Synthetic chemistry methodology

[ad_1]

Carbon-based frameworks constitute the backbone of organic molecules that are used in applications ranging from pharmaceuticals to materials. Thus, the modular and efficient construction of C–C bonds is one of the ultimate goals in organic synthesis. Traditionally, cross-coupling reactions between an organometallic reagent and an electrophile have been used for this task, leading to the synthesis of numerous essential molecules^2,3. However, one of the coupling partners usually needs to be pre-functionalized as an organometallic nucleophilic reagent, increasing the step count and requiring the handling of potentially sensitive intermediates. By contrast, recently discovered cross-electrophile coupling reactions (XECs) bypass the use of organometallic reagents by directly coupling two electrophilic coupling partners^1,4,5. XECs are most often catalysed by transition metals and rely on an external stoichiometric reductant for catalytic turnover. Often, these reactions require at least one coupling partner with C(sp²) hybridization at the reactive site to control chemoselectivity^1,6. Forming C(sp³)–C(sp³) bonds by XECs is notably more difficult^7,8,9,10,11, but highly desirable owing to their ubiquity in organic molecules. For example, the fraction of C(sp³) atoms in molecules has been positively correlated with the clinical success of drug candidates, creating an even stronger demand for C(sp³)–C(sp³) bond-forming reactions^12,13,14. A challenge in transition-metal-catalysed C(sp³)–C(sp³) XECs is the propensity to form homocoupling side products instead of the desired heterocoupling products¹⁵ (Fig. 1a). This unwanted reactivity can be partially reduced by using a large excess of one of the coupling partners, leading to additional waste¹⁶. Combined with the negative economic and environmental impact of transition metals, the development of alternative, transition-metal-free strategies is highly desirable. This would not only avoid potential issues with metal contamination for subsequent applications but could also improve the scope of C(sp³)–C(sp³) XECs to include functional groups that are reactive under many metal-catalysed conditions or that could poison a metal catalyst. In this context, two seminal reports emerged recently that used either electrochemical or enzymatic platforms to enable this highly challenging reactivity^17,18 (Fig. 1b). These approaches rely on the formation of radical or anionic intermediates from one of the substrates that can then be engaged in C–C bond formation with the second coupling partner. To avoid undesired side reactions of these highly reactive intermediates, stabilizing substituents (for example, amides or aromatics) are required, inherently restricting these approaches in their reaction scope and preventing the use of fully unactivated substrates. Therefore, a transition-metal-free XEC of two fully unactivated and unstabilized C(sp³) coupling partners has so far remained elusive, clearly highlighting the limitations of current approaches. There is, thus, an urgent need for the design of strategically distinct mechanistic manifolds that can address these challenges and thereby unlock the immense potential of XECs for the assembly of C(sp³)–C(sp³) linkages ubiquitously found across the molecular sciences.

Here, we report the discovery and investigation of an unprecedented coupling related to XEC in which two completely unactivated and non-organometallic C(sp³) fragments are combined in a transition-metal-free protocol. Alkylphosphonium salts are reacted with alkyl halides in the presence of a sterically hindered base as the sole reagent to give rise to the coupled products (Fig. 1d). Mechanistic studies suggest that the reaction is enabled by an unusual single-electron transfer (SET) in a frustrated ion pair that is reminiscent of frustrated Lewis pair radical reactivity^{19,20,21,22,23,24,25,26}.

Phosphonium salts are key reagents in organic synthesis. They are readily available and widely used, for example, in the Wittig reaction to construct alkenes²⁷. Recently, their intriguing reactivity has also been harnessed in unconventional manners, including the selective functionalization of heteroarenes^28,29. Moreover, the use of phosphonium salts as C(sp²) electrophiles in transition-metal-catalysed cross-coupling reactions has been investigated^30,31,32. Our group has a longstanding interest in diverting the reactivity of phosphonium salts towards unusual transformations, such as C–P metathesis and related transformations for late-stage phosphine functionalization^33,34. During our recent studies³⁵, we discovered conditions under which the P–C(sp³) bond of the phosphonium salt is activated in addition to the more reactive P–C(sp²) bond if the phosphonium salt contained a tertiary alkyl group (Fig. 1c). This peculiar hydro-dephosphination sparked our curiosity as it occurred in the absence of any conventional reductant. We reasoned that this unexpected bond activation likely arose from a different mechanism than oxidative addition to the metal catalyst. Our hypothesis was confirmed by subsequent control experiments, which demonstrated that the reaction does not require a transition metal catalyst to yield alkane products and is mediated by a sterically hindered base. Although the activation of C–P bonds in phosphonium salts is well known to occur by basic hydrolysis, it is documented to favour the cleavage of aromatic substituents over alkyl substituents if the alkyl group is not activated³⁶. Indeed, subjecting tert-alkyltriphenylphosphonium salts to known hydrolysis conditions³⁷ did not lead to the formation of the desired alkane products (Supplementary Information section 5.8), implying that the discovered hydro-dephosphination process occurred by a different reaction pathway with distinct selectivity. Although the nature of this bond activation was unclear at this stage, we aimed to leverage this new reactivity to design an innovative approach to challenging C(sp³)–C(sp³) bond formation. We hypothesized that an α-tertiary alkylphosphonium salt could be formed in situ by the alkylation of a phosphorus ylide, which can be generated through the deprotonation of a phosphonium salt (Fig. 1d). Subsequent base-mediated hydro-dephosphination would enable an overall coupling between these two C(sp³) electrophiles that does not require the presence of a transition metal catalyst.

Initial experiments using phosphonium salts and alkyl iodides as the alkylating agent indeed confirmed our hypothesis, as the alkane product was obtained in moderate yields using sterically hindered bases. After careful optimization (Supplementary Tables 1–5), we developed a highly efficient transition-metal-free coupling between unactivated sec-alkylphosphonium salts and unactivated alkyl halides. This reaction is mediated by lithium hexamethyldisilazide (LiHMDS) as the only added reagent.

As the reaction does not contain an obvious reductant and the mode of C–P bond activation is unconventional, we became interested in understanding the origin of this unique reactivity. As a first finding, we noticed that triphenylphosphine (5), the expected byproduct from a traditional reductive C–P bond activation, was formed only in small amounts (9% nuclear magnetic resonance (NMR) yield), whereas dibenzophosphole 4 was the main byproduct of the reaction in 86% NMR yield. This suggested that the phosphonium moiety serves as a reductant because of the oxidation of two phenyl C–H bonds to a C–C bond in dibenzophosphole 4 (Fig. 2a). Reaction monitoring by NMR spectroscopy also showed that the phosphonium salt 1 is rapidly deprotonated to its ylide in the presence of LiHMDS (Supplementary Fig. 8). When the phosphonium salt 1, LiHMDS, and alkyl iodide 2 are reacted at room temperature, the tert-alkylphosphonium salt 3 is formed as the main species after a few minutes. The identity of 3 was confirmed by an independent synthesis (Supplementary Information, section 6.3). The C–C bond formation is, thus, fast and proceeds through the alkylation of a phosphorus ylide, as set out in our reaction design. These results also highlight that the unique reactivity of the phosphonium electrophile is crucial for the coupling reactivity as it enables the C–C bond formation through a deprotonation-mediated philicity switch to the nucleophilic ylide and also acts as an internal reductant for the unusual oxidation to the dibenzophosphole 4 (refs. ^24,25). Subjecting the tert-alkylphosphonium salt 3 to different bases demonstrated that product formation is strongly dependent on the size of the base (Fig. 2b). Although LiHMDS provided a high yield of product, smaller amide bases gave much lower yields that decreased with the base size. This demonstrates that the size of the base is not only important for limiting undesired S_N2 reactions of the base with the alkyl halide substrate but also has an important role in the cleavage of the C–P bond. To further study this key step, which is at the core of this new reaction, we used adamantyltriphenylphosphonium bromide (9), a phosphonium salt that already contains a tertiary alkyl group, as a model substrate. When 9 was reacted with LiHMDS, adamantane (8) was formed in a similar yield as the coupling products under the model reaction conditions (Supplementary Information, section 6.5). However, adding 2,2,6,6-tetramethylpiperidinyloxyl (TEMPO) as a radical trap to this reaction shut down the formation of adamantane (8). Instead, the TEMPO adduct 7 was isolated in 62% yield, suggesting the formation of tert-alkyl radicals in the reaction (Fig. 2c). Interestingly, dibenzophosphole 4 was formed in a similar yield compared with the standard reaction conditions, and its formation was, therefore, not influenced by the presence of TEMPO. As the formation of dibenzophosphole 4 from the PPh₃ moiety in the starting phosphonium salts requires the loss of two aryl hydrogen atoms, we next subjected phosphonium salt 9-d₁₅, containing perdeuterated phenyl groups, to the reaction conditions (Fig. 2d). NMR analysis indicated that the adamantane isolated from this reaction contained 72% deuterium at one of the tertiary positions (8-d).

With these results in hand, we propose a plausible reaction mechanism (Fig. 2e). The starting phosphonium salt is deprotonated by LiHMDS, and the resulting ylide is rapidly alkylated by the alkyl halide to yield the α-tertiary phosphonium salt A. Next, salt metathesis between A and LiHMDS takes place first, leading to a sterically encumbered ion pair B. Experiments using different alkali HMDS salts showed that the presence of lithium ions is important for the reaction, possibly because the formation of LiX (X = Br or I) serves as a driving force for the salt metathesis step. Similarly, the addition of 12-crown-4 to the reaction, which selectively binds lithium ions³⁸, shut down the reactivity (Supplementary Table 12). The key step enabling the coupling reaction is a SET from the HMDS anion to the phosphonium cation in the frustrated ion pair B. Such SET processes in phosphonium halide salts are well known to occur photochemically by ultraviolet irradiation or also by visible-light activation in the case of phosphonium salts containing strongly electron-withdrawing substituents^{39,40,41,42,43}. By contrast, the thermal process has only been postulated for the reaction of tetraphenylphosphonium chloride with lithium amides based on preliminary data^24,25. At the time this study was carried out, no applications or detailed studies of such a process existed. We however would like to point out a complementary report⁴⁴, using a SET step in a frustrated ion pair for C–H activation, that appeared shortly after an initial version of our work was deposited as a preprint.

In the lowest-energy isomer of the salt B, the HMDS anion forms a C–H⋯anion interaction with a phenyl group of the phosphonium ion and an additional π–anion interaction with a second phenyl group (Fig. 2f). Isomers with the direct interaction of the phosphonium P atom and the anion in a phosphorane-type geometry (isomer B′), which are typical for smaller nucleophiles (Supplementary Information section 7.3), are much higher in energy²⁸. We performed distortion–interaction analyses on the different phosphonium HMDS ion pair geometries and a related smaller ion pair containing the N(i-Pr)₂ anion (structure H) to understand the origin of the unusual isomer preference (Fig. 2f). The analysis showed that the phosphorane-like isomer B′ of the HMDS ion pair is destabilized because the interaction energy is substantially lower than in the case of the smaller N(i-Pr)₂ ion pair H (−75.7 kcal mol^–1 and −107.3 kcal mol^–1, respectively). This is caused by the large steric bulk of the two ions, which leads to a longer P–N distance in the HMDS ion pair B′ (2.10 Å) than in the N(i-Pr)₂ ion pair H (1.87 Å). Because of this steric frustration, the formation of the atypical phosphonium amide ion pair B containing interactions between the HMDS anion and the phosphonium phenyl groups becomes favoured. Analysis of the frontier orbitals of B shows that the highest occupied molecular orbital is located on the HMDS anion whereas the lowest unoccupied molecular orbital is delocalized on the phosphonium cation, in line with the direction of the proposed electron transfer. Calculations of the redox potentials of the ions in B show that the proposed SET is accessible under the experimental conditions (Supplementary Table 14). The SET step would be preceded by the formation of a charge-transfer complex^45,46. Indeed, strong charge-transfer bands at around 380 nm and 515 nm, potentially arising from such an interaction, were visible by ultraviolet–visible spectroscopy when the phosphonium salt 9 was mixed with LiHMDS (Supplementary Fig. 9). A calculated ultraviolet–visible spectrum of B shows a charge-transfer band at 348 nm, closely matching one of the experimentally observed bands. By contrast, conformer B′ is predicted to have little charge-transfer character (272 nm). This suggests that B is more likely to be the active ion pair leading to SET, and its formation is enabled by the steric frustration in the phosphonium HMDS ion pair.

The SET step leads to the formation of a phosphoranyl radical C and an HMDS radical. The α-scission of C forms the tertiary alkyl radical D and triphenylphosphine (5). The highly reactive HMDS radical can abstract a hydrogen atom from triphenylphosphine (5), resulting in the aryl radical E, which can undergo cyclization to the cyclic radical F. The weak alkyl C–H bond in F can be abstracted by the previously generated alkyl radical D to aromatize the phosphorus-containing ring system to the experimentally observed byproduct 4 and leading to the alkane product G. The proposed pathway is supported by additional density functional theory calculations (Supplementary Information, section 7).

Having gained an increased understanding of the reaction mechanism, we explored the scope of the reaction (Fig. 3). Reacting phosphonium salts with an almost equimolar amount of alkyl halides in the presence of LiHMDS provided a large variety of C(sp³)–C(sp³)-coupled products without forming homocoupling side products that are problematic under transition-metal-catalysed conditions. Although the reaction works best using alkyl iodides as coupling partners, alkyl bromides provide the product in a similar yield (79% versus 81%, respectively, for 12a). Alkyl tosylates could also be engaged and gave the product in a fair yield (6, 62%), providing an opportunity to use alcohol starting materials in the reaction after facile derivatization to the corresponding tosylate. Alkyl chlorides participated in the reaction as well, although in lower yields (6, 32%). Ethers, silyl-protected alcohols, and acetal moieties were well tolerated (12b–d). Substrates containing amides or tert-butyloxycarbonyl (Boc)-protected amines provided the desired coupled products in moderate yields (12e and 12f). Groups that would be reactive under many transition-metal-catalysed conditions, namely, aryl halides (12g–i) and aryl boronate esters (12j), gave rise to the desired products. The reaction also tolerates the presence of heterocycles such as phenoxazines (12k), indoles (12l), and pyrimidines (12m). When the homobenzylic iodide 11n was engaged as a coupling partner, the spirocyclopropane 12n was isolated in 44% yield instead of the coupling product. We hypothesize that the corresponding styrene of 11n is formed by a fast E2 elimination to which the in situ formed phosphorus ylide can be added. This would give rise to a benzylic carbanion that can undergo 1,3-elimination with the phosphonium moiety to form the cyclopropane ring⁴⁷. In contrast to 11n, the benzylic bromide 11o formed the desired coupling product 12o in a moderate yield, demonstrating that the reaction can be used for the coupling of electron-rich activated alkyl halides in addition to the unactivated ones. The reaction can furthermore be carried out in the presence of acidic moieties such as alcohols (12p and 12s), carboxylic acids (12q), and unprotected azaindoles (12r) when additional LiHMDS is used in the reaction to account for the deprotonation of the acidic groups. Moreover, substrates containing the hormone oestradiol (11s) and the drug molecules nevirapine and metaxalone (11t and 11u) were successfully engaged in the coupling reaction. Phosphonium salts containing differently sized cyclic alkyl groups afforded the alkane products (12v–z). Smaller-ring phosphonium salts provided higher yields (12v and 12w) than larger-ring ones (12x–z). We noticed increased levels of alkene side products in these reactions that might arise from a Hofmann-type elimination of the alkylated phosphonium salt intermediate. A range of acyclic phosphonium salts containing functional groups such as aryl halides or heterocycles provided the corresponding products (12aa–12ag). Notably, the phosphonium salt 11ad was also engaged with iodomethane to afford the methylated product 12ae in the same yield as product 12ad, which contained a longer-chain alkyl group.

**Fig. 3: Scope of the coupling reaction.**

To further evaluate the functional group tolerance of the reaction, we carried out a compatibility screen with different additives⁴⁸ (Supplementary Fig. 7). Several functional groups that are known to be reactive or inhibitory under many transition-metal-catalysed reactions, such as terminal alkenes, tertiary amines, pyridines, and epoxides, did not interfere with the reaction. Limitations of the reaction include nitro compounds and most carbonyl groups such as ketones, which can undergo Wittig-type side reactions. The addition of an enol ether, however, did not noticeably inhibit the reaction, indicating that protected ketones may be viable functionalities in substrates.

Using the insights gained from the mechanistic study, we were able to leverage this new reactivity beyond the coupling reaction in a series of preliminary results (Fig. 4). On the basis of the knowledge that the C–C bond is formed by the alkylation of a phosphorus ylide, we also devised a formal [1+n] annulation of n-alkylphosphonium salt 13 and 1,ω-dibromoalkane 14 (Fig. 4a). In this reaction, NaH-mediated phosphorus ylide alkylation occurs twice⁴⁹, leading to the same tert-alkylphosphonium salt intermediate 3 as in the standard reaction that can then undergo C–P fragmentation in the presence of LiHMDS. Besides this two-electron process leveraging the phosphorus ylide reactivity, we also used the one-electron reactivity of the alkyl radical intermediate for further functionalization. First, we were able to intercept it with styrene 16 to construct a quaternary centre in 15, forming two new C(sp³)–C(sp³) bonds in a single step (Fig. 4b). Similarly, the addition of heterocycle 17 to the standard conditions led to the formation of an additional C(sp³)–C(sp²) bond in product 18. Leveraging the hydrogen-atom transfer step of the tert-alkyl radical with the phosphonium phenyl groups, the coupling of a phosphonium salt containing perdeuterated phenyl groups (1-d₁₅) led to the formation of the monodeuterated product 12m-d. The deuteration is regioselective for the position at which the phosphorus was bound in the starting material over multiple other sites that would be prone to deuteration reactions using conventional approaches⁵⁰. In the described reactions, the radical arising from the key frustrated ion pair ultimately engages in product formation. We envisioned that this radical could also be used to mediate the formation of other open-shell intermediates through atom transfer processes, thereby extending the applicability of this chemistry. When the alkyl iodide 19 and alkene 20 were reacted in the presence of tetraphenylphosphonium bromide and LiHMDS, C–C bond formation occurred and product 21 was formed. We believe that this reaction could proceed through the formation of a frustrated ion pair between the PPh₄ cation and the HMDS anion, which leads to a phenyl radical that can react in a halogen-atom transfer step with alkyl iodide 19.

**Fig. 4: Mechanistically informed extensions of the reaction scope.**

[ad_2]

Source link

Fluorspar to fluorochemicals upon low-temperature activation in water

[ad_1]

Innovations with an impact on the large-scale production of chemicals are in demand considering the pressing global challenges faced by the manufacturing industry. These include the availability and management of raw materials, transportation disruptions, complex supply chains and climate change pressure¹⁰. This urgency applies to fluorochemicals, a class of molecules that remain in increasing demand owing to their critical role as pharmaceuticals to cure diseases, agrochemicals for food security and in the production of lithium-ion batteries, which is expected to soar over the coming decade^1,2,3,11. At present, the starting point of the entire fluorine industry is the treatment of fluorspar (fluorite, CaF₂) with concentrated sulfuric acid (H₂SO₄) at high temperature (>300 °C) to produce dangerous hydrogen fluoride (HF)^4,5. Subsequently, HF is used as is, or converted into various fluorinating reagents for the synthesis of fluorochemicals (Fig. 1a). For decades, other routes to HF have been studied, all featuring inorganic acids, such as hydrochloric acid (HCl), and requiring harsh reaction conditions¹². We recently disclosed an activation strategy whereby fluorochemicals as diverse as sulfonyl, benzyl, allyl and alkyl fluorides can be accessed with a fluorinating reagent directly prepared from acid-grade fluorspar (AGF; >97% CaF₂)¹³. This solid reagent obtained by ball-milling CaF₂ with dipotassium phosphate (K₂HPO₄) consists of two crystalline components identified as K₃(HPO₄)F and K_2−xCa_y(PO₃F)_a(PO₄)_b. This mechanochemical reaction stands out because it bypasses the production, storage and complex transport chain of HF. The reactivity of this fluorinating reagent for well-established transformations unavoidably requires independent investigation, and preliminary work has revealed some limitations. One of the challenges that we encountered is its broad application to the synthesis of fluoroarenes frequently used for the production of pharmaceuticals and agrochemicals. Also, mechanochemical reactions require specialized equipment that is available in only a few laboratories. This state of play encouraged the development of an alternative strategy to prepare fluorochemicals directly from fluorspar applying mild conditions for its activation. Departing from solid-state chemistry induced by mechanical energy, we surmise that activation of fluorspar in solution, and specifically in nature’s solvent water, would represent a highly attractive approach (Fig. 1b).

**Fig. 1: Strategies to access fluorochemicals from fluorspar.**

In search for a suitable manifold to activate CaF₂ in water, we considered mild acidic conditions that provide direct access to frequently used fluorinating reagents other than HF. We propose that the synergistic use of a Brønsted acid and a fluorophilic Lewis acid would activate CaF₂, with Ca²⁺ capture as an insoluble salt and simultaneous HF capture in a form suitable for subsequent fluorination processes. A report by Lennox and Lloyd-Jones provided guidance by highlighting the beneficial influence of tartaric acid on the equilibrium in solution between aryl boronic acids, potassium fluoride and potassium trifluoroborate salts through precipitation of potassium bitartrate¹⁴. This knowledge prompted us to select a suitable Brønsted acid for AGF activation that enables concurrent Ca²⁺ sequestration as an insoluble salt. Oxalic acid (C₂H₂O₄, H₂Ox) stood out as a suitable candidate because calcium oxalate (CaOx) is poorly soluble in water (0.0061 mg ml⁻¹ (H₂O, 20 °C))^15,16. This choice was also driven by the appearance of economically viable and environmentally friendly methods for its synthesis, such as bio-based production in fast-growing microorganisms, or through carbon dioxide (CO₂) capture^6,7. The upsurge of interest in oxalic acid results from the ability of this solid organic acid to extract rare-earth elements from primary or waste ores and possibly replace existing recovery processes that depend on inorganic acids such as H₂SO₄ (ref. ¹⁷). Mechanistically, we surmise that the dissolution of AGF in water with oxalic acid in the presence of fluorophilic boric acid (B(OH)₃) may be driven by precipitation of CaOx and immediate capture of HF as a strong B−F bond (732 kJ mol⁻¹) with water as a by-product¹⁵. We prioritized B(OH)₃ as the Lewis acid prompted by the kinetic studies of Wamser on the reaction of hydrofluoric acid and B(OH)₃ in water¹⁸, and the target Balz–Schiemann reaction for the synthesis of fluoroarenes. Fluoride capture from AGF with B(OH)₃ was found to be highly effective in the presence of H₂Ox when the reaction was carried out in water for 15 h at 50 °C (Fig. 2). The [B−F] products were identified by ¹⁹F and ¹¹B nuclear magnetic resonance (NMR) spectroscopy in deuterium oxide (D₂O; Fig. 3a). By ¹⁹F NMR spectroscopy, HBF₄ (quartet, chemical shift (δ) = −150.3 parts per million (ppm), coupling constant ¹J_B–F = 1.1 Hz) and HBF₃OH (quartet, δ = −145.3 ppm, ¹J_B–F = 10.7 Hz) were identified as the major products¹⁹. In addition, a broad singlet (δ = −152.3 ppm) characteristic of difluoro(oxalate)borate species HOxBF₂ was observed in trace amount (<1%)²⁰. Notably, H₂Ox afforded a higher yield (96%) of [B−F] products than tartaric acid (2%), H₂SO₄ (69%) or HCl (60%) under these reaction conditions (Fig. 2). An investigation on fluorspar activation with a range of Brønsted acids gave insight on the interplay between acidity, denticity and solubility of the Ca²⁺ salt by-product. For monoacids (2 equiv. versus CaF₂), we noted a correlation between acidity and fluoride release in a pK_a (negative logarithm of the acid dissociation constant K_a) range between 5 and –0.5 (Supplementary Table 3). Multifunctional activators (1 equiv. versus CaF₂) were also investigated (Supplementary Table 4). Organic acids leading to five-membered Ca²⁺ chelates stood out with H₂Ox (pK_a1 = 1.3 and pK_a2 = 4.1 (where pK_a1 and pK_a2 are the pK_a values associated with the dissociation of the first and second proton of the diacid, respectively)) being the most effective activator followed by croconic acid (pK_a1 = 0.8 and pK_a2 = 2.2) and squaric acid (pK_a1 = 1.5 and pK_a2 = 3.4). Oxalic acid dihydrate (H₂Ox·2H₂O), which is more cost-effective than H₂Ox, gave the [B–F] products HBF₄ and HBF₃OH with efficacy similar to H₂Ox (total yield of 98%) upon treatment of AGF with B(OH)₃ at 50 °C for 15 h (Supplementary Table 12). Subsequent studies were therefore performed with H₂Ox·2H₂O.

**Fig. 2: Identification of oxalic acid for cooperative activation of fluorspar.**

**Fig. 3: Fluorinating reagents from fluorspar and mechanistic insight.**

It was also found that H₂Ox·2H₂O was a suitable activator for AGF when combined with silicon dioxide (SiO₂) in water at 50 °C for 15 h, with fluoride release as [Si–F] products (Si–F bond 577 kJ mol⁻¹) in 97% total yield¹⁵ (Fig. 3b). The formation of H₂SiF₆ (broad singlet, δ = −129.6 ppm), which in water exists in equilibrium with H₂SiF₅(OH) (broad singlet, δ = −128.8 ppm), was evidenced by ¹⁹F NMR spectroscopy of the reaction mixture²¹. Two triplets were also observed (δ = −124.5 ppm and −135.9 ppm, ²J_F–F = 8.9 Hz) and assigned to H₂OxSiF₄ (supported by ²⁹Si NMR spectroscopy; Supplementary Figs. 14 and 15)²². Subsequent studies focused on the synthesis of frequently used fluorinating reagents from these [Si–F] products (Fig. 3b). For this purpose, AGF (1.1 equiv.) was reacted with H₂Ox·2H₂O (1 equiv.) and SiO₂ (0.4 equiv.) in water at 50 °C for 15 h. The reaction mixture was filtered and treated with potassium hydroxide (KOH). The insoluble by-product of this filtration was unambiguously characterized as CaOx·H₂O by powder X-ray diffraction (Supplementary Fig. 17). Neutralization with KOH (2 equiv.) led to the formation of K₂SiF₆ (Supplementary Fig. 20), whereas treatment with excess KOH (6 equiv.) afforded KF^AGF (85% yield, calculated from AGF; purity analysis is included in Supplementary Information)²³. A similar protocol gave NaF^AGF (85% yield, 94% purity) and CsF^AGF (89% yield, 96% purity) from NaOH and CsOH·H₂O, respectively. Alternatively, treatment of the filtered reaction mixture with tetramethylammonium hydroxide (6 equiv.) afforded tetramethylammonium fluoride hydrate, which was converted to tetramethylammonium tert-amyl alcohol fluoride [Me₄NF·(^tAmOH)] ^AGF (88% yield), a fluorinating reagent well documented for nucleophilic aromatic fluorination (S_NAr)²⁴. This strategy also enabled the preparation of tetrabutylammonium fluoride hydrate, which was converted to the bench-stable reagent tetrabutylammonium tetra(tert-butyl alcohol) fluoride [Bu₄NF·(^tBuOH)₄]^AGF (71% yield calculated from AGF)²⁵. Mechanistic investigations by ¹⁹F NMR spectroscopy in D₂O established whether HF is formed upon dissolution of CaF₂ with H₂Ox (Supplementary Fig. 6). In the absence of a fluorophilic Lewis acid, HF is indeed observed (singlet, δ = −166.0 ppm), and an equilibrium is established with the amount of HF plateauing after 3 h at approximately 10% (ref. ²⁶; Fig. 3c). In the presence of either B(OH)₃ or SiO₂, the equilibrium is displaced via the precipitation of highly insoluble CaOx and immediate HF capture by the Lewis acid. Under these conditions, the singlet diagnostic of HF was not detected by ¹⁹F NMR spectroscopy during the entire course of the reaction. As anticipated, the reaction of AGF with fluorophilic Lewis acid in the absence of H₂Ox resulted in no fluoride release (Supplementary Information). These data highlight how Brønsted and Lewis acid cooperativity allows for fluorspar activation under mild conditions, prevents HF from building up, and enables access to HBF₄ (aq.), KF, NaF, CsF, Me₄NF·^tAmOH and ⁿBu₄NF·(^tBuOH)₄, directly from fluorspar.

Further studies compared the reactivity of H₂Ox·2H₂O and H₂SO₄ with AGF. At 50 °C, the combination of AGF, concentrated H₂SO₄ and B(OH)₃ increased the yield of [B−F] products to 97% when the reaction time was extended to 24 h (versus 69% yield after 15 h). This reaction was also effective at 25 °C, offering [B−F] products in 94% yield (average of 2 runs) after a reaction time of 48 h. For comparison, H₂Ox·2H₂O afforded [B−F] products in 98% yield under these conditions (25 °C, 48 h). When SiO₂ served as the fluorophilic Lewis acid, activation of AGF with H₂Ox·2H₂O was effective at 25 °C, albeit requiring a prolonged reaction time of 72 h. After neutralization with KOH, KF was indeed isolated in 74% yield. Replacement of oxalic acid for H₂SO₄ did not give satisfactory results. At 50 °C for 24 h, the reaction of AGF, concentrated H₂SO₄ and SiO₂ gave only partial fluoride release as H₂SiF₆ and H₂SiF₅OH with a total [Si−F] product yield of 48% (quantified by ¹⁹F NMR spectroscopy). At 25 °C for 72 h, the total [Si−F] product yield was reduced to 29%. Notably, the treatment of the [Si−F] solutions with KOH afforded a solid material containing both KF and K₂SO₄ (46% KF (50 °C) and 36% KF (25 °C)) as evidenced by powder X-ray diffraction analysis, two salts that are challenging to separate (KF, 1,020 mg ml⁻¹; K₂SO₄ 120 mg ml⁻¹ (H₂O, 25 °C))¹⁵. The formation of K₂SO₄ is indicative of the incomplete reaction between AGF, H₂SO₄ and SiO₂. The overall superior reactivity of H₂Ox·2H₂O compared with H₂SO₄ correlates with the solubility of the calcium by-product formed upon AGF activation (CaOx.H₂O, 0.0061 mg ml⁻¹ versus CaSO₄.2H₂O, 2.1 mg ml⁻¹ (H₂O, 20 °C))¹⁵.

With an effective strategy to convert AGF into HBF₄ (aq.), KF and Me₄NF·^tAmOH, we had a duty to demonstrate that these AGF-derived reagents react as expected, with the synthesis of industrially valuable fluoroarenes and a focus on those not accessible via mechanochemical activation of AGF using a phosphate salt. For Balz–Schiemann chemistry, a two-step procedure followed the preparation of HBF₄ (aq.); addition of tert-butyl nitrite to a solution of aryl amine and aqueous HBF₄ led to the precipitation of the corresponding aryl diazonium tetrafluoroborate salt, which was isolated and subsequently heated to liberate the desired fluoroarene²⁷ (Fig. 4a). The protocol was validated first with 4-bromoaniline, AGF-derived HBF₄ (HBF₄^AGF; 1.1 equiv.) and tert-butyl nitrite (2 equiv.). Heating the resulting diazonium salt at 90 °C in chlorobenzene gave 4-bromofluorobenzene in 98% yield (as measured by ¹⁹F NMR spectroscopy), a key intermediate featured in the synthesis of the antidepressant citalopram²⁸. This chemistry was subsequently applied to prepare multiple fluoroarenes frequently used as building blocks in the synthesis of various organo-fluorine-containing drugs in up to 87% yield (dediazoniation yield). Examples include the precursors of lipitor (cholesterol lowering), norfloxacin (antibiotic), raltegravir (HIV), eravacycline (antibiotic), rosuvastatin (cardiovascular disease), flurbiprofen (anti-inflammatory), flunarizine (vertigo) and ezetimibe (cholesterol-lowering drug)^{29,30,31,32,33,34,35,36}. The methodology was also suitable for the preparation of fluoropyridines (4, 8 and 13) that are building blocks for drugs such as MK2 inhibitors (autoimmune diseases) and vericiguat (heart failure), and agrochemicals including the herbicide clodinafop^37,38,39. For diazonium salts prone to decomposition, we developed a one-pot protocol using tert-butyl nitrite and LiBF₄ enabling access to fluoroarenes (4 and 9) without the necessity to isolate the diazonium salt. For this purpose, LiBF₄ was prepared from fluorspar upon treatment of HBF₄^AGF with Li₂CO₃ (Supplementary Information).

**Fig. 4: Scope of fluoroarenes prepared from fluorspar.**

Next, AGF-derived KF (KF^AGF) (90% purity) was found to achieve high-yielding fluorination of chloroarene substrates using Me₄NCl (5 mol%) and DMSO as solvent to provide fluoroarenes 14 to 17 (Fig. 4b). We noted that the performance of KF^AGF was comparable to commercial KF (99% purity; Supplementary Table 22). Aromatic fluorodenitration using AGF-derived Me₄NF·^tAmOH proceeded in DMSO (30−80 °C) to access 2,6-difluorobenzonitrile (18), 4-fluoronitrobenzene (19) and 2-fluorobenzonitrile (20) in high yield.

The successful cooperative activation of AGF in water and its application to the synthesis of fluoroarenes encouraged an investigation on the reactivity of lower-grade metallurgical fluorspar (metspar). These studies were performed with materials sourced from China (Metspar^I CaF₂ (85%), SiO₂ (10%), CaCO₃ (<5%), S (0.12%), P (0.1%)), and Mexico (Metspar^II from CaF₂ (88.98%), SiO₂ (5.43%), CaCO₃ (4.02%), Al₂O₃ (0.41%), Fe₂O₃ (0.24%), S (0.011%), P (0.023%), Pb (<0.001%)). Both metspar^I and metspar^II yielded the [B–F] products HBF₄ and HBF₃OH with an overall yield of 83% each, upon activation with B(OH)₃ and H₂Ox·2H₂O at 50 °C for 15 h (Supplementary Information). The reaction of metspar with H₂Ox·2H₂O and SiO₂ also enabled the preparation of metspar-derived KF (KF^M) (53% (KF^M(I)) and 63% (KF^M(II)) yield, calculated from metspar I or II, respectively). Fluoroarenes 1, 14 and 16 were prepared in good yield using HBF₄^M(I) or KF^M(I) despite the reduced purity of metspar.

[ad_2]

Source link

Direct radical functionalization of native sugars

[ad_1]

Widely distributed across the three domains of cellular life forms, carbohydrates play pivotal parts in many biological processes^4,5,6,7. Nature often provides greatly altered function simply through the attachment of a glycosyl moiety. Because of their importance, substantial efforts have been devoted to accessing these saccharides and their conjugates to better understand their properties, functions and potential disease-related roles and to enable the discovery of sugar-based therapeutics^8,9,10. The difficulty of extracting notable quantities of pure samples from nature has prompted chemists to secure most saccharides by synthetic means. To this end, non-enzymatic chemical glycosylation^{11,12,13,14,15} represents the cornerstone of carbohydrate chemistry by offering a reliable avenue to assemble a vast array of natural and non-natural glycoside entities. However, unlike enzymatic machineries that can mediate glycosylation by using unprotected polyhydroxylated glycosyl donors with excellent regiocontrol^16,17, established chemical glycosylation methodologies are less precise and typically require cumbersome protecting-group strategies^{3,11,12,13,14,15} to overcome the problem of site selectivity. These complications are highlighted in the existing synthetic routes to C-glycosyl compounds¹³, a parallel carbohydrate class that is rarer in nature but has gained increasing prominence as robust and often more biologically potent surrogates of O-glycosides in developing medications to treat cancer, diabetes and other illnesses^18,19.

In contrast to the highly site-selective nature of enzymatic C-glycosylation (Fig. 1a), the state-of-the-art advances in non-enzymatic chemical C-glycosylation often require multi-step reaction sequences (hydroxyl group protection, functionalization and deprotection) involving delicate control and/or harsh reaction conditions to transform fully unprotected native sugars (the most abundant form in nature) into tailored glycosyl precursors containing anomeric leaving groups such as halides^{20,21,22,23,24}, esters^25,26,27, sulfoxides^28,29 or sulfones^30,31,32, setting the stage for the ensuing carbon–carbon bond-forming reaction to deliver the desired unprotected C-glycosyl compound only after eventual deprotection (Fig. 1b). The practical drawbacks and inefficiencies of these approaches consequently limit their use in synthetic glycochemistry and prevent further applications under intricate biological conditions. Thus, enacting a regime that allows direct coupling of native sugars for broad-scope glycosylation³³ to access stereoisomerically pure C-glycosyl compounds and other hydrolytically stable and medicinally important variants (such as S– and Se-glycosides)^34,35,36 as well as C-linked glycoproteins is a longstanding goal in glycoscience research. However, this has remained unknown owing to numerous challenges associated with efficiency, selectivity and biocompatibility.

**Fig. 1: Design of a protecting-group-free ‘cap and glycosylate’ blueprint for direct functionalization of native sugars to robust glycosides.**

Inspired by reports of biological S-glycosylation in which S-glycosyltranferases mediate the formation of stable S-glycosidic linkages using unprotected nucleotide sugars generated from their native variants by regioselective anomeric phosphorylation^37,38, we reasoned that a biomimetic approach could be adopted to preferentially activate and substitute the anomeric hydroxyl group (hemiacetal) in a native sugar in its cyclic form (capping). This would afford a thioglycoside intermediate that, under suitable conditions, could undergo stereocontrolled desulfurative cross-coupling^39,40,41 with an appropriate reagent in a single operation (glycosylation). Just as in nature, the activated glycosyl donor that was temporarily generated remains traceless. However, several challenges have to be addressed for the success of this ‘cap and glycosylate’ strategy. First, the multiple hydroxyl groups must be distinguished to ensure selective masking of the hemiacetal to form a transient thioglycosyl donor. Second, the donor must be sufficiently reactive to participate in cross-coupling without competitive interference or reaction on other hydroxyl sites, which would otherwise result in undesired reactions and intractable mixtures. Added to these is the challenge of controlling the stereochemical outcome of anomeric functionalization in the context of a complex, polyhydroxylated carbohydrate residue.

Here we report a metal- and protecting-group-free blueprint that enables the direct anomeric functionalization of unprotected monosaccharides and oligosaccharides in their native forms by radical-based cross-coupling with various electrophiles under mild photoirradiation conditions (Fig. 1c). This ‘cap and glycosylate’ approach eliminates the need for pre-installation and removal of protecting groups, solving an enduring problem in the field and providing a general platform to accelerate the preparation of robust carbo-, thio- and selenoglycosyl compounds as well as O-glycosides in high regio- and stereoselectivity. We also show that the protocol is amenable to the direct chemical synthesis of unprotected C-glycosylproteins in a post-translational manner that is complementary to the analogous biological O– and N-glycosylation processes.

Considering the susceptibility of certain transition metals to inhibition by polar hydroxyl groups⁴², we sought to engineer a metal-free protocol that harnesses the reactivity of electron-deficient alkyl sulfides to undergo desulfurative transformations on photoactivation^39,40,41. We first evaluated reaction parameters that promote regioselective nucleophilic substitution (capping) using d-glucose 1 as the model substrate (Supplementary Table 3). Taking advantage of the greater acidity of the anomeric OH with respect to other hydroxyl units⁴³, various activating agents (R−LG) were examined to convert 1 to its bench-stable 2,3,5,6-tetrafluoropyridine-4-thioglycoside derivative 2 under weakly basic conditions (Fig. 2a). In the presence of commercially available 2-chloro-1,3-dimethylimidazolinium chloride (DMC) as activator and triethylamine as base, 2 was obtained in 85% yield (72% isolated yield) and more than 95:5 β:α ratio at 0 °C within 2 h. In our hands, 2 (white solid) could be stored in air on the bench over months without noticeable decomposition. It is worth noting that the C1 stereochemistry of this S-glycosyl donor is inconsequential as it will be transformed into a glycosyl radical species during the course of C−C bond formation in the subsequent step (Fig. 3a). Other analogues of DMC (3 and 4) led to markedly diminished yields, whereas other commonly used reagents such as chlorophosphonium salt 5 and a combination of 2-chloro-4,6-dimethoxy-1,3,5-triazine (CDMT) and N-methylmorpholine (NMM) failed to promote the reaction.

With DMC identified as the most effective activator, we used the nucleophilic substitution conditions to synthesize not only 2 but also a range of unprotected (hetero)aryl thioglucosides (6−9) for comparison. To drive glycosylation, we subjected the thioglucosides to a reaction with acrylate 10 under visible light illumination³⁰. After an extensive survey of conditions (Supplementary Table 4), we discovered that 2 underwent desulfurative C−C coupling to deliver unprotected C-alkyl glucoside 11 in 96% yield (82% isolated yield) and more than 95% α selectivity using a combination of Hantzsch ester as reductant, 1,4-diazabicyclo[2.2.2]octane (DABCO) and dimethyl sulfoxide (DMSO) as solvent under blue LED irradiation at ambient temperature (Fig. 2b). To our knowledge, this reaction represents the first successful use of 2,3,5,6-tetrafluoropyridine-4-thioglycoside as a new class of glycosyl donor in chemical glycosylation.

By contrast, poor conversion was observed with the less redox-active S-glucosides derived from other less electron-withdrawing (hetero)aryl thiols (6−9), highlighting the importance of the fluorinated heteroaromatic moiety for photoinduced cross-coupling^39,40,41. This was supported by cyclic voltammetry studies showing that 2 has the least negative reduction potential (Supplementary Figs. 2–6), which is comparable to that of a redox-active heteroaryl glycosyl sulfone³⁰. By contrast, excluding the light source, Hantzsch ester or DABCO was detrimental to the reaction, and changing the base or solvent led to lower yields. To demonstrate the power of the ‘cap and glycosylate’ approach by traceless activation, we showed that α-11 could be generated from 1 in a single sequence without the need for isolating the S-glycosyl intermediate 2. The overall step efficiency and yield (64% yield and 52% isolated yield) offer marked advantages over previous chemical C-glycosylation approaches that require multiple steps (Fig. 1b).

Experiments were conducted to gain insight into the individual processes for native sugar activation and cross-coupling. As shown in Fig. 2a, nucleophilic substitution of d-glucose 1 afforded 2,3,5,6-tetrafluoropyridine-4-thioglycoside 2 in 85% yield (72% isolated yield) and more than 95:5 β:α ratio. On the contrary, we found that the corresponding thioglycoside 13 was secured in 44% yield (30% isolated yield) and more than 95:5 α:β ratio from d-maltose 12 under the same established conditions (Fig. 3a). In solution, the α and β anomers of native sugars (1, 12) can interconvert and exist in equilibrium; each anomer individually reacts with DMC before undergoing stereoinvertive nucleophilic displacement⁴³ by the thiol (Supplementary Fig. 7). Alternatively, the 2-OH group of the DMC-activated β anomeric intermediate may engage in neighbouring group participation by an intramolecular nucleophilic attack to generate a 1,2-anhydro species⁴³, which is susceptible to site-selective ring cleavage by the thiol nucleophile. This pathway is probably insignificant in the reaction leading to 13, given that β-13 was detected in minor amounts. For other saccharides (Fig. 4), the various pathways for nucleophilic substitution may be favoured to different extents in the reaction system⁴³. The structure of the 2,3,5,6-tetrafluoropyridine-4-thioglycoside derived from d-mannose was confirmed by X-ray crystallographic analysis (Supplementary Information section 7).

**Fig. 4: Scope of the reaction with various native sugars.**

Subjecting 2 and 13 separately to standard cross-coupling conditions with an acrylate gave 11 and 15, respectively, both of which possess the same sense of anomeric selectivity (Fig. 3a). This notably implies that, unlike heterolytic glycosylations, the C1 stereochemistry of the S-glycosyl donor is inconsequential, highlighting the distinct advantage of the present strategy in transforming mixtures of unprotected native sugar isomers, through their thioglycoside derivatives, into stereoisomerically pure glycosides in a streamlined fashion. In a separate study, the addition of exogenous (2,2,6,6-tetramethylpiperidin-1-yl)oxyl (TEMPO) inhibited the photoinduced transformation of 2 to 11 (Fig. 3b). High-resolution mass spectrometry (HRMS) analysis revealed the formation of a complex that can be ascribed to a TEMPO-glycoside adduct 16, providing evidence that a sufficiently long-lived glycosyl (anomeric) radical species is generated during the process. These processes are in contrast to heterolytic glycosylations that essentially lack the formation of a clear intermediate species (for example, glycosyl cation).

We further explored the nature of the photoinduced reaction (using 2 as the model substrate) through ultraviolet–visible absorption (UV–vis) spectroscopy (Fig. 3c). Independent absorption spectra of 2 and DABCO revealed bands largely in the UV region, and a mixture of these two components led only to a small redshift that extends into the visible region (>400 nm). By contrast, a DMSO solution of Hantzsch ester exhibited strong absorption in the visible region, but no noticeable changes were observed with a mixture of Hantzsch ester and DABCO. A mixture of 2 and Hantzsch ester showed a slight bathochromic shift, which was amplified when 2, Hantzsch ester and DABCO were combined together in solution. These results indicate the generation of a putative ternary complex^44,45 between 2, Hantzsch ester and DABCO, which is proposed to absorb visible light and undergo fragmentation to the glycosyl radical.

The studies presented here support a mechanism as shown in Fig. 3d. Site-selective capping of the more acidic anomeric hydroxyl motif by DMC forms an activated leaving group that undergoes facile nucleophilic attack by 2,3,5,6-tetrafluoropyridine-4-thiol under basic conditions, driven by concomitant generation of 1,3-dimethylimidazolidin-2-one (DMI) as a by-product⁴³. Formation of a 1,2-anhydro species before nucleophilic substitution could also occur and cannot be completely ruled out (Supplementary Fig. 7). The resulting thioglycoside intermediate is postulated to associate with Hantzsch ester and DABCO in solution, affording a ternary complex that can absorb visible light to trigger photoinduced electron transfer (PET)⁴⁶. Consistent with previously documented reactions^39,40,41, the highly electrophilic nature of the fluorinated heteroaryl motif renders the thioglycoside sufficiently redox-active for PET. This delivers dihydropyridine radical 17 and a radical anion 18, which is prone to desulfurative fragmentation to give a glycosyl radical species and 2,3,5,6-tetrafluoropyridine-4-thiolate (the conjugate acid was detected in the reaction mixture). Subsequent reaction of the glycosyl radical with an electrophilic cross-coupling partner, facilitated by 17, proceeds in a stereoselective manner under kinetic control^30,47,48 to give the desired unprotected glycoside.

The generality of our protecting-group-free protocol was highlighted by the wide spectrum of native mono- and oligosaccharides that could be reliably transformed into fully unprotected C-alky glycosyl compounds (Fig. 4) through their 2,3,5,6-tetrafluoropyridine-4-thioglycoside precursors, which were either isolated or generated in situ and used (without purification) for cross-coupling. Representative examples include pyranoside products constructed from biomass-derived monosaccharides (19−21, 24), rare sugars (22, 23) and non-natural l-glucose (25). More complex glycans from natural sources also served as effective substrates to deliver the corresponding C-alkyl glycosyl compounds (15, 26−29) in good efficiency. Across the board, good to excellent stereocontrol was observed.

