Directed γ-C(sp3)-H Alkylation of Carboxylic Acid Derivatives through Visible Light Photoredox Catalysis

Article abstract of DOI:10.1021/jacs.7b09306

Visible light photoredox catalysis enables direct γ- C(sp³)-H alkylation of saturated aliphatic carbonyl compounds. Electron-deficient alkenes are used as the coupling partners in this reaction. Distinguished site selectivity is controlled by the predominant 1,5-hydrogen atom transfer of an amidyl radical generated in situ.

Full text of DOI:10.1021/jacs.7b09306

Communication
pubs.acs.org/JACS
Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX-XXX
Directed γ‑C(sp3)−H Alkylation of Carboxylic Acid Derivatives
through Visible Light Photoredox Catalysis
Dian-Feng Chen,† John C. K. Chu,‡ and Tomislav Rovis*,†
†Department of Chemistry, Columbia University, New York, New York 10027, United States
‡Department of Chemistry, Colorado State University, Fort Collins, Colorado 80523, United States
S
* Supporting Information
ABSTRACT: Visible light photoredox catalysis enables
direct γ- C(sp3)−H alkylation of saturated aliphatic
carbonyl compounds. Electron-deficient alkenes are used
as the coupling partners in this reaction. Distinguished site
selectivity is controlled by the predominant 1,5-hydrogen
atom transfer of an amidyl radical generated in situ.
trategies that selectively functionalize unactivated C(sp3)−
S
H bonds in the presence of multiple other sites are of
intense current interest.1 Directed reactions are common in
this milieu, often via transition-metal catalysis.2 Among diverse
functionality capable of directing C−H activation, carbonyls
are attractive due to their ubiquity and versatility. β-
Functionalization of aliphatic carbonyl compounds has been
facilitated by robust organo-3 or transition-metal catalysis.4
Although some progress has been made, γ-C−H functional-
ization remains a formidable challenge. A literature survey
disclosed two strategies for addressing this issue. First, Corey,5
Chatani,6 Chen,7 and Yu8 pioneered primary γ-C−H arylation,
alkynylation, olefination, and intramolecular amination reac-
tions based on sterically controlled formations of six-
membered metallacycles (Figure 1a). Methylene and methine
groups are not amenable to these reactions. Second,
intramolecular hydrogen atom abstraction by an amidyl
Figure 1. Remote C(sp3)−H functionalization.
radical,9 in a 1,5-fashion (i.e., modified Hofmann−Loffler−
̈
Freytag reaction10), generates a γ-carbon-centered radical.
These radicals largely end up as halogenated11 or cyclized
product,12 due to high reactivity of the excess halogen
reagents within the systems.
used in our amine-directed functionalization, could not be
used for carboxyl-directed functionalization. At the outset of
this work, it was not clear what functionality would be
required since it would have to fulfill several criteria in the
successful generation/translocation of nitrogen radicals. It is
well documented16 that the high electrophilic character of the
amidyl radical9 is crucial for hydrogen atom abstraction.
Therefore, an appropriate protecting group (usually electron-
deficient) on nitrogen is required for three reasons: (1) to
acidify the N−H proton to enable mild deprotonation, (2) to
give the resulting amidyl anion reasonable oxidation potential
for single electron transfer (SET), and (3) to provide a higher
N−H bond dissociation energy (BDE)17 relative to that of
C(sp3)−H bonds, ultimately allowing for efficient abstraction.
An initial survey of common N-protecting groups on 4-
methylpentanamide revealed ethoxycarbonyl as being optimal
(see SI for details). Methyl methacrylate (MMA) was used as
Very recently, we13 and Knowles14 reported selective
C(sp3)−H bond abstraction and alkylation with electron-
deficient alkenes relying on photoredox catalysis15 with
selectivity governed by a 1,5-hydrogen atom transfer
(HAT)16 reaction (Figure 1b). The reacting center was thus
delivered from a pendant amine functionality which was
oxidized by the excited state of the photocatalyst. The net
transformation may be viewed as remote functionalization of
an aliphatic amine. With this success, we turned our attention
to remote functionalization of other aliphatic systems using
complementary functional groups to deliver selectivity.
We anticipated that the formation of an amide bond on a
carboxylic acid would afford an opportunity to create a
different N-centered radical after oxidation, but it still might
allow for selective 1,5-HAT to activate remote sites. The
immediate challenge was that a trifluoroacetamide, which we
Received: August 31, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.7b09306
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
a b
,
the alkene coupling partner, and Ir[(dF-CF3)ppy]2(dtbbpy)]-
PF6 ((dF-CF3)ppy = 2-(2,4-difluorophenyl)-5-(trifluoro-
methyl)pyridine; dtbbpy = 4,4′-di-tert-butyl-2,2′-bipyridine)18
was chosen as the photocatalyst. We were delighted to find
that the reaction of 1a in the presence of saturated aq. K3PO4
and blue light provided the desired product 3aa in 76% NMR
yield (Table 1, entry 1). Superior yield (74% isolated, entry 5)
Scheme 1. Alkene Scope
a
Table 1. Optimization
b
entry
base
K3PO4
solvent
yield (%)
1
2
3
4
DMF
DMF
DMF
DMF
76
66
0
Cs2CO3
quinuclidine
c
K3PO4
83
d
c
e
5
df
,
K3PO4
DMF/t-AmylOH (1:1)
DMF/t-AmylOH (1:1)
84(74 )
c
a
6
K3PO4
0
Conditions: Ir(dF-CF3) (2.0 mol%), K3PO4 (1.1 equiv), 1 (1.0
equiv), 2 (1.2−3.0 equiv), 1:1 mixed DMF/t-AmylOH, 34 W blue
a
Ir(dF-CF3) (2.0 mol%), aq. K3PO4 (2.0 equiv), 1a (1.0 equiv), 2a
b
b
LED, ∼40 °C, 12 h. Isolated yields are given.
(5.0 equiv), DMF, 34 W blue LED, ∼40 °C, 12 h. Yield determined
by H NMR using trimethoxybenzene as the internal standard. 1.1
equiv. 3.0 equiv of 2a. Isolated yield. No photocatalyst. Similarly,
there was no product formation when either light or base was
excluded.
c
1
d
e
f
alkene starting materials. A pharmacologically relevant γ-
aminobutyric acid derivative gives monoalkylated product 3kl
in moderate yield (38%). Unactivated secondary C−H bonds
remain a challenge (3hi, 14% yield). Next, amide substrates
possessing two or more potential hydrogen abstraction sites
were examined (Scheme 2c). Gratifyingly, 1,5-HAT process
was found to outcompete other pathways.19 As a result,
products 3lf−3rf are formed in moderate to good yields.
Functional groups such as chloride (3ee), BocN-H (3ne), and
acetal (3se) are also tolerated. A steroidal derivative with
multiple stereogenic centers and tertiary C−H bonds
exclusively furnishes product 3ue in moderate yield. It is
noteworthy that a glucose-derived radical intermediate
undergoes alkylation with modest diastere oselectivity (3tl).
We further interrogated the selectivity of 1,5-HAT when
presented with two distinct tertiary C−H bonds (eqs 1−3). In
all cases investigated, regioselectivity is high. With substrates
1v and 1w, HAT occurs on the C−H bond distal to the σ-
withdrawing oxygen functionality, presumably for electronic
reasons. Substrate 1x, on the other hand, has two electroni-
cally similar tertiary C−H bonds (eq 3). HAT is still
completely selective, delivering product 3xe in 54% yield. The
two competitive HAT reactions presumably proceed via
bicyclic transition states ([4.4.0] and [3.3.1]), with the former
being preferred for reasons of ring strain.20
was obtained by using 1.1 equiv of K3PO4 and 1:1 mixed
DMF/t-AmylOH. Control experiments suggested that the
photocatalyst, base, and light were all essential for a successful
γ-C−H alkylation reaction of carbamate 1a (entry 6).
With optimized conditions in hand, we next investigated
the alkene scope with amide 1a or 1b as the substrate
(Scheme 1). Reactions of 1a with acrylates bearing α-
substituents, including alkyl and aryl groups, proceed well,
delivering products 3ab−3ad in good yields (68−72%).
Acrylates without α-substituents couple with 1a providing
products 3ae and 3af in slightly lower yields, likely a
consequence of competitive aza-Michael addition as well as
oligomerization (see SI for details). Allyl acrylate engages in
reactivity as the desired product 3ag is exclusively obtained in
62% isolated yield. Methacrylonitrile is also a suitable partner
for this reaction (3ah, 62% yield). This protocol was then
successfully applied to other electron-deficient alkenes, such as
methyl vinyl ketone, acrylamide, vinyl sulfone, and vinyl
phosphonate, when amide 1b is used. Products 3bi− 3bl were
isolated in moderate to good yields. β-Substituents on more
electron-deficient alkenes (e.g., dimethyl maleate) result in an
incomplete reaction, as 3bm was obtained in only 47% yield.
The amide scope was examined next (Scheme 2).
Substrates with tertiary γ-C−H bonds (BDE ≈ 95 kcal/
mol) are well tolerated, giving alkylated products 3cf−3ge in
good yields (Scheme 2a). Tunable mono- or dialkylation of
substrates bearing activated secondary C−H bonds (BDE ≈
90 kcal/mol) is also possible (3ie−3je′). This could be
achieved by simply altering the ratio between the amide and
A possible mechanistic pathway is shown in Figure 2.
Deprotonation of 1a (pKa ≈ 11.0 in H2O, pKa ≈ 16.5 in
DMSO)21 with K3PO4 (pKa ≈ 12.4 in H2O)22 provides
potassium amide I. A Stern−Volmer experiment (see SI for
details) reveals the feasibility of single-electron oxidation of I
red
(E1/2 = +1.04 V vs saturated calomel electrode (SCE) in
DMF) by the excited state of Ir[(dF-CF3)ppy]2(dtbbpy)]PF6
B
DOI: 10.1021/jacs.7b09306
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
a b c
, ,
Scheme 2. Amide Scope
a
Conditions: Ir(dF-CF3) (2.0 mol%), K3PO4 (1.1 equiv), 1 (1.0 equiv), 2 (1.2−3.0 equiv), 1:1 mixed DMF/t-AmylOH (1.0 mL), 34 W blue LED,
b
c
d
N2, ∼40 °C, 12 h. Isolated yields are given. dr was determined by the crude 1H NMR. 1.2 equiv of amide and 1.0 equiv of benzyl acrylate were
used. 2.0 equiv of benzyl acrylate was used. 4.0 mmol% of Ir(dF-CF3) was used. Ir(dF-CF3) = Ir[(dF-CF3)ppy]2(dtbbpy)PF6.
e
f
18
(E1/2 [*IrIII/IrII] = +1.21 V vs SCE in MeCN), which
furnishes amidyl radical II. A subsequent energetically
favorable 1,5-HAT (N−H BDE ≈ 105.5 kcal/mol23 vs
tertiary γ-C−H BDE ≈ 95 kcal/mol) gives nucleophilic γ-
carbon radical III, which is then trapped by electron-deficient
alkenes (e.g., methyl methacrylate), leading to the formation
of a new carbon radical IV. Single-electron reduction of IV
red
(E1/2 = −0.73 to −0.59 V vs SCE in MeCN)24 by the IrII
red
species (E1/2red[IrIII/IrII] = −1.37 vs SCE in MeCN),18
followed by protonation, affords product 3aa, thereby also
regenerating the ground-state IrIII catalyst.25
In summary, we have developed a visible light photoredox-
catalyzed γ-alkylation of saturated aliphatic carboxylic acid
derivatives with a variety of electron-deficient alkenes. Efficient
in situ generation of the amidyl radical, and subsequent
translocation in a 1,5-fashion, ensures excellent site-selectivity.
We have further delineated substrate scope and shown that
synthetically useful levels of regioselectivity may be attained
for more complex substrates.
Figure 2. Proposed mechanism.
C
DOI: 10.1021/jacs.7b09306
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Rankic, D. A.; Martin, D. B. C.; MacMillan, D. W. C. Science 2013,
339, 1593.
(5) Reddy, B. V. S.; Reddy, L. R.; Corey, E. J. Org. Lett. 2006, 8,
3391.
(6) Ano, Y.; Tobisu, M.; Chatani, N. J. J. Am. Chem. Soc. 2011, 133,
12984.
(7) (a) He, G.; Zhang, S.-Y.; Nack, W. A.; Li, Q.; Chen, G. Angew.
Chem., Int. Ed. 2013, 52, 11124. (b) He, G.; Zhang, S.-Y.; Nack, W.
A.; Pearson, R.; Rabb-Lynch, J.; Chen, G. Org. Lett. 2014, 16, 6488.
(8) (a) Li, S.; Chen, G.; Feng, C.-G.; Gong, W.; Yu, J.-Q. J. Am.
Chem. Soc. 2014, 136, 5267. (b) Li, S.; Zhu, R.-Y.; Xiao, K.-J.; Yu, J.-
Q. Angew. Chem., Int. Ed. 2016, 55, 4317.
(9) (a) Zard, S. Z. Chem. Soc. Rev. 2008, 37, 1603. (b) Sutcliffe, R.;
Griller, D.; Lessard, J.; Ingold, K. U. J. Am. Chem. Soc. 1981, 103,
624.
(10) For a review, see: Wolff, M. E. Chem. Rev. 1963, 63, 55.
(11) Liu, T.; Myers, M. C.; Yu, J.-Q. Angew. Chem., Int. Ed. 2017,
56, 306.
(12) (a) Liu, T.; Mei, T.-S.; Yu, J.-Q. J. Am. Chem. Soc. 2015, 137,
5871. (b) Richers, J.; Heilmann, M.; Drees, M.; Tiefenbacher, K. Org.
Lett. 2016, 18, 6472. (c) Wappes, E. A.; Fosu, S. C.; Chopko, T. C.;
Nagib, D. A. Angew. Chem., Int. Ed. 2016, 55, 9974.
(13) Chu, J. C. K.; Rovis, T. Nature 2016, 539, 272.
(14) Choi, G. J.; Zhu, Q.; Miller, D. C.; Gu, C. J.; Knowles, R. R.
Nature 2016, 539, 268.
(15) For excellent reviews, see: (a) Prier, C. K.; Rankic, D. A.;
MacMillan, D. W. C. Chem. Rev. 2013, 113, 5322. (b) Skubi, K. L.;
Blum, T. R.; Yoon, T. P. Chem. Rev. 2016, 116, 10035.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.7b09306.
̌
̌
(16) Cekovic, Z. J. Serb. Chem. Soc. 2005, 70, 287.
́
■
(17) Relative homolytic bond dissociation energies (ΔBDEs)
compared to acetamide (pKa = 25.5 in DMSO, Eox(N−) = +0.73
V vs Ag/AgCl in MeCN): ΔBDEs(N−H) = 1.37ΔpKa(N−H) +
23.06ΔEox(N−). Bausch, M. J.; David, B.; Prasad, V.; Wang, L.-H.;
Vaughn, A. J. Phys. Org. Chem. 1992, 5, 1.
S
Experimental details, characterization data, and NMR
spectra (PDF)
(18) Lowry, M. S.; Goldsmith, J. I.; Slinker, J. D.; Rohl, R.; Pascal,
R. A., Jr.; Malliaras, G. G.; Bernhard, S. Chem. Mater. 2005, 17, 5712.
(19) For other pathways, see: (a) Nechab, M.; Mondal, S.;
Bertrand, M. P. Chem. - Eur. J. 2014, 20, 16034. (b) Chen, K.;
Richter, J. M.; Baran, P. S. J. Am. Chem. Soc. 2008, 130, 7247.
(20) Compare the strain energies of a [4.4.0] bicyclic system (cis-
decalin, 4.1 kcal/mol) with that of a [3.3.1] system (9.6 kcal/mol);
see: Greenberg, A.; Liebman, J. F. Strained Organic Molecules,
Organic Chemistry: A Series of Monographs 38; Academic Press:
New York, 1978; p 72.
(21) (a) Arnett, E. M.; Harrelson, J. A., Jr. J. Am. Chem. Soc. 1987,
109, 809. (b) Bordwell, F. G.; Harrelson, J. A., Jr.; Lynch, T.-Y. J.
Org. Chem. 1990, 55, 3337. (c) Bordwell, F. G.; Fried, H. J. Org.
Chem. 1991, 56, 4218.
(22) Chiong, H. A. Potassium Phosphate. e-EROS Encyclopedia of
Reagents for Organic Synthesis 2010, 10.1002/047084289X.rn01172
(23) Eox(I) = 1.18 V vs Ag/AgCl in DMF (see SI for details).
(24) Bortolamei, N.; Isse, A. A.; Gennaro, A. Electrochim. Acta
2010, 55, 8312.
AUTHOR INFORMATION
Corresponding Author
* tr2504@columbia.edu
ORCID
Tomislav Rovis: 0000-0001-6287-8669
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
■
We thank NIGMS (GM125206) for support. J.C.K.C.
acknowledges a Croucher Scholarship from the Croucher
Foundation, Hong Kong. We thank Matthew Shores (CSU)
for the use of his actinometer.
REFERENCES
(1) White, M. C. Science 2012, 335, 807.
(2) He, J.; Wasa, M.; Chan, K. S. L.; Shao, Q.; Yu, J.-Q. Chem. Rev.
2017, 117, 8754.
■
(25) A low quantum yield (6.3%) and an on/off study suggest that
the chain propagation mechanism is unlikely to be dominant in this
reaction; see SI.
(3) Fu, Z.; Xu, J.; Zhu, T.; Leong, W. W. Y.; Chi, Y. R. Nat. Chem.
2013, 5, 835.
(4) For excellent examples, see: (a) Giri, R.; Maugel, N.; Li, J.-J.;
Wang, D.-H.; Breazzano, S. P.; Saunders, L. B.; Yu, J.-Q. J. Am.
Chem. Soc. 2007, 129, 3510. (b) Wasa, M.; Engle, K. M.; Yu, J.-Q. J.
Am. Chem. Soc. 2009, 131, 9886. (c) He, J.; Li, S.; Deng, Y.; Fu, F.;
Laforteza, B. N.; Spangler, J. E.; Homs, A.; Yu, J.-Q. Science 2014,
343, 1216. (d) Zhang, F.-L.; Hong, K.; Li, T.-J.; Park, H.; Yu, J.-Q.
Science 2016, 351, 252. (e) Chen, G.; Gong, W.; Zhuang, Z.; Andra,
̈
M. S.; Chen, Y.-Q.; Hong, X.; Yang, Y.-F.; Liu, T.; Houk, K. N.; Yu,
J.-Q. Science 2016, 353, 1023. (f) Wu, Q.-F.; Shen, P.-X.; He, J.;
Wang, X.-B.; Zhang, F.; Shao, Q.; Zhu, R.-Y.; Mapelli, C.; Qiao, J. X.;
Poss, M. A.; Yu, J.-Q. Science 2017, 355, 499. (g) Pirnot, M. T.;
D
DOI: 10.1021/jacs.7b09306
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX