“This document is the Accepted Manuscript version of a Published Work that appeared in final form in ACS Catal. 2017, 7, 5478-5482, copyright © American Chemical Society after peer review and technical editing by the publisher. To access the final edited and published work see DOI: 10.1021/acscatal.7b01748. This article may be used for non-commercial purposes in accordance with the ACS guidelines published at http://pubs.acs.org/page/policy/articlesonrequest/index.html].” 1 Substrate Triggered Stereoselective Preparation of Highly Substituted Organic Carbonates Victor Laserna,† Eddy Martin,† Eduardo C. Escudero-Adán† and Arjan W. Kleij*,†,‡ † Institute of Chemical Research of Catalonia (ICIQ), the Barcelona Institute of Science and Technology, Av. Països Catalans 16, 43007 – Tarragona (Spain) ‡ Catalan Institute for Research and Advanced Studies (ICREA), Pg. Lluis Companys 23, 08010 – Barcelona (Spain). 2 ABSTRACT: Trisubstituted cyclic organic carbonates with multiple though well-defined stereochemical configurations are difficult to prepare. Here we present a conceptual design towards these CO2 based synthons using hydroxyl-substituted cyclic epoxide precursors and their catalytic conversion, to afford these challenging target compounds with fused ring sizes up to eight under excellent stereo-control. The observed stereochemistry of the organic carbonates combined with various control experiments revealed that these compounds are formed through a mechanistic manifold that involves a depolymerization reaction within an oligomeric carbonate induced by a pendent hydroxyl nucleophile. This manifold therefore provides an alternative approach towards CO2 valorization into functional, cyclic carbonate scaffolds of use in synthetic chemistry. Keywords: carbon dioxide, cyclic organic carbonates, depolymerization, homogenous catalysis, stereoselectivity 3  INTRODUCTION Carbon dioxide offers a cheap and renewable carbon feedstock for fine-chemical synthesis1 and recent progress clearly testifies the imminent role of homogeneous catalysis to turn this waste into more complex molecules.2 In this respect, cyclic organic carbonates are increasingly used as useful synthetic intermediates if appropriate substituted with functional groups that, after activation, furnish reactive intermediates for a wide variety of transformations.3 Whereas mono-substituted organic carbonates are fairly easily prepared and methodologies for di-substituted versions4 have recently become available, it remains a huge challenge to prepare more densely substituted/functionalized organic carbonate scaffolds5 which ultimately may limit further exploration of the synthetic potential of these structures. We recently reported the synthesis of bicylic organic carbonates which have carbon dioxide incorporated as a temporary protecting group and deliver cis-diols upon hydrolysis.6 In addition, we communicated the use of acyclic epoxy alcohols that can be converted into organic carbonates through a substrate-controlled CO2 activation process.7 Inspired by these results and the need to develop efficient procedures for highly substituted and functional organic carbonates prompted us to consider an unexplored approach towards the coupling of CO2 and cyclic syn-configured epoxyalcohols. Stereoselective conversions in cyclic carbonate synthesis have been subject of various recent investigations8 but represent a relatively new area in CO2 valorization catalysis. We envisioned that various cyclic carbonate scaffolds with different stereochemistry (Scheme 1) could be produced from a common syn-configured cyclic epoxy alcohol precursor (i.e, stereodivergence) through previously reported reaction manifolds (Scheme 1). A double inversion pathway (Scheme 1) retains the original stereochemistry