Visible-light photoredox catalysis: Dehalogenation of vicinal dibromo-, α-halo-, and α,α-dibromocarbonyl compounds

Article abstract of DOI:10.1021/jo102239x

vic-Dibromo-, α-halo-, or α,α-dibromocarbonyl compounds can be efficiently dehalogenated using catalytic tris(2,2′-bipyridyl) ruthenium dichloride (Ru(bpy)₃Cl₂) in combination with 1,5-dimethoxynaphthalene (DMN) and ascorbate as sacrificial electron donor. For this process, a visible light promoted photocatalytic cycle is proposed that involves the reduction of carbon halogen bonds via free radical intermediates.

Full text of DOI:10.1021/jo102239x

pubs.acs.org/joc
molecules to absorb in the range between 400 to 800 nm.
Utilizing photosensitizers and photocatalysts3 solved this
long-standing problem.
Visible-Light Photoredox Catalysis:
Dehalogenation of Vicinal Dibromo-, r-Halo-,
and r,r-Dibromocarbonyl Compounds
In accord with our interest to develop new chemical
transformations going through radical intermediates,4 we
wanted to utilize alkyl and vinyl bromides as radical pre-
cursors being liberated by photoredox catalysis. Recent
reports from the groups of MacMillan5 and Stephenson6
have focused on the use of photoredox catalyst Ru(bpy)3Cl2
to promote enantioselective alkylations of aldehydes or
reductive dehalogenations, respectively, starting from acyl
bromides or tertiary benzyl halides. In all of these cases,
photoexcited *Ru2þ receives an electron from a sacrificial
electron donor such as a tertiary amine, which in turn
transfers that electron to a substrate. When we applied those
conditions to a variety of primary acyl bromides, we noticed,
however, their undesired direct reaction with the tertiary
amine by a nucleophilic substitution, thus greatly reducing
the yield of the desired radical processes.
Tapan Maji, Ananta Karmakar, and Oliver Reiser*
Institut fu€r Organische Chemie, Universita€t Regensburg
€
Universitatsstrasse 31, 93053 Regensburg, Germany
oliver.reiser@chemie.uni-regensburg.de
Received November 14, 2010
In the search for an alternative photocatalytic system we
came across the combination of 1,5-dimethoxynaphthalene
(DMN) as a primary and ascorbic acid as a sacrificial
electron donor,7 as applied by Pandey et al.8 for the cycliza-
tion of aldehydes and ketones onto tethered R,β-unsaturated
esters. We report here that these reagents in combination
with Ru(bpy)3Cl2 form an excellent photocatalytic system
allowing the visible light mediated reductive debromination
of vicinal dibromides as well as R-halocarbonyl compounds
in high yields.
Irradiation of vicinal dibromocarbonyl compounds 1 by a
blue LED (455 ( 10 nm) with Ru(bpy)3Cl2 in the presence of
DMN and ascorbic acid smoothly gave rise to the corre-
sponding R,β-unsaturated carbonyl compounds 2 (Scheme 1
and Table 1), with quantum yields ranging between 0.01 and
0.02. Optimization of the reaction conditions revealed that
a methanol-water mixture (10:1) is the best choice of sol-
vent for the process,9 employing 2 mol % of ruthenium
catalyst and 50 mol % of DMN. Reducing the amount of the
latter was possible but resulted in slower conversion of the
substrates (cf. entries 1 and 2). However, DMN could be
vic-Dibromo-, R-halo-, or R,R-dibromocarbonyl com-
pounds can be efficiently dehalogenated using catalytic
tris(2,20-bipyridyl)ruthenium dichloride (Ru(bpy)3Cl2) in
combination with 1,5-dimethoxynaphthalene (DMN)
and ascorbate as sacrificial electron donor. For this
process, a visible light promoted photocatalytic cycle is
proposed that involves the reduction of carbon halogen
bonds via free radical intermediates.
Bromination-debromination sequences are widely used in
organic synthesis for the protection-deprotection of olefins.
Although functionalization of olefins by simple bromination
generally proceeds smoothly and stereospecifically in high
yields, reversing the process to the parent olefin via a debromi-
nation step is more challenging. There are a number of reagents
known for this transformation;1 however, the necessity to
employ stoichiometric amounts of strongly reducing agents
might cause problems of selectivity and functional group
compatibility. Moreover, the toxicity of some reagents such
as the widely used organotin compounds makes the develop-
ment of more sustainable alternatives desirable.
In recent years, there has been increasing interest in the
use of visible light to drive organic reactions because of its
infinite availability, ease of handling, and promising applica-
tion in industry.2 However, the use of visible light in organic
reactions is limited due to the inability of many organic
(3) (a) Kuczkowski, R. L. J. Am. Chem. Soc. 1963, 85, 3047–3048.
(b) Baldwin, J. E.; Greeley, R. H. J. Am. Chem. Soc. 1965, 87, 4514–4516.
(c) Salomon, R. G.; Kochi, J. K. J. Am. Chem. Soc. 1974, 96, 1137–1144.
(d) Zeitler, K. Angew. Chem., Int. Ed. 2009, 48, 9785–9789. (e) Narayanam,
J. M. R.; Stephenson, C. R. J. Chem. Soc. Rev. 2010, 39.
(4) (a) Prediger, I.; Weiss, T.; Reiser, O. Synthesis 2008, 14. (b) Gheorghe,
A.; Schulte, M.; Reiser, O. J. Org. Chem. 2006, 71, 2173–2176. (c) Lin, H.;
Schall, A.; Reiser, O. Synlett 2005, 2603–2606. (d) Jezek, E.; Schall, A.;
Kreitmeier, P.; Reiser, O. Synlett 2005, 915–918.
(5) Nicewicz, D. A.; MacMillan, D. W. C. Science 2008, 322, 77–80.
(6) Narayanam, J. M. R.; Tucker, J. W.; Stephenson, C. R. J. J. Am.
Chem. Soc. 2009, 131, 8756–8757.
(7) Hamada, T.; Nishida, A.; Yonemitsu, O. J. Am. Chem. Soc. 1986, 108,
140–145.
(1) (a) Ranu, B. C.; Guchhait, S. K.; Sarkar, A. Chem. Commun. 1998,
2113–2114. (b) Schmidt, A.; Snovydovych, B.; Gjikaj, M. Synthesis 2008,
2798–2804. (c) Ranu, B. C.; Jana, R. J. Org. Chem. 2005, 70, 8621–8624. (d)
Butcher, T. S.; Detty, M. R. J. Org. Chem. 1998, 63, 177–180. (e) Li, W.; Li, J.;
Lin, M.; Wacharasindhu, S.; Tabei, K.; Mansour, T. S. J. Org. Chem. 2007,
72, 6016–6021.
(8) (a) Pandey, G.; Hajra, S.; Ghorai, M. K.; Kumar, K. R. J. Org. Chem.
1997, 62, 5966–5973. (b) Pandey, G.; Hajra, S.; Ghorai, M. K.; Kumar, K. R.
J. Am. Chem. Soc. 1997, 119, 8777–8787.
(9) Solvent combinations had to be employed in which the ruthenium
catalyst and DMN are soluble. The reaction proceeded well in acetonitrile or
methanol, although with considerably longer reaction times (24 h instead of
5 h for full conversion), while DMF resulted in an unclean reaction with the
formation of a number of unidentified side products.
(2) Sala, X.; Romero, I.; Rodrıguez, M.; Escriche, L.; Llobet, A. Angew.
´
Chem., Int. Ed. 2009, 48, 2842–2852.
736 J. Org. Chem. 2011, 76, 736–739
Published on Web 12/30/2010
DOI: 10.1021/jo102239x
r
2010 American Chemical Society

