Oxidative cyclization reactions: Controlling the course of a radical cation-derived reaction with the use of a second nucleophile

Article abstract of DOI:10.1002/anie.201308739

Construction of new ring systems: Oxidative cyclizations (see picture; RVC=reticulated vitreous carbon) have been conducted that use two separate intramolecular nucleophiles to trap an enol ether-derived radical cation intermediate. The reactions provide

Full text of DOI:10.1002/anie.201308739

Angewandte
Chemie
DOI: 10.1002/anie.201308739
Anodic Cyclizations
Oxidative Cyclization Reactions: Controlling the Course of a Radical
Cation-Derived Reaction with the Use of a Second Nucleophile**
Alison Redden, Robert J. Perkins, and Kevin D. Moeller*
The oxidative generation of reactive radical cation inter-
mediates can serve as a powerful tool for the construction of
new ring systems.[1,2] For example, substrates with electron-
rich olefins can be oxidized to generate radical cations that
trigger cyclizations with a variety of electron-rich groups.[3]
Enol ethers, vinylsulfides, ketene derivatives, electron-rich
aryl rings, and styrenes have all been oxidized to form radical
cations, whereas enol ethers, allyl and vinylsilanes, aryl rings,
styrenes, alcohols, amides, sulfonamides, and amines have all
been used to trap the radical cation. The reactions have led to
the synthesis of fused and bicyclic ring skeletons and are often
compatible with the formation of tetrasubstituted carbons. In
addition, they have served to help us gain a better under-
standing of radical cation intermediates.[4]
The failure of reactions like those highlighted in Scheme 1
suggests that an alternative strategy is needed that will allow
for oxidative cyclization reactions to be accomplished when
they involve slower ring formation. To do so requires that one
slows down the competitive “cationic” decomposition path-
ways while pushing the intermediate toward the desired
cyclization. We report here that this can be accomplished with
the use of a second intramolecular nucleophile for trapping
the radical cation.
The basic idea is a simple one. It is illustrated in Scheme 2
as a potential solution to the first problematic cyclization
shown in Scheme 1. In this oxidative cyclization, the radical
However, not all oxidative cyclizations work well. Radical
cations are very reactive intermediates. If a cyclization
reaction is too slow, then alternative pathways compete.
Two examples are shown in Scheme 1. In both examples,
a slow cyclization reaction led to side reactions involving an
elimination step after formation of the radical cation.[5,6] This
type of “cationic” decomposition of the radical cation is
common in anodic reactions that fail.
Scheme 2. A plan for avoiding elimination reactions.
cation (2) would initially be trapped by an alcohol nucleophile
to make a five-membered ring acetal (3). The five-membered
ring cyclization between an enol ether radical cation and an
alcohol-trapping group is known to be very fast, a situation
that should reduce the chance for competing elimination
reactions.[4,7]
Scheme 1. Failed anodic cyclization reactions.
The result of the initial cyclization would be the formation
of a radical intermediate that could then go on to complete
the desired cyclization while avoiding the unwanted elimi-
nation reaction. Oxidation of a second electron and elimi-
nation of the silyl group would then complete the formation
of product (4).
Work on the project was started by first establishing the
feasibility of the general plan. To this end, substrate 5 was
synthesized and exposed to the anodic oxidation reaction
(Scheme 3).[8] The oxidative cyclization was conducted in an
undivided cell with the use of a reticulated vitreous carbon
(RVC) anode, a carbon-rod cathode, 2,6-lutidine as a proton
scavenger, a 0.1m LiClO4 in 20% MeOH/CH2Cl2 electrolyte
solution, and a constant current of 8 mA. The reaction was
allowed to proceed until 2.1 FmolÀ1 of charge had been
[*] A. Redden, R. J. Perkins, Prof. Dr. K. D. Moeller
Department of Chemistry, Washington University in St. Louis
St. Louis, MO 63130 (USA)
E-mail: moeller@wustl.edu
[**] We thank the National Science Foundation (grant number CHE-
1151121) for their generous support of our work. We also gratefully
acknowledge the Washington University High Resolution NMR
facility, partially supported by the NIH (grant numbers RR02004,
RR05018, and RR07155), and the Washington University Mass
Spectrometry Resource Center, partially supported by the NIH
(grant number RR00954), for their assistance.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201308739.
Angew. Chem. Int. Ed. 2013, 52, 12865 –12868
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12865

