RES EARCH

◥ In this research, we report six C–C and four

RESEARCH ARTICLE C–O bonds of (–)-cylindrocyclophane A gener-
ated by means of 10 C–H functionalizations with
ORGANIC CHEMISTRY high site selectivity and stereoselectivity made
possible by rhodium- and palladium-catalyzed
Total synthesis of (−)-cylindrocyclophane A processes originating from within the CCHF.
Although catalytic amounts of these precious
facilitated by C−H functionalization metals are required in these key transforma-
tions, related palladium-catalyzed processes
Aaron T. Bosse1, Liam R. Hunt2†, Camila A. Suarez1,2†, Tyler D. Casselman2, are widely used in industry (28), and we have pre-
Elizabeth L. Goldstein2, Austin C. Wright2, Hojoon Park3, Scott C. Virgil2, Jin-Quan Yu3, viously demonstrated that dirhodium-catalyzed
Brian M. Stoltz2*, Huw M. L. Davies1* C–H functionalizations can be conducted effi-
ciently with loadings as low as 0.0005 mol % (29).
(−)-Cylindrocyclophane A is a 22-membered C2-symmetric [7.7]paracyclophane that bears bis-resorcinol
functionality and six stereocenters. We report a synthetic strategy for (−)-cylindrocyclophane A that Synthetic plan

Downloaded from [Link] at Indian Institute of Technology Bombay on November 11, 2024
uses 10 C−H functionalization reactions, resulting in a streamlined route with high enantioselectivity and By incorporating C–H functionalization logic into
efficiency (17 steps). The use of chiral dirhodium tetracarboxylate catalysis enabled the C–H our retrosynthetic analysis of (–)-cylindrocyclophane
functionalization of primary and secondary positions, which was complemented by palladium-catalyzed A (1), we considered that we could effect sev-
C(sp2)–C(sp2) cross-couplings, resulting in the rapid formation of the macrocyclic core and all eral unusual transformations to rapidly sim-
stereocenters with high regio-, diastereo-, and enantioselectivity. The use of a late-stage palladium- plify the molecule (Fig. 2). We posited that
catalyzed fourfold C(sp2)–H acetoxylation installed the bis-resorcinol moieties. This research 1 could be derived from bis-amide 2 by the
exemplifies how multilaboratory collaborations can produce substantial modernizations of complex addition of an appropriate nucleophile into
total synthesis endeavors. each Weinreb amide. We envisioned introduc-
ing the 2,6-resorcinol functionality at a late

