Organic Letters
Letter
Lowering the catalyst loading did not improve the yield of
the reaction as only 42% conversion of 1a was observed (entry
6). Monitoring the reaction with NMR in deuterated solvent
revealed that the reaction was fully converted in 10 min
providing 2a in 80% yield with trace amounts of 3a and 4a
(entry 7). The reaction run in only MeCN-d3 led to poor
conversion and yields (entry 8). Lowering the concentration to
0.033 M did not affect the yield of the reaction (entry 9).
Using the same conditions but with nondeuterated solvents
showed a similar yield of 2a and a slight increase of 3a (16%;
likely due to hydrogen atom abstraction from the solvent), but
no 4a was observed (entry 10). Finally, control experiments
showed that the photocatalyst and light were needed to
perform the reaction (entries 11−13).
the compound patterns listed above are compatible with our
methodology. Having monosubstituted olefins and disubsti-
tuted terminal olefins was also compatible with our method-
ology (Scheme 3A). The cyclization of malonate-based
substrates 1a and 1b gave the desired compounds 2a and 2b
in 82% and 55% yields, respectively. In both cases, small
amounts of the reduced side products 3a (16%) and 3b (29%)
were observed. Moving onto the NTs-based substrates, the use
of monosubstituted alkenes 1c provided the desired product 2c
in 82% yield with trace amounts of its reduced counterpart
(>5%), while disubstituted olefins 1d gave bromocycloalkane
2d in 50% yield along with 3c (21%). These results suggest
that paths A or B are affected by steric hindrance leading to the
formation of competitive processes. Switching to allyl ethers
1e−i, the corresponding 5- and 6-membered rings 2e−i were
obtained in moderate to good yields. Volatile bromoalkanes
were also employed, and their results were recorded using
NMR spectroscopy (2j−2l). Using starting material 1j
provided a mixture of 1j and the desired product 2j (82%,
1:4 1j/2j). Furthermore, the opening of 1k and 1l was also
compatible with the methodology (Scheme 3B). Linear alkyl
bromides 2k and 2l were formed in good yields (72% and 55%
respectively). These results provide support for the formation
of free alkyl radicals in solution. Unfortunately, meaningful
data was not obtained when varying the concentrations of gold
catalyst, bromoalkane, or solution to obtain absolute rates for
the transformation using radical clocking methods. Given the
formation of 2k through a radical ring opening, we envisaged
an intermolecular photoredox cascade reaction between 1k and
cyclohex-2-enone. Under standard conditions, the desired
bicyclic compound 4 was obtained in 41% as a mixture of
diastereomer (1:1).
With optimized reaction conditions in hand, we explored the
reaction scope using allyl-functionalized NTs− and malonate-
tethered bromoalkanes as well as pyran- and furan-based
compounds (Scheme 3). The investigation showed that all of
Scheme 3. Intra- and Intermolecular Formal Bromine Atom
Transfer Radical Addition Reaction Scope
Next, we tested iodoarenes to see if similar reactivity to the
bromoalkanes could be achieved (Scheme 3C). It should be
noted that iodoalkanes are known to undergo chain reaction
mechanisms and as such, were excluded from this study. In
past methodologies, the use of iodoarenes resulted in products
from β-hydride elimination when not using a blocking group;
therefore, we were curious if our methodology was mild
enough to isolate the corresponding iodoalkanes. Starting with
O-tethered iodoarenes, the desired product was obtained in
good yields. Having a monosubstituted olefin 1m and
disubstituted olefin 1n gave the desired products 2m and 2n
in 62% and 45% yields, respectively. Nitrogen-based substrates
using Ac- and Ts- protecting groups were also compatible with
the methodology, giving the desired product in good yields.
Compared to other methods pertaining to this transformation,
the described protocol showed no need of a blocking group to
obtain the primary alkyl iodide.
The scope of the formal bromine transfer radical reaction
was extended to substituted alkenes (Scheme 4). Interestingly,
under the usual reaction conditions, the cyclization of alkene
1r led to the formation of the desired bicyclic bromide 2r in
29% (dr = 1:1) along with the elimination products 2r′ and
2r′′ in 54% yield (5:1). Similarly, the conversion of
trisubstituted 1s gave the desired product 2s in 41% yield
(dr = 2:1) along with the Heck-like compound 2s′’ in 58%
yield. However, the photoredox cyclization of nonterminal
olefin such as 1t led to the exclusive formation 2t and 2t′ in
75% yield as a mixture (2:1). This could be attributed to a fast
elimination of the tertiary bromine in the reaction medium.
In conclusion, we have developed a mild and operationally
simple intramolecular bromine atom transfer radical addition
a
Yields in parentheses correspond to the reduced product X = H.
b
c
Isolated yield at 1 mmol scale. NMR yield using mesitylene as
d
internal standard. 15 min reaction time.
C
Org. Lett. XXXX, XXX, XXX−XXX