Besides α,β-unsaturated carbonyl compounds, other alkenes were investigated as cross-coupling partners (Fig. 5a). Densely functionalized acrylates and acrylamides conjugated to biologically active compounds (30, 31), an aminosalicylate (32), an amino sugar (33) and oligopeptides (34−36) were compatible substrates, providing access to highly polar C-glycosylated conjugates bearing multiple acidic and basic sites. This offers a straightforward way to glycosylate complex molecules with native sugars for various applications, including the design of sugar-based peptidomimetics^23,28. Other Michael acceptors such as vinyl sulfone (37), vinylphosphonate (38) and vinylboronate (40) as well as less electrophilic vinyl silane (39) and allyl acetate (41) also underwent efficient reaction to furnish the desired C-alkyl glycosyl adducts bearing functional groups that could serve as useful synthetic handles for further manipulations. Of particular note, cross-coupling was found to proceed even in the presence of a less-activated alkyl-substituted alkene (42). Metabolically stable pseudo-oligosaccharide⁴⁹ building blocks such as C-glycosidic disaccharide 43 featuring two newly formed stereocentres could be expeditiously assembled with complete stereocontrol through reaction with an exo-glucal as radical acceptor.

**Fig. 5: Synthesis of diverse classes of robust glycosides and glycoconjugates.**

To showcase the versatility of the ‘cap and glycosylate’ approach in securing other categories of unprotected saccharides, we replaced the alkene coupling partner with other electrophilic reagents that could participate as radical acceptors. Using a haloalkene reagent (Fig. 5b), a C-alkenyl glycosyl compound (44) was successfully secured in high anomeric selectivity; this process is postulated to proceed through a glycosyl radical addition–reduction–β halide elimination pathway²⁴. C-Heteroaryl glycosylation could also be realized by direct coupling with heteroarenes under acid-free conditions, delivering unprotected 45−47 selectively at the most electron-deficient sites, which is congruent with a previous report involving fully protected glycosyl radicals⁵⁰. Our heteroarylation approach is complementary to a previously reported metallaphotoredox-enabled deoxygenative strategy (incompatible with native sugars) that involved pre-activation of an exposed anomeric hydroxyl followed by cross-coupling with a heteroaryl halide⁵¹.

Beyond C-glycosylation, we extended the protecting-group-free reaction manifold to the preparation of other glycomimetics such as selenoglycosides (Fig. 5c) and thioglycosides (Fig. 5d). Along with C-glycosyl compounds, these entities have found many applications as robust substitutes of the naturally occurring O-saccharides, thus efficient ways to synthesize them in high stereochemical purity are highly desirable. Both unprotected Se-glycosides (48, 49) and S-glycosides (50−54) were accessible through reaction with diselenide or disulfide reagents⁵², respectively, comparing favourably with previous protocols that relied on laborious preparation of glycosyl precursors. It is to be noted that 50−54 were exclusively isolated as α anomers (compared with β anomers from nucleophilic substitution in Fig. 2). Similar to the C-glycosyl cases in Figs. 2 and 4, the observed stereochemical outcome for 48−54 could be rationalized by the stabilizing orbital interaction between the nonbonding electron pair of the ring oxygen and the σ* of the incipient bond at the anomeric carbon in the transition state, as the glycosyl radical reacts with the electrophile^47,48 (Fig. 3d). O-glycosylation⁵³ with phenols could also be achieved by tuning the photoinduced cross-coupling conditions using iodide as reductant^40,45 (Extended Data Fig. 1).

Encouraged by our successful efforts in small-molecule glycosyl compound synthesis, we attempted to test our ‘cap and glycosylate’ protocol in the synthesis of glycoproteins, which are known to mediate numerous essential biological processes. In nature, glycoproteins are typically formed by linking sugar units to O– or N-containing side chains of amino acid residues serine, threonine or asparagine using glycosyltransferases, such as the attachment of O-linked-β-d–N-acetylglucosamine (O-GlcNAc) to serine or threonine residues by O-GlcNAc transferase. However, this glycosylation can be reversed by intracellular glycosidases, and such a write-and-erase dynamic process makes it challenging to probe the biological functions of glycoproteins. In this context, chemical approaches to generate non-cleavable glycoproteins (such as C-glycosylproteins) offer alternative and promising strategies for systematically investigating glycoprotein functions. Nevertheless, post-translational chemical glycosylation of proteins, particularly the attachment of sugar units to proteins by direct anomeric functionalization, is largely unexplored in synthetic carbohydrate and protein chemistry⁵⁴. This may be ascribed to the lack of suitable unprotected glycosyl precursors that are stable yet sufficiently reactive, as well as the stringent requirements for biocompatible conditions, including water compatibility (which quenches heterolytic chemical glycosyl donors), ability to remain non-destructive to biological substrates and low reactivity towards the biogenic functional groups that are present in most biological environments⁵⁵. Owing to the insolubility of Hantzsch ester in the necessary aqueous medium, photoinduced cross-coupling conditions were instead based on the formation of charge-transfer complexes between 2,3,5,6-tetrafluoropyridine-4-thioglycoside and bis(catecholato)diboron (B₂Cat₂) as reductant⁴¹ (Supplementary Tables 6–8).

After identifying the optimal conditions (500 equiv. of B₂Cat₂, 4 °C, 1 h, pH 8.0 in Tris buffer) as shown in Fig. 6, three mammalian glycoprotein sugars (d-mannose, d-galactose and N-acetylglucosamine) were selected to react with dehydroalanine (Dha)-tagged proteins^56,57 with varying architectures and functions, including histone H3 (a small α-helical nuclear protein), PanC (Mycobacterium tuberculosis pantothenate synthetase enzyme)⁵⁸, PstS (a bacterial phosphate transport protein)⁵⁹ and SsβG (an αβ₈ TIM (triose-phosphate isomerase) barrel enzyme)⁶⁰. In the event, all the examined proteins were found to be competent glycosyl radical acceptors under the established conditions, with the desired C-alkyl glycosylproteins secured in good to excellent yields across the board regardless of their size and fold. The stereochemistry of the newly formed C−C bond at the anomeric carbon (α selectivity) is presumed to be identical to that of small-molecule glycosylation (Fig. 4). Notably, histone H3–GlcNAc–Ala10 generates a non-cleavable mimetic of the reported epigenetic mark GlcNAc–Ser10 (ref. ⁶¹); access to this glycoprotein conjugate may shed light on the poorly understood biological role of this post-translational modification process. Similar efficiencies were also observed in the reactions of different thioglycosyl donors with each given protein (about 85% conversion for eH3–Dha9, about 80% conversion for H3–Dha10, about 55% conversion for TEV H3–Dha2, about 80% conversion for PanC–Dha44 and about 70% conversion for PstS-Dha57). About 5% of a minor product featuring two units of GlcNAc addition was detected for PstS–GlcNAc–Ala57, which we ascribe to non-specific glycosylation of lysine residues⁶² (Supplementary Table 9 and Supplementary Fig. 15). Similar to the examples in Figs. 2b, 4 and 5, traceless activation by in situ formation of the S-glycosyl intermediates could be implemented without compromising on protein glycosylation efficiency, thereby exemplifying the power of our protecting-group-free ‘cap and glycosylate’ approach allowing native sugars to be directly used for glycosylating proteins post-translationally.

**Fig. 6: Application to direct post-translational chemical glycosylation of proteins.**

[ad_2]

Source link

Dynamic stereomutation of vinylcyclopropanes with metalloradicals

[ad_1]

Whereas the reactivity principles of organic free radicals are well established and have found widespread implementation in synthetic methodology developments⁵, there is comparably much less application of metalloradical-based reactivity, especially with regard to homogeneous catalysis and the discovery of potentially unique and enabling reactivity modes. Metalloradical catalysis operates via open-shell metal complexes and intermediates throughout the entire catalytic cycle^1,2,3. Exemplary manifestations of this concept in synthesis include the radical relay catalysis of Cu^(I), Ti^(III) (refs. ^7,8), porphyrin-type Co^(II) or Fe^(III) complexes⁹, which are based on the transposition of the radical character from the metal to the organic moiety through covalent bonding (Fig. 1). These catalytically generated radicals have been shown to give tamed reactivity¹⁰ compared with organic free radicals that do not have the metal bound—a feature also harnessed in several metalloenzymes. However, compared with omnipresent homogeneous metal catalysis based on closed-shell species, there is comparably much less exploitation and implementation of metalloradical catalysis in synthesis. To meet the ever increasing demands for greater sustainability, and to capitalize on the frequent occurrence of odd-oxidation states in non-precious metals, a greater understanding of the principles and potential of metalloradical catalysis would therefore be greatly enabling.

**Fig. 1: Reactivity modes of (di)vinylcyclopropanes.**

We herein disclose a metalloradical catalysis mode that is initiated by non-covalent interaction via π-coordination, which leads to reactivity that is fundamentally in contrast to organic free radical reactivity and enables a reversible cis/trans-isomerization of vinylcyclopropanes under chiral inversion.

Vinylcyclopropanes serve as valuable synthetic precursors for cycloadditions or rearrangements and are also key structural units in natural products and bioactive compounds, such as the antivirals Simeprevir (against hepatitis C) or Danoprevir (against COVID-19)^11,12. The geometry of the vinylcyclopropane, that is, cis versus trans, impacts its overall function, and stereoselective synthetic approaches to these motifs have been developed accordingly^13,14. Although remarkable developments have been achieved, the selectivity of cyclopropane construction can be substrate-specific or involve multistep processes, especially in efforts to reach high enantioselectivity along with cis/trans selectivity, which is well addressed for higher substituted cyclopropanes, but still challenging for 1,2-disubstituted analogues^13,15,16,17. A complementary strategy would be the unselective (and potentially enantiopure) synthesis of cis/trans mixtures of vinylcyclopropane derivatives, followed by their isomerization towards a single isomer. Ideally, this process is steerable towards either isomer without compromises on enantiopurity. However, reversible cis/trans isomerization without loss of enantiopurity is fundamentally unknown. If realizable, it could find numerous applications, for example, in strategic uses of divinylcyclopropanes. Such species are more (and sometimes solely) stable in their trans-geometry and feature in several natural products, while the cis-isomer is prone to undergo rearrangement^6,18. This orbital-symmetry-controlled stereo-retentive electrocyclic reaction has served as a key strategic transformation in total syntheses, but frequently requires the cis-isomer to be made in situ (owing to its high reactivity), which was achieved predominantly through direct construction of the cyclopropane ring. Construction of the more stable trans-divinylcyclopropane, followed by isomerization to cis requires high temperature (approximately 200 °C) and occurs under racemization^19,20,21,22. Such thermally or photochemically induced homolytic scissions towards diradical intermediates²³ can also trigger the cis-to-trans isomerization of alternatively substituted cyclopropanes and, with appropriate metal complexation for specific substrates, also at lower temperature (80 °C)²⁴, albeit under loss of enantiopurity towards the racemic, thermodynamic trans-isomer. Isomerization may also occur in a polar mechanism via zwitterionic intermediates, oxidatively²⁵, reductively or under Lewis acid catalysis for specific substrates with the analogous challenges on enantiopurity (that is, its loss)¹⁸, although remarkable strides towards photochemically assisted chiral resolution with a chiral catalyst have been made recently for specifically substituted cyclopropyl ketones^26,27. There are isolated reports of cis/trans isomerization with a precious metal (Au, Rh)^28,29,30,31 at elevated temperature for which the wider potential (that is, scope) and stereospecificity have not (yet) been established. On the other hand, transition metal catalysis based on Ni, Pd, Rh, Ir or Fe has been explored for their addition to vinylcyclopropanes and subsequent follow up chemistry (for example, rearrangement, cycloaddition) towards products of decreased ring-strain^32,33,34,35.

We reported previously that the N-heterocyclic carbene (1,3-bis(2,6-diisopropylphenyl)-1,3-dihydro-2H-imidazol-2-ylidene; IPr)-derived Ni^(I) dimer 1 interacts with a double bond to trigger homolytic Ni–Ni scission and formation of an olefin-bound Ni^(I) metalloradical complex³⁶. The radical character in this π-coordinated metalloradical is located primarily at the Ni-centre, and this metalloradical consequently differs from covalently attached metalloradicals as introduced above, where the radical character is transposed primarily to the organic moiety via π-bond cleavage and formation of a C_sp2-hybridized carbon-centred radical. The Ni^(I) metalloradical–olefin complex was found previously to have an inherent driving force for reorganization, which was leveraged in selective olefin migration³⁶. Our preliminary tests with a specific diphenyl substituted vinylcyclopropane indicated a partial isomerization instead of opening, which fundamentally questions the validity of vinylcyclopropanes as mechanistic probes to test for metal-based radical intermediates.

To date, neither the factors that dictate isomerization versus opening have been elucidated, nor the wider scope and potential of this reactivity in synthesis. We therefore set out to resolve these fundamental questions with a view to devise a selective and general isomerization for vinylcyclopropanes that proceeds without any loss of enantiopurity.

We initially set out to study the reactivity of Ni^(I) dimer 1 with the 4-methoxyphenyl-substituted vinylcyclopropane 2 and evaluated various reaction parameters, which ultimately revealed that, with only 1 mol% of catalyst at room temperature, the cis-cyclopropane 2 was fully transformed to the corresponding trans-isomer within 5 min (in 90:10 trans:cis selectivity and 98% yield) in dioxane (or tetrahydrofuran). Such mildness and extreme speed under such low catalyst loading was remarkable, considering that the Ni-free isomerization requires heating at over 200 °C (ref. ²⁰).

As outlined above, the thermal isomerization is postulated to proceed via homolytic scission and diradical intermediates that is accompanied with loss of stereochemical information if a chiral starting material is used. We therefore next examined the isomerization for its potential to retain enantiopurity. To this end, we prepared (R,R)-cis–2 in >99.9% enantiomeric excess (e.e.) and subjected it to the same isomerization conditions (Fig. 2a). The corresponding trans-product 2 was generated under retention of enantiopurity in 98.8% e.e. (in 88% yield) within 5 min at room temperature using 1 mol% of Ni^(I) dimer 1. To the best of our knowledge, such a mild and rapid cis/trans isomerization of vinylcyclopropanes without any loss of enantiopurity is unprecedented to date.

Our closer inspection of the absolute stereochemistry of the product revealed that the carbon distal from the vinyl substituent was inverted selectively during isomerization, having generated (R,S)-trans-2 (Fig. 2a). Ni^(I) therefore must have a profound effect on the overall reaction path, in locking the stereocentre adjacent to the vinyl substituent, while also influencing the overall driving force to favour the strained, ring-closed product over ring-opened/rearranged products.

To gain greater insight, we undertook computational studies at CPCM (1,4-dioxane) M06L/def2-TZVP//MN15/6-31G(d)(SDD) level of theory, following a comprehensive methods assessment (see Supplementary Information). The calculations suggest that, owing to the relatively weak Ni–Ni bond in dimer 1, homolytic scission towards the Ni^(I) monomer and π-coordination to the olefin occurs readily. The vinylcyclopropane serves solely as π-ligand to the Ni^(I) and no radical character is induced to the organic moiety, residing solely on the [Ni^(I)Cl(IPr)] with 88% of the radical being located in the d_xy orbital of Ni^(I) (Fig. 2c). However, a geometric change from tetrahedral to square-planar arrangement of this metalloradical–olefin complex, which is 4.3 kcal mol⁻¹ higher in free energy, leads to partial induction of radical character onto the vinylcyclopropane (see Supplementary Information for details). Subsequent cyclopropane C–C bond cleavage proceeds with an overall activation free energy barrier of 15.8 kcal mol⁻¹ and is endergonic by 7.4 kcal mol⁻¹. The ring-opened intermediate Int2 is square-planar in geometry and the energetically favoured state is of broken-symmetry with two electrons in parallel spin located on the Ni^(II)-centre and a radical located on the remote carbon, which resulted from homolytic scission of the C–C bond in the cyclopropane. Our examination of a potential alternative Ni^(I)/Ni^(III) oxidative addition directly into the cyclopropane ring suggests that, although a higher energy transition state can be located for the direct addition by Ni^(I), no Ni^(III) intermediate is subsequently formed. Instead, the intrinsic reaction coordinate leads also to Int2 (see Supplementary Information for additional details). Rotation of intermediate Int2 is facile, proceeding with an activation free energy barrier of ΔG^‡ = 2.2 kcal mol⁻¹, and subsequent ring closure is thermodynamically downhill relative to the opened intermediate (Fig. 2c). Only carbon (b) is rotatable; the configuration of carbon (a) is locked through the formation of a Ni^(II)–π–allyl complex. Our calculations suggest that any alternative isomer (or spin state) of Int2 that could allow for carbon (a) inversion lies 7–25 kcal mol⁻¹ higher in energy than the transition state for ring closure TS3 (see Supplementary Information for additional details). Consequently, the lack of racemization observed under Ni^(I)-catalysed isomerization (as opposed to thermal isomerization) means that, although formally in both processes a homolytic scission of the central cyclopropane bond takes place, the Ni-coordination locks one stereocentre through a π–allyl complex, leaving only one centre to rotate (and invert).

Another notable difference to organic free radical reactivity is that addition of an organic free radical, upon ring opening, delivers a more stabilized radical, both in terms of strain release and also the resonance delocalization gained through the aromatic substituent. In contrast, for the Ni^(I) metalloradical, the radical is clearly more stabilized on the Ni-centre in the closed cyclopropane form, in which the vinylcyclopropane serves predominantly as a π-ligand. So called ‘radical clocks’ should hence be used with caution to test for the intermediacy of metal-based radicals.

Finally, the computed profile also suggests that the Ni^(I)-bound cis– and trans-vinylcyclopropanes have only minute differences in energies, which suggests full reversibility of the isomerization process. The equilibrium is shifted ultimately through coordination/decoordination of the Ni^(I) so as to reach the overall cis:trans ratio that reflects the inherent cis/trans preference of the substrate itself. For compound 4, the observed ratio indeed matches the substrate’s free energy difference of the cis– and trans-isomer, which is 1.8 kcal mol⁻¹. This suggests that the overall selectivity of the isomerization is readily predictable upon assessment of the inherent energy difference of the cis- and trans-isomer.

To further confirm the metalloradical nature of the process, we examined whether a separately synthesized IPr-ligand-derived Ni^(I) monomer would trigger the analogous reactivity as observed when starting from dimer 1. To this end, we synthesized the known pyridine-coordinated (IPr)Ni^(I) monomer 3 (ref. ³⁷) and subjected 5 mol% thereof to cis–2 in dioxane at room temperature for 60 min, which resulted in the analogous isomerization to 89% of trans–2 (Fig. 2b). Our electron paramagnetic resonance (EPR) measurements of 3 itself in dioxane or with added cis–2 substrate gave distinct signals characteristic of the intermediacy of a metal-based radical. While the Ni^(I) dimer 1 in dioxane is EPR silent, consistent with its dimeric low-spin configuration, upon addition of the substrate cis–2, the EPR spectrum showed a distinct signal characteristic of a metal-based radical, which would be in line with our above calculations and proposed pathway.