C
stage by a fourfold Weinreb amide–directed
–H functionalization has become an in- goal has been to demonstrate how the use of C–H acetoxylation of paracyclophane 3, re-
creasingly viable and accessible strategy C–H functionalization can provide diverse ducing the complexity and modulating the
over recent years, which has led to its routes toward challenging natural products reactivity of the aryl groups. The acetates at
amplified use toward more challenging and fully alter the manner in which we con- C1 and benzylic amide at C7 of macrocycle 3
bond constructions in complex settings ceptualize making architecturally and stereo- could be traced back to an orthogonally pro-
(1–3). Instead of relying on the established chemically complex targets. In this research tected macrocyclic tetra-ester 4. We identified
approach of a series of functional group trans- we illustrate our application of asymmetric, that rhodium(II)-catalyzed secondary C(sp3)–H
formations, the key focus becomes the strate- catalyst-controlled C–H functionalization to functionalizations could set the vicinal stereo-
gic and site-selective functionalization of the total synthesis, a strategy that constructs chal- centers in the natural product and serve as
traditionally unactivated C–H bonds. Many lenging bonds with high regio-, diastereo-, and the linchpin strategy for C–C bond construc-
highly effective methods have been devel- enantioselectivity derived from the stereo- tion and macrocyclization. These transforma-
oped for selective C–H functionalization, re- electronic precision of specialized transition tions could reveal diazoester 5, which could
lying on different tactics, such as the use of metal–catalyzed systems. be prepared by a palladium-catalyzed C–H func-
directing groups, radical transformations, and The target compound is the [7.7]paracyclophane tionalization of a diazoacetate with aryl halide
catalyst-controlled group transfer reactions natural product, (–)-cylindrocyclophane A (1) 6. Last, we anticipated that aryl halide 6 could
(4–8). (Fig. 1), a compound that has garnered exten- arise from the enantioselective rhodium(II)-
To date, C–H functionalization has been used sive synthetic interest because of its distinctive catalyzed primary C−H functionalization of a
as a critical transformation in several total molecular architecture (18–25). The previous syn- synthetic equivalent of hexane with a donor-
syntheses, often used as a late-stage diversifi- theses of 1 applied venerable chemical trans- acceptor rhodium-carbenoid derived catalyt-
cation strategy or to achieve a key disconnec- formations such as the olefin metathesis, ically from an aryl diazoacetate.
tion that was not previously achievable (9–12). Horner–Wadsworth–Emmons reaction, and
For example, the Li and Lei groups’ total syn- Ramberg–Bäcklund cyclodimerization strat- Asymmetric C–H functionalization
thesis of (–)-incarviatone A demonstrates an egies to construct the macrocycle and several We initiated our synthesis with a selective pri-
impressive illustration of sequential C–H func- other well-known transformations to forge mary C–H functionalization of known aryl di-
tionalization reactions conducted (13). As mem- each stereocenter (Fig. 1A and figs. S6 to S10), azoacetate 7 and trans-2-hexene 8 (Fig. 3A)
bers of the National Science Foundation (NSF) whereas this effort uses C–H functionaliza- (15, 26). The reaction proceeded with high
Center for Selective C−H Functionalization tion methodologies developed in our groups regio- and enantioselectivity using the ster-
(CCHF) and the Catalysis Innovation Consor- to tackle these C–C and C–O bond formations. ically hindered Rh2(R-p-PhTPCP)4 catalyst (Fig.
tium (CIC), we have been exploring ways to Our approach differs from the previous syn- 3B), designed to favor reactions at less crowded
combine the development of powerful C–H theses because it centers on catalyst-controlled C–H bonds, affording 9 in good yield and high
functionalization synthetic methods alongside asymmetric C–H functionalization (Fig. 1B). In enantioselectivity at C20 [73% yield, 96% en-
their application in total synthesis (14–17). Our our previous study, we reported a method that antiomeric excess (ee)] (26–30, 31). Hydro-
1
forges C–C bonds through asymmetric C–H genation of trans-olefin 9 was performed by
Department of Chemistry, Emory University, Atlanta, GA
functionalization of distal, unactivated meth- using Crabtree’s catalyst, affording iodide 10 in
30322, USA. 2Warren and Katherine Schlinger Laboratory of
Chemistry and Chemical Engineering, California Institute of ylene sites by using 1,2,2-triphenylcyclopropane quantitative yield (32, 33). Attempts to directly
Technology, Pasadena, CA 91125, USA. 3The Scripps carboxylate (TPCP) ligands specifically in access 10 through functionalization of n-hexane
Research Institute, La Jolla, CA 92037, USA. our Rh2(2-Cl-5-Br)TPCP4 catalyst and demon- produced inseparable mixtures from primary
*Corresponding author. Email: hmdavie@[Link] (H.M.L.D.);
stoltz@[Link] (B.M.S) strated its utility to generate a simplified [7.7] and secondary C–H functionalization reactions.
†These authors contributed equally to this work. paracyclophane model (26, 27). Subjecting iodide 10 to a palladium-catalyzed

Bosse et al., Science 386, 641–646 (2024) 8 November 2024 1 of 6

RES EARCH | R E S E A R C H A R T I C L E

Downloaded from [Link] at Indian Institute of Technology Bombay on November 11, 2024
B

Fig. 1. Strategies toward (–)-cylindrocyclophane A. (A) Reported synthetic strategies of cylindrocyclophane A. (B) Our C–H functionalization-based synthetic
strategy. Ac, acetyl; Bu, butyl; CBS, Corey-Bakshi-Shibata; Me, methyl; TES, triethylsilyl; Ms, mesyl.

Fig. 2. Retrosynthesis of (−)-cylindrocyclophane A.

C(sp2)–C(sp2) cross-coupling with diazoester functionalization sequence to construct the Direct cyclodimerization
11 smoothly delivered the aryl diazoacetate [7.7]paracyclophane framework (Fig. 3A), With aryl diazoacetate 12 we investigated the
12 through C–H functionalization of the ace- which was guided by our earlier work in syn- cyclodimerization. After initial disappointing
tate. We then turned our attention to the C−H thesizing the model macrocycle (26). results, we found that treating aryl diazoacetate

Bosse et al., Science 386, 641–646 (2024) 8 November 2024 2 of 6

RES EARCH | R E S E A R C H A R T I C L E

A B

Downloaded from [Link] at Indian Institute of Technology Bombay on November 11, 2024

Fig. 3. Synthesis of the [7.7]paracyclophane framework. (A) Detailed forward synthesis of a direct cyclodimerization and sequential C−H functionalization
approach to construct the cylindrocyclophane core. MS, molecular sieves; Ph, phenyl; Et, ethyl; SM, supplementary materials; XRD, x-ray diffraction. (B) Structures of
the chiral catalysts used to enable selective primary and secondary C–H functionalizations.

Bosse et al., Science 386, 641–646 (2024) 8 November 2024 3 of 6

RES EARCH | R E S E A R C H A R T I C L E

12 with Rh2(R-2-Cl-5-BrTPCP) delivered mac- centers in a single step. The direct cyclodimer- put and developed a stepwise approach to
rocycle 15 in a moderate yield, given the com- ization of diazoacetate 12 had limitations—in assemble macrocycle 15 on the basis of our
plexity of the transformations [19% yield, 6:1 particular, material throughput and the need published model study (27). Treatment of di-
diastereomeric ratio (dr)] (Fig. 3 and fig. S3). for high-performance liquid chromatography azoacetate 12 with Rh2(R-2-Cl-5-BrTPCP)4 and
The absolute and relative configuration of the purification. three equivalents of 10 generated aryl iodide
macrocyclic tetra-ester 15 was confirmed with 13 with high stereo- and regiocontrol to gen-
x-ray diffraction analysis. This one-pot reaction Stepwise macrocyclization erate the C1–C2 stereodiad (68% yield, 19:1 dr).
assembles the [7,7]paracyclophane core of the To complete the synthesis, we preferred to de- There is notably high regioselectivity in this
natural product along with four new stereo- velop a strategy with higher material through- unusual transformation for the most sterically

Downloaded from [Link] at Indian Institute of Technology Bombay on November 11, 2024

Fig. 4. Final C–H functionalizations toward (–)-cylindrocyclophane A. DIPEA, N,N-diisopropylethylamine; DMF, N,N′-dimethylformamide; HATU, hexafluorophosphate
azabenzotriazole tetramethyl uronium; HFIP, hexafluoroisopropanol; PIDA, phenyliodine(III) diacetate; THF, tetrahydrofuran; Mes, mesityl; DIBAL–H, diisobutylaluminum hydride.

Bosse et al., Science 386, 641–646 (2024) 8 November 2024 4 of 6

RES EARCH | R E S E A R C H A R T I C L E

accessible and electronically stabilized C–H We envisioned that treatment of bis-Weinreb 13. B. Hong et al., J. Am. Chem. Soc. 137, 11946–11949 (2015).
bond [(>20:1 regioisomeric ratio (rr)], and amide 2 with excess propyl Grignard would 14. S. A. Loskot, D. K. Romney, F. H. Arnold, B. M. Stoltz, J. Am.
Chem. Soc. 139, 10196–10199 (2017).
high relative diastereocontrol (dr) for the two effect nucleophilic addition as well as global 15. N. A. Falcone et al., Org. Lett. 23, 9393–9397 (2021).
new stereogenic centers generated (>30:1 dr), deacylation. Despite extensive efforts, the result- 16. T. A. Bedell et al., Angew. Chem. Int. Ed. 55, 8270–8274
as well as a high level of asymmetric induction ing diketone was impervious to deoxygenation (2016).
17. A. D. Yamaguchi, K. M. Chepiga, J. Yamaguchi, K. Itami,
(19:1 dr), and it affords the product in good (fig. S4). The next approach centered around
H. M. L. Davies, J. Am. Chem. Soc. 137, 644–647 (2015).
yield (68% yield). The bis-arene intermediate the reaction of macrocycle 2 with excess di- 18. B. S. Moore et al., J. Am. Chem. Soc. 112, 4061–4063
13 was subjected to another palladium-catalyzed isobutylaluminum hydride (fig. S5). Unfortunate- (1990).
cross-coupling by means of C–H functional- ly, the reduction provided a complex mixture of 19. D. Berthold, B. Breit, Synlett 32, 436–446 (2021).
20. T. R. Hoye, P. E. Humpal, B. Moon, J. Am. Chem. Soc. 122,
ization and afforded the macrocyclization pre- products that could not be elaborated to the 4982–4983 (2000).
cursor 14. The macrocyclization upon treatment final natural product (1) in acceptable yield. 21. A. B. Smith, S. A. Kozmin, C. M. Adams, D. V. Paone, J. Am.
of diazoester 14 with Rh2(R-2-Cl-5-BrTPCP)4 We hypothesized that the poor results were Chem. Soc. 122, 4984–4985 (2000).
22. A. B. Smith III, C. M. Adams, S. A. Kozmin, D. V. Paone, J. Am.
proceeded with very high asymmetric induc- due to the in situ generation of phenoxides Chem. Soc. 123, 5925–5937 (2001).
tion (>30:1 dr) but now with lower relative di- cyclizing onto the Weinreb amide; therefore, 23. H. Yamakoshi et al., Org. Biomol. Chem. 7, 3772–3781 (2009).
astereocontrol (8:1 dr) to generate the second we elected to transform the phenolic acetates 24. K. C. Nicolaou, Y.-P. Sun, H. Korman, D. Sarlah, Angew. Chem.

Downloaded from [Link] at Indian Institute of Technology Bombay on November 11, 2024
Int. Ed. 49, 5875–5878 (2010).
stereodiad (C14–C15) of macrocycle 15 (70% into methyl ethers. 25. K.-Y. Chen, H.-Q. Wang, Y. Yuan, S.-B. Mou, Z. Xiang, Angew.
yield). The lower overall diastereoselectivity for We found that chemoselective deacylation Chem. Int. Ed. 62, e202307602 (2023).
the macrocyclization in comparison with the of the phenolic acetates followed by methyla- 26. L. Fu, J. D. Mighion, E. A. Voight, H. M. L. Davies, Chemistry 23,
formation of aryl iodide 13 is due to Horeau’s tion delivered tetra-methyl ether 18 (39). Treat- 3272–3275 (2017).
27. W. Liu et al., J. Am. Chem. Soc. 140, 12247–12255 (2018).
effect, in which imperfect asymmetric induction ment of macrocycle 18 with DIBAL reductively 28. C. Torborg, M. Beller, Adv. Synth. Catal. 351, 3027–3043
generates diastereomeric rather than enantio- cleaved the remaining acetates and reduced (2009).
meric mixtures. Using the stepwise sequence both Weinreb amides to the corresponding al- 29. B. Wei, J. C. Sharland, D. G. Blackmond, D. G. Musaev,
H. M. L. Davies, ACS Catal. 12, 13400–13410 (2022).
enabled the preparation of >1.2 mmol of macro- dehydes (40), which were directly subjected to 30. K. Liao et al., ACS Catal. 8, 678–682 (2018).
cycle 15 in a single pass. Additionally, at this a Wittig olefination to deliver bis-olefin 19 as a 31. C. Qin, H. M. L. Davies, J. Am. Chem. Soc. 136, 9792–9796
scale, macrocycle 15 could be isolated as a single single set of alkene isomers (41). Last, hydro- (2014).
32. R. H. Crabtree, Chem. Rev. 115, 127–150 (2015).
diastereomer through recrystallization from the genation of bis-olefin 19 delivered (–)-tetra- 33. Y. Xu, D. M. P. Mingos, J. M. Brown, Chem. Commun. 14,
crude reaction mixture. O-methyl-cylindrocyclophane A (20), which 199–201 (2008).
Our synthesis of macrocycle 15 through was then demethylated by using the vigor- 34. R. B. Woodward et al., J. Am. Chem. Soc. 88, 852–853 (1966).
35. S. Senaweera, K. C. Cartwright, J. A. Tunge, J. Org. Chem. 84,
either a sequence of the direct cyclodimeriza- ous conditions of Hoye and coworkers to af-
12553–12561 (2019).
tion (12% yield over four steps) or stepwise ford (–)- cylindrocyclophane A (1) (20). 36. G. Li et al., J. Am. Chem. Soc. 137, 4391–4397 (2015).
macrocyclization (23% yield over six steps) Overall, our synthesis involves either a shorter 37. H. Park, N. Chekshin, P.-X. Shen, J.-Q. Yu, ACS Catal. 8,
from aryl diazoacetate 7 allowed for the effi- 17-step or more scalable 19-step sequence from 9292–9297 (2018).
38. H. Park, Y. Li, J.-Q. Yu, Angew. Chem. Int. Ed. 58, 11424–11428
cient construction of the macrocyclic core and commercial starting material by using 10 C–H (2019).
six stereocenters en route to the target. This functionalization reactions that forge six C–C 39. B. P. Bandgar, L. S. Uppalla, V. S. Sadavarte, S. V. Patil, New J.
approach of establishing the core macrocycle bonds and four C–O bonds. Specifically, the Chem. 26, 1273–1276 (2002).
40. S. Azeez, P. Sureshbabu, S. Sabiah, J. Kandasamy, Org. Biomol.
of the natural product relatively early in the route showcases four catalyst-controlled enan- Chem. 20, 2048–2053 (2022).
synthesis is distinct from previous approaches. tioselective and diastereoselective C–H func- 41. N. C. Dwulet, Z. Chahine, K. G. Le Roch, C. D. Vanderwal,
Although we have a shorter route to the macro- tionalizations to generate all six stereogenic J. Am. Chem. Soc. 145, 3716–3726 (2023).
cycle, we elected to use the stepwise sequence centers of the natural product, as well as two AC KNOWLED GME NTS
because of greater material throughput. palladium-catalyzed C−H functionalizations of This study was facilitated by constructive discussions within the
With the macrocyclic framework in hand, diazocarbonyl compounds and four amide- NSF CCHF and CIC. The authors thank J. Bacsa (Emory University)
the trichloroethyl esters of the macrocycle 15 directed C−H acetoxylations. This research and M. Takase (California Institute of Technology) for x-ray
structure determination, D. VanderVelde (California Institute of
were chemoselectively converted to the bis- exemplifies the power of multi-institutional Technology) for nuclear magnetic resonance (NMR) expertise, and
Weinreb amide 16 (Fig. 4) (34). The remaining collaboration and C–H functionalization as M. Shahgholi and J. Barbor (California Institute of Technology)
trifluoroethyl esters were then hydrolyzed to an enabling technology to selectively transform for mass spectrometry assistance. The authors thank S. Osler for
afford the bis-carboxylic acid 17, which was low-cost materials into highly functionalized assistance in gel permeation chromatography analysis. Any
opinions, findings, and conclusions or recommendations expressed
subjected to photocatalytic decarboxylative and stereochemically complex building blocks. in this material are those of the author(s) and do not necessarily
acetoxylation to deliver bis-benzylic acetate 3 reflect the views of the National Science Foundation. Funding:
RE FERENCES AND NOTES This work was supported by the National Science Foundation under
(35). At this stage, we were poised to conduct
1. N. Y. S. Lam, K. Wu, J.-Q. Yu, Angew. Chem. Int. Ed. 60, the CCI CCHF (CHE-1700982). B.M.S.’s portion of the work was
the final C−H functionalization reaction in the 15767–15790 (2021). also supported by the NIH-NIGMS (R35GM145239), Heritage
synthesis, namely a tetra-C(sp2)–H acetoxyla- 2. J. Yamaguchi, A. D. Yamaguchi, K. Itami, Angew. Chem. Int. Ed. Medical Research Investigators Program, and the California
tion of macrocycle 3 with a 5-(trifluoromethyl) 51, 8960–9009 (2012). Institute of Technology. H.M.L.D.’s portion of this work was also
3. D. J. Abrams, P. A. Provencher, E. J. Sorensen, Chem. Soc. Rev. supported by the National Science Foundation (CHE-1956154) and
pyridine-3-sulfonic acid ligand, a transforma- 47, 8925–8967 (2018). the NIH-NIGMS (R01GM099142). E.L.G was supported by the
tion only made possible by the conditions de- 4. N. Holmberg-Douglas, D. A. Nicewicz, Chem. Rev. 122, 1925–2016 NSF GRFP. L.R.H was supported by the University of Auckland
veloped by the Yu group (36–38). In a single (2022). doctoral scholarship award. Author contributions: Conceptualization:
5. M. C. White, J. Zhao, J. Am. Chem. Soc. 140, 13988–14009 (2018). B.M.S. and H.M.L.D. Methodology: A.T.B., L.R.H., C.A.S., T.D.C.,
transformation, amide-directed tetra-C(sp2)–H 6. C. Sambiagio et al., Chem. Soc. Rev. 47, 6603–6743 (2018). E.L.G., A.C.W., H.P., S.C.V., J.-Q.Y., B.M.S., and H.M.L.D. Investigation:
oxidation of 3 delivered macrocycle 2 in 60% 7. H. M. L. Davies, D. Morton, Chem. Soc. Rev. 40, 1857–1869 (2011). A.T.B., L.R.H., C.A.S., T.D.C., E.L.G., A.C.W., H.P., S.C.V., B.M.S.,
yield. This final linchpin transformation signi- 8. D. A. Colby, A. S. Tsai, R. G. Bergman, J. A. Ellman, Acc. Chem. and H.M.L.D. Visualization: A.T.B., L.R.H., and C.A.S. Funding
Res. 45, 814–825 (2012). acquisition: B.M.S. and H.M.L.D. Project administration: B.M.S. and
fied the completion of all the desired C−H func- H.M.L.D. Supervision: B.M.S. and H.M.L.D. Writing – original draft:
9. L. Guillemard, N. Kaplaneris, L. Ackermann, M. J. Johansson,
tionalization steps toward the natural product. Nat. Rev. Chem. 5, 522–545 (2021). A.T.B., L.R.H., C.A.S., B.M.S., and H.M.L.D. Writing – review and
10. T. Cernak, K. D. Dykstra, S. Tyagarajan, P. Vachal, S. W. Krska, editing: A.T.B., L.R.H., C.A.S., T.D.C., E.L.G., A.C.W., H.P., S.C.V.,
Completion of the synthesis Chem. Soc. Rev. 45, 546–576 (2016). J.-Q.Y., B.M.S., and H.M.L.D. Competing interests: H.M.L.D. is a
11. J. H. Docherty et al., Chem. Rev. 123, 7692–7760 (2023). named inventor on a patent entitled “Dirhodium catalyst compositions
The last challenge to complete the synthesis 12. P. Bellotti, H. M. Huang, T. Faber, F. Glorius, Chem. Rev. 123, and synthetic processes related thereto” (US 8.974.428, issued
was the installation of the two propyl side chains. 4237–4352 (2023). 10 March 2015). The other authors declare that they have no

Bosse et al., Science 386, 641–646 (2024) 8 November 2024 5 of 6

RES EARCH | R E S E A R C H A R T I C L E

competing interests. Data and materials availability: reserved; exclusive licensee American Association for the Figs. S1 to S10
Crystallographic parameters for compounds 15, 17, 20, and 1 are Advancement of Science. No claim to original US government works. Tables S1 to S30
available free of charge from the Cambridge Crystallographic [Link] NMR Spectra
Data Centre under CCDC 2364844, 2364845, 2364843, and Reference (42)
2364846, respectively. Spectra, materials and methods, and substrate SUPPLEMENTARY MATERIALS
and reagent details are available in the supplementary materials. [Link]/doi/10.1126/science.adp2425 Submitted 14 March 2024; accepted 2 October 2024
License information: Copyright © 2024 the authors, some rights Materials and Methods 10.1126/science.adp2425

Downloaded from [Link] at Indian Institute of Technology Bombay on November 11, 2024

Bosse et al., Science 386, 641–646 (2024) 8 November 2024 6 of 6