Maji et al.
JOCNote
TABLE 1. Reductive Debromination of Vicinal Dibromocarbonyl
Compounds Leading to Disubstituted Alkenesa
TABLE 2. Reductive Debromination of Vicinal Dibromocarbonyl
Compounds
a4 mol % of Ru(bpy)3Cl2 was employed.
alkenes (Table 1). A carbonyl substituent, e.g., a ketone,
ester, lactone, or oxazolidinone, adjacent to the bromide is
required for the debromination to take place (entries 1-9),
e.g., no reaction took place with 1,2-dibromo-1,2-diphenyl-
ethane (entry 10). Moreover, alkyl-substituted dibromides
react sluggishly in most cases (Table 1, entries 8 and 9; Table 2,
entries 2 and 3). Only (E)-configured disubstituted alkenes
wereobtained from acyclic precursors inall cases, even when a
43:57 mixture of erythro and threo vic-dibromide 1c (Table 1,
entry 3) was subjected to the above conditions. The cyclic
dibromide 1d also underwent debromination smoothly
(Table 1, entry 4), giving rise to enone 2d predetermined by
the ring size of the cycle.
The E/Z-selectivity in debrominations of compounds that
give rise to trisubstituted alkenes (Table 2) seems to be
governed by the relative stability of the products, allowing
in some cases the conversion of stereomixtures of dibromides
to one geometrically pure alkene (Table 2, entries 1 and 3)
but might also lead to E/Z-mixtures of alkenes despite
starting from diastereomerically pure dibromide (Table 2,
entry 5).
aReaction conditions: vic-dibromo compound (1.0 equiv), Ru(bpy)3Cl2
(2 mol %), DMN (50 mol %), ascorbic acid (2 equiv), MeOH/H2O = 10:1,
b
blue LED. Isolated yield after purification by chromatography on SiO2.
cReaction performed with 25 mol % of DMN keeping other conditions
unaltered. dErythro/threo =43:57. eRu(bpy)3Cl2 (4 mol %). fNo reaction.
These observations confirm that the mechanism does not
proceed by a stereospecific E2-type anti-elimination but
rather involves a relatively stable radical or anion intermedi-
ate which breaks down to the alkenes according to their
thermodynamic stability.
SCHEME 1. Photocatalyzed Reductive Debromination of
Vicinal Dibromocarbonyl Compounds
The synthesis of an alkyne from the corresponding vic-
dibromoalkene was also achieved successfully utilizing the
same reaction conditions as previously applied for the de-
bromination of the vic-dibromoalkanes (Scheme 2).
Using 1a (Table 1, entry 1) as a representative substrate, a
series of control experiments were performed. Exclusion of
any one of the reaction components ((Ru(bpy)3Cl2), DMN,
or ascorbic acid) yielded only a negligible amount of the
product, even after 48 h.
recovered at the end of the reaction and reused, making the
employment of higher amounts more practical but never-
theless economically acceptable.
A representative range of vic-dibromides 1 underwent de-
bromination by this procedure to provide the corresponding
J. Org. Chem. Vol. 76, No. 2, 2011 737

