Cycloadditions by oxidative visible light photocatalysis

Article abstract of DOI:10.1021/ja103934y

Photochemical reactions are remarkable for their ability to easily assemble cyclobutanes and other strained ring systems that are difficult to construct using other conventional synthetic methods. We have previously shown that Ru(bpy)₃²⁺ is an efficient photocatalyst that promotes the [2+2] cycloadditions of electron-deficient olefins with visible light. Here, we show that Ru(bpy)₃²⁺ is also an effective photocatalyst for the [2+2] cycloaddition of electron-rich olefins. This transformation is enabled by the versatile photoelectrochemical properties of Ru(bpy) ₃²⁺, which enables either one-electron reduction or one-electron oxidation of interesting organic substrates under appropriate conditions.

Full text of DOI:10.1021/ja103934y

Published on Web 06/08/2010
[2+2] Cycloadditions by Oxidative Visible Light Photocatalysis
Michael A. Ischay, Zhan Lu, and Tehshik P. Yoon*
Department of Chemistry, UniVersity of WisconsinsMadison, 1101 UniVersity AVenue, Madison, Wisconsin 53706
Received May 8, 2010; E-mail: tyoon@chem.wisc.edu
Cyclobutanes are prominent structural features of many bioactive
natural products.1 Arguably the most straightforward method for
the preparation of cyclobutane rings is the [2+2] photocycloaddition
of olefins, and the utility of this prototypical photochemical reaction
has been demonstrated in numerous synthetic applications.2 Nev-
ertheless, the requirement for irradiation with high energy UV light
is a disadvantage of this reaction in terms of the cost, scalability,
and environmental impact of the methodology.3 We recently
reported a new approach to [2+2] enone cycloadditions catalyzed
by Ru(bpy)3Cl2 upon irradiation with low-intensity visible light.4
Several other groups5 have also recently become interested in
similar strategies for utilizing the well-studied photoredox properties
of metal polypyridyl complexes6 in various synthetically useful
transformations.
sponding radical cations. This reactivity was first described by
Ledwith7 in 1969 and has subsequently been shown to be
accessible both by chemical oxidants and by photoinduced
electron transfer with organic sensitizers.8 Second, the photo-
2+
chemistry of Ru(bpy)3 has been extensively investigated for
solar energy applications,6a and the most well-studied among
these systems rely on an oxidative quenching cycle (Figure 1,
2+
Path B) in which Ru*(bpy)3 reacts with an electron accept-
or (e.g., methyl viologen, MV2+).9 The resulting oxidized
3+
Ru(bpy)3 complex is turned over by one-electron reduction
using a sacrificial electron-rich organic species such as a tertiary
amine base.
Given these precedents, it seemed logical that the photogenerated
3+
Ru(bpy)3 complex generated upon visible light irradiation of
Notably, each of the methods recently developed by us,4
MacMillan,5a,b Stephenson,5c-e and Akita5f has taken advantage
of a reductive quenching photoredox cycle (Figure 1, Path A). In
our method for [2+2] cycloaddition of enones, for example, the
photoexcited state (Ru*(bpy)32+), generated upon visible light
irradiation of the photocatalyst, abstracts an electron from a
relatively electron-rich tertiary amine base (i-Pr2NEt); the resulting
Ru(bpy)3+ complex is a strong reductant that reduces an aryl enone
to the key radical anion intermediate involved in the [2+2]
cycloaddition. This mechanism implies that a fundamental limitation
of our strategy is the requirement for an alkene that is sufficiently
electron-deficient to undergo efficient one-electron reduction by
Ru(bpy)3+; indeed, electron-rich olefins (e.g., styrenes) do not react
under the conditions we previously reported.
Ru(bpy)32+ in the presence of MV2+ should also oxidize electron-
rich styrenes, affording a radical cation that would undergo
subsequent [2+2] cycloaddition. Indeed, under optimized condi-
tions, bis(styrene) 1 undergoes efficient intramolecular cycloaddition
2+
upon irradiation in the presence of 5 mol % Ru(bpy)3 and 15
mol % MV2+, affording cyclobutane 2 in 89% yield with excellent
2+
diastereoselectivity. As with our previously described Ru(bpy)3
-
catalyzed reactions, this transformation can be conducted using a
standard household light bulb and does not require specialized
photochemical equipment.
Scheme 1. Photooxidative [2+2] Cycloaddition
Table 1. Modification of Experimental Parameters
entry
variation from optimized conditions
yielda
1
No change
89%b
0%
0%
0%
0%
2c
3
Conventional photolysis (Xe arc lamp, no Ru)
No light
2+
4
5
No Ru(bpy)3
No MV2+
6
7
8
9
10
11
12
1,4-Dinitrobenzene instead of MV2+
1,4-Benzoquinone instead of MV2+
1 atm of O2 instead of MV2+
MeCN instead of MeNO2
Acetone instead of MeNO2
DMSO or DMF instead of MeNO2
No MgSO4
13%
14%
0%
36%
11%
0%
Figure 1. Reductive and oxidative photocatalytic cycles.
We therefore became interested in designing a complementary
method for photooxidative electron transfer catalysis that could
engage electron-rich olefins in productive [2+2] cycloadditions.
Our design plan draws upon two well-established precedents.
First, electron-rich olefins are known to participate in [2+2]
cycloadditions upon one-electron oxidation to afford the corre-
73%
a Yields determined by 1H NMR spectroscopy using an internal
standard, unless noted. b Isolated yield. c A solution of 1 in MeNO2 was
irradiated with a xenon arc lamp for 3.5 h.
9
8572 J. AM. CHEM. SOC. 2010, 132, 8572–8574
10.1021/ja103934y  2010 American Chemical Society

C O M M U N I C A T I O N S
Table 2. Representative [2+2] Cycloadditionsa
Table 1 summarizes the importance of each of the experimental
parameters to the efficiency of the reaction. First, we observe no
cycloaddition upon irradiation of 1 with UV light under conven-
tional photolytic conditions (entry 2). This result suggests that the
reaction does not involve direct photoexcitation of the styrenic
substrate. The observation that visible light, Ru(bpy)32+, and MV2+
are each required for successful cycloaddition (entries 3-5) is
instead consistent with the photoinduced one-electron oxidation
mechanism that we have proposed. Other known oxidative quench-
ers of Ru*(bpy)32+ such as nitroarenes and quinones10 also promote
cycloaddition but were not as effective as MV2+ (entries 6 and 7).
Oxygen was an ineffective co-oxidant (entry 8).11 We observed a
pronounced solvent dependency on the efficiency of the reaction
(entries 9-11); reactions conducted in MeCN and acetone produced
poorer yields of the cycloadduct than those using MeNO2, and we
observed no conversion at all in more polar solvents such as DMSO
and DMF. Finally, the reaction showed a modest sensitivity to
adventitious water, and we found that the addition of MgSO4
provided slightly higher and more reproducible yields (entry 12).
Experiments probing the scope of [2+2] cycloadditions using
the Ru(bpy)32+/MV2+ system are outlined in Table 2. In most cases,
we found that 1 mol % of the Ru photocatalyst is sufficient for
successful cycloaddition. Our mechanistic design plan is validated
by the observation that at least one styrene must bear an electron-
donating substituent at the para or ortho position (entries 1 and
2); meta-substituted and unsubstituted styrenes are presumably not
electron-rich enough to undergo one-electron oxidation to afford
the key radical cation intermediate (entries 3 and 4).12 However, a
number of electron-donating para substituents are effective activa-
tors of the styrene (entries 5-6), and the presence of an electron-
withdrawing substituent at the meta position does not seem to
significantly decrease the efficiency of the cycloaddition (entry 7).
Aliphatic olefins are not suitable reaction partners (entry 8). On
the other hand, both electron-rich and electron-poor styrenes react
smoothly with the photogenerated radical cation (entries 9-11).
Substitutents at the R-position of the styrene are tolerated (entry
10), which enables access to all-carbon quaternary stereocenters
on the cyclobutane framework, although ꢀ-substituents significantly
retard the rate of reaction. Substituents on the tether are also
tolerated, and these modifications can induce good levels of facial
selectivity in the cycloaddition (entries 11 and 12, 5:1 and 7:1 d.r.,
respectively). The identity of the tether seems to be critical; while
both oxygen and nitrogen-containing tethers give good yields (entry
13), we have been unable to identify all-carbon tethers that promote
efficient cycloaddition, and intermolecular cycloadditions are im-
practically slow.
a Unless otherwise noted, reactions conducted using
1 mol %
Ru(bpy)3(PF6)2 and 15 mol % MV(PF6)2. b Yields represent the averaged
results of two reproducible experiments. c Conducted using 5 mol %
Ru(bpy)3(PF6)2 and 15 mol % MV(PF6)2. d 5:1 d.r. e 7:1 d.r.
studies of intermolecular [2+2] cycloadditions chemically initiated
by an aminium radical cation.13a It is also consistent with the
observation that the minor trans isomers produced from cycload-
dition of 3 and of 4 are different and are consistent with the
stereoretentive suprafacial cycloadditions of each of these (E,Z)
substrates.
Substrates 3 and 4, the (E,Z) bis(styrenes) isomeric to the model
(E,E) substrate 1 were also prepared and irradiated with visible
light in the presence of Ru(bpy)32+ and MV2+ (Scheme 2). In both
cases, the major product observed is the same cis diasteromer
obtained from cycloaddition of 1, indicating that the stereochemical
integrity of the olefins is lost over the course of the reaction. To
better understand the origins of the stereoconvergency, we moni-
tored the cycloaddition of 4 by GC (see Supporting Information).
During the course of this reaction, 4 undergoes isomerization to 1
at a rate competitive with that of cycloaddition. As the reaction
proceeds, the ratio of cis-4 to the isomeric trans cycloadduct
increases from 1:1 at 30 min to 5:1 upon completion of the reaction.
We conclude from these studies that the [2+2] cycloaddition step
is itself stereospecific, as predicted from previous theoretical and
experimental studies of radical cation cyclobutanations13 but that
the rate of cycloaddition is relatively slow compared to the rate of
olefin isomerization. This conclusion is consistent with Bauld’s
Thus, the experimental evidence suggests that cycloaddition using
the Ru(bpy)32+/MV2+ catalyst system indeed involves a radical
Scheme 2. Stereoconvergent [2+2] Cycloadditions
9
J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010 8573

C O M M U N I C A T I O N S
cationic intermediate whose reactivity is identical to those generated
using other methods. Photoinduced electron transfer has previously
been used to initiate similar radical cation [2+2] cycloadditions of
electron-rich olefins,14 but these reactions have generally required
mercury arc lamps and relatively high loadings of an aromatic nitrile
photosensitizer. Consistent with these reports, we find that when
the cycloaddition of 1 is conducted using 5 mol % of 9,10-
dicyanoanthracene (DCA)14b,c in place of Ru(bpy)32+, the reaction
is considerably slower and produces cyclobutane 2 in only 19%
yield after 3.5 h under otherwise identical conditions. The faster
Acknowledgment. We thank Chris Shaffer and Prof. Robert
McMahon for their assistance performing the direct photolysis
experiment (Table 1, entry 2). We thank the Sloan Foundation,
Beckman Foundation, and Research Corporation for financial
support. The NMR facilities at UW-Madison are funded by the
NSF (CHE-9208463, CHE-9629688) and NIH (RR08389-01).
Supporting Information Available: Experimental procedures and
spectral data for all new compounds are provided. This information is
available free of charge via the Internet at http://pubs.acs.org.
2+
reaction rates using Ru(bpy)3 may be attributable to its longer
excited state lifetime (600 ns vs 15 ns), its larger extinction
coefficients (13 000 vs 11 500 M-1 cm-1), and its broader absorption
in the visible range compared to DCA.6,15
References
(1) (a) Hansen, T. V.; Stenstrøm, Y. Naturally Occurring Cyclobutanes. In
Organic Synthesis: Theory and Applications; Hudlicky, T., Ed.; Elsevier:
Oxford, U.K., 2001; Vol. 5, pp 1-38. (b) Dembitsky, V. M. J. Nat. Med.
2008, 62, 1–33.
(2) For recent reviews, see: (a) Crimmins, M. T. Chem. ReV. 1988, 88, 1453–
1473. (b) Bach, T. Synthesis 1998, 683–708. (c) Lee-Ruff, E.; Mladenova,
G. Chem. ReV. 2003, 103, 1449–1483. (d) Hoffman, N. Chem. ReV. 2008,
108, 1052–1103.
As in the photoreductive enone cycloaddition we previously
reported,4 this method for photooxidative cycloaddition does not
require the use of any specialized photochemical equipment, and
our reactions are typically conducted using a standard household
light bulb. To highlight the efficiency of the Ru(bpy)32+/MV2+
system in promoting the radical cation mediated cycloaddition, we
conducted the cycloaddition of 5 on a gram scale in a laboratory
window using ambient sunlight as the only source of irradiation
(Scheme 3). The cycloaddition still proceeded to completion in 2.5 h
and provided a nearly identical yield of the cyclobutane product as
smaller-scale experiments under more controlled conditions. In
addition, the larger-scale reaction was conducted in undistilled
nitromethane and without rigorous degassing of the solvent. Thus,
these conditions provide a powerful and operationally facile method
to perform photochemical cycloadditions using convenient sources
of visible light including ambient sunlight.
(3) Esser, P.; Pohlmann, B.; Scharf, H.-D. Angew. Chem., Int. Ed. Engl. 1994,
33, 2009–2023.
(4) (a) Ischay, M. A.; Anzovino, M. E.; Du, J.; Yoon, T. P. J. Am. Chem. Soc.
2008, 130, 12886–12887. (b) Du, J.; Yoon, T. P. J. Am. Chem. Soc. 2009,
131, 14604–14605.
(5) (a) Nicewicz, D.; MacMillan, D. W. C. Science 2008, 322, 70–80. (b) Nagib,
D. A.; Scott, M. E.; MacMillan, D. W. C. J. Am. Chem. Soc. 2009, 131,
10875–10877. (c) Narayanam, J. M. R.; Tucker, J. W.; Stephenson, C. R. J.
J. Am. Chem. Soc. 2009, 131, 8756–8757. (d) Tucker, J. W.; Narayanam,
J. M. R.; Krabbe, S. W.; Stephenson, C. R. J. Org. Lett. 2010, 12, 368–
371. (e) Condie, A. G.; Gonza´lez-Go´mez, J. C.; Stephenson, C. R. J. J. Am.
Chem. Soc. 2010, 132, 1464–1465. (f) Koike, T.; Akita, M. Chem. Lett.
2009, 38, 166–167.
(6) For reviews, see: (a) Kalyanasundaram, K. Coord. Chem. ReV. 1982, 46,
159–244. (b) Juris, A.; Balzani, V.; Barigelletti, F.; Campagna, S.; Belser,
P.; von Zelewsky, A. Coord. Chem. ReV. 1988, 84, 85–277. (c) Flamigni,
L.; Barbieri, A.; Sabatini, C.; Ventura, B.; Barigelletti, F. Top. Curr. Chem.
2007, 281, 143–203.
(7) (a) Bell, F. A.; Crellin, R. A.; Fugii, N.; Ledwith, A. J. Chem. Soc., Chem.
Commun. 1969, 251–252. (b) Carruthers, R. A.; Crellin, R. A.; Ledwith,
A. J. Chem. Soc., Chem. Commun. 1969, 252–253.
Scheme 3. Gram-Scale Cycloaddition with Ambient Sunlight
(8) For reviews, see: (a) Ledwith, A. Acc. Chem. Res. 1972, 5, 133–139. (b)
Bauld, N. L.; Bellville, D. J.; Harirchian, B.; Lorenz, K. T.; Pabon, R. A.,
Jr.; Reynolds, D. W.; Wirth, D. D.; Chiou, H. S.; Marsh, B. J. Acc. Chem.
Res. 1987, 20, 371–378. (c) Bauld, N. L. Tetrahedron 1989, 45, 5307–
5363.
(9) For example: (a) Kalyanasundaram, K.; Kiwi, J.; Gra¨tzel, M. HelV. Chim.
Acta 1978, 61, 2720–2730. (b) Kirch, M.; Lehn, J.-M.; Sauvage, J.-P. HelV.
Chim. Acta 1979, 62, 1345–1383.
(10) (a) Bock, C. R.; Meyer, T. J.; Whitten, D. G. J. Am. Chem. Soc. 1975, 97,
2909–2911. (b) Darwent, J. R.; Kalyanasundaram, K. J. Chem. Soc.,
Faraday Trans. 2 1981, 77, 373–382.
(11) Oxygen is generally believed to quench Ru*(bpy)32+ by energy transfer in
organic solvents. See: (a) Mulazzani, Q. G.; Sun, H.; Hoffman, M. Z.; Ford,
W. E.; Rodgers, M. A. J. J. Phys. Chem. 1994, 98, 1145–1150. (b) Tanielian,
C.; Wolff, C.; Esch, M. J. Phys. Chem. 1996, 100, 6555–6560.
Thus, we have shown that Ru(bpy)32+ is a powerful photocatalyst
for the [2+2] cycloaddition of both electron-rich and electron-
deficient olefins. The versatility of this catalyst arises from the
ability to access either photooxidative or photoreductive reactivity
by choosing the appropriate oxidative or reductive quencher,
respectively. In both regimes, the photophysical properties of
Ru(bpy)32+ enable a variety of inexpensive, readily available sources
of visible light to be utilized, including sunlight. In addition, there
exists a vast wealth of electrochemical literature that describes
synthetically useful organic transformations initiated by one-electron
redox processes. We expect that photocatalytic systems exploiting
3+
(12) The oxidation potential of Ru(bpy)3 is +1.3 V (ref 6a). This should be
sufficient to oxidize 4-methoxy-ꢀ-methylstyrene (+1.2 V) but not ꢀ-me-
thylstyrene itself (+1.7 V). Values from: Yueh, W.; Bauld, N. L. J. Phys.
Org. Chem. 1996, 9, 529–538.
(13) (a) Bauld, N. L.; Pabon, R. J. Am. Chem. Soc. 1983, 105, 633–634. (b)
Bauld, N. L.; Bellville, D. J.; Pabon, R.; Chelsky, R.; Green, G. J. Am.
Chem. Soc. 1983, 105, 2378–232. (c) Bauld, N. L.; Yang, J. Org. Lett.
1999, 1, 773–774. (d) Wiest, O. J. Phys. Chem. A 1999, 13, 7907–7911.
(14) For a recent review, see: (a) Fagnoni, M.; Dondi, D.; Ravelli, D.; Albini,
A. Chem. ReV. 2007, 107, 2725–2756. For examples using a 9,10-
dicyanoanthracene photosensitizer, see: (b) Mizuno, K.; Kagano, H.; Otsuji,
Y. Tetrahedron Lett. 1983, 24, 3849–3850. (c) Lewis, F. D.; Kojima, M.
J. Am. Chem. Soc. 1988, 110, 8664–8670.
2+
the reactivity of Ru(bpy)3 should also be able to efficiently
promote similar reactivity. The exploration of this reactivity will
continue to be a focus of research in our lab.
(15) Gould, I. R.; Ege, D.; Moser, J. E.; Farid, S. J. Am. Chem. Soc. 1990, 112,
4290–4301.
JA103934Y
9
8574 J. AM. CHEM. SOC. VOL. 132, NO. 25, 2010