Journal of the American Chemical Society
Communication
C−H bond dissociation energies render the initial HAT step
facile, the stability of the resultant radical led us to initially
question the favorability of its reaction with sulfur dioxide.
Gratifyingly, toluene derivatives were highly effective in this
transformation, producing the expected benzylic sulfinic acids
in good yields across a broad range of aryl functionality (28−
33, 67−82% yield), including ortho substitution (33, 81%
yield). Despite greater stabilization of the resultant radical,
secondary benzylic substrates also performed well in this
transformation (34 and 35, 74% and 46% yield, respectively).
Notably, many of these benzylic substrates contain protic
functional groups, such as sulfonamides (30), amides (32 and
33), and boronic acids (35), which prove problematic for
traditional approaches requiring a strong base or organo-
metallic nucleophiles.34 Further investigation revealed that
heterobenzylic substrates were also effective in this reaction,
with selectivity observed for functionalization at the (hetero)-
benzylic C−H bonds (36 and 37, 58% and 56% yield,
respectively).
Scheme 2. Computational Study of Radical Addition to SO2
To demonstrate the utility of this methodology for late-stage
functionalization, natural products and pharmaceuticals were
converted to the corresponding sulfinates in a single step
(Table 2). Notably, natural amino acids leucine and GABA
afforded the corresponding benzyl sulfones in synthetically
useful yields and excellent selectivity (38 and 39, 58% and 24%
yield, respectively, 91% selective and single regioisomer,
respectively). The monoterpenoid fenchone was functionalized
with good selectivity for the most electron-rich, sterically
accessible C−H bond (40, 56% yield, 63% selectivity), and
pregabalin was converted to the corresponding benzyl sulfone
with excellent regioselectivity for the tertiary position (41, 54%
yield, 85% selectivity). Finally, two drugs bearing benzylic C−
H bonds, celecoxib and prilocaine, were derivatized with
complete selectivity observed for functionalization at the
benzylic position (42 and 43, 70% and 73% yield,
respectively).
As an illustration of the broad utility of this platform for the
synthesis of diverse organosulfur compounds, a range of one-
pot procedures for the divergent functionalization of tricyclic
imide 44 were developed. As shown in Table 2, the sulfinic
acid intermediate was successfully converted to a range of alkyl
sulfone derivatives (45−47, 56−93% yield). Introduction of
heteroatoms also proved facile, with the corresponding sulfonic
acid (48, 82% yield), primary sulfonamide (49, 65% yield),
sulfonyl fluoride (50, 66% yield), and sulfonyl chloride (51,
57% yield) all generated with good efficiency. Additionally,
two-pot protocols were developed for the conversion of
celecoxib to a diverse range of sulfonamides via the
intermediacy of a sulfonyl chloride, generated via chlorination
of the C−H sulfinylation product without intermediate
purification. Alkyl amines (52 and 53, 62% and 53% yield,
respectively), anilines (54, 56% yield), and N-heterocycles (55,
31% yield) all reacted to afford the desired sulfonamide
products.
continuous increase in energy, possibly indicating a barrierless
process (Figure S7). In order to further investigate the nature
of the C−S bond-forming step, we calculated the transition
state energies for the addition of a series of stabilized radicals
(e.g., benzylic) into SO2 in the gas phase, conditions under
which sulfonyl radical formation is predicted to be markedly
less favorable than in the presence of polar solvent.
Remarkably, however, low barriers to radical capture with
SO2 were determined (7.5 and 6.8 kcal/mol for primary and
secondary benzylic, respectively), consistent with the observed
efficiencies in experiments involving aliphatic radicals and
sulfur dioxide.
In summary, we have developed a perfectly atom-
economical protocol for the photocatalytic conversion of
C(sp3)−H bonds into the corresponding alkyl sulfinic acids,
thereby enabling unprecedented access to a broad array of
valuable organosulfur products. Furthermore, these studies
clearly illustrate the importance of sulfur dioxide as an efficient
reagent for the formation of C−S bonds from a diverse range
of aliphatic radicals and, as such, should inform the
development of related transformations that proceed via this
key elementary step.
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
Experimental and characterization data, UV/vis and
computational studies, and spectral data (PDF)
As a preliminary investigation into the mechanism of this
transformation, we computationally studied the coupling of a
range of alkyl radicals with sulfur dioxide (Scheme 2). Notably,
with aliphatic radicals, a significant negative free energy of
reaction was observed for this trapping in water ((U)-
ωB97XD/6-31+G(d,p), SMD solvent model). Moreover,
efforts to identify a transition state in the case of unstabilized
aliphatic radicals proved unsuccessful, with stretching of the
C−S bond of the sulfonyl radical product resulting in a
AUTHOR INFORMATION
■
Corresponding Author
David W. C. MacMillan − Merck Center for Catalysis at
Princeton University, Princeton, New Jersey 08544, United
9741
J. Am. Chem. Soc. 2021, 143, 9737−9743