670.
(26) Molecular weight loss of the corresponding acetyl-, benzyl- and
benzoyl-protected 1-phenyl mannose were theoretically calculated as
41% (MW change: 408.40 to 240.26), 60% (MW change: 600.76 to
240.26), and 63% (MW change: 656.69 to 240.26), respectively.
(27) A similar observation was noted in the nickel-catalyzed cross-
coupling of δ-olefinic 1-bromo alkane with PhMgCl; see: Breitenfeld,
J.; Wodrich, M. D.; Hu, X. Organometallics 2014, 33, 5708.
(28) For formation of an alkyl radical intermediate from alkyl halides
in iron-catalyzed cross-coupling reactions, see ref 14c, i, and 14m.
(29) Recently, we reported a bimetallic (out-of-cage) mechanism for
iron-catalyzed enantioselective cross-coupling reactions of α-chlor-
oesters with aryl Grignard reagents (see ref 14a and references cited
therein). This does not exclude the possibility of competition of an in-
cage mechanism (see ref 14d and 14m).
(30) Bimetallic mechanisms are reported for nickel catalysts; see:
(a) Schley, N. D.; Fu, G. C. J. Am. Chem. Soc. 2014, 136, 16588.
(b) Choi, J.; Martín-Gago, P.; Fu, G. C. J. Am. Chem. Soc. 2014, 136,
12161. (c) Biswas, S.; Weix, D. J. J. Am. Chem. Soc. 2013, 135, 16192.
For a discussion of related nickel-catalyzed bimetallic mechanisms, see
ref 27.
M.; Nagashima, H. J. Am. Chem. Soc. 2009, 131, 6078. (e) Furstner, A.;
̈
Martin, R.; Krause, H.; Seidel, G.; Goddard, R.; Lehmann, C. W. J. Am.
Chem. Soc. 2008, 130, 8773. (f) Nakamura, M.; Matsuo, K.; Ito, S.;
Nakamura, E. J. Am. Chem. Soc. 2004, 126, 3686. For reactions with
aryl zinc reagents, see: (g) Bedford, R. B.; Carter, E.; Cogswell, P. M.;
Gower, N. J.; Haddow, M. F.; Harvey, J. N.; Murphy, D. M.; Neeve, E.
C.; Nunn, J. Angew. Chem., Int. Ed. 2013, 52, 1285. (h) Lin, X.; Zheng,
F.; Qing, F.-L. Organometallics 2012, 31, 1578. (i) Hatakeyama, T.;
Kondo, Y.; Fujiwara, Y.; Takaya, H.; Ito, S.; Nakamura, E.; Nakamura,
M. Chem. Commun. 2009, 45, 1216. (j) Nakamura, M.; Ito, S.; Matsuo,
K.; Nakamura, E. Synlett 2005, 1794. For reactions with aryl aluminum
reagents, see: (k) Kawamura, S.; Kawabata, T.; Ishizuka, K.; Nakamura,
M. Chem. Commun. 2012, 48, 9376. (l) Bedford, R. B.; Brenner, P. B.;
Carter, E.; Clifton, J.; Cogswell, P. M.; Gower, N. J.; Haddow, M. F.;
Harvey, J. N.; Kehl, J. A.; Murphy, D. M.; Neeve, E. C.; Neidig, M. L.;
Nunn, J.; Snyder, B. E. R.; Taylor, J. Organometallics 2014, 33, 5767.
For reactions with organoboron reagents, see: (m) Hatakeyama, T.;
Hashimoto, T.; Kondo, Y.; Fujiwara, Y.; Seike, H.; Takaya, H.;
Tamada, Y.; Ono, T.; Nakamura, M. J. Am. Chem. Soc. 2010, 132,
10674. (n) Bedford, R. B.; Brenner, P. B.; Carter, E.; Carvell, T. W.;
Cogswell, P. M.; Gallagher, T.; Harvey, J. N.; Murphy, D. M.; Neeve, E.
C.; Nunn, J.; Pye, D. R. Chem. - Eur. J. 2014, 20, 7935.
(31) For the mechanism for the formation of Fe(I) species in cross-
coupling reactions, see: (a) Adams, C. J.; Bedford, R. B.; Carter, E.;
́
Gower, N. J.; Haddow, M. F.; Harvey, J. N.; Huwe, M.; Cartes, M. A.;
Mansell, S. M.; Mendoza, C.; Murphy, D. M.; Neeve, E. C.; Nunn, J. J.
Am. Chem. Soc. 2012, 134, 10333. (b) Bedford, R. B. Acc. Chem. Res.
2015, 48, 1485. (c) See also ref 14l.
(15) Isobe, M.; Kondo, S.; Nagasawa, N.; Goto, T. Chem. Lett. 1977,
6, 679.
(16) Decomposition of starting chlororibofuranose 1 competed with
the coupling reaction under the reaction conditions regardless of the
difference on the ligands and reagents used, resulting in the poor mass
balance.
