Palladium-catalyzed sulfination of aryl and heteroaryl halides: Direct access to sulfones and sulfonamides

Article abstract of DOI:10.1021/ol403072r

A novel palladium-catalyzed sulfination of aryl and heteroaryl halides is described. This reaction operates under mild conditions and provides access to a wide range of aryl and heteroaryl sulfinates, a useful and versatile class of synthetic intermediates. Capitalizing on this sulfination reaction, one-pot protocols allowing direct access to sulfones and sulfonamides have also been developed. The practicality of these transformations is illustrated with the parallel synthesis of analogues of the drug Viagra.

Full text of DOI:10.1021/ol403072r

ORGANIC
LETTERS
2013
Vol. 15, No. 24
6226–6229
Palladium-Catalyzed Sulfination of Aryl
and Heteroaryl Halides: Direct Access to
Sulfones and Sulfonamides
Andrei Shavnya,* Steven B. Coffey, Aaron C. Smith, and Vincent Mascitti*
Pfizer Inc., Eastern Point Road, Groton, Connecticut 06340, United States
vincent.mascitti@pfizer.com; andrei.shavnya@pfizer.com
Received October 24, 2013
ABSTRACT
A novel palladium-catalyzed sulfination of aryl and heteroaryl halides is described. This reaction operates under mild conditions and provides
access to a wide range of aryl and heteroaryl sulfinates, a useful and versatile class of synthetic intermediates. Capitalizing on this sulfination
reaction, one-pot protocols allowing direct access to sulfones and sulfonamides have also been developed. The practicality of these
transformations is illustrated with the parallel synthesis of analogues of the drug Viagra.
Aryl and heteroaryl [(het)Ar] sulfinate salts are useful
and versatile synthetic intermediates that provide access to
structural motifs of relevance to numerous fields of
chemistry;1 sulfones, sulfonamides, and structures arising
from CÀC bond formation, such as ArÀ(het)Ar, ArÀCOR,
and ArÀCN, can all be accessed from a common sulfinate
intermediate (Figure 1). In particular, sulfinates are among
the intermediates of choice in drug discovery where analo-
gue generation based on versatile synthetic intermediates
helps promote rapid exploration of structureÀactivity re-
lationships (SAR).2 Indeed, such an approach allows for the
efficient synthesis of analogues containing sulfone and
sulfonamide functionalities which, as illustrated by the
medicines bicalutamide (Casodex) and sildenafil (Viagra),
are two pharmacophores often found in therapeutic agents
(Figure 1). However, the scope for using sulfinates as
intermediates in synthesis has been limited by their modest
commercial availability (∼200 compounds)3 and the practi-
Figure 1. Aryl sulfinates as versatile synthetic intermediates and
some medicines containing an arylsulfonyl motif.
(1) Schubart, R. Sulfinic Acids and Derivatives. Ullmann’s Encyclo-
pedia of Industrial Chemistry 2000, 677–693.
(2) In this context, we refer to “a versatile synthetic intermediate” as
one that can be diversely further functionalized and effectively reacted
with another broad set of reactants.
(3) Results from a search conducted using the ACD and eMolecules
databases (total of 7.9 million compounds exemplified). In comparison,
a search for boronic acids/esters returned ∼15 000 examples commer-
cially available.
(4) Aryl sulfinates have been traditionally accessed by the following:
(i) Reaction of organometallic reagents with SO2 gas or a surrogate; see:
Woolven, H.; Gonzalez-Rodriguez, C.; Marco, I.; Thompson, A. L.;
Willis, M. C. Org. Lett. 2011, 13, 4876. (ii) Reduction of sulfonyl halides;
see ref 1. (iii) Reaction of aryl halides with the SO22À surrogate SMOPS;
see: Baskin, J. M.; Wang, Z. Tetrahedron Lett. 2002, 43, 8479.
cality of existing preparative methods used to access them.1,4
As a result, there remains a need to develop novel synthetic
transformations that will allow direct access to aryl and
heteroaryl sulfinate salts from a large pool of readily avail-
able reactants.
(5) (a) Wojcicki, A. Adv. Organomet. Chem. 1974, 12, 31. (b) Pelzer,
G.; Herwig, J.; Keim, W.; Goddard, R. Russ. Chem. Bull. 1998, 47, 904.
r
10.1021/ol403072r
Published on Web 11/20/2013
2013 American Chemical Society

