Oxalates as Activating Groups for Alcohols in Visible Light Photoredox Catalysis: Formation of Quaternary Centers by Redox-Neutral Fragment Coupling

Article abstract of DOI:10.1021/jacs.5b07678

Alkyl oxalates are new bench-stable alcohol-activating groups for radical generation under visible light photoredox conditions. Using these precursors, the first net redox-neutral coupling of tertiary and secondary alcohols with electron-deficient alkenes is achieved.

Full text of DOI:10.1021/jacs.5b07678

Communication
pubs.acs.org/JACS
Oxalates as Activating Groups for Alcohols in Visible Light
Photoredox Catalysis: Formation of Quaternary Centers by Redox-
Neutral Fragment Coupling
Christopher C. Nawrat,† Christopher R. Jamison,‡ Yuriy Slutskyy,‡ David W. C. MacMillan,*,†
and Larry E. Overman*,‡
†Merck Center for Catalysis at Princeton University, Princeton, New Jersey 08544, United States
‡Department of Chemistry, University of California, Irvine, California 92697-2025, United States
S
* Supporting Information
sensitivity of the N-phthalimidoyl oxalate moiety can present
ABSTRACT: Alkyl oxalates are new bench-stable alcohol-
activating groups for radical generation under visible light
photoredox conditions. Using these precursors, the first
net redox-neutral coupling of tertiary and secondary
alcohols with electron-deficient alkenes is achieved.
complications for purification of the requisite intermediates
because N-phthalimidoyl oxalates are not stable to aqueous
workup or flash chromatography. As such, identifying an
activating group for tertiary alcohols that is both stable and
capable of generating and coupling radicals in a redox-neutral
fashion (i.e., without the need for supplementary reductants or
oxidants) would be a desirable goal. The MacMillan group
reported a net redox-neutral method to generate carbon radicals
and couple them with Michael acceptors by visible-light-
promoted decarboxylation of carboxylic acids using the photo-
catalyst Ir[dF(CF3)ppy]2(dtbbpy)PF6 (1) (eq 2).6 In contrast to
the reductive coupling of N-phthalimidoyl oxalates, this method
produces no byproduct other than CO2. We questioned whether
it would be possible to combine the salient features of both the
Overman and MacMillan radical generation methods to identify
a new activating-group strategy for tertiary alcohols that proceeds
via a redox-neutral manifold using stable, easily handled
intermediates. Here we report that simple oxalate salts of tertiary
alcohols undergo high-yield photocatalytic coupling with
electron-deficient alkenes in the presence of 1 and visible light
(eq 3).
ucleophilic tertiary carbon radicals are useful intermediates
N
for combining structurally complex carbon fragments with
the formation of new quaternary carbons.1 These alkyl radicals
are formed most commonly from halide precursors; however,
competing elimination and rearrangement reactions can under-
mine the preparation of structurally complex tertiary halides.2 In
contrast, tertiary alcohols would be ideal precursors of tertiary
radicals because they can be prepared by various reliable methods
and are widely commercially available. Inspired by Barton’s
introduction of tert-alkyl N-hydroxypyridine-2-thionyl oxalates
for generating carbon radicals from alcohols,2b Overman et al.