Ni(cod)₂/IPr has been used previously for the rearrangement of vinylcyclopropanes^32,38. Calculations support the ring-opened/rearranged products to be favoured for 2 with Ni⁽⁰⁾, but the process is calculated to be at least 10 kcal mol⁻¹ higher in overall activation barrier than the Ni^(I)-based cis/trans isomerization (see Supplementary Information). Our tests with Ni(cod)₂/IPr showed diminished reactivity with strong dependence on solvent, substrate and temperature, overall suggesting divergent catalyst speciations. Indeed, we were able to detect an EPR signal characteristic of in situ formed Ni^(I). In other words, Ni(cod)₂/IPr can engage in rearrangement and/or in situ formation of Ni^(I) (ref. ³⁹), overall leading to product mixtures (see Supplementary Information).

To test the wider synthetic potential of the isomerization, we set out to examine alternative vinylcyclopropane motifs. Variation of the aromatic-substituents had no marked impact; isomerization was equally rapid (5 min, room temperature) and efficient, yielding approximately 90:10 trans:cis selectivity with 82–98% yield with methoxy (2), alkyl (5–6) or trifluoromethyl-substitution (7) as well as heterocyclic motifs, such as a benzofuran-substituted vinylcyclopropane (9). The same final cis:trans ratio resulted irrespective of whether the exclusive cis-isomer or a cis/trans mixture was subjected to isomerization. Beyond aromatic substitution, vinylcyclopropyl ketones (aromatic and aliphatic, 10–12), esters (13–15), amides (21–22) as well as Weinreb amides (23) also isomerized efficiently to the corresponding trans-products in high selectivity and yield, albeit under slightly longer reaction times of 15 min to 2 h at room temperature. Even the free carboxylic acid was well tolerated and delivered 17 within 15 min at room temperature using 5 mol% of catalyst 1. As there was no by-product formation, purification was achieved by straightforward filtration.

We next studied whether isomerization of N-methyliminodiacetyl boronate (BMIDA)-, pinacol boronic ester (BPin)- and germyl-substituted vinylcyclopropanes would also be feasible. Given the enabling value of these modular platforms to access a wide range of compounds upon (stereospecific) derivatization⁴⁰, we envisioned that the straightforward construction of these building blocks as cis/trans mixtures, followed by Ni^(I)-catalysed enrichment to the trans-isomer would be of value. Indeed, we observed facile isomerization to 24–26, albeit under slightly more forced conditions (60 °C and/or extended reaction time). Capitalizing on the rich stereospecific diversification chemistry of the boronic ester unit can thus allow formation of substituted trans-vinylcyclopropanes with rests that might either be incompatible with Ni^(I) catalysis or for which the inherent cis/trans stability difference of the substrate itself may not allow discrimination or induced selectivity in the reversible Ni^(I)-catalysed isomerization process. For example, this is the case for alkyl substituted versions, for which only bulky substituents (20) deliver high trans-thermodynamic preference and selectivity.

We subsequently examined whether additional substitution on the vinyl motif could also be tolerated and synthesized alkenyl cyclopropanes (19, 28–36) that contain internal alkenes in E or Z geometry. Although longer reaction times and elevated temperature were found to be necessary for these cases, high trans-yields were seen in all cases under full retention of the olefin (Z or E) geometry. There was no isomerization of the olefin; solely the cyclopropane isomerized, in line with our mechanism in Fig. 2c and the locked π–allyl configuration in Int2. Similarly, 1,1-disubstituted olefins (19, 35–36) and higher substitution at the cyclopropane ring (18, 19) were also tolerated and, in the absence of the vinyl group, the starting material was fully recovered (27).

The scalability of the process was next examined. To this end, we employed 1 g of cyclopropane 16; since it is a liquid, we omitted any solvent and attempted the isomerization of the neat compound, adding solely 0.5 mol% of Ni^(I) catalyst 1. After 3.5 h at room temperature, a 92:8 trans:cis mixture of 16 was generated and isolated in 93% yield. Neither additives nor solvent nor heating were necessary, only minute amounts of the non-precious Ni^(I) metalloradical was required and full atom-economy was retained in the process.

As discussed above, the ultimate cis:trans ratio upon stereomutation depends on the inherent energetic prerequisite of the substrate itself. In light of the reversibility of the isomerization, any removal of the trans-isomer should then lead to further enrichment towards more trans-isomer of the remaining mixture, if multiple rounds of Ni^(I) based isomerization are conducted. To test for the possibility of such an ‘iterative thermodynamic resolution’, we studied a 50:50 cis/trans mixture of the Weinreb amide 23, which gave a 91:9 trans/cis mixture after 1 h Ni^(I) catalysis, as shown in Fig. 3a. We found that 81% of the trans-product was separated readily by column chromatography and the remaining mixture re-subjected to another round of Ni^(I) isomerization. This sequential separation/isomerization was then repeated. After three rounds of this thermodynamic resolution, trans–23 was obtained in 97% yield and more than 99:1 diastereoselectivity (Fig. 3). Given the existing methodological challenges in accessing 1,2-disubstituted cyclopropanes in high e.e. and diastereometric ratio (d.r.), the metalloradical-based iterative thermodynamic enrichment is a powerful downstream manipulation. Beyond 1,2-disubstitution, we further considered the enantiopure trisubstituted 37, which has the status of an essential pharmacophore as it is not only the key building block of the anti-viral drugs Simeprevir and Danoprevir (against hepatitis C and COVID-19, respectively) but overall featured in a dozen drugs that are either in clinical trial or already approved⁴¹. Large-scale industrial processes to 37 have been developed⁴¹. We envisioned that the Ni-based iterative thermodynamic enrichment to the opposite enantiomer could greatly advance the preparation, discovery and large-scale synthesis (based on otherwise existing processes) of new anti-viral drugs. Indeed, three rounds of Ni^(I)-based enrichment converted 1 g of enantiopure (1R,2S)-37 (1:99 d.r., more than 99% e.e.) to the single enantiomer (1S,2S)-38 in 99:1 d.r., more than 99% e.e. (in 91% yield). Both separation techniques, that is, crystallization and column chromatography, proved effective in this context.

**Fig. 3: Reactivity and selectivity investigations.**

To further harness the reversibility of the isomerization, its unique capability to retain stereochemical information and its exceptional mildness, we next considered to extend our studies to divinylcyclopropanes. We considered the identified mechanistic features an ideal match to facilitate this strategic transformation, as the reversible nature of the Ni^(I)-based isomerization should cause any formed cis-1,2-divinylcyclopropane to undergo stereospecific Cope-type rearrangement and hence be removed from the equilibrium. Overall, the first counter-thermodynamic trans-to-cis isomerization of a divinylcyclopropane under retention of enantiopurity should result, followed by Cope-type rearrangement to generate the enantiomerically pure product.

As proof-of-principle, we undertook the enantioselective synthesis of (−)-dictyopterene A (39, 88% e.e.; Fig. 4a), which is a class of natural products featured in marine algae. We isolated the corresponding seven-membered ring, that is, (+)-dictyopterene C′ (40) in 84% yield as the exclusive product in 86% e.e., following the Ni^(I) catalysed isomerization-Cope sequence at 45 °C over 24 h. The isomerization proceeded under exclusive inversion of the stereocentre distal from the unsubstituted vinyl substituent, in line with our identified mechanism in Fig. 2 and the higher reactivity that we observed for non-substituted vinylcyclopropanes to associate the Ni^(I) (cf. Figs. 3a and 3b). Such a mild and enantiopure isomerization-Cope sequence is, to our knowledge, unprecedented. The established corresponding thermal process that starts with enantiopure trans–39 is known to require heating at 165 °C over 48 h and proceeds under loss of stereochemistry¹⁸. Indeed, we also obtained the seven-membered ring in modest 10% e.e. (enantiomer-40, Fig. 4b).

**Fig. 4: Tandem *trans*-to-*cis* isomerization/Cope rearrangement.**

The phenyl substituted analogue 41 proceeded equally efficiently to 42 under Ni^(I) catalysis. Similarly, the silyl enol ether 43 cyclized smoothly to 44, which eventually yields the ketone 45 upon desilylation. The analogous Ni-free, thermal process has been reported to occur under heating at 230 °C (ref. ⁴²) and has found applications in natural product syntheses⁶.

We next considered variants that contain one of the vinyl substituents in a cyclic geometry that would, upon contra-thermodynamic trans-to-cis isomerization followed by Cope-type rearrangement, result in a fused bicyclic product. Such processes, but thermally induced (more than 130 °C), have been used strategically in synthesis, including of natural products (for example, (+/−)-beta-himachalene⁴³ or karahanaenone⁴⁴). We observed that, under metalloradical catalysis, the analogous process happened cleanly at 60 °C starting from the trans-1-(2-vinylcyclopropyl)cyclohex-1-ene and delivered fused bicycle 46 in 97% yield (Fig. 4c). The same process was equally effective with a larger ring (47) or one containing heteroatoms (48–50).

We next considered the synthesis of spirocycles starting from trans-divinylcyclopropanes (Fig. 4d). We obtained spirocycle 52 in 68% under Ni^(I)-metalloradical catalysis at 80 °C over 24 h. Silyl enol ethers also proceeded smoothly (at 60 °C), and we generated the corresponding spirocyclic ketones (54, 56) after deprotection of the corresponding silyl enol ethers (53, 55). Rearrangements of such silyl enol ethers were previously thermally conducted at 230 °C (refs. ^18,42).

In summary, this report disclosed the distinct reactivity of a Ni^(I)-metalloradical, which, in contrast to organic free radicals or many closed-shell metal catalysts, does not lead to ring opening of a (di)vinylcyclopropane under strain release but instead triggers its reversible isomerization under selective inversion of a single (stereo)centre under mild conditions. In light of the pronounced prevalence of odd-oxidation states in non-precious metal species, this study further manifests their potential to enhance the synthetic toolbox and to meet global demands for greater sustainability and lower energy usage in chemical processes.

[ad_2]

Source link

Tag: Synthetic chemistry methodology