(32) Several pathways have been proposed for the carbon−halogen
(C−X) bond cleavage in iron-catalyzed cross-coupling reactions: the
iron(II) species having two aryl groups, or one aryl and one halide
group, with a SciOPP ligand can react with alkyl halides to generate the
radical intermediate and to give the corresponding cross-coupling
product. The reaction rates of the coupling reactions depend on the
nature of aryl groups and alkyl halides. See: (a) Daifuku, S. L.;
Kneebone, J. L.; Snyder, B. E. R.; Neidig, M. L. J. Am. Chem. Soc. 2015,
137, 11432. (b) Takaya, H.; Nakajima, S.; Nakagawa, N.; Isozaki, K.;
Iwamoto, T.; Imayoshi, R.; Gower, N.; Adak, L.; Hatakeyama, T.;
Honma, T.; Takagi, M.; Sunada, Y.; Nagashima, H.; Hashizume, D.;
Takahashi, O.; Nakamura, M. Bull. Chem. Soc. Jpn. 2015, 88, 410. This
process may proceed in the first catalytic cycle of the reaction before
the formation of the highly reactive iron(I) species.31
(33) We do not exclude the possibility of the mechanism via the
formation of iron (III) intermediate and following reductive
elimination of the carbon−carbon bond formation. We consider that
the reductive elimination from iron(III)-glycosyl-aryl complex B
should be rapid due to instability of the iron(III) intermediate and
can be stereospecific to reflect the stereochemistry of the stable radical
conformer. Our ongoing extensive DFT study on the mechanism of
iron-catalyzed enantioselective coupling suggests potential competition
of both mechanisms via B or TS C in Figure 1. The results will be
reported in due course. Insightful comments on this point by a
reviewer are gratefully acknowledged.
(17) Details are provided in the Supporting Information.
(18) Qing reported that the combination of TMEDA and DPPP as a
ligand improved the Fe-catalyzed Negishi-type cross-coupling reaction
of alkyl halides bearing β-fluorine. See ref 14h.
1
(19) The stereochemistry of the anomers was determined by H
NMR analysis and confirmed by single crystal X-ray analysis. X-ray
structures are shown in the Supporting Information.
(20) (a) Mayer, M.; Welther, A.; von Wangelin, A. J. ChemCatChem
2011, 3, 1567. (b) Jennerjahn, R.; Jackstell, R.; Piras, I.; Franke, R.;
Jiao, H.; Bauer, M.; Beller, M. ChemSusChem 2012, 5, 734. (c) Chen,
C.; Dugan, T. R.; Brennessel, W. W.; Weix, D. J.; Holland, P. L. J. Am.
Chem. Soc. 2014, 136, 945. (d) Bai, X.-F.; Song, T.; Deng, W.-H.; Wei,
Y.-L.; Li, L.; Xia, C.-G.; Xu, L.-W. Synlett 2014, 25, 417.
(21) Berger, S.; Langer, F.; Lutz, C.; Knochel, P.; Mobley, T. A.;
Reddy, C. K. Angew. Chem., Int. Ed. Engl. 1997, 36, 1496.
(22) The reaction of glucopyranosyl chloride 3′ under the conditions
of 0 °C, 16 h gave 3a in only 32% yield (α/β = 70/30). In addition,
The FeCl3−DPPP catalyst system provided it in 67% yield by the
reaction of 3′ with the phenylzinc reagent; see the Supporting
Information (Table S3). The preparation of DPPP derivatives is an
ongoing process in our laboratory, in order to improve the efficiency of
the cross-coupling reaction of glycosyl chlorides. A patent application
has been filed: Nakamura, M.; Hatakeyama, T.; Hashimoto, T.;
Nakajima, S.; Nakagawa, N. Jpn. Kokai Tokkyo Koho JP 2013180991,
2013.
(23) (a) Nolen, E. G.; Ezeh, V. C.; Feeney, M. J. Carbohydr. Res.
2014, 396, 43. (b) Tatina, M.; Kusunuru, A. K.; Yousuf, S. K.;
Mukherjee, D. Chem. Commun. 2013, 49, 11409.
(24) (a) Yamago, S.; Miyazoe, H.; Yoshida, J.-i. Tetrahedron Lett.
1999, 40, 2343. (b) Togo, H.; He, W.; Waki, Y.; Yokoyama, M. Synlett
(34) (a) Giese, B. Angew. Chem., Int. Ed. Engl. 1989, 28, 969.
(b) Giese, B.; Dupuis, J.; Leising, M.; Nix, M.; Lindner, H. J. Carbohydr.
Res. 1987, 171, 329. (c) Dupuis, J.; Giese, B.; Ruegge, D.; Fischer, H.;
̈
Korth, H.-G.; Sustmann, R. Angew. Chem., Int. Ed. Engl. 1984, 23, 896.
(d) See also ref 24b.
(35) Shuto and Matsuda proposed the transition states of the reaction
of glucosyl radical, where the selectivity of products was shown to be
mainly controlled by kinetic anomeric effect. Therefore this effect may
also be considered; see: Abe, H.; Shuto, S.; Matsuda, A. J. Am. Chem.
Soc. 2001, 123, 11870.
(36) For the steric effect on product selectivity, see: (a) Araki, Y.;
Endo, T.; Tanji, M.; Nagasawa, J.; Ishido, Y. Tetrahedron Lett. 1988, 29,
351. (b) Barton, D. H. R. Pure Appl. Chem. 1988, 60, 1549. (c) Giese,
́
1998, 1998, 700. (c) Mazeas, D.; Skrydstrup, T.; Doumeix, O.; Beau, J.-
M. Angew. Chem., Int. Ed. Engl. 1994, 33, 1383. (d) Stork, G.; Suh, H.
S.; Kim, G. J. Am. Chem. Soc. 1991, 113, 7054.
(25) Conventional methods for deprotection of acetyl groups may
provide lower yields due to a tedious workup procedure, but
transesterification reactions using dibutyltin oxide in methanol provide
B.; Gonzal
23, 69.
́
ez-Gom
́
ez, J. A.; Witzel, T. Angew. Chem., Int. Ed. Engl. 1984,
I
DOI: 10.1021/jacs.7b03867
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX