Metal-Free-Visible Light C-H Alkylation of Heteroaromatics via Hypervalent Iodine-Promoted Decarboxylation

Article abstract of DOI:10.1021/acs.orglett.8b01085

A metal-free photoredox C-H alkylation of heteroaromatics from readily available carboxylic acids using an organic photocatalyst and hypervalent iodine reagents under blue LED light is reported. The developed methodology tolerates a broad range of functional groups and can be applied to the late-stage functionalization of drugs and drug-like molecules. The reaction mechanism was investigated with control experiments and photophysical experiments as well as DFT calculations.

Full text of DOI:10.1021/acs.orglett.8b01085

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Metal-Free-Visible Light C−H Alkylation of Heteroaromatics via
Hypervalent Iodine-Promoted Decarboxylation
Julien Genovino,*,† Yajing Lian,† Yuan Zhang,† Taylor O. Hope,‡ Antoine Juneau,‡ Yohann Gagne,‡
́
Gajendra Ingle,† and Mathieu Frenette*,‡
†Pfizer Worldwide Medicinal Chemistry, Eastern Point Road, Groton, Connecticut 06340, United States
‡Dep
́ ́ ́ ̀ ́ ́ ́
artement de Chimie, Universite du Quebec a Montreal, Case Postale 8888, Succursale Centre-Ville, Montreal, Quebec, H3C
3P8, Canada
S
* Supporting Information
ABSTRACT: A metal-free photoredox C−H alkylation of heteroaromatics from readily available carboxylic acids using an
organic photocatalyst and hypervalent iodine reagents under blue LED light is reported. The developed methodology tolerates a
broad range of functional groups and can be applied to the late-stage functionalization of drugs and drug-like molecules. The
reaction mechanism was investigated with control experiments and photophysical experiments as well as DFT calculations.
Scheme 1. Hypervalent Iodine Dicarboxylates in Minisci
Transformations
he C−H alkylation of heteroaromatics remains a
Tformidable challenge in synthetic organic chemistry.
Such functionalization is nonetheless remarkably useful, and it
has become one of the most sought after strategies for the rapid
late-stage diversification of pharmaceuticals.1 The most
ubiquitous method for C−H alkylation of heteroaromatics
was pioneered by Minisci who used carboxylic acids to generate
alkyl radicals in the presence of silver nitrate, ammonium
persulfate, and sulfuric acid at elevated temperatures.2 Due to
the harsh conditions, variable yields, and practical limitations in
library format, the past decade has seen a surge in alternative
thermal reaction conditions showcasing a variety of coupling
partners.3
Photoredox catalysis has become a powerful methodology in
synthetic organic chemistry.4 New photoredox Minisci
conditions that take advantage of the mild generation of alkyl
radicals at room temperature have arisen.5 These protocols
however still display recurrent limitations such as the use of
expensive metal photocatalysts (iridium or gold),5b,c,e−j a large
excess of coupling partner (up to 10 equiv), and the
requirement to prepare either the coupling partner5d,e,g,i,j or
the photocatalyst.5f To our surprise, no photoredox protocol
had been developed for carboxylic acids, a broadly available
monomer set. During the preparation of this manuscript,
however, decarboxylative couplings were reported5h,l,m using,
for instance, an iridium catalyst in the presence of a persulfate
oxidant under visible light. Instead, our method uses the more
soluble and under-utilized hypervalent iodine reagents.6,7 These
reagents were utilized in Minisci transformation under refluxing
benzene or ultraviolet light (mercury lamp, Scheme 1a) albeit
with a large excess of the heteroaromatic and the carboxylic acid
coupling partners (3−6 equiv).7c
Herein, we report the first metal-free and visible light-
promoted C−H alkylation of heteroaromatics using hyper-
valent iodine dicarboxylates at room temperature. Our method
leverages the broad range of commercially available carboxylic
acids and a catalytic amount of an organic dye: 9-mesityl-10-
methyl acridinium (1, MesAcr, Scheme 1b).5d,8,9 Our
mechanistic studies suggest a mechanism that is distinct from
recent literature.5h,m
At the outset of our studies, we selected lepidine (2) as the
limiting substrate (1 equiv) and isobutyric acid (3, 5 equiv) to
Received: April 6, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b01085
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
optimize the reaction conditions under blue LED light
irradiation (λmax = 455 nm). Bis(acetoxy)iodobenzene (BAIB,
3 equiv) was used as a starting point along with the use of
trifluoroacetic acid (10 equiv) in acetonitrile (Table 1; see
Scheme 2. Carboxylic Acid Scope
Table 1. Reaction Conditions Optimization
a
General conditions: lepidine 2 (1 equiv), isobutyric acid 3 (3 equiv),
oxidant (2 equiv), MeCN (0.1 M), 32 W blue LED light, rt, 16 h.
b
c
TFA (5 equiv) was added. NMR yields using mesitylene as external
d
e
standard. Isolated yields are shown in parentheses. No TFA was
a
b
added.
51% recovered starting material. 40% recovered starting material.
c
d
AcOH was used as solvent, and PhI(OAc)2, as oxidant. MeOH was
e
used as solvent. 40% recovered starting material and less than 10% of
Supporting Information (SI) for the full optimization study).
Our initial screening tested a range of well-known metal
photocatalysts (iridium-, ruthenium-, and copper-based). We
were pleased to see the formation of desired isopropylated
lepidine 4a in 29% yield with Ru(bpz)3(PF6)2 (entry 1).
The conversion of lepidine improved significantly with
iridium catalysts (up to 78%), even though incomplete
conversion remained an issue (entry 2). We then turned our
attention to organic dyes (fluorescein, Eosin Y, and acridinium
derivatives), and we were pleased to find that 9-mesityl-10-
methyl acridinium 1 (MesAcr) increased the yield to 86%
(entry 3). In order to avoid the competing methylation
pathway, due to the radical decomposition of the acetoxy-
ligated iodine reagent, we explored other oxidants such as
oxone, TBHP, persulfate, and other hypervalent iodine
derivatives (entries 4−7). Iodosobenzene led to poor
conversion (9%, entry 4) while IBX derivative 5 improved
the yield to 91% (entry 5). However, the poor solubility of the
latter in acetonitrile and the requirement for its preparation
made this oxidant a less attractive option. As expected,
PhI(OCOt-Bu)2 led to a mixture of the desired product (4a)
and tert-butylated lepidine analogue (4b; see Scheme 2) in 48%
and 52% yields, respectively. Finally, we decided to use
bis(trifluoroacetoxy)iodo benzene (PIFA) to avoid the
complication of forming undesired alkylated products and to
bypass adding additional trifluoroacetic acid to the reaction
mixture. This modification led to full conversion of lepidine to
isopropylated lepidine 4a in 94% isolated yield (97% GC yield,
entry 7). Solvent screening confirmed that the best results were
obtained with acetonitrile.
alkylated product were observed.
isobutyric acid, other α-branched carboxylic acids that form
secondary alkyl radicals reacted smoothly under the reaction
conditions. Cyclobutyl, cyclopentyl, and cyclohexyl analogues
(4c−e) were isolated in 81%, 73%, and 72% yields, respectively.
In contrast, addition of the smaller cyclopropyl group suffered
from poor conversion and afforded cyclopropylated lepidine 4f
in a low 33% yield with 38% of unreacted lepidine. This
reactivity is consistent with the known instability of cyclopropyl
radicals, hence resulting in the unproductive consumption of
the carboxylic acid coupling partner. Cyclic ethers (tetrahy-
drofuran 4g and tetrahydropyrane 4h) were installed efficiently
(80%, 86%) while the presence of an unsaturation (cyclo-
pentenyl) decreased the yield slightly (4i, 56% versus 4d, 73%).
Finally, an acyclic secondary radical also reacted well to form 4j
(85%). Tertiary alkyl groups such as methyl cyclobutyl (4k),
methyl oxetane (4l), tert-butyl (4b), and adamantyl (4m) were
incorporated to lepidine in similar good yields. As expected
from the superior relative stability of tertiary radicals over
secondary radicals, the methyl cyclopropyl group was installed
in an increased 59% isolated yield (versus 33% for 4f) along
with 40% recovered lepidine.
Finally, the more challenging linear carboxylic acids that form
primary radicals were used successfully, leading to excellent
yields considering their high reactivity. A solvent and oxidant
swap (acetonitrile to acetic acid; PIFA to BAIB) allowed us to
maximize the formation of methylated lepidine 4o obtained in
an excellent 43% yield. In addition, our reaction conditions in
acetonitrile worked smoothly for longer chains: ethyl (4p,
76%), phenyl ethyl (4q, 65%), and phenyl propyl (4r, 67%). It
is worth noting that the good yield of 4q demonstrates the
advantage of using the carboxylic acid monomer set since other
monomers such as homoallylic bromides or iodides used by
With the optimized conditions in hand, we explored the
substrate scope with commercially available carboxylic acids
that were selected based on the type of alkyl radicals they
generate (e.g., primary, secondary, tertiary radicals; Scheme 2).
Consistent with our preliminary results obtained with
B
DOI: 10.1021/acs.orglett.8b01085
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
recent photoredox protocols decompose readily under the
reaction conditions to form styrenes.5a,c,d As expected, the more
electrophilic difluoromethyl radical reacted in a lower yield (4s,
26%). Another solvent swap (acetonitrile to methanol) enabled
hydroxymethylation to occur smoothly (4t, 66%). We
hypothesized that the alkyl radical generated from isobutyric
acid abstracted the methanol α-oxy proton by analogy to the
mechanism reported by DiRocco in the photocatalyzed
hydroxymethylation of heteroarenes.5c,10 Finally, acylation of
lepidine was exemplified by using 2-oxo-2-(tetrahydro-2H-
pyran-4-yl)acetic acid to give ketone 4u in 35% isolated yield.
We then selected a range of simple and complex electron-
deficient heteroaromatics in order to explore functional group
tolerance (Scheme 3).
(5f, 5g), phthalazines (5h, 5i), pyrimidine (5j), and pyrazine
(5k) were also obtained in good yields. Three drugs
(voriconazole, varenicline, and quinine) were selected to test
our protocol as a late-stage C−H functionalization. To our
delight, the isopropyl analogue (5l) of antifungal voriconazole
was isolated in 77% isolated yield along with 11% of the bis-
alkylated drug analogue (not shown; second alkylation
occurred at the C2 position of the pyrimidine). The conversion
of the trifluoroacetamide precursor of varenicline, a smoking
cessation drug, successfully gave isopropylated analogue 5m in
17% yield. The antimalarial drug quinine also yielded the
desired alkylated product (5n) in 76% yield. Together, these
results demonstrate that this catalytic system tolerates a variety
of functional groups such as secondary alcohols, olefins, tertiary
amines, aryl ethers, and other heterocycles.
Unlike the photoredox Minisci mechanism proposed for
iridium or ruthenium photocatalysts,5h,m the photoactivated
catalyst *(MesAcr+) is unable to reduce the hypervalent iodine
Scheme 3. Scope of the Heteroaromatics and Drugs
dicarboxylate adduct due to its high redox potential (E*red
>
+2.0 V), leaving the origin of this reduction elusive. To better
understand the reactivity of *(MesAcr+) under the reaction
conditions, we carried out a series of steady-state and time-
resolved Stern−Volmer quenching experiments (see the SI).
While the direct oxidation of an alkyl carboxylate by
*(MesAcr+) under basic conditions is well precedented,11 it
has not been reported under acidic conditions and cannot be
assumed under our reaction conditions. The species responsible
for the reduction of *(MesAcr+) thus far remains unclear. We
believe inefficient oxidation of trace carboxylate or neutral
heteroaromatic may be sufficient to generate small amounts of
MesAcr· as part of an initiation step. The inefficient nature of
this oxidation could explain the low quantum efficiency of this
process (vide infra). Once reduced, the MesAcr· is able to
transfer an electron to the hypervalent iodine adduct (Scheme
4), to initiate the proposed catalytic cycle and regenerate the
MesAcr+ photocatalyst. The hypervalent iodine radical anion,
[PhI(O2CR)2]•−, is expected to rapidly collapse leading to the
formation of the alkyl radical R·. Using high-level DFT
calculations (CPCM(ACN)/B3LYP/6-311++g-
(2d,2p)+LanL2DZ with extra diffuse and polarization for
iodine; see the SI for details), the collapse of the radical anion
[PhI(O2CR)2]•− was found to be highly favorable with no
significant energy barriers found at this level of theory. Overall,
the proposed catalytic cycle is thermodynamically favorable
(ΔGcycle = −83.15 kcal/mol) and each individual step is
thermodynamically feasible.
