Organic Letters
Letter
a
b
Table 1. Optimization of C−C Cross-Coupling
Scheme 3. Scope of C-Arylation
a
b
87% isolated yield with 0.15 mmol of 13a. For all compounds
tested, only one anomer (α: β > 99: 1) was observed by 1H NMR in
crude reaction mixtures. Reagents and conditions taken from Table 1,
entry 6. Aryl iodides used for all entries unless stated otherwise.
substituted phenol gave the corresponding glycoside 22e in
an acceptable 69% yield similar to electron-poor (22f, 22h)
and electron-rich (22g) meta-substituted functional groups,
which provided clean conversion into products in 73−91%.
These results led us to the investigate more difficult pyridine-
based couplings, such as 2-iodopyridine (22i) and 2-fluoro-4-
iodopyridine (22j), which gave the expected products
establishing a simple one-step protocol for the synthesis of
pyridine glycals. Following these results, we focused on more
challenging substrates such as the ones with a formyl group
(22k, 82%) and 4-iodophenyalanine (22l, 73%). For all
compounds shown in Scheme 3, exclusive α selectivity was
a
General conditions: 13a (0.15 mmol), (Ph-I) (0.10 mmol), PC3
(0.025 equiv), NiCl2·DME (0.05 equiv), dtbpy (0.05 equiv), CsF (1.5
equiv), 1,4-dioxane (2.0 mL), blue LED (40W), 23 °C, 24 h, N2.
1
showed reduced yields with an increase in elimination.
Although additional studies on the role of the base in
photoredox reactions is needed, we also established that the
counterion combination is essential, and fluoride ion itself is
not sufficient to suppress elimination, as demonstrated by the
results with KF and LiF (entries 11 and 12, respectively).
Finally, we explored solvents other than 1,4-dioxane−acetone
(entry 13) and tetrahydrofuran (entry 14) produced
comparable although slightly diminished yields. For all
conditions shown in Table 1, a single anomer (α) was
observed.
With suitable conditions in hand (Table 1, entry 6), our
focus was to determine the extent of functional group tolerance
(Scheme 3). Electron-rich 4-iodoanisole afforded 22a in 88%
with no loss of stereoselectivity. Electron-deficient cyano
(22b) and ester (22c) groups gave almost quantitative yields
(96% and 94%, respectively). These results suggest that
electron withdrawing groups furnish better yields likely
because of increasing the rate of reductive elimination and
the ability to stabilize generated aryl radical.23 Unprotected 4-
iodobenzyl alcohol produced 22d cleanly in 72% yield
signifying that fully deprotected saccharides could be tolerated
with these reaction conditions (vide infra). The meta-
observed (>99:1 α:β, H NMR).
After uncovering the broad functional-group tolerance of an
assortment of iodoarenes, we proceeded to apply our
conditions to incorporate other sugars (Scheme 4). D-Glucose
with a free hydroxyl group (24a) or protected as an acetate
(24b) is a viable substrate for photoredox C(sp3)−C(sp2)
cross-coupling. Similarly, 2-deoxy-D-galactose (24c) and
dideoxysugars derived from D-arabinose (24e) and L-fucose
(24f) furnished high selectivities in modest to high yields. To
our delight, disaccharides are also competent reagents in
photocatalytic C-arylation and afforded 24d in 63%. Both, D-
galactose (24g) and D-glucose (24h, 24i) regardless of the
nature of the protective group located at C2 resulted in a
mixture of α:β products, with the smaller hydroxyl groups
resulting in slightly improved axial preference (2.3−3.7:1).
Because the proposed activation method proceeds through a
radical species, we also wondered if C1 trifluoroborates can be
enaged in C−S24 and C−Se25 bond-forming reactions
(Scheme 5). To our delight, the formation of both ethers 26
and 28 proceeded smoothly from 13a and eletrophilic sulfur
(25) or selenium (27) sources without a nickel cocatalyst.
In summary, we introduced here a new class of C1-
trifluoroborates that are competent reagents for the formation
4291
Org. Lett. 2021, 23, 4289−4293