of the substrate would afford a syn/cis configured product,9 whereas under appropriate reaction conditions a metal-alkoxide driven depolymerization of an in 4 situ prepared polycarbonate species takes place with inversion of configuration at one of the epoxide-carbons giving a syn/trans isomer. An alcohol mediated activation of carbon dioxide and intra-molecular ring-opening leads to yet another diastereo-isomer (anti/trans). Whereas these known manifolds thus lead to diastereoisomeric syn/cis, syn/trans and anti/trans configured organic carbonate products, these approaches are not able to provide the corresponding anti/cis isomer (Scheme 1, at the right). Scheme 1. The Conversion of Different Ring-Size Cyclic Epoxy-alcohols into Functionalized Organic Carbonate Scaffolds with various Stereochemical Configurations. Therefore, we set out to design a new conceptual approach towards such anti/cis configured carbonate scaffolds to further widen the synthetic potential of these epoxy-alcohol substrates while controlling the stereoselectivity, functionality and scope of these transformations. Here we report 5 a general, catalytic and stereoselective approach towards the formation of functional anti/cis bicyclic carbonates through a new and substrate-triggered mechanism. The latter conveniently allows the preparation of challenging trisubstituted organic carbonates, and their post-synthetic potential is also reported.  RESULTS AND DISCUSSION First, we examined the use of a simple and synthetically accessible epoxy-alcohol substrate based on a cyclohexyl skeleton (Table 1, 2; see Supporting Information (SI) for details on this type of substrate synthesis). Various reaction conditions including nucleophilic additive, Al-catalyst, temperature and pressure were varied to examine the influence on the compositions of the reaction mixture (see also Table S1). In general, we found that three principal products were formed in these experiments (syn/cis-3, anti/cis-3 and triol 4) which were unambiguously identified by 1D and 2D NMR spectroscopy, and for syn/cis 3 and anti/cis 3 also by X-ray diffraction (see inserts in the Figure to Table 1).10 The syn/cis isomer of 3 (82% yield) was formed as the major reaction component by using only the nucleophilic additive (Table 1, entries 16; NBu4Br), and its synthesis is the result of a double inversion mechanism.8b However, in the presence of Al-complexes 1a1c (see Scheme 1; Table 1, entries 711, see also Table S1) the diastereoselectivity could be tuned towards the formation of anti/cis 3 (entry 8; 88% yield) whose stereochemistry cannot be explained by the known manifolds detailed in Scheme 1. 6 Table 1. Screening of Reaction Conditions towards the Stereoselective Formation of Carbonates 3. MEK is Methylethyl Ketone, see for Al-complexes 1a-1c Scheme 1, Nu Stands for Nucleophile.a Entry [Al, (mol %) Nu (mol %) Solvent Conv. (%)b Sel. syn/anti/4 (%)b 1  Br, 2.5 MEK 51 38 : 54 : 6 2  Br, 5.0 MEK 73 50 : 41 : 9 3  Br, 10 MEK >99 68 : 23 : 9 4c  Br, 25 MEK >99 82 : 13 : 5d 5  Br, 10 MeOH >99 2 : 13 : 85 6  Br, 10 Tol >99 35 : 35 :30 7 1a, 0.5 Br, 1.0 MEK >99 0 : 71 : 29 8 1a, 1.0 Br, 2.5 MEK >99 0 : 89 : 11e 9f 1a, 0.5 Cl, 0.5 Tol >99 0 : 91 : 9 10 1b, 0.5 Br, 1.0 MEK >99 0 : 83 : 17 11 1c, 1.0 Br, 2.5 MEK >99 11 : 64 : 25 General conditions: 0.50 mmol of 2, p(CO2)º = 10 bar, 70ºC, solvent (200 L), amounts of [Al] and Nu indicated. Br is NBu4Br and Cl stands for PPNCl, see Figure. bDetermined by 1H NMR (CDCl3). cAt 50ºC. dIsolated yield of syn/cis 3 was 82%. eIsolated yield of anti/cis 3 was 88%. fAt 40 bar of CO2. a 7 Therefore, in order to gain more insight into the operating mechanism leading to these anti/cis configured bicyclic carbonates, we decided to consider various hydroxyl-substituted cyclic epoxides (Scheme 2, 5ad; for their synthesis see the SI). The conversion of five-, six- and sevenmembered hydroxyl-substituted cyclic epoxides 5a5c and the conditions leading to the anti/cis bicyclic carbonates 6a6c (yield: 7187%) had to be optimized individually (details on the screening conditions in Tables S2S4), whereas the synthesis towards anti/cis 6d was not feasible. Scheme 2. Preparation of Different Diastereoisomeric Bicyclic Carbonates 68. X-ray of 8d a X-ray of 8e General conditions: 0.50 mmol substrate, 40 bar CO2, 70ºC, 18 h, solvent (200 L), reported yields are of the isolated compound after chromatographic purification. Specific conditions: (i) 1a (1.0 mol%), PPNCl (1.0 mol%), MEK; (ii) 1b (0.50 mol%), NBu4Br (1.0 mol%), MEK; (iii) 1a (5.0 mol%), DMAP (5.0 mol%), Tol; (iv) NBu4Br (25 mol%), MEK; (v) 1a (1.0 mol%), PPNCl (5.0 mol%), Tol; (vi) 1a (2.5 mol%), PPNCl (2.5 mol%), MEK; (vii) 1a (5.0 mol%), DMAP (5.0 mol%), Tol. 8 The synthesis of the isomeric syn/cis carbonates 7a7d was simply mediated by using high loadings of the nucleophilic additive NBu4Br, and is pertinent to the aforementioned double inversion pathway.8b In contrast, the manifold leading to 6a6c is distinct, and the applied experimental conditions favor a backbiting process of an oligocarbonate intermediate produced in situ from cyclic epoxides as previously reported.8 Standard depolymerization11 of oligo/polycarbonates would produce a trans configured cyclic carbonate product through a metalalkoxide terminus (alkoxide backbiting, see Scheme 3), and consequently a syn/trans configured product. Alternatively, carbonate backbiting (Scheme 3) in a polymer with a metal-carbonate endgroup would furnish a syn/cis carbonate product. Therefore the mechanism accountable towards the formation of the anti/cis configured carbonates 6a6c is distinct from these previous reported depolymerization pathways. We propose that the hydroxyl-unit is actively involved in the depolymerization process as outlined in Scheme 3 (OH-assisted backbiting). The OH may be readily activated towards nucleophilic attack onto the adjacent carbonate unit through Hbonding and likely this process is favored over the standard depolymerization process for substrates of type 5a‒b (with n = 5 or 6). An OH-assisted depolymerization nicely fits the experimentally observed formation of anti/cis carbonate products 3 and 6a6c. This latter mechanistic hypothesis was further challenged by consideration of seven- and eightmembered cyclic epoxide substrates as these are generally less prone towards copolymerization with CO2.12 The experimental conditions were therefore varied such that copolymerization would be favored.8,11 The larger ring-size, bicyclic carbonate products 8c and 8d (assumed to be generated via a standard alkoxide backbiting process) could be isolated though their yields were modest/low due to competitive pathways leading to other stereoisomers as shown in Scheme 3; (optimized conditions from Tables S3 and S4, SI). 9 Scheme 3. Mechanistic Pathways to the Differently Configured Bicyclic Carbonate Products Observed Experimentally. Al Stands for an Aluminum Catalyst such as Complexes 1a-c. Such trans configured carbonates are typically not formed from smaller bicyclic epoxides (n = 5 or 6) as the intrinsic ring strain in the bicylic carbonate product would lead to decarboxylation and/or decomposition.13 The anti/trans carbonate 8e (86%) was obtained in high yield in the presence of 1a/DMAP and its configuration suggests the occurrence of substrate-assisted (and not an OH-mediated depolymerization) activation of CO2 (Scheme 3) facilitated by the Al-complex 1a as previously described by us for acyclic epoxy-alcohols.7,14 Thus, it seems that for larger ring-size hydroxysubstituted cycloalkanes the formation of an anti/cis configured bicyclic carbonate is less likely to 10 occur due to competitive pathways that lead to other diastereoisomers. As far as we know, the formation of 8d and 8e represent rare examples of bicyclic carbonates with fused eight-membered rings, and their structures were unambiguously confirmed by X-ray analyses (inserts to Scheme 3; optimized reaction conditions from Table S4, SI). In order to examine whether the formation of the unusual anti/cis configured bicyclic carbonates could be extended to more functional derivatives, we then turned our focus (Scheme 4) on the conversion of substituted versions of hydroxy-cyclohexene oxides A–N (Scheme 4) under similar conditions as reported in Table 1, entry 9. Gratifyingly, the carbonate products 922 could be generally prepared in good isolated yields of up to 85% and with wide functional group diversity allowing for the introduction of synthetically useful alkyne (9 and 13), vinyl/olefin (12 and 14) and para- and/or meta-substituted aromatic fragments (17 and 22). Cyclohexene oxide substrates with ortho-substituted aryl groups display more sluggish reactivity, and only a moderate yield for 21 (41%) was obtained after 66 h indicating some degree of steric impediment in this transformation. Intriguingly, the straightforward and high yield synthesis of a series of trisubstituted functional bicyclic carbonates could be easily achieved which is known to be extremely challenging in the area of CO2/epoxide couplings.5 The presence of a vinyl group in the cyclic carbonate structure (such as in 12) conveniently allows for post-modification chemistry as recently demonstrated.15 Beside the formation of tri-substituted, functional carbonates 922, anti/cis configured 3 could be easily converted in high yield and selectivity into bicyclic carbonate scaffolds 23ae: these latter compounds incorporate rather common and synthetically useful protecting groups including benzoyl, mesyl and a bulky silyl. It should be noted that the R fragments from the cyclohexene oxide substrates are not attached to 11 Scheme 4. Substrate Scope for the Conversion of Various Hydroxy-Substituted Cyclohexene Oxides into Anti/Cis Configured Bicyclic Carbonates 922 and Protected Derivatives 23ae.a X-ray 19 a Conditions: 0.50 mmol epoxide substrate, Al-complex 1a (1.0 mol%), PPNCl (5.0 mol%), toluene (200 L), 70ºC, 40 bar CO2. Reported yields are isolated ones after chromatographic purification. b Reaction time was 66 h; Bz = benzoyl, Ac = acetyl, Ms = mesyl, Ts = tosyl, TBDMS = tert- butyldimethylsilyl. 12 the same carbon center as the (unprotected) alcohol group in the carbonate products anti/cis 3 and 922, and the X-ray structure determined for 19 (Scheme 4) further confirming the general spectroscopic assignments. These results further support that the original alcohol unit of the epoxide substrate is incorporated into the cyclic carbonate ring of the product and thus implies, as suggested, that this OH unit is actively involved towards the formation of the bicyclic carbonate product. Scheme 5. Synthetic Potential of Carbonate 12 and conversion into its derivatives 24‒28. Finally, we used vinyl-substituted bicyclic carbonate 12 as a starting point to demonstrate the potential of these functionalized carbonates in organic synthesis and to access other useful precursors (Scheme 5). Treatment of 12 with morpholine gave access to carbamate 24 in good yield (85%) and with high regio-selectivity (90:10).3e A Dess-Martin oxidation of 12 furnished 13 the ketone derivative 25 in 84% yield. Simple hydrolysis under basic conditions converted 8 into vinyl substituted triol product 26 (89%) of potential use in natural product synthesis.16 A decarboxylative amination17 gave access to stereoselective formation of tetrasubstituted olefin 27 (64%) while the alcohol group in 12 could be protected by a mesyl group to afford 28 in 91% yield. It should be noted that these transformations are rather general and should be applicable to the other bicyclic carbonates reported in Scheme 4.  CONCLUSION In summary, we here present a unique manifold to access stereoselectively anti/cis configured bicyclic carbonates that are the result of a hydroxy-mediated depolymerization process of in situ prepared oligocarbonate precursor. This new manifold affords a wide range of challenging trisubstituted bicyclic carbonates that cannot be accessed through any previously reported methodology. Thus, this new depolymerization manifold provides a new valorization approach for carbon dioxide and its conversion into useful precursors for synthetic chemistry as demonstrated herein. Currently we are examining the use of other, related epoxy alcohol scaffolds in various coupling reactions with a focus on synthetic applications. AUTHOR INFORMATION Corresponding Author akleij@iciq.es Notes The authors declare no competing financial interest. 14 ASSOCIATED CONTENT Supporting Information. Copies of analytical data/spectra, and full characterization data for all new compounds. This material is available free of charge via the Internet at http://pubs.acs.org.  ACKNOWLEDGMENTS We thank ICREA, the CERCA Program/Generalitat de Catalunya, and the Spanish Ministerio de Economía y Competitividad (MINECO) through project CTQ-2014–60419-R and the Severo Ochoa Excellence Accreditation 2014–2018 through project SEV-2013–0319. Dr. Noemí Cabello is acknowledged for providing the mass analyses. V. L. thanks the MINECO for a FPU predoctoral fellowship.  REFERENCES (1) (a) Liu, Q.; Wu, L.; Jackstell, R.; Beller, M. Nat. Commun. 2015, 6, 5933. (b) Kielland, N.; Whiteoak, C. J.; Kleij, A. W. Adv. Synth. Catal. 2013, 355, 21152138. (c) Gui, Y.-Y.; Zhou, W.J.; Ye, J.-H.; Yu, D.-G. ChemSusChem 2017, 10, 13371340. (d) Klankermayer, J.; Wesselbaum, S.; Beydoun, K.; Leitner, W. Angew. Chem. Int. Ed. 2016, 55, 72967343. (e) Yu, B.; He, L.-N. ChemSusChem 2015, 8, 5262. (2) For a selection of recent examples: (a) Zhang, Z.; Liao, L.-L.; Yan, S.-S.; Wang, L.; He, Y.Q.; Ye, J.-H.; Li, J.; Zhi, Y.-G.; Yu, D.-G. Angew. Chem. Int. Ed. 2016, 55, 70687072. (b) Carry, B.; Zhang, L.; Nishiuara, M.; Hou, Z. Angew. Chem. Int. Ed. 2016, 55, 62576260. (c) Ren, X.; Zheng, Z.; Zhang, L.; Wang, Z.; Xia, C; Ding, K. Angew. Chem. Int. Ed. 2017, 56, 310313. (d) 15 Thenert, K.; Beydoun, K.; Wiesenthal, J.; Leitner, W.; Klankermayer, J. Angew. Chem. Int. Ed. 2016, 55, 1226612269. (3) For recent examples: (a) Gómez, J. E.; Guo, W.; Kleij, A. W. Org. Lett. 2016, 18, 60426045. (b) Guo, W.; González-Fabra, J.; Bandeira, N. A. G.; Bo, C.; Kleij, A. W. Angew. Chem. Int. Ed. 2015, 54, 1168611690. (c) Guo, W.; Martínez-Rodríguez, L.; Martín, E.; Escudero-Adán, E. C.; Kleij, A. W. Angew. Chem. Int. Ed. 2016, 55, 1103711040. (d) Guo, W.; Laserna, V.; Martin, E.; Escudero-Adán, E. C.; Kleij, A. W. Chem.–Eur. J. 2016, 22, 1722–1727. (e) Sopeña, S.; Laserna, V.; Guo, W.; Martin, E.; Escudero-Adán, E. C.; Kleij, A. W. Adv. Synth. Catal. 2016, 358, 21722178. (4) For recent contributions to this area: (a) Castro-Osma, J. A.; Lamb, K. J.; North, M. ACS Catal. 2016, 6, 5012–5025. (b) Rintjema, J.; Kleij, A. W. ChemSusChem 2017, 10, 12741282. (c) Qin, J.; Wang, P.; Li, Q.; Zhang, Y.; Yuan, D.; Yao, Y. Chem. Commun. 2014, 50, 10952– 10955. (d) Della Monica, F.; Vummaleti, S. V. C.; Buonerba, A.; De Nisi, A.; Monari, M.; Milione, S.; Grassi, A.; Cavallo, L.; Capacchione, C. Adv. Synth. Catal. 2016, 358, 3231–3243. (e) Sopeña, S.; Martín, E.; Escudero-Adán, E. C.; Kleij, A. W. ACS Catal. 