Maji et al.
JOCNote
SCHEME 2. Photocatalytic Reductive Debromination of a
vic-Dibromoalkene
TABLE 3. Photocatalyzed Reductive Dehalogenation of r-Haloketonesa
SCHEME 3. Photocatalytic Reductive Dehalogenations of
r-Halocarbonyl Compounds
The conditions described here for the debromination of
vic-dibromoalkanes and vic-dibromoalkenes were also suc-
cessfully employed in the dehalogenation of a variety of
R-halocarbonyl compounds (Scheme 3, Table 3).
Irradiating 3a (Table 3, entry 1) with blue LED light gave
rise to 4a within 5 h in excellent yield (93%). No side product
was observed, whereas DMN could be reisolated almost
quantitatively during purification of the product on silica. A
control experiment excluding Ru(bpy)3Cl2 gave a very neg-
ligible amount of 4a, whereas decreasing the amount of
DMN (10 mol %) requires 40 h to complete the reaction.
Both primary and secondary R-bromocarbonyl compounds
can be reduced in high yield (Table 3, entries 1-8) under the
reaction conditions. This protocol is also amenable to the
corresponding R-chlorocarbonyl compounds; however,
longer reaction times are necessary to achieve good conversion
(Table 3, entries 1, 6, and 8). Good functional group tolerance,
notably for those that could in principle also undergo reduction
such as aromatic nitro or bromo moieties (Table 3, entries 3, 9,
and 10), was observed.
The protocol described here is also useful for the selective
debromination of R,R-dibromo to their corresponding
monobromo ketones (Table 3, entries 10-12), a transforma-
tion that is difficult to achieve with conventional reducing
reagents.10 By simple control of the reaction time, the
selective photocatalytic reduction of R,R-dibromoketones
to either monobromo ketones or to the doubly debrominated
parent ketones (Table 3, entries 9-12) becomes possible.
Mechanistically, we propose that the dehalogenation involves
the activation of the substrate to form a radical intermediate by
transferring an electron from Ru(I) (Scheme 4) following
literature precendent.6 It is well established11 that Ru(II) readily
accepts a photon from a variety of light sources (here from a
blue LED) to populate the activated *Ru(II) metal-to-ligand
charge transfer excited state. DMN acting as the primary donor
transfers an electron to initiate the catalytic cycle by generating a
Ru(I) complex. As Ru(I) is a potent reductant, it is able to
transfer a single electron to the R-halocarbonyl substrate, thus
aReaction conditions: organohalide (1.0 equiv), Ru(bpy)3Cl2 (2 mol %),
DMN (50 mol %), ascorbic acid (2 equiv), MeOH/H2O = 10:1, blue LED.
bIsolated yield after purification by chromatography on SiO2. c10 mol %
DMN was employed. dThe reaction was stopped (TLC monitoring) when
2-fold debrominated product just started to form.
furnishing a radical anion that rapidly eliminates a halide. The
resulting R-acyl radical then could be further reduction by the
photocatalyst to give an enolate that is ultimately undergoes
protonation or halide elimination to the products observed.
In conclusion, we have developed a conceptually new
photocatalytic system with visible light to drive sequential
electron transfer processes for reductive debromination
of vicinal dibromo compounds to the corresponding
(E)-alkenes and alkynes, gem-dibromides to monobromides,
(10) Ranu, B. C.; Chattopadhyay, K.; Jana, R. Tetrahedron 2007, 63,
155–159.
(11) (a) Juris, A.; Balzani, V.; Barigelletti, F.; Campagna, S.; Belser, P.;
von Zelewsky, A. Coord. Chem. Rev. 1988, 84, 85–277. (b) Inagaki, I.; Akita,
M. Coord. Chem. Rev. 2010, 254, 1220–1239.
738 J. Org. Chem. Vol. 76, No. 2, 2011

Maji et al.
JOCNote
SCHEME 4. Mechanistic Proposal for the Photocatalyzed
Debromination Processes
plastic septum and magnetic stir bar was charged with tris-
(2,20-bipyridyl)ruthenium(II) chloride hexahydrate (2.5 mg,
2 mol %), the corresponding halide (0.17 mmol, 1.0 equiv),
1,5-dimethoxynaphthalene (16 mg, 0.09 mmol, 0.5 equiv), and
ascorbic acid (61 mg, 0.35 mmol, 2 equiv). The flask was purged
with a stream of nitrogen and 1 mL of solvent (MeOH/
H2O =10:1) was added. The resultant mixture was degassed
for 10 min by nitrogen sparging and placed at a distance of
∼0.5-1.0 cm away from a blue LED lamp. After the reaction
was completed (as judged by TLC analysis), the mixture was
diluted with 3 mL of methanol and concentrated in vacuo. The
residue was purified by chromatography on silica gel, using
hexanes/ethylacetate 9:1 as the eluent. 1,5-Dimethoxynaphtha-
lene (13 mg, 82%) was reisolated during the column purification
(Rf 0.63).
In a typical experiment (2S*,3R*)-dibromo-1,3-diphenyl-
propan-1-one (64 mg, 0.17 mmol, 1.0 equiv) (1a) was con-
verted to (E)-1,3-diphenylprop-2-en-1-one (2a, 33 mg, 92%,
Rf (EtOAc/hexanes 1:9) 0.41) along with reisolation of DMN
(13 mg, 82%, Rf 0.63). 2a:12 1H NMR (300 MHz; CDCl3) δ
7.36-7.70 (m, 9H), 7.84 (d, J = 15.6 Hz, 1H), 7.98-8.10 (m,
2H); 13C NMR (75 MHz; CDCl3) δ 122.1, 128.5 (relative
intensity 2), 128.6, 128.9, 130.5, 132.8, 134.9, 138.2, 144.8,
190.6.
and monobromides or chlorides to their dehalogenated
counterparts. As a key structural feature, a carbonyl group
positioned alpha to one halogen atom is required for the
success of these transformation. The photosystem is also
highlighted by its chemoselectivity, resulting in reduced
products in high yields under mild reaction conditions.
Acknowledgment. This work was supported by the
DAAD (fellowship for T.M.), Alexander von Humboldt
Foundation (fellowship for A.K.), the DFG Graduierten-
kolleg 1626 “Photocatalysis”, and the Fonds der Che-
mischen Industrie.
Experimental Section
General Procedure for the Photoredox-Catalyzed Transforma-
tion of Organohalides. An oven-dried 10 mL vial equipped with a
Supporting Information Available: Experimental proce-
dures, detailed studies, and spectral characterization of all
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
(12) Mahesh, M.; Murphy, J. A.; Wessel, H. P. J. Org. Chem. 2005, 70,
4118–4123.
J. Org. Chem. Vol. 76, No. 2, 2011 739