We experimentally validated the proposed single electron
transfer (SET) step between lepidine’s neutral α-amino radical
and PIFA by cyclic voltammetry (see the SI). In acetonitrile,
the reduction potential of protonated lepidine and PIFA were
−1.34 V and −0.527 V vs ferrocene, respectively. Both half
reactions therefore predict a very favorable SET reaction with a
+2.05 V driving force.
a
b
49% recovered starting material. Minor amount of bis-alkylated
c
product was observed. The product was isolated as a 4:1 mixture in
favor of the mono- versus the bis-alkylated product. Monoalkylated
d
Since this reaction appears to proceed via a chain reaction,
the quantum yield of the reaction was measured using a
Ru(bpy)3 and diphenylanthracene (DPA) actinometer accord-
ing to the literature.12 While our analysis suggests a chain
reaction, we obtained a low overall quantum yield of ∼0.18 at
the start of the reaction. This discrepancy is likely due to an
inefficient initiation step (e.g., inefficient oxidation of a
carboxylate) and/or inefficient chain propagation. A rapid
termination was ruled out since the quantum yield was slightly
higher at higher light intensity.
and bis-alkylated pyrimidine analogues were isolated in 77% and 11%
isolated yields, respectively. 55% recovered starting material.
e
For instance, the C4 position of quinaldine and the most
electrophilic C2 positions of benzothiazole, benzimidazole, and
2,6-dichloro purine provided alkylated products 5a, 5b, 5c, 5d,
and 5e in moderate to good yields (40−69%). Incomplete
conversion of the starting heterocycles accounted for the
modest yields. Other heteroaromatic products such as pyridines
C
DOI: 10.1021/acs.orglett.8b01085
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
discussions. M.F. thanks NSERC, FRQNT, and NanoQAM
for generous funding.
Scheme 4. Proposed Mechanism and DFT-Calculated ΔGrx
(ΔHrx in Parentheses) for Each Step of the Catalytic Cycle
REFERENCES
■
(1) (a) Cernak, T.; Dykstra, K. D.; Tyagarajan, S.; Vachal, P.; Krska,
S. W. Chem. Soc. Rev. 2016, 45, 546. (b) Genovino, J.; Sames, D.;
́
Hamann, L. G.; Toure, B. B. Angew. Chem., Int. Ed. 2016, 55, 14218.
(2) (a) Tauber, J.; Imbri, D.; Opatz, T. Molecules 2014, 19, 16190.
(b) Duncton, M. A. J. MedChemComm 2011, 2, 1135. (c) Minisci, F.;
Bernardi, R.; Bertini, F.; Galli, R.; Perchinummo, M. Tetrahedron 1971,
27, 3575.
́
(3) (a) Gutierrez-Bonet, A.; Remeur, C.; Matsui, J. K.; Molander, G.
A. J. Am. Chem. Soc. 2017, 139, 12251. (b) Galloway, J. D.; Mai, D. N.;
Baxter, R. D. Org. Lett. 2017, 19, 5772. (c) Wang, J.; Wang, S.; Wang,
G.; Zhang, J.; Yu, X.-Q. Chem. Commun. 2012, 48, 11769. (d) Seiple, I.
B.; Su, S.; Rodriguez, R. A.; Gianatassio, R.; Fujiwara, Y.; Sobel, A. L.;
Baran, P. S. J. Am. Chem. Soc. 2010, 132, 13194.
(4) (a) Prier, C. K.; Rankic, D. A.; MacMillan, D. W. C. Chem. Rev.
2013, 113, 5322. (b) Matsui, J. K.; Lang, S. B.; Heitz, D. R.; Molander,
G. A. ACS Catal. 2017, 7, 2563. (c) Shaw, M. H.; Twilton, J.;
MacMillan, D. W. C. J. Org. Chem. 2016, 81, 6898. (d) Romero, N. A.;
Nicewicz, D. A. Chem. Rev. 2016, 116, 10075.
(5) (a) Nuhant, P.; Oderinde, M. S.; Genovino, J.; Juneau, A.; Gagne,
́
Y.; Allais, C.; Chinigo, G. M.; Choi, C.; Sach, N. W.; Bernier, L.;
Fobian, Y. M.; Bundesmann, M.; Khunte, B.; Frenette, M.; Fadeyi, O.
O. Angew. Chem., Int. Ed. 2017, 56, 15309. (b) Jin, J.; MacMillan, D.
W. C. Nature 2015, 525, 87. (c) Huff, A. A.; Cohen, R. D.; Dykstra, K.
D.; Streckfuss, E.; DiRocco, D. A. J. Org. Chem. 2016, 81, 6980.
(d) Matsui, J. K.; Primer, D. N.; Molander, G. A. Chem. Sci. 2017, 8,
3512. (e) Li, G.−X.; Morales-Rivera, C. A.; Wang, Y.; Gao, F.; He, G.;
Liu, P.; Chen, G. Chem. Sci. 2016, 7, 6407. (f) McCallum, T.; Barriault,
L. Chem. Sci. 2016, 7, 4754. (g) DiRocco, D. A.; Dykstra, K.; Krska, S.;
Vachal, P.; Conway, D. V.; Tudge, M. Angew. Chem., Int. Ed. 2014, 53,
4802. (h) Garza-Sanchez, R. A.; Tlahuext-Aca, A.; Tavakoli, G.;
Glorius, F. ACS Catal. 2017, 7, 4057. (i) Klauck, F. J. R.; James, M. J.;
Glorius, F. Angew. Chem., Int. Ed. 2017, 56, 12336. (j) Cheng, W.-M.;
Shang, R.; Fu, M.-C.; Fu, Y. Chem. - Eur. J. 2017, 23, 2537. (k) Liu, P.;
Liu, W.; Li, C.-J. J. Am. Chem. Soc. 2017, 139, 14315. (l) Sherwood, T.
C.; Li, N.; Yazdani, A. N.; Dhar, T. G. M. J. Org. Chem. 2018, 83, 3000.
(m) Yang, B.; Yu, D.; Xu, X.-H.; Qing, F. L. ACS Catal. 2018, 8, 2839.
(6) Yoshimura, A.; Zhdankin, V. V. Chem. Rev. 2016, 116, 3328.
(7) (a) He, Z.; Bae, M.; Wu, J.; Jamison, T. F. Angew. Chem., Int. Ed.
2014, 53, 14451. (b) Wu, T.; Zhang, H.; Liu, G. Tetrahedron 2012, 68,
5229. (c) Togo, H.; Aoki, M.; Kuramochi, T.; Yokoyama, M. J. Chem.
Soc., Perkin Trans. 1 1993, 20, 2417.
We report herein a metal-free photoredox C−H functional-
ization of heteroarenes using an acridinium photocatalyst and
hypervalent iodine reagents under blue LED light. This light-
initiated chain reaction occurs under mild reaction conditions
and uses the broad and readily available carboxylic acid
monomer set. A range of functional groups are well tolerated
(olefins, alcohols, ethers, tertiary amines), and the method is
applicable to late-stage functionalization of complex molecules.
(8) (a) Fukuzumi, S.; Ohkubo, K. Org. Biomol. Chem. 2014, 12, 6059.
(b) Nicewicz, D. A.; Nguyen, T. M. ACS Catal. 2014, 4, 355.
(c) Nguyen, T. M.; Manohar, N.; Nicewicz, D. A. Angew. Chem., Int.
Ed. 2014, 53, 6198.
(9) (a) Fukuzumi, S.; Kotani, H.; Ohkubo, K.; Ogo, S.; Tkachenko,
N. V.; Lemmetyinen, H. J. Am. Chem. Soc. 2004, 126, 1600.
(b) Benniston, A. C.; Harriman, A.; Li, P.; Rostron, J. P.; van
Ramesdonk, H. J.; Groeneveld, M. M.; Zhang, H.; Verhoeven, J. W. J.
Am. Chem. Soc. 2005, 127, 16054.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.8b01085.
■
S
Full experimental details (PDF)
(10) For an early precedent of radical hydroxymethylation, see:
Schenck, G. O.; Koltzenburg, G.; Grossmann, H. Angew. Chem. 1957,
69, 177.
(11) Griffin, J. D.; Zeller, M. A.; Nicewicz, D. A. J. Am. Chem. Soc.
2015, 137, 11340.
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: jgenovino@hotmail.com.
*E-mail: frenette.mathieu@uqam.ca.
ORCID
(12) Pitre, S. P.; McTiernan, C. D.; Vine, W.; DiPucchio, R.; Grenier,
M.; Scaiano, J. C. Sci. Rep. 2015, 5, 16397.
Julien Genovino: 0000-0001-5879-2491
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank Dr. David Piotrowski (Pfizer), Dr. Joseph Tucker
(Pfizer), and Dr. Martins Oderinde (Pfizer) for helpful
■
D
DOI: 10.1021/acs.orglett.8b01085
Org. Lett. XXXX, XXX, XXX−XXX