2017, 7, 35323539. (f) Wu, X.; North, M. ChemSusChem 2017, 10, 74–78. (g) Martínez, J.; Castro-Osma, J. A.; Alonso-Moreno, C.; Rodríguez-Diéguez, A.; North, M.; Otero, A.; Lara-Sánchez, A. ChemSusChem 2017, 10, 1175–1185. (h) Martínez, J.; Castro-Osma, J. A.; Earlam, A.; Alonso-Moreno, C.; Otero, A.; LaraSánchez, A.; North, M.; Rodríguez-Diéguez, A. Chem.–Eur. J. 2015, 21, 9850–9862. For a recent overview: (i) Martín, C.; Fiorani, G.; Kleij, A. W. ACS Catal. 2015, 5, 13531370. (5) The preparation of trisubstituted cyclic organic carbonates has been rarely reported and with important limitations in scope and yield, see for instance: (a) Fiorani, G.; Stuck, M.; Martín, C.; 16 Martínez-Belmonte, M.; Martin, E.; Escudero-Adán, E. C.; Kleij, A. W. ChemSusChem 2016, 9, 13041311. (b) Maeda, C.; Shimonishi, J.; Miyazaki, R.; Hasegawa, J.-Y.; Ema, T. Chem.–Eur. J. 2016, 19, 6556–6563. (c) Bähr, M.; Bitto, A.; Mülhaupt, R. Green Chem. 2012, 14, 14471454. (6) Laserna, V.; Fiorani, G.; Whiteoak, C. J.; Martin, E.; Escudero-Adán, E. C.; Kleij, A. W. Angew. Chem. Int. Ed. 2014, 53, 1041610419. (7) Rintjema, J.; Epping, R.; Fiorani, G.; Martín, E.; Escudero-Adán, E. C.; Kleij, A. W. Angew. Chem. Int. Ed. 2016, 55, 39723976. (8) (a) Liu, Y.; Ren, W.-M.; He, K.-K.; Zhang, W.-Z.; Li, W.-B.; Wang, M.; Lu, X.-B. J. Org. Chem. 2016, 81, 89598966. (b) Whiteoak, C. J.; Martin, E.; Escudero-Adán, E. C.; Kleij, A. W. Adv. Synth. Catal. 2013, 355, 22332239. See also refs. 1b and 3d. (9) Note that anti/cis and syn/cis notations refers to the relative position of the OH group with respect to a cis-configured cyclic carbonate unit. (10) Further details can be found in the Supporting Information. (11) For seminal mechanistic contributions concerning polycarbonate depolymerization: (a) Darensbourg, D. J.; Yeung, A. D. Macromolecules 2013, 46, 8395. (b) Darensbourg, D. J.; Wei, S.-H. Macromolecules 2012, 45, 59165922. (c) Wu, G.-P.; Wei, S.-H.; Ren, W.-M.; Lu, X.-B.; Xu, T.-Q.; Darensbourg, D. J. J. Am. Chem. Soc. 2011, 133, 1519115199. (12) Kruper and Dellar reported the formation of a trans-configured carbonate product in high yield derived from cycloheptene oxide, and this may the result of an in situ depolymerization process. See: Kruper, W. J.; Dellar, D. V. J. Org. Chem. 1995, 60, 725727. 17 (13) For decarboxylation/decomposition of trans configured (oligo)carbonates derived from cyclopentene/cyclohexene oxides at elevated temperatures: (a) Darensbourg, D. J.; Wei, S.-H.; Yeung, A. D.; Ellis, W. C. Macromolecules 2013, 46, 58505855. (b) Darensbourg, D. J.; Wei, S.-H.; Wilson, S. J. Macromolecules 2013, 46, 32283233. (c) Darensbourg, D. J.; Bottarelli, P.; Andreatta, J. R. Macromolecules 2007, 40, 77277729. (14) Note that acyclic epoxy alcohols have distinct reactivity compared to cyclic analogues, and much lower tendencies towards in situ copolymerization with CO2, and therefore favor the alcoholassisted CO2 activation pathway. (15) Vinyl cyclic carbonates have recently shown to be useful in fine-chemical targeted synthesis: (a) Cai, A.; Guo, W.; Martínez-Rodríguez, L.; Kleij, A. W. J. Am. Chem. Soc. 2016, 138, 1419414197. (b) Khan, A.; Zheng, R.; Kan, Y.; Ye, J.; Xing, J.; Zhang, Y. J. Angew. Chem. Int. Ed. 2014, 53, 64396442. (16) Cyclic triol scaffolds are of importance in the Synthesis of various natural products, see for instance: (a) Vadivel, S. K.; Vardarajan, S.; Duclos, R. I. Jr.; Wood, J.-A. T.; Guo, J.; Makriyannis, A. Bioorg. Med. Chem. Lett. 2007, 17, 59595963. (b) Pingli, L.; Vandewalle, M. Tetrahedron 1994, 50, 70617074. (c) Crowley, P. J.; Fawcett, J.; Griffith, G. A.; Moralee, A. C.; Percy, J. M.; Salafia, V. Org. Biomol. Chem. 2005, 3, 32973310. (17) Guo, W.; Martínez-Rodríguez, L.; Kuniyil, R.; Martin, E.; Escudero-Adán, E. C.; Maseras, F.; Kleij, A. W. J. Am. Chem. Soc. 2016, 138, 1197011978. 18 SYNOPSIS/TOC: 19