.
Angewandte
Communications
Scheme 3. The initial experiment.
passed through the cell. Initially, the reaction led to a mixture
of two main products (both a mixture of stereoisomers): the
expected cyclic acetal product 6 and a mixed acetal product 7
that was derived from 6. The mixed acetal was converted back
to the cyclic acetal with toluenesulfonic acid and 4 ꢀ
molecular sieves to afford an 82% isolated yield of the
desired 6 from substrate 5.
Clearly, inclusion of the second nucleophile did not
interfere with the success of the electrolysis in any way. But
did the reaction really lead to a radical intermediate like that
proposed in Scheme 2?
Scheme 5. The compatibility of the cyclization with alternative trapping
groups.
Insight into this question was gained by examining the
intermolecular trapping reaction illustrated in Scheme 4. The
oxidation was conducted using identical electrolysis condi-
leading to the formation of a cation. Formation of the cation
led to products 9, 10, and 11.
With a mechanistic premise in place, the scope of the
reactions was examined (Scheme 5). Three trapping groups
were selected for this study: an enol ether because the
coupling reaction would afford a chance to make a bis-acetal
product with the ends differentiated, a furan because furans
are synthetically very useful coupling partners for oxidative
cyclization reactions,[9] and a second alcohol because we
wondered if an intramolecular alcohol trapping group might
channel a reaction like the one shown in Scheme 4 to a single
product.
Scheme 4. Intermolecular trapping.
In each case, the electrolysis conditions were identical to
those used previously, and in each case the cyclization
proceeded nicely. In the first reaction [Scheme 5, Eq. (1)],
the oxidative cyclization led to an 85% yield of a product
having two distinct acetal groups. The acyclic acetal was then
converted in 85% yield to an aldehyde affording the new ring
with the two ends of the oxidative cyclization clearly differ-
entiated. The second cyclization [Scheme 5, Eq. (2)] showed
the compatibility of the approach with the use of a furan
coupling partner. As in the earlier reaction with substrate 5,
the cyclization led to a mixture of cyclic and acyclic acetal
products. The furan was oxidized and trapped with methanol
as we have seen in all previous furan-based cyclizations.[9] The
use of toluenesulfonic acid and 4 ꢀ molecular sieves again
converted the mixture to the desired cyclic acetal, while
simultaneously regenerating the aromatic furan ring.
The oxidative cyclizations resulting from substrates 18a
and 18b were interesting since one might anticipate that the
formation of a radical intermediate like 20 (Figure 1) would
preclude formation of the second ring and lead to hydrogen
atom abstraction products like 12 above. However, unlike the
trapping of an enol ether derived radical cation to form
a carbon-carbon bond that has been shown to give rise to
kinetic product formation,[10] the alcohol trapping of an enol
ether radical cation has been shown to be reversible.[4a]
tions to the reaction shown in Scheme 3 and led to the
formation of four products in an overall yield of 65%. The
products were a mixture of molecules that contained either
a five-membered ring acetal or a mixed acetal derived from
methanol opening of the five-membered ring cyclic acetal.
The formation of product 12 was consistent with an initial
cyclization to form a five-membered ring acetal derivative
analogous to intermediate 3 in Scheme 2 followed by hydro-
gen atom abstraction from solvent by the radical left at the b-
carbon of the radical cation. The possibility that the product
was derived from a simple methanolysis of the starting
material was ruled out because of the success of the oxidation
shown in Scheme 3. Substrates 5 and 8 were oxidized using
identical reaction conditions. It is highly unlikely that one
reaction led to methanolysis of the substrate and the other did
not. Instead, it is more likely that both reactions led to the
same radical intermediate. In one case (substrate 5), that
intermediate was presented with an efficient intramolecular
trapping group. The result was a high yield of cyclization. In
the other case (substrate 8), no intramolecular trapping group
was present. In the absence of a fast cyclization, the reaction
led to a competition between hydrogen atom abstraction and
the formation of product 12 and oxidation of the radical
12866 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 12865 –12868

