Cooperative iodine and photoredox catalysis for direct oxidative lactonization of carboxylic acids

Article abstract of DOI:10.1039/c8cc08594c

A new method for the formation of γ- and δ-lactones from carboxylic acids through direct conversion of benzylic C-H to C-O bonds is described. The reaction is conveniently induced by visible light and relies on a mild cooperative catalysis by the combination of molecular iodine and an organic dye.

Full text of DOI:10.1039/c8cc08594c

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Cooperative iodine and photoredox catalysis for
direct oxidative lactonization of carboxylic acids†
Cite this: Chem. Commun., 2019,
55, 933
ab
ac
Thomas Duhamel
and Kilian Muniz
*
˜
Received 26th October 2018,
Accepted 14th December 2018
DOI: 10.1039/c8cc08594c
rsc.li/chemcomm
A new method for the formation of c- and d-lactones from carboxylic photoredox catalysis with organic dyes enables an efficient path-
acids through direct conversion of benzylic C–H to C–O bonds is way to benzylic radicals via initial oxidation of the phenyl ring
described. The reaction is conveniently induced by visible light and (Scheme 1a).11 Consequently, our strategy was to develop a new
relies on a mild cooperative catalysis by the combination of molecular route that would allow selective visible light-initiated benzylic
iodine and an organic dye.
oxidation and provide conditions under iodine co-catalysis to
subsequently intercept the benzylic radical, thereby promoting a
Lactones are abundant structural units in natural products and unique pathway to lactonization from the corresponding benzyl
important building blocks in the pharmaceutical sciences.1 In iodide intermediate (Scheme 1b).12
particular, significant biological activities have been reported
Here, we present the development of a mild, metal-free
for certain benzylic g-lactone derivatives.2 Four representative cooperative catalysis for the synthesis of such g- and d-lactones
examples are depicted in Scheme 1. There exist various synthetic starting from the corresponding free carboxylic acids under mild
approaches for the formation of lactones. Among the diverse conditions using an organic dye as the photoredox catalyst.
methodologies, recent work in the area has dealt with the
As a conceptual approach, the modular combination of distinct
development of economic lactonization reactions combining organocatalysts enabling consecutive reactions to be performed in
the concept of direct C–H oxygenation3 with free carboxylic acids. one pot has become a powerful tool in organic synthesis.13
Examples to this end involve oxidative C–H functionalization Previously, we developed a new cooperative light-activated catalysis
concepts such as 1,5-hydrogen atom transfer (HAT) within the based on molecular iodine and TPT (2,4,6-triphenylpyrylium tetra-
4
¨
Hofmann–Loffler manifold, by copper single electron transfer or fluoroborate) to perform the intramolecular amination of Csp3–H
by the use of stoichiometric amounts of iodine(III) reagents.5
Visible light-mediated photoredox catalysis has attracted
more and more attention over the last few years because it renders
new transformations accessible via single electron transfer (SET)
processes.6,7 However, so far, photoredox catalysts have been used
in association with carboxylic acids only to promote visible
light-mediated decarboxylation at the outset of arylation or
alkylation reactions.8,9 To the best of our knowledge, no metal-
free direct oxidative cyclisation of carboxylic acids under the
conceptual control of photocatalysis has been reported to date.
In previous work, Nicewicz explored the decarboxylation of free
carboxylic acids through organic photoredox catalysis.10 As an
interesting starting point, work by Wu has demonstrated that
bonds.14,15 In this process, hypoiodite was generated as the
a Institute of Chemical Research of Catalonia (ICIQ), The Barcelona Institute of
¨
Science and Technology, Av. Paısos Catalans 16, 43007 Tarragona, Spain.
E-mail: kmuniz@iciq.es
b
c
´
Universidad de Oviedo, C/Julian Claveria, 33006 Oviedo, Spain
´
Scheme 1 Representative lactone-containing natural products and the
ICREA, Pg. Lluıs Companys 23, 08010 Barcelona, Spain
design of a strategy for oxidative cyclisation.
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c8cc08594c
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun., 2019, 55, 933--936 | 933

View Article Online
Communication
ChemComm
Table 1 Optimisation of the reaction conditions for the cooperative
catalysis
yield (entries 13 and 14). Here, the working hypothesis suggests
that the presence of water is increasing the yield due to the
formation of hypoiodite in wet medium,17,18 which is an important
acceleration factor of the overall reaction (vide supra). As expected,
the reaction is not working in the absence of light irradiation
(entry 15). On a preparative 1 mmol scale, 80% isolated yield is
obtained (entry 16).
Entrya
x
y
Solvent
Time [h]
Conversion [%]
With the optimised conditions in hand, we then investigated
the scope of the reaction to produce structurally diversified
g-lactone derivatives (Table 2).
1
2
3
4
5
6
7
8
2
—
5
10
5
5
5
DCE/HFIP
DCE/HFIP
DCE/HFIP
DCE/HFIP
DCE/HFIP
DCE/HFIP
DCE/HFIP
DCE/HFIP
DCE
18
18
18
18
18
18
18
24
18
18
18
18
18
18
18
18
21
—
—
1
2
2.5
2
2
2
2
2
2
2
2
2
nrb
nrb
70
95 [94]c
64
53
77
56
44
52
25
30
As a starting point, we focused on the oxidative cyclisation of
aliphatic carboxylic acids (2a–g, 21–94%). Different substituents
in the para-position are well tolerated (2a–d, 51–94%). As the
only example, para-chlorinated 1d showed a decreased yield (2d,
51%) due to the quenching of the photoredox catalyst. We also
noticed that the para-methyl derivative 1b provided a compar-
ably lower yield due to the potential competition between the
formation of the two possible benzylic radicals (2b, 65%). As a
result of the radical conditions, no acyclic stereocontrol was
observed in the formation of 2f, while cyclic stereocontrol was
indeed possible, providing 2g as a single diastereoisomer.
Heteroatoms are well tolerated in this process as observed from
7.5
5
5
5
5
5
5
5
5
5
9
10
11
12
13
14
15
16
HFIP
C6H5Cl/HFIP
ACN/HFIP
DCE dried
DCE/HFIP dried
DCE/HFIP
DCE/HFIP
60
nrb,d
[80]c,e
2
a
Reaction conditions: 0.3 mmol 1a in 3 mL of solvent under blue light
b
c
irradiation. nr = no reaction. The number in brackets refers to the
isolated yield after purification by column chromatography. d Reaction
performed in the absence of light. e Reaction performed with 1 mmol 1a. the oxidation-labile thiophene derivative 2m (60% yield). The
g,g-diphenylated derivative 1e provided a reduced yield of 21%
active species, which promoted the prerequisite N-iodination of for 2e, which is a consequence of the slow lactone formation.
the sulphonamide as the key step of this catalytic Hofmann– The product 2h is particularly interesting as it proves that it is
¨
Loffler variant.
possible to cyclise benzoic acid derivatives onto a fused aliphatic
We decided to extend the study of such a cooperative catalysis chain. Moreover, the g-lactone 2h constitutes the major struc-
to the case of carboxylic acids for the realisation of the corres- tural unit of the natural product miltiorin D.1b We were also able
ponding C–H oxidation reaction using both TPT and molecular to access isobenzofurane derivatives in moderate to excellent
iodine. Gratifyingly, the initial experiment using only 2 mol% yields (2i–l, 31–74%). Due to the higher stability of the double
TPT and the carboxylic acid 1a in a mixture of DCE/HFIP (1,2- benzylic radical intermediate, the process is faster and the
dichloroethane/hexafluoropropan-2-ol, 1/1, v/v) already yielded reaction time can be reduced from 18 h to 12 h. The substituents
21% conversion (Table 1, entry 1) without any observable side in the meta or para position do not affect the cyclisation process to
products.16 Under the conditions of cooperative catalysis starting the final lactones, although an ortho-substituent led to a decrease
from 1 mol% photoredox catalyst and 5 mol% molecular iodine,
70% conversion was obtained (entry 4). An increase in the
amount of the photoredox catalyst to 2 mol% led to an increased Table 2 Oxidative cyclisation of carboxylic acids 1a–u for the formation
of lactones 2a–ua,b
94% of isolated 2a (entry 5). In the absence of the photoredox
catalyst, the reaction does not proceed (isolated starting material
only, entries 2 and 3). This observation suggests that TPT in the
photocatalysis not only acts as the terminal oxidant for the
regeneration of the molecular iodine, but plays a more direct
role in the functionalization of the substrate. It is important to
note that an adequate ratio of the two catalysts is decisive for the
effectiveness of the cooperative catalysis as a whole. A relative
increase of TPT or molecular iodine results in a significant
decrease of the conversion (entries 6 and 7). In particular, the
latter causes undesirable absorption of visible light and prevents
the photochemical activation of the dye. It is also crucial to
control the reaction time to avoid product decomposition, which
was noticed when the mixture remained for an extended period
a
of 24 h (entry 8). Solvent investigations were performed to
finally conclude that the use of a 1/1-mixture of DCE/HFIP
provides the best conditions (entries 9–12). The transformation
still occurs with dry solvents, although with a decrease in the
Reaction conditions: 0.3 mmol substrate in 3 mL of DCE/HFIP (1 : 1)
b
for 18 h at room temperature under blue LED irradiation. Isolated
c
yields after purification by column chromatography. 495% based on
d
the recovered starting material. Product obtained as a 1 : 1 mixture of
e
diastereoisomers. 12 h reaction time.
934 | Chem. Commun., 2019, 55, 933--936
This journal is ©The Royal Society of Chemistry 2019

View Article Online
ChemComm
Communication
Scheme 2 Control experiment and proposed mechanism for the iodine–
photoredox cooperative catalysis.
Fig. 1 Cyclic voltammetry of 1a (3 mM, blue line), 4 (3 mM, yellow line)
and 5 (3 mM, green line). The voltammogram for the blank electrolyte is
shown for comparison (red line). Working electrode: glassy carbon;
counter electrode: platinum wire; reference electrode: Ag/0.01 M AgNO3
in 0.1 M Bu4NClO4; and scan rate: 50 mV sÀ1
.
in the yield (2k, 31%, 495% based on the recovered material), which
may be due to steric hindrance. Using this process, we are also able
to achieve decorated d-lactones (2n–u, 35–96%). Interestingly,
2-phenethylbenzoic acid 1u exclusively yields the 6-membered cyclohexadienyl radical cation I.11,22 To further advocate for the
ring isochromanone 2u.
arene oxidation, we carried out cyclic voltammetry (Fig. 1). We
Regarding the reaction mechanism (Scheme 2), the most observed that both butylbenzene 4 and substrate 1a have the
noteworthy detail of the present transformation is the marked same oxidation potential (1.8 V vs. Ag/AgNO3), which is lower
absence of the decarboxylation pathway, particularly in the presence than that of cyclohexylbutanoic acid 5 (2.2 V vs. Ag/AgNO3).
of molecular iodine. As a general possibility at the outset, one Therefore, the aromatic core should be oxidised first. TPT, in its
should consider the formation of an intermediary acyl hypoiodide triplet exited state TPT*, has an oxidation potential Et(TPT*/
RCO2I from iodine and the substrate. Under irradiation, such a TPTꢀÀ) of 2.3 V. Hence, TPT* can oxidise the aromatic core by a
compound should be prone to undergo the corresponding homo- single electron transfer to form the p-C–H radical cation I.3d At
lytic O–I cleavage followed by a Hunsdiecker decarboxylation.19 The this stage, the acidity of the benzylic proton is dramatically
absence of decarboxylation products suggests that such a pathway is increased; the estimated pKa of a toluene radical-cation in
not operative, however, the initial formation of a carboxylate radical acetonitrile is between À9 and À13, thus making it an extremely
remains a possibility. Such an intermediary species would then strong acid that deprotonates readily, even in moderately acidic
need to be involved in a 1,5-hydrogen atom abstraction, in a step media.24 Moreover, we determined a KIE of 1 for a competition
that would be analogous of related 1,5-HAT with amide-based experiment between 1a and its mono-deuterated benzylic derivative,
radicals. To rule out such a possibility, we investigated the reactivity which is in perfect agreement with an initial arene oxidation. A
of [bis(4-phenylbutanoxy)iodo]benzene 3 in the presence of subsequent additional TPT oxidation to IV may be contributing to
molecular iodine. Upon in situ formation of the O-iodinated the product formation, but it is not the major pathway. In the
intermediate, irradiation does not provide the desired product presence of molecular iodine, the process needs to proceed through
2a.5c,20 Only traces of the decarboxylation side-product and free interplay between photoredox and iodine catalysis. Following the
carboxylic acid 1a were observed after 18 h at rt.21 This outcome formation of the benzylic radical II, interception of the benzyl
provides strong evidence that the phenyl oxidation occurs radical by molecular iodine generates the benzylic iodide III.
selectively at the outset and is not due to a putative carboxylate Formation of product 2a can occur directly by nucleophilic
radical. The formation of product 2a from a reaction with substitution liberating HI, which is then re-oxidised by the
involvement of only TPT (Table 1, entry 1) can be rationalised photoredox system.
as follows: TPT in its excited state, TPT*, can only be oxidised by
An alternative mechanism is based on the observation that
the SET mechanism. Substrate 1a contains two oxidation labile water traces are a requirement in the present catalysis. In
sites, which are the carboxylic acid and the phenyl substituent. agreement with our earlier work,15 this points to an involvement
In previous work on photoredox-based decarboxylation,8,10 it of hypoiodite. We used Raman spectroscopy to clearly identify
has been predicted that initial oxidation of the carboxylic acid by the the rapid formation of hypoiodite in the wet medium by
dye should generate an alkane. Since no traces of decarboxylation disproportionation of the molecular iodine.21,25 Hypoiodide
side-products were observed in any of the investigated cases and no should then be involved in the formation of the corresponding
reactivity was obtained for molecular iodine without TPT, we iodoso benzyl derivative V, which provides sufficient leaving group
conclude that no carbonyloxy radical is involved.23 Instead, the ability to foster rapid reductive C–O bond formation.26 This
benzylic radical II should be formed from arene oxidation with a sequence to 2a releases HOI, which is thus a steady component
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun., 2019, 55, 933--936 | 935

View Article Online
Communication
ChemComm
7 For reviews on C–H bond functionalization in combination with
photoredox catalysis, see: (a) D. C. Fabry and M. Rueping, Acc. Chem.
Res., 2016, 49, 1969; (b) Y. Qin, L. Zhu and S. Luo, Chem. Rev., 2017,
117, 9433.
under these conditions. While we are currently unable to
differentiate the individual cycles, the observed selectivity of
the overall transformation proves the superior performance
(synthetic potential) of dual iodine/photoredox catalysis. As a
mechanistic consequence, all oxidation is exclusively promoted
by the dye within multiple events. These include initial oxidation of
the organic substrate and the regeneration of the molecular iodine.
In addition, a competing, albeit kinetically less competent pathway
based on photoredox catalysis alone may be involved.
We have discovered a new oxidative process for the lactoni-
zation of benzylic C–H bonds. The reaction combines catalytic
amounts of molecular iodine and a dye and relies on photo-
redox catalysis as the terminal oxidation. The present protocol
broadens molecular iodine catalysis and should be instructive
for the development of related C–H oxidation reactions.
8 (a) Y. Miyake, K. Nakajima and Y. Nishibayashi, Chem. Commun., 2013,
49, 7854; (b) C. P. Johnston, R. T. Smith, S. Allmendinger and
D. W. C. MacMillan, Nature, 2016, 536, 322; (c) W.-M. Cheng, R. Shang,
M.-C. Fu and Y. Fu, Chem. – Eur. J., 2017, 23, 2537; (d) K. Xu, Z. Tan,
H. Zhang, L. Liu, S. Zhang and Z. Wang, Chem. Commun., 2017, 53, 10719;
(e) L. Candish, M. Freitag, T. Gensch and F. Glorius, Chem. Sci., 2017,
8, 3618; ( f ) H. Zhang, P. Zhang, M. Jiang, H. Yang and H. Fu, Org. Lett.,
2017, 19, 1016; (g) H. Huang, K. Jia and Y. Chen, ACS Catal., 2016, 6, 4983;
(h) G. H. Lovett and B. A. Sparling, Org. Lett., 2016, 18, 3494.
9 For non-photoredox based decarboxylative functionalization: (a) Z. Wang,
L. Zhu, F. Yin, Z. Su, Z. Li and C. Li, J. Am. Chem. Soc., 2012, 134, 4258;
(b) J. Prakash and S. Roy, J. Org. Chem., 2002, 67, 7861; (c) C. Kuang,
Q. Yang, H. Senboku and M. Tokuda, Synthesis, 2005, 1319.
10 J. D. Griffin, M. A. Zeller and D. A. Nicewicz, J. Am. Chem. Soc., 2015,
137, 11340.
11 R. Zhou, H. Liu, H. Tao, X. Yu and J. Wu, Chem. Sci., 2017, 8, 4654.
12 (a) C. Giordano and A. Belli, J. Org. Chem., 1980, 45, 345; (b) D. J. Wilger,
J.-M. M. Grandjean, T. R. Lammert and D. A. Nicewicz, Nat. Chem., 2014,
6, 720; (c) D. S. Hamilton and D. A. Nicewicz, J. Am. Chem. Soc., 2012,
134, 18577.
Conflicts of interest
13 (a) R. C. Wende and P. R. Schreiner, Green Chem., 2012, 14, 1821;
(b) A. E. Allen and D. W. C. MacMillan, Chem. Sci., 2012, 3, 633.
There are no conflicts to declare.
´
14 M. A. Miranda and H. Garcıa, Chem. Rev., 1994, 94, 1063.
˜
15 P. Becker, T. Duhamel, C. J. Stein, M. Reiher and K. Muniz, Angew.
Notes and references
Chem., Int. Ed., 2017, 56, 8004.
1 (a) I. J. Collins, J. Chem. Soc., Perkin Trans. 1, 1998, 1869; (b) A. Hirata, 16 The beneficial effect of HFIP might be shifting the disproportiona-
S.-Y. Kim, N. Kobayakawa, N. Tanaka and Y. Kashiwada, Fitoterapia, tion equilibrium and stabilising the hypoiodite species.
2015, 102, 49; (c) T. K. Devon and A. I. Scott, Handbook of Naturally 17 M. Uyanik, H. Hayashi and K. Ishihara, Science, 2014, 345, 291.
Occurring Compounds, Academic Press, New York, 1975, vol. 1, p. 249. 18 For non-related hypoiodide catalysis in the a-lactonization of carbonyls:
2 (a) L. E. Johnson and A. Dietz, Appl. Microbiol., 1968, 16, 1815;
(b) J. A. Katzenellenbogen, R. Rai and W. Dai, Bioorg. Med. Chem.
Lett., 1992, 2, 1399; (c) J. D. Lambert, J. E. Rice, J. Hong, Z. Hu and
(a) M. Uyanik, D. Suzuki, T. Yasui and K. Ishihara, Angew. Chem., Int. Ed.,
2011, 50, 5331; (b) M. Uyanik, D. Suzuki, M. Watanabe, H. Tanaka,
K. Furukawa and K. Ishihara, Chem. Lett., 2015, 44, 387.
C. S. Yang, Bioorg. Med. Chem. Lett., 2005, 15, 873; (d) E.-I. Hwang, 19 (a) H. Hunsdiecker and C. Hunsdiecker, Chem. Ber., 1942, 75, 291;
B.-S. Yun, Y.-K. Kim, B.-M. Kwon, H.-G. Kim, H.-B. Lee, W.-J. Jeong
and S.-U. Kim, J. Antibiot., 2000, 53, 903.
(b) R. G. Johnson and R. K. Ingham, Chem. Rev., 1956, 56, 219;
(c) C. V. Wilson, Org. React., 1957, 9, 341.
´
3 (a) A. Molnar, G. A. Olah and G. K. S. Prakash, Hydrocarbon 20 (a) K. Kiyokawa, R. Ito and S. Minakata, Chem. Commun., 2018,
Chemistry, Wiley, New York, 2017, ch. 9; (b) T.-L. Ren, B.-H. Xu,
S. Mahmood, M.-X. Sun and S.-J. Zhang, Tetrahedron, 2017, 73, 2943;
(c) S. Zhang, L. Li, H. Wang, Q. Li, W. Liu, K. Xu and C. Zeng, Org.
54, 7609; (b) M. Yudasaka, T. Maruyama, E. Yamaguchi, N. Tada and
A. Itoh, Eur. J. Org. Chem., 2018, 550; (c) C. Tian, X. Yao, W. Ji,
Q. Wang, G. An and G. Li, Eur. J. Org. Chem., 2018, 5972; (d) K. Jia
and Y. Chen, Chem. Commun., 2018, 54, 6105; (e) G.-X. Li, X. Hu,
G. He and G. Chen, Chem. Sci., DOI: 10.1039/C8SC04134B.
21 For more details please see the ESI†.
˜
Lett., 2018, 20, 252; (d) S. Herold, D. Bafaluy and K. Muniz, Green
Chem., 2018, 20, 3191; (e) F. Rusch, J.-C. Schober and M. Brasholtz,
ChemCatChem, 2016, 8, 2881.
4 J. Richers, M. Heilmann, M. Drees and K. Tiefenbacher, Org. Lett., 22 N. A. Romero, K. A. Margrey, N. E. Tay and D. A. Nicewicz, Science,
2016, 18, 6472. 2015, 349, 1326.
5 (a) S. Sathyamoorthi and J. Du Bois, Org. Lett., 2016, 18, 6308; 23 6-Membered ring product 2u further corroborates this mechanistic
(b) H. Togo, T. Muraki, Y. Hoshina, K. Yamagushi and M. Yokoyama,
J. Chem. Soc., Perkin Trans. 1, 1997, 787; (c) T. Dohi, N. Takenaga,
A. Goto, A. Maruyama and Y. Kita, Org. Lett., 2007, 9, 3129;
proposal. If the intermediate II was formed by a 1,5 or 1,6 HAT from
the carbonyloxy radical, the corresponding formation of 5- and
6-membered rings should be observed.
(d) S. Banerjee, S. Sathyamoorthi, J. Du Bois and R. N. Zare, Chem. 24 (a) M. Schmittel and A. Burghart, Angew. Chem., Int. Ed., 1997, 36, 2550;
Sci., 2017, 8, 7003.
6 (a) N. A. Romero and D. A. Nicewicz, Chem. Rev., 2016, 116, 10075;
(b) A. Albini, M. Mella and M. Freccero, Tetrahedron, 1994, 50, 575;
(c) A. M. P. Nicholas and D. R. Arnold, Can. J. Chem., 1982, 60, 2165.
(b) M. H. Shaw, J. Twilton and D. W. C. MacMillan, J. Org. Chem., 25 (a) M. Uyanik, D. Suzuki, M. Watanabe, H. Tanaka, K. Furukawa and
2016, 81, 6898; (c) Q. Qin, H. Jiang, Z. Hu, D. Ren and S. Yu, Chem.
Rec., 2017, 7, 754; (d) K. L. Skubi, T. R. Blum and T. P. Yoon, Chem.
Rev., 2016, 116, 10035; (e) J. Xuan and W.-J. Xiao, Angew. Chem., Int. 26 For related mechanisms: (a) C. Martınez and K. Muniz, Angew.
K. Ishihara, Chem. Lett., 2015, 44, 387; (b) M. Uyanik and K. Ishihara,
ChemCatChem, 2012, 4, 177.
´
˜
Ed., 2012, 51, 6828; ( f ) J. M. R. Narayanam and C. R. J. Stephenson,
Chem. Soc. Rev., 2011, 40, 102.
Chem., Int. Ed., 2015, 54, 8287; (b) P. Becker, T. Duhamel,
˜
C. S. Stein, M. Reiher and K. Muniz, ACS Catal., 2018, 8, 3918.
936 | Chem. Commun., 2019, 55, 933--936
This journal is ©The Royal Society of Chemistry 2019