recently introduced N-phthalimidoyl oxalate derivatives of
tertiary alcohols for the reductive coupling of tertiary radicals
with Michael acceptors using visible light and [Ru(bpy)3](PF6)2
(eq 1).3−5 Though this method was shown to possess a wide
substrate scope, its mechanism necessitates the use of a
stoichiometric reductant to produce the tertiary alkyl radical
and forms phthalimide as a byproduct. Also, the inherent
The mechanistic details of the proposed coupling reaction are
outlined in Scheme 1. Irradiation of heteroleptic photocatalyst
Ir[dF(CF3)ppy]2(dtbbpy)PF6 (1) [dF(CF3)ppy = 2-(2,4-
difluorophenyl)-5-trifluoromethylpyridine, dtbbpy = 4,4′-di-t-
Bu-2,2′-bipyridine] with visible light leads to the formation of a
long-lived (τ = 2.3 μs) excited state *IrIII 2, which is a strong
oxidant (E1/2red [*IrIII/IrII] = +1.21 V vs SCE in CH3CN).7 On
the basis of this, we hypothesized that oxidation of the conjugate
red
base of 3 (E1/2 = +1.28 V vs SCE in CH3CN for t-
BuOCOCO2Cs)8 by *IrIII (2) via single-electron transfer
(SET) should be thermodynamically feasible, generating 4
following the stepwise loss of two molecules of CO2.2b,3a
Nucleophilic C-centered radical 4 should rapidly undergo
addition to electron-deficient alkenes such as 5. Finally, we
red
expected that reduction of resulting adduct radical 6 (E1/2
=
−0.59 to −0.73 V vs SCE in MeCN)9 by SET from available IrII
7
species 3 (E1/2 [IrIII/IrII] = −1.37 V vs SCE in CH3CN)
red
should yield 7 after protonation and regenerate ground-state
photocatalyst 1, completing the proposed catalytic cycle.
Received: July 22, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.5b07678
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
substrates containing sensitive functional groups. Fortunately, it
was found that the preformed Cs salts of the starting acids were
also competent in the reaction (entry 5). In contrast to the parent
acid (and, indeed, to most activating groups for tertiary alcohols
used for radical generation),2,3 alkyl cesium oxalates were bench-
stable, nonhygroscopic, and easy to isolate and handle.11 Adding
10 equiv of water was found to be beneficial when utilizing the
preformed oxalate salt, giving the coupled product in an excellent
95% yield (entry 6). Presumably, water both assists in
solubilizing oxalate salt and provides a proton source to quench
the intermediate cesium enolate after radical coupling and
reduction. Additionally, oxalates bearing other alkali counterions,
such as Li (entry 7), performed comparably to cesium oxalates in
the reaction.12 The use of a 26 W CFL bulb in place of a 34 W
blue LED lamp resulted in a diminished but still useful yield
(entry 8). Finally, it was observed that control experiments run in
the absence of photocatalyst or a visible light source did not
generate any of the desired 1,4-addition product (entries 9 and
10).
Scheme 1. Proposed Mechanism for Redox-Neutral Radical-
Coupling Reaction Using Alkyl Oxalates and Michael
Acceptors
Having identified optimal conditions, we examined the scope
of the acceptor component. As shown in Table 2, a wide range of
a
Table 2. Acceptor Scope with Alkyl Cesium Oxalate 9
Table 1. Initial Studies and Reaction Optimization
a
entry
X
solvent
DMF
base
yield (%)
1
2
3
4
5
H
K2HPO4
CsF
64
74
82
91
65
95
93
66
0
H
DMF
H
DME
CsF
H
3:1 DME/DMF
3:1 DME/DMF
3:1 DME/DMF
3:1 DME/DMF
3:1 DME/DMF
3:1 DME/DMF
3:1 DME/DMF
CsF
Cs
Cs
Li
none
none
none
none
none
none
b
6
b
7
bc
,
8
Cs
Cs
Cs
bd
,
9
be
,
10
0
a
Reactions on a 0.2 mmol scale using 1.0 equiv of acceptor and 1.1
equiv of oxalate. Yields determined by 1H NMR using mesitylene as an
internal standard. Water added = 10 equiv. Reaction carried out with
26 W CFL. Reaction carried out without photocatalyst. Reaction
carried out in the absence of light.
b
c
d
e
a
We first explored the proposed decarboxylative alkylation
reaction using the alkyl hydrogen oxalate derived from 1-methyl-
1-cyclohexanol as the radical precursor in the presence of benzyl
acrylate as an archetypal Michael acceptor (Table 1). Using 1 as
photocatalyst and dipotassium phosphate as base,6 we were
delighted to obtain the desired product in moderate yield (entry
1). Further optimization revealed CsF to be a more competent
base and a 3:1 mixture of DME/DMF to be the optimal solvent
(entries 2−4).
Though tert-alkyl hydrogen oxalates clearly function as viable
radical precursors, many were observed to be intrinsically
unstable species that disproportionate into a mixture of the
corresponding dialkyl oxalate and oxalic acid, even during storage
at −18 °C.10 Also, it was apparent that the presence of a highly
acidic hydrogen oxalate would likely preclude the preparation of
Isolated yields using optimized conditions from Table 1 with 1.0
equiv of acceptor and 1.1 equiv of oxalate (Supporting Information).
b
c
Carried out with 1.5 equiv of cesium oxalate. Carried out in 100%
d
e
f
DME. Run at 22 °C. Isolated as trans isomer. Isolated as cis isomer.
electron-deficient alkenes can be used in the reaction. As
expected, various acrylates were capable acceptors in the reaction
(10 and 11, 88 and 86% yield, respectively); the presence of α
substitution was well-tolerated (17−20, 85−96% yield). Also,
α,β-unsaturated acids can be used as coupling partners, owing to
the low basicity of the oxalate salt (20, 85% yield). This
procedure could be applied to a range of other electron-deficient
alkenes (e.g., enones, enals, acrylamides, vinyl phosphonates, and
vinyl sulfones; 12−16, 68−86% yield). Surprisingly, acrylonitrile
produced little product (11% yield),13 whereas methacrylonitrile
B
DOI: 10.1021/jacs.5b07678
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
a
Table 3. Scope of Cesium Oxalate Salts with Benzyl Acrylate as the Acceptor
a
b
Isolated yields using optimized conditions from Table 1 with 1.0 equiv of acceptor and 1.1 equiv of oxalate (Supporting Information). Carried out
c
with 1.5 equiv of oxalate. Using 100% DME.
proved to be a much more capable acceptor (17, 85% yield).
Substitution at the β position was tolerated for more electron-
deficient alkenes such as dimethyl fumarate and dimethyl
ethylidene- and benzylidenemalonate, furnishing the expected
adducts with good efficiency (21−23, 70−99% yield). As
expected, 4-vinylfuran-2-one gave exclusively the 1,6-addition
product in excellent yield (24, 89% yield). In the case of
acceptors harboring existing stereogenic centers, high levels of
diastereoselectivity were obtained (25 and 26, 73−90% yield, >
20:1 dr).
Next, we investigated the scope of cesium oxalate (Table 3).
Owing to the long forming bond (2.2−2.5 Å) in the transition
state of carbon radical conjugate addition14 and the poor
solvation of carbon radicals,1b the reaction proved quite
insensitive to steric hindrance around the site of radical
generation, with adjacent isopropyl and tert-butyl groups not
greatly reducing the efficiency of the reaction (29 and 31, 73−
93% yield). Cyclopentanol-derived oxalates also underwent
coupling in good yield (33 and 39, 85−92% yield), but very low
conversion was observed for 1-methylcyclopropanol- and 1-
methylcyclobutanol-derived oxalates. Heterocycles (e.g., pyrro-
lidines, piperidines, tetrahydrofurans, pyridines, and indoles)
were well-tolerated in the reaction (35, 37, 41, 59, and 63, 54−
77% yield). Underscoring the utility of this method for
constructing quaternary stereocenters in complex molecules,
natural product-derived oxalates also performed well, with good
levels of diastereoselectivity being observed (41, 43, 45, 47, and
49, 85−96% yield). Indeed, high yields were obtained even for
the formation of vicinal quaternary stereocenters (47, 85%
yield). Also, a number of acyclic tert-alkyl oxalates also undergo
the coupling with high levels of efficiency (51−63, 54−93%
yield).
The reaction was examined with several secondary cesium
oxalates; two representative examples are shown in Table 4.
Although still synthetically useful, lower yields of coupled
products were obtained with these substrates, and the product of
trapping of the intermediate alkoxyacyl radical was also isolated
(65 and 66). For more stabilized benzylic radicals, these side
products were not observed; the yields remained moderate (68).
a
Table 4. Examples of Secondary Oxalate Salts
a
Isolated yields using optimized conditions from Table 1 with 1.0
equiv of acceptor and 1.5 equiv of oxalate (Supporting Information).
C
DOI: 10.1021/jacs.5b07678
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Scheme 2. Six-Step Synthesis of a trans-Clerodane Natural
Product
REFERENCES
■
(1) (a) Giese, B. Angew. Chem., Int. Ed. Engl. 1983, 22, 753. (b) Barton,
D. H. R.; Sas, W. Tetrahedron 1990, 46, 3419. (c) Russell, G. A.; Shi, B.
Z.; Jiang, W.; Hu, S.; Kim, B. H.; Baik, W. J. Am. Chem. Soc. 1995, 117,
3952. (d) Togo, H.; Taguchi, R.; Yamaguchi, K.; Yokoyama, M. J. Chem.
Soc., Perkin Trans. 1 1995, 2135. (e) Schnermann, M. J.; Overman, L. E.
Angew. Chem., Int. Ed. 2012, 51, 9576. (f) Lo, J. C.; Yabe, Y.; Baran, P. S.
J. Am. Chem. Soc. 2014, 136, 1304.
(2) In addition to halides, a number of activating groups have been
reported that allow tertiary alcohols to be used as radical precursors. For
existing methods that employ oxalate or xanthate derivatives, see
(a) Dolan, S. C.; MacMillan, J. J. Chem. Soc., Chem. Commun. 1985,
1588. (b) Barton, D. H. R.; Crich, D. J. Chem. Soc., Perkin Trans. 1 1986,
1603. (c) Coppa, F.; Fontana, F.; Lazzarini, E.; Minisci, F.; Pianese, G.;
Zhao, L. Chem. Lett. 1992, 1295. (d) Crich, D.; Fortt, S. M. Synthesis
1987, 1987, 35. (e) Coppa, F.; Fontana, F.; Lazzarini, E.; Minisci, F.
Chem. Lett. 1992, 1299. (f) Togo, H.; Matsubayashi, S.; Yamazaki, O.;
Yokoyama, M. J. Org. Chem. 2000, 65, 2816.
(3) (a) Lackner, G. L.; Quasdorf, K. W.; Overman, L. E. J. Am. Chem.
Soc. 2013, 135, 15342. (b) Lackner, G. L.; Quasdorf, K. W.; Pratsch, G.;
Overman, L. E. J. Org. Chem. 2015, 80, 6012.
(4) For recent reviews on photoredox catalysis, see (a) Tucker, J. W.;
Stephenson, C. R. J. J. Org. Chem. 2012, 77, 1617. (b) Prier, C. K.;
Rankic, D. A.; MacMillan, D. W. C. Chem. Rev. 2013, 113, 5322. (c) Xie,
J.; Jin, H.; Xu, P.; Zhu, C. Tetrahedron Lett. 2014, 55, 36. (d) Schultz, D.
M.; Yoon, T. P. Science 2014, 343, 1239176.
(5) For selected reviews on radical conjugate additions in general, see
(a) Renaud, P.; Gerster, M. Angew. Chem., Int. Ed. 1998, 37, 2562.
(b) Sibi, M. P.; Porter, N. A. Acc. Chem. Res. 1999, 32, 163. (c) Sibi, M.
P.; Manyem, S.; Zimmerman, J. Chem. Rev. 2003, 103, 3263.
(d) Srikanth, G. S. C.; Castle, S. L. Tetrahedron 2005, 61, 10377.
(6) Chu, L.; Ohta, C.; Zuo, Z.; MacMillan, D. W. C. J. Am. Chem. Soc.
2014, 136, 10886.
(7) Lowry, M. S.; Goldsmith, J. L.; Slinker, J. D.; Rohl, R.; Pascal, R. A.;
Malliaras, G. G.; Bernhard, S. Chem. Mater. 2005, 17, 5712.
(8) See the Supporting Information for further details about CV
measurements.
(9) (a) Bortolamei, N.; Isse, A. A.; Gennaro, A. Electrochim. Acta 2010,
55, 8312. (b) See the Supporting Information for evidence of an enolate
intermediate.
(10) This disproportionation is unique to tert-alkyl hydrogen oxalates.
Alkyl hydrogen oxalates derived from primary and secondary alcohols
are stable and do not readily disproportionate.
(11) The cesium oxalates were most conveniently prepared via
hydrolysis of the corresponding tert-alkyl methyl oxalates with aqueous
CsOH. This route avoids the potentially unstable alkyl hydrogen oxalate
moiety; the tert-alkyl methyl oxalate intermediates are stable to silica
column chromatography and aqueous work up. The cesium oxalates
may also be prepared in a simple one-pot procedure directly from
tertiary alcohols (Supporting Information).
(12) Lithium, sodium, potassium, and cesium oxalates all gave similar
results (Supporting Information). However, cesium oxalates were
chosen for development under the assumption that they were likely to
have favorable physical properties over a wide range of oxalate
substrates.
The reaction also enabled a short synthesis of 70,15 a member
of the trans-clerodane family of natural products (Scheme 2).16
Activation of the tertiary alcohol of known intermediate 6917 is
particularly challenging because the trans-decalin ring system
places the tertiary alcohol in a 1,3-diaxial relationship with the
angular methyl substituent. This severe steric interaction had
previously prevented the preparation of the N-phthalimidoyl
oxalate derivative.17 However, acylation of 69 with methyl
chlorooxoacetate proceeded in excellent yield. In situ hydrolysis
with aqueous CsOH allowed pure cesium oxalate 44 to be
isolated in one step and high yield from alcohol 69 without the
use of chromatography. Coupling of oxalate 44 (1.5 equiv) with
commercially available 4-vinylfuran-2-one (1.0 equiv) proceeded
with perfect diastereo- and regioselectivity in 98% yield to give
trans-clerodane 70, a natural product that is a versatile precursor
of many other members of the trans-clerodane family.17
We have developed a new visible light photoredox-catalyzed
method for the generation of alkyl radicals from secondary and
tertiary alcohols and shown its use in the redox-neutral formation
of quaternary carbon centers through alkylation with electron-
deficient alkenes. The intermediate alkyl cesium oxalates are
bench-stable, easily handled, and provide a notably convenient
means to activate alcohols for radical generation.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.5b07678.
■
S
Experimental procedures and data. (PDF)
AUTHOR INFORMATION
■
(13) Rapid reaction of the coupled radical with additional acrylonitrile
is likely responsible for the low yield in this case.
Corresponding Authors
*dmacmill@princeton.edu
*leoverma@uci.edu
(14) (a) Damm, W.; Giese, B.; Hartung, J.; Hasskerl, T.; Houk, K. N.;
Hueter; Zipse, H. J. Am. Chem. Soc. 1992, 114, 4067. (b) Arnaud, R.;
Postlethwaite, H.; Barone, V. J. Phys. Chem. 1994, 98, 5913.
(15) Hara, N.; Asaki, H.; Fujimoto, Y.; Gupta, Y.; Singh, A.; Sahai, M.
Phytochemistry 1995, 38, 189.
(16) For general information on the clerodane terpenoids, see
(a) Merritt, A. T.; Ley, S. V. Nat. Prod. Rep. 1992, 9, 243.
(b) Tokoroyama, T. Synthesis 2000, 2000, 611.
Author Contributions
C.C.N., C.R.J., and Y.S. contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
(17) Muller, D. S.; Untiedt, N. L.; Dieskau, A. P.; Lackner, G. L.;
̈
Financial support at UC Irvine was provided by the NSF
(CHE1265964) and the NIHGMS (R01-GM098601) and at
Princeton by the NIHGMS (R01-GM078201).
Overman, L. E. J. Am. Chem. Soc. 2015, 137, 660.
D
DOI: 10.1021/jacs.5b07678
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX