Organic Letters
Letter
Scheme 1. Select Optimization Results and Control
a
Studies
Figure 1. Selection of acetal-containing drugs.
can be excited to form an aryl radical and an iodine radical by
an energy transfer (EnT) process with a newly formed gold
complex. On the basis of this principle, we assumed that THF
acetals might be accessible by the reaction of simple
iodobenzene with THF. The formed aryl radical should
quickly be trapped by THF to form an α-heteroatom-stabilized
alkyl radical. The latter could then recombine with the iodine
radical to form 2-iodotetrahydrofuran. A nucleophilic attack
then might give rise to the desired products, a strategy different
from that in Toste’s report.
a
Yield determined by 1H NMR analysis using 1,2,4,5-tetramethyl-
b
c
benzene as the internal standard. n.d. = not detected. Isolated yield.
3e, and 3g) gave moderate yields. No racemization was
observed for substrates with chiral centers (3c and 3h).
Differently substituted phenylethanol derivatives bearing
halogens at different positions (3i and 3j) and a
trifluoromethyl (3l) group were all tolerated. A wide variety
of benzylic alcohols (3m−3s) reacted well, showing no large
difference in yield. The less nucleophilic phenol also reacted
with THF to the desired product in 35% isolated yield, but this
compound decomposed after storage at 4 °C for 2 days. Even
medically relevant heteroaromatic alcohols, which are notori-
ously problematic in many photoredox protocols, performed
well under our reaction conditions, including thiophene, furan,
and pyridine (3u−3w, respectively; 45−65% yields).
We next evaluated other commonly used cyclic ethers
instead of THF for this acetalization. Tetrahydropyran also
reacted adjacent to the oxygen atom in 48% yield (3x). The
diminished reactivity compared to THF might result from
stereoelectronic factors. On the other hand, we found different
aryl iodides influence the yield, maybe because of polarity
match. On the basis of condition optimization, 2-chloroiodo-
benzene was better than others. In the case of 1,4-dioxane
(3y), despite a comparable steric demand, the second oxygen
atom was not functionalized. For 2-methyltetrahydrofuran,
only the less hindered position reacted to give 3z.
In our initial experiments, we found 4-iodoanisole to be a
suitable precursor for this transformation in the presence of the
gold catalyst, but the yield was only 38% (Scheme 1, entries 1,
2, and 5) after irradiation for 24 h. A catalyst screening using
different counteranions (entries 3−4) enabled an increase in
the yield to 60%, with OTf− being the best choice. A slight
increase in the amount of 4-iodoanisole (entries 6−8) and an
extended irradiation time (entry 12) further improved the
reaction. Attempts to lower the catalyst load failed (entries 10
and 11). We found the condition shown in entry 12 to be the
optimum conditions {1.0 equiv of NaHCO3, 1.5 equiv of 4-
iodoanisole, and 1.0 mol % [Au2(μ-dppm)2](OTf)2 irradiated
by blue LED lights for 36 h}. On the basis of entries 7 and 9,
we found reducing the amount of THF does not influence the
yield, but using less ether could reduce cost and is beneficial
for large scale reactions. We used 0.2 mL of THF for further
experiments. The use of simple PPh3AuCl or other widely
applied photocatalysts, like [Ir(ppy)2(dtbbpy)]PF6, [Ru-
(bpy)3](PF6)2, Eosin Y, or Mes-Acr-Me+BF4 , instead of the
−
[Au2(μ-dppm)2]Cl2 complex, resulted in none of the desired
product (entries 13−17, respectively). Decreased yields were
obtained upon irradiation with CFL or UVA light (entries 20
and 21). The control experiments shown in entries 1, 5, 18,
and 19 prove that all of the parameters are significant for a
successful transformation.
With the optimal conditions in hand, we then explored the
scope of this reaction. Different kinds of primary and
secondary alcohols underwent the desired C(sp3)−H acetal-
ization (Scheme 2). Both linear and cyclic allylic alcohols (3d,
Next we expanded the application of this C−H functional-
ization strategy toward the synthesis of thioacetals and α-
alkoxypyrrolidines (Scheme 3). Like the sulfur analogues of
acetals, thioacetals have potential applications in pharmaceut-
ical development and alcohol protection.12 A variety of
alcohols, including simple alkyl alcohol (4a), cinnamyl alcohol
(4b), benzyl alcohols (4d and 4e), and heterocyclic benzyl
alcohol (4f), all directly reacted with thiophene in our new
photoredox C−H activation (functionalization) route in
moderate yields (33−62%). The oxidation of pyrrolidine to
B
Org. Lett. XXXX, XXX, XXX−XXX