Angewandte
Chemie
À788C (Scheme 6). Under these conditions, a 65% yield of
the desired cyclic product could be obtained. The isolated
stereoisomers of the product were assigned as having a cis
ring fusion based upon analogy to the earlier reactions that
coupled two enol ethers.[5a]
The temperature of the reaction was dropped because the
initial oxidation led to, along with the small amount of
product, an unidentifiable mixture of elimination and poly-
mer byproducts that we have come to recognize as the
hallmarks of radical cation decomposition. As mentioned
earlier, we have shown that the alcohol trapping of a radical
cation intermediate can be reversible.[4a] If this is the case for
the oxidation of 22, then intermediates 2 and 3 (Scheme 2)
would be in equilibrium with each other. In this way, the
presence of the second nucleophile would reduce the effective
concentration of radical cation 2, but it would still be present.
Subsequent radical cation decomposition would be slowed
but not completely avoided. Fortunately, the same studies that
showed alcohol-trapping reactions to be reversible also
showed them to be exothermic. The reactions could be
driven to the cyclic product with lower temperature.[4a] In the
same manner, we hoped that lowering the temperature for the
oxidation of 22 would push the alcohol trapping reaction
toward the formation of cyclic intermediate 3, reduce the
concentration of the radical cation, and more effectively
channel the reaction toward the radical cyclization pathway.
This turned out to be the case, and lowering the temperature
of the reaction did dramatically improve the cleanliness of the
transformation.
Figure 1. A potential equilibration of intermediates.
Hence, the formation of 20 can reverse, regenerating the
original radical cation 19 and giving rise to an opportunity to
equilibrate acetal 20 with cyclic ether 21.[11] The alternative
cyclization to form 21 would be a relatively fast pathway when
compared to the decomposition pathways available to the
intermolecular trapping reaction (Scheme 4). The formation
of cyclic ether 21 would place the radical next to an oxygen,
a scenario that would dramatically lower its oxidation
potential and give rise to formation of the bicyclic product.
If the oxidation is slower than the equilibration between 20
and 21, then product formation would be governed by the
Curtin–Hammett Principle.[6] This turned out to be the case,
and the reactions proved to be compatible with the formation
of both five- and a six-membered ring ether products in good
yield (Scheme 5).
Next, attention was turned toward a demonstration that
the method would allow us to overcome the problems
encountered earlier with slow cyclization reactions. For this
reason, substrate 22 (Scheme 6) was synthesized.[8] This
The success of the cyclization originating from the
oxidation of 22 relative to the reaction that originated from
the nearly identical methoxy enol ether substrate [Scheme 1,
Eq. (1)] further supports the suggestion that the cyclization
reaction involves radical 3 and helps to rule out an alternative
mechanism where a second oxidation step converts radical 3
to a cation prior to the cyclization reaction. While it is
certainly possible that a second oxidation step occurs prior to
the cyclization, it is unlikely that a radical cation intermediate
with cationic character at the b-carbon of the enol ether
would lead to cation-based elimination reactions [Scheme 1,
Eq. (1)] and no cyclization while the formation of a full cation
at the same b-carbon would lead to the complete opposite
selectivity.
Scheme 6. Application to a prior failed cyclization.
substrate was selected so that the new method could be
directly compared with a previous cyclization that had failed
[Scheme 1, Eq. (1)].
Oxidation of 22 using the conditions employed for the
cyclizations in Schemes 3 and 5 led to a small amount of
slightly impure product (ca. 10%) along with general
decomposition of the starting material. The result was
encouraging because oxidation of the methoxy enol ether
substrate under similar conditions led to none of the desired
product. The presence of the second nucleophile in 22 did
indeed push the reaction toward the desired direction.
Changing to reaction conditions used for the initial failed
cyclization attempts with the methoxy enol ether substrate
(K2CO3, 0.5m LiClO4 in 50% MeOH/THF) led to an
increased yield (ca. 25%).[5a] However, the isolated product
was again slightly impure and decomposition of starting
material was still observed.
In conclusion, we have found that trapping both ends of an
enol ether radical cation is an effective tool for completing
oxidative cyclization reactions. The chemistry expands the
utility of enol ether—enol ether coupling reactions by differ-
entiating the ends of the cyclization, is compatible with the
use of a variety of trapping groups, and provides a method to
accomplish previously unsuccessful cyclizations.
Received: October 7, 2013
Published online: November 19, 2013
The anodic cyclization resulting from substrate 22 could
be optimized by dropping the temperature for the reaction to
Keywords: anodic cyclizations · electrochemistry · radical ions ·
reactive intermediates
.
Angew. Chem. Int. Ed. 2013, 52, 12865 –12868
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12867

.
Angewandte
Communications
[4] See for example: a) J. M. Campbell, H.-C. Xu, K. D. Moeller, J.
Am. Chem. Soc. 2012, 134, 18338; b) F. Tang, K. D. Moeller,
Tetrahedron 2009, 65, 10863; c) Y. Sun, K. D. Moeller, Tetrahe-
dron Lett. 2002, 43, 7159.
[5] a) L. V. Tinao-Wooldridge, K. D. Moeller, C. M. Hudson, J. Org.
Chem. 1994, 59, 238; b) For similar results with a bridged bicyclic
system see: S. H. K. Reddy, K. Chiba, Y. Sun, K. D. Moeller,
Tetrahedron 2001, 57, 5183.
[1] For radical cation-initiated cyclizations derived from chemical
oxidations see: a) D. Crich, K. Ranganathan, S. Neelamkavil, X.
Huang, J. Am. Chem. Soc. 2003, 125, 7942; b) D. Crich, V. Shirai,
F. Brebion, S. Rumthao, Tetrahedron 2006, 62, 6501; c) D. Crich,
K. Ranganathan, J. Am. Chem. Soc. 2005, 127, 9924; d) D. Crich,
M. Shirai, S. Rumthao, Org. Lett. 2003, 5, 3767; e) J. C. Conrad, J.
Kong, B. N. Laforteza, D. W. C. MacMillan, J. Am. Chem. Soc.
2009, 131, 11640; f) N. T. Jui, E. C. Y. Lee, D. W. C. MacMillan, J.
Am. Chem. Soc. 2010, 132, 10015; g) S. Rendler, D. W. C.
MacMillan, J. Am. Chem. Soc. 2010, 132, 5027; h) D. S.
Hamilton, D. A. Nicewicz, J. Am. Chem. Soc. 2012, 134, 18577;
i) D. A. Wilger, N. J. Gesmundo, D. A. Nicewicz, Chem. Sci.
2013, 4, 3160.
[2] For general reviews of radical cation-initiated cyclizations
derived from electrochemical oxidations see: a) K. D. Moeller,
Tetrahedron 2000, 56, 9527; b) J. B. Sperry, D. L. Wright, Chem.
Soc. Rev. 2006, 35, 605; c) J. Yoshida, K. Kataoka, R. Horcajada,
A. Nagaki, Chem. Rev. 2009, 109, 2265.
[3] For a recent review see: K. D. Moeller, Synlett 2009, 8, 1208; For
earlier work see reference [2a]. For selected recent examples
see: a) Y. Ashikari, T. Nokami, J. Yoshida, Org. Biomol. Chem.
2013, 11, 3322; b) H. C. Xu, K. D. Moeller, Angew. Chem. 2010,
122, 8176; Angew. Chem. Int. Ed. 2010, 49, 8004; c) R. J. Perkins,
H.-C. Xu, J. M. Campbell, K. D. Moeller, Beilstein J. Org. Chem.
2013, 9, 1630; d) L. A. Anderson, A. Redden, K. D. Moeller,
Green Chem. 2011, 13, 1652.
[6] A. Redden, K. D. Moeller, Org. Lett. 2011, 13, 1678.
[7] G. Xu, K. D. Moeller, Org. Lett. 2010, 12, 2590, and references
therein.
[8] Details are included in the Supporting Information.
[9] For selected examples see: a) D. G. New, Z. Tesfai, K. D.
Moeller, J. Org. Chem. 1996, 61, 1578; b) D. L. Wright, C. R.
Whitehead, C. H. Sessions, I. Ghiviriga, D. A. Frey, Org. Lett.
1999, 1, 1535; c) J. Mihelcic, K. D. Moeller, J. Am. Chem. Soc.
2004, 126, 9106; d) C. C. Hughes, A. K. Miller, D. Trauner, Org.
Lett. 2005, 7, 3425; e) A. K. Miller, C. C. Hughes, J. J. Kennedy-
Smith, S. N. Gradl, D. Trauner, J. Am. Chem. Soc. 2006, 128,
17057; f) H. Wu, K. D. Moeller, Org. Lett. 2007, 9, 4599.
[10] D. A. Frey, S. H. K. Reddy, K. D. Moeller, J. Am. Chem. Soc.
1998, 120, 2805.
[11] Calculations suggest that deprotonation of the alcohol nucleo-
phile occurs during the transition state for the cyclization. See:
a) J. H. Horner, L. Bagnol, M. Newcomb, J. Am. Chem. Soc.
2004, 126, 14979; b) D. Mangion, D. R. Arnold, Acc. Chem. Res.
2002, 35, 297; as well as reference [4a]. For this reason, the
second proton is not included in products 20 and 21.
12868 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 12865 –12868