Table 1. Summary of Reaction Optimizationa
entry
X
ligand(s)
PPh3
additive
solvent
yieldb
1
2
3
4
5
I
À
DMF
38%
66%
69%
71%
61%
55%c
21%
56%
3%
I
I
I
I
PPh3; phen
PPh3; phen
PPh3; phen
PPh3; phen
À
DMF
TBAB
TBAB
TBAB
DMF
DMSO
MeCN
Figure 2. Approaches to metal-catalyzed sulfination reactions.
6
7
8
Br
Br
Br
PPh3; phen
PPh3; phen
PPh3; phen
TBAB
TBAB
TBAB
DMF
DMSO
MeCN
Interestingly, even though the behavior of SO2 as a
ligand and its insertion into metalÀcarbon bonds has been
long known,5 only one catalytic sulfination reaction has
been reported to date.6 In this approach reported by Keim,
the palladium-catalyzed reaction of phenyl diazonium
tetrafluoroborate salts with SO2 and hydrogen gas under
high pressure/high dilution conditions produces the corre-
sponding phenyl sulfinic acid (Figure 2).6 Unfortunately,
the use of phenyl diazonium salts as reactants and the
ready disproportionation of sulfinic acids severely limit the
scope of this transformation. Concurrent with the sub-
mission of our work, Willis also reported a method that
allows indirect access to aryl sulfinates by capitalizing on a
palladium-catalyzed N-aminosulfonylation of aryl iodides
previously developed in his laboratories.7,8 However, un-
der these conditions, sulfinates are not directly obtained
from the catalytic reaction but instead are generated in
situ by subsequent treatment of the N-aminosulfonamides
(products from the catalytic reaction) with an excess of an
alkylating agent under basic conditions. Thus, in contrast
to the plethora of metal-catalyzed methods available for
carbonylation there is still a notable lack of mild, catalytic,
and general methods for effecting sulfination reactions.
Herein we report the first palladium-catalyzed reaction
of aryl and heteroaryl halides and triflates in the presence
of potassium metabisulfite (K2S2O5) and sodium formate
(NaO2CH) to directly produce sulfinate synthetic inter-
mediates. As shown in Figure 2, these can be used without
isolation to form sulfones and sulfonamides.
a ArX (0.58 mmol), K2S2O5 (2 equiv), TBAB (1.1 equiv), NaO2CH
(2.2 equiv), solvent (2 mL), 70 °C then MeI (1.5 equiv), solvent (2 mL), 23
°C. b Isolated yield of sulfone 1. c Sulfinate isolated as a solid was used in
alkylation step.
formation of the corresponding sulfinate, which was con-
verted to sulfone 1 (38% yield) by in situ alkylation with
iodomethane (Table 1, entry 1). Our choice of Pd(OAc)2
was based on its excellent bench and air stability; thus,
based on the success of this initial experiment, subsequent
optimization was conducted retaining this palladium
source. This effort focused on identification of various
ligands and additives that could stabilize the palladium-
derived catalytic species and prevent their possible pre-
mature decomposition.11 From this work, we found that
addition of 15 mol % of the bidentate ligand 1,10-phenan-
throline (phen) led to the isolation of 1 in 66% yield (entry
2). Addition of tetrabutyl ammonium bromide (TBAB)
further increased the yield slightly (entry 3). DMSO and
MeCN also proved to be effective solvents for this trans-
formation providing 1 in 71% and 61% yield respectively
(entries 4À5). Interestingly, the one-pot conversion of the
crude sulfinate intermediate formed in situ led to similar
yields of 1 compared to when the crude sulfinate was
isolated as a solid and subsequently reacted with iodo-
methane (61% vs 55%, entry 5); the same trend was
observed when benzyl bromide was used as the alkylating
agent (benzyl sulfone obtained in 79% and 71% yield
respectively). In contrast, when 1-bromo-4-methoxyben-
zene was used, DMSO emerged as the solvent of choice
giving 1 in 56% yield (entries 6À8). The reaction did not
proceed in the absence of a catalyst or hydrogen donor
source, and among the ones we tested, sodium formate
gave the best results. We used potassium metabisulfite as a
source of SO2, as it offers the advantage of buffering the
pH of the reaction mixture, thus minimizing any risk of
decomposition of the sulfinate by disproportionation of
In the first experiment demonstrating the feasibility of
our approach, 1-iodo-4-methoxybenzene was treated with
K2S2O5 and NaO2CH (SO2 and hydrogen donors
respectively)9,10 in the presence of Pd(OAc)2 and PPh3 in
DMF at 70 °C. As desired, this reaction led to the
(6) Pelzer, G.; Keim, W. J. Mol. Catal. A.: Chem. 1999, 139, 235.
(7) Richards-Taylor, C. S.; Blakemore, D. C.; Willis, M. C. Chem.
Sci. 2013, DOI:10.1039/c3sc52332b.
(8) Nguyen, B.; Emmett, E. J.; Willis, M. C. J. Am. Chem. Soc. 2010,
132, 16372.
(9) For a recent example of application of K2S2O5 in a catalytic
process, see: Ye, S.; Wu, J. Chem. Commun. 2012, 48, 10037.
(10) Pri-Bar, I.; Buchman, O. J. Org. Chem. 1984, 49, 4009.
(11) For a postulated catalytic cycle operating in the palladium-
mediated sulfination of aryldiazonium salts, see ref 5b.
Org. Lett., Vol. 15, No. 24, 2013
6227

the corresponding sulfinic acid.1 Based on all of the above
observations, we selected Pd(OAc)2/PPh3/phen in the pre-
sence of K2S2O5, NaO2CH, and TBAB in DMSO at 70 °C
as general conditions to use in further evaluating the scope
of the transformation.
Table 3. Representative Set of Electrophilesa
Table 2. Representative Set of (het)Aryl Halides/Triflatesa
a Reaction conditions: see Table 2 and Supporting Information.
b Isolated yield. c Each step run in MeCN for 20 h. d Step 2 run at 50 °C.
e Step 2 was run at 60 °C for 48 h. f For step 2: sulfinate isolated as a solid was
employed, and water was used as solvent; H2NOSO3H (4 equiv).
entry 5 vs 8) in MeCN, treatment of 1-bromo-4-iodobenzene
in this solvent led to the chemoselective reaction of the aryl
iodide functionality to provide 2a, leaving the bromide
unreacted and available for subsequent cross-coupling if
desired. Also of note, an aryl triflate provided sulfone 2g in
57% yield (entry 10) whereas an aryl tosylate or mesylate
proved unreactive. Of particular interest to medicinal chem-
istry, where control of the lipophilicity is critical during
optimization of drug-like properties, the method is compa-
tible with heteroaryl halides (entries 11À14). Thus, heteroaryl
bromides and chlorides produced the corresponding sulfones
2hÀk in yields ranging from 40% to 68%. As is the case for
aryl halides (entry 2 vs 1), the yields of sulfones produced from
these substrates can be conveniently increased by using the
corresponding heteroaryl iodide (entry 11 vs 15, and 16). The
compatibility with heterocycles of medicinal chemistry rele-
vance was nicely demonstrated by the preparation of sildenafil
sulfone analogue 2l from the corresponding bromoaryl
and iodoaryl pyrazolopyrimidones (entry 16).12 Substrates
containing functional groups that are prone to reduction
(entry 10) or oxidation (entries 11À15) are compatible with
the reaction conditions thereby allowing for the rapid synth-
esis of sulfones that otherwise may have been difficult to
directly access using currently available methods.
a Reaction conditions: (het)ArX, K2S2O5 (2 equiv), TBAB
(1.1 equiv), NaO2CH (2.2 equiv), Pd(OAc)2 (5 mol %), PPh3 (15 mol %),
phen (15 mol %), DMSO, 70 °C, 3 h; then MeI (1.5 equiv), 23 °C, 18 h.
b Isolated yield. c Step 1 was run for 2 h. d Step 1 was run for 20 h. e Each step
was run in MeCN for 20 h. f TEAB used instead of TBAB.
In further exploration of our methodology, we decided
to focus on the synthesis of sulfones and sulfonamides in
view of the importance of these functionalities in chemis-
try. The scope of the sulfination reaction as it relates to aryl
and heteroaryl halides and triflates is illustrated in Table 2.
As noted previously with 1-bromo-4-methoxybenzene
(Table 1, entry 7), aryl bromides are competent substrates
for the reaction. This is beneficial for future applications,
as the bromides are more widely commercially available
than the iodides by an order of magnitude (∼500 000 vs
50000 compounds).3 Mindful of this observation, empha-
sis was placed on bromides in the evaluation of the reaction
scope. When submitted to a two-step, one-pot protocol
(i.e., alkylatingagent added directlytothe reaction mixture
arising from the sulfination step), electron-rich, -neutral,
and -deficient aryl bromides produced the corresponding
sulfone 1, 2aÀf in 53À72% yield (entries 2 and 4À9).
Unless activated, aryl chlorides did not react (entry 3 vs 13).
Taking advantage of the difference in reactivity between
1-iodo-4-methoxybenzene and its bromo analogue (Table 1,
As described in Table 3, the crude sulfinates could be
used without isolation and treated in situ with a wide range
of electrophiles to provide the corresponding sulfones or
sulfonamides in yields similar to those reported in litera-
ture for equivalent transformations.13 Primary alkyl ha-
lides and tosylates, secondary alkyl iodides, and benzyl
(12) TEAB used instead of TBAB to facilitate the purification of 2l.
(13) For representative examples of reactions of sulfinate anions with
electrophiles, see: Meek, J. S.; Fowler, J. S. J. Org. Chem. 1968, 33, 3422
and ref 7 (alkylation). Culvenor, C. C. J.; Davies, W.; Heath, N. S.
J. Chem. Soc. 1949, 278 (epoxide opening).
6228
Org. Lett., Vol. 15, No. 24, 2013

Table 4. Two-Step, One-Pot Sulfonamide Synthesisa
Table 5. Rapid Synthesis of Analogues of Sildenafil
as demonstrated by the synthesis of sulfone and sulfona-
mide analogues of sildenafil using a representative set of
alkylating agents and amines (Table 5). Beyond medicinal
chemistry, this example highlights the potential offered by
these two-step, one-pot protocols for future application in
any field where the rapid synthesis of sulfone and sulfona-
mide derivatives is needed.
In conclusion, we have developed a novel palladium-
catalyzed sulfination of aryl and heteroaryl halides and
triflates. The sulfinate synthetic intermediates thus formed
can be used without isolation and converted to the corre-
sponding sulfones and sulfonamides in situ, providing
unprecedented direct and practical access to these deriva-
tives from aryl and heteroaryl halides. The method has a
broad scope, is well suited for parallel chemical synthesis,
and is compatible with nitrogen-containing aromatic het-
erocycles of medicinal chemistry relevance.
a Reaction conditions: see Table 2 and Supporting Information. b Iso-
lated yield of 4. c DMF used as solvent, and NCS used instead of NBS.
d TEAB used instead of TBAB; 10 and 30 mol % of catalyst and ligands used.
bromide provided (alkylsulfonyl)benzene derivatives 1,
3aÀe in 31À79% yield (entries 1À8). When methyl glyci-
date was used as the electrophile, epoxide opening led to the
formation of the corresponding 3-phenylsulfonylpropanoate
derivative 3f (entry 9), offering direct access to the carbon
framework found in the drug bicalutamide (Figure 1). Lastly,
upon treatment with hydroxylamine-O-sulfonic acid, pri-
mary sulfonamide 3g was obtained in 37% yield (entry 10).
We next sought to identify a two-step, one-pot protocol
that would provide access to N-substituted sulfonamides.
Thus, when an amine was added to the precooled (0 °C)
reaction mixture from the sulfination step, followed by a
solution of N-bromosuccinimide (NBS) in THF, the cor-
responding sulfonamide 4a was isolated in 52% yield
(Table 4, entry 1). The same yield of 4a was obtained when
DMF and N-chlorosuccinimide (NCS) were used instead
of DMSO and NBS respectively. Additional examples are
presented in Table 4.14 Notably, heteroarylbromides again
proved to be suitable substrates, in this case affording the
corresponding sulfonamides in useful yields. The transfor-
mation also provides a convenient way to rapidly synthe-
size sildenafil and, by inference, sulfonamide analogues of
this drug as well (entry 5). The overall yields could be
increased by using the corresponding heteroaryl iodide as
the substrate (entry 5 and also 2 vs 1).
Acknowledgment. This article is dedicated to Professor
E. J. Corey on the occasion of his 85th birthday and for his
extensive contributions to organic synthesis. We thank Dr.
Ralph Robinson, Dr. Kevin Hesp, and Dr. Spiros Liras for
reviewing this manuscript. We also thank Jacquelyn Klug-
McLeod for her assistance in conducting searches in the
ACD and eMolecules databases, Jason Ramsay for high
resolution mass spectrometry determination, and Bha-
gyashree Khunte and Jaclyn Siderewicz for their assistance
with the purification of the analogues made in parallel
synthesis fashion. All colleagues acknowledged are from
Pfizer Inc., Groton, CT, USA.
Finally, reflecting their robustness and scope, all these
protocolscouldbecarried outina parallel synthesisformat
Supporting Information Available. Experimental pro-
cedures and characterization data. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
(14) Novel transformations have also been recently reported that
allow access to N-aminosulfonamides from aryl iodides (refs 8, 9) and to
sulfonamides from arylboronic acids ( DeBergh, J. R.; Niljianskul, N.;
Buchwald, S. L. J. Am. Chem. Soc. 2013, 135, 10638).
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 24, 2013
6229