Soluble individual metal atoms and ultrasmall clusters catalyze key synthetic steps of a natural product synthesis

The study addresses whether soluble individual metal atoms (single-atom catalysts in solution) and few-atom metal clusters can maintain their exceptional catalytic activity when applied to long, multi-step total syntheses of complex molecules. Over the last decade, single-atom catalysts (SACs) and subnanometric clusters have shown high activity and metal efficiency across hydrogenation, oxidation, hydroaddition and cross-coupling reactions, often outperforming traditional ligand-based organometallic complexes. Despite this progress, their implementation across entire synthetic routes, especially at late stages with sterically and electronically complex substrates, has been scarcely explored. Neolignans, particularly dehydrobenzofurane-containing members such as Licarin B, possess diverse biological activities, yet modern catalytic syntheses are limited. Previous total syntheses of (±)-Licarin B relied on non-catalytic methods and extensive protecting-group manipulations, giving low overall yields (~2.5% over ~10 steps). The authors propose an 11-step, catalyst-driven route leveraging Pd, Pt, and Cu individual atoms and clusters in solution, along with a late-stage intermolecular carbonyl-olefin metathesis to forge a trans-alkene, aiming to establish the utility of these ultrasmall catalytic species in complex settings.

The paper situates its work within advances in SACs and few-atom clusters for organic synthesis (hydrogenation, oxidation, hydroaddition, cross-couplings), highlighting their high turnover numbers and ligand-free operation. It references robust characterization and performance of Pd, Pt, and Cu subnanometric species, both in solution and supported, and notes comparable activity under heterogeneous conditions. In natural product synthesis, (neo)lignans are a prolific family with significant pharmacological properties, yet few have modern synthetic routes; Licarin B’s last reported syntheses date back decades with low yields and limited catalytic methods. Prior methodologies (e.g., Nozaki–Hiyama–Kishi, Wittig-type olefinations) have limitations in selectivity or sustainability, motivating exploration of alternative catalytic disconnections (e.g., carbonyl-olefin metathesis) and the integration of SAC/cluster-catalyzed steps in total synthesis.

Design and retrosynthesis: An 11-step linear synthesis of (±)-Licarin B was designed incorporating individual metal atoms (Pd1) and few-atom clusters (Cu2–7, Pd2–3, Pt3–5) for key bond constructions, along with a terminal intermolecular carbonyl-olefin metathesis to form a trans-alkene. Catalyst preparation and characterization: Pd1 was generated in situ from Pd(OAc)2 via mild reduction by benzyl alcohol under O2. Few-atom clusters (Cu2–7, Pd2–3, Pt3–5) were prepared by mild reduction in amide solvents (DMF/NMP) at 120–150 °C (with 1 wt% H2O to stabilize Pd2–3). Characterization included UV–vis absorption/emission spectroscopy, MALDI-TOF MS, and comparison with previously reported spectra; representative AC-HAADF-STEM imaging and XANES/EXAFS for Pd single atoms and ORBITRAP HRMS for Pt clusters are referenced. Catalyst speciation post-reaction was not re-characterized in this work but in prior studies. Analytical methods: Products were characterized by GC, GC–MS, HR ESI–MS, 1H/13C NMR, DEPT. Metal content during NaBH4 hydrogenation was assessed by ICP–OES. XAS (XANES/EXAFS) at the Pd K-edge was performed at ALBA. Synthetic sequence and conditions (selected): - Step 1: Pd1-catalyzed aerobic oxidation of 4-bromobenzyl alcohol 2 to 4-bromobenzoic acid 3 (neat, O2 4 bar, 150 °C) in 83% yield. - Step 2: Di-iodination (NIS/H2SO4) of 3 to 4 (>95%). - Step 3: Cu2–7-catalyzed mono-hydroxylation (NaOH, H2O, 85 °C) of 4 followed by methylation (K2CO3/DMS) to give 6 in 43% over two steps. - Step 4: Pd2–3 cluster-catalyzed Cu-free Sonogashira coupling of 6 with TMS-acetylene (NMP, KOAc, 150 °C), affording 7 in low yield (17% by GC; conventional PdCl2(PPh3)2/CuI gave >95%). One-pot desilylation (K2CO3/MeOH) to terminal alkyne 8 in 89% over two steps. - Step 5: Hydrometallation/activation of alkyne 8: Pt3–5-catalyzed Markovnikov hydrosilylation with HSiEt3 in toluene at 110 °C gave 9 (30%). Alternatively, Pd(PPh3)2Cl2-catalyzed hydrostannylation with HSnBu3 yielded 10 (>95%). Several attempted couplings with piperonal/piperonyl derivatives failed under literature conditions. - Steps 6–7: Pd1-catalyzed aerobic oxidation of piperonyl alcohol 11 to acid 12, then SOCl2 chlorination to acyl chloride 13 (71%). - Step 8: Pd(PPh3)4-catalyzed Stille-type acyl coupling of 10 with 13 in toluene at 110 °C furnished enone 14 in 92%. Attempts with Pd2–3 clusters failed to deliver 14. - Step 9: NaBH4 reduction (THF:MeOH 1:1, rt) of 14 unexpectedly hydrogenated both the enone double bond and carbonyl, affording alcohol 15 in 75%. ICP–OES indicated extremely low transition metal content; varying Sn content did not enhance alkene hydrogenation, suggesting intrinsic NaBH4 reactivity under these conditions. - Step 10: Intramolecular C–O coupling of 15 catalyzed by Cu clusters (CuI, Cs2CO3, DMF, 130 °C) gave benzofuran 16 in >95%. - Step 11: LiAlH4 reduction of the methyl ester in 16 to alcohol 17 (99%), followed by MnO2 oxidation to aldehyde 18 (66%). Final formation of the trans-alkene via BF3·OEt2-catalyzed intermolecular carbonyl-olefin metathesis with 2-ethyl-2-phenyl-1,3-dioxolane 19 in DCE (70 °C) delivered (±)-Licarin B 1 in 46% yield (anti:syn ~5:1). As confirmation, Takai olefination of 18 with 1,1-diiodoethane/CrCl2 provided 1 in 93% yield with >20:1 trans/cis. Control/failed experiments: Intermolecular Cu-catalyzed C–O coupling on 8 (C–Br) failed; hydrometallation/coupling with Et2AlH/Ti(OiPr)4 and phosphazene-promoted additions were unsuccessful; Nozaki–Hiyama–Kishi approach after vinyl iodide formation failed.

- Demonstrated an 11-step linear total synthesis of (±)-Licarin B with an overall yield of 13.1%, surpassing prior reports (~2.5%). - Six key steps were catalyzed by soluble individual metal atoms or few-atom clusters: • Pd1: aerobic oxidation of benzyl alcohols to acids (2→3, 83%; 11→12). Ligand-, solvent-, and additive-free conditions. • Cu2–7: selective C–O cross coupling leading to 6 (43% over two steps) and quantitative intramolecular C–O cyclization to benzofuran 16 (>95%). • Pd2–3: Cu-free Sonogashira coupling (low isolated yield via clusters vs >95% with conventional system) but with far higher Pd metal efficiency (TON ≈ 566 for clusters vs 20 for PdCl2(PPh3)2). • Pt3–5: Markovnikov hydrosilylation of terminal alkyne 8 to 9 (30% yield), confirming cluster-catalyzed regioselectivity. - Pd(PPh3)4-catalyzed Stille-type acylation (10+13→14) proceeded in 92% yield when Pd2–3 clusters were ineffective for this step. - Discovery of an apparently metal-free NaBH4 hydrogenation of a conjugated enone (14) that reduced both C=C and C=O to furnish 15 in 75%. ICP–OES indicated very low transition metal content; increasing Sn traces slowed alkene hydrogenation, supporting intrinsic NaBH4 reactivity under the conditions. - Late-stage intermolecular carbonyl-olefin metathesis (18+19, BF3·OEt2) forged the trans-alkene of (±)-Licarin B in 46% yield (GC conversion ~50%), delivering an anti:syn ≈ 5:1 regioisomer ratio consistent with natural Licarin B enrichment. - Takai olefination corroborated the trans geometry with 93% yield and >20:1 trans/cis ratio. - Full spectroscopic comparison (FT-IR, GC, UV–vis, 1H/13C NMR) matched a commercial sample; minor solvent traces and a small cis signal were noted in Takai-derived material.

The work validates that soluble individual metal atoms and subnanometric clusters can be deployed effectively in a demanding, late-stage total synthesis context. Pd1 delivered robust, scalable, and green aerobic oxidations of benzyl alcohols directly to benzoic acids without ligands or solvents, aligning with sustainability goals and operational simplicity. Cu2–7 clusters enabled both challenging selective mono-hydroxylation/C–O coupling in a polyhalogenated substrate and a high-yield intramolecular C–O cyclization to a dehydrobenzofuran, demonstrating cluster selectivity and reactivity control in sterically encumbered systems. Pd2–3 clusters mediated a Sonogashira coupling at orders-of-magnitude lower Pd loading than conventional systems, indicating superior metal efficiency despite lower yield in this substrate setting. Pt3–5 clusters provided Markovnikov hydrosilylation confirming their regiocontrol, although alternative Pd-catalyzed hydrometallation/stannylation was ultimately used to progress the route. The unexpected NaBH4-mediated alkene hydrogenation, faster than ketone reduction under the reported conditions, reveals a practical and potentially general method for enone saturation without added metal catalysts. Finally, implementing an intermolecular carbonyl-olefin metathesis at a late stage furnished the trans-alkene with good selectivity, providing an alternative to less sustainable methods (e.g., Takai olefination) and expanding the toolbox for constructing trans-alkenes in complex settings. Collectively, these results support the central hypothesis that individual atoms and few-atom clusters in solution retain high catalytic competence with complex substrates and offer complementary reactivity to traditional organometallic catalysts in total synthesis.

An 11-step linear synthesis of (±)-Licarin B was achieved with an overall 13.1% yield, incorporating six steps catalyzed by soluble individual metal atoms or few-atom clusters. Key features include Pd1-catalyzed aerobic oxidations, Cu2–7-catalyzed C–O couplings including a high-yield intramolecular cyclization, Pd2–3-facilitated Sonogashira coupling at low Pd loadings, Pt3–5-driven Markovnikov hydrosilylation, a late-stage intermolecular carbonyl-olefin metathesis forming a trans-alkene, and an apparently metal-free NaBH4 hydrogenation of an enone. The route is adaptable to other dehydrobenzofurane neolignans and to diastereomerically pure isomers. These findings open avenues for integrating individual atoms and clusters into complex synthetic sequences and highlight new, potentially general transformations suitable for late-stage functionalization. Future work may focus on mechanistic elucidation of the NaBH4 alkene hydrogenation, expanding substrate scope, optimizing yields (e.g., metathesis step), and exploring heterogeneous analogues for greener, scalable processes.

- Some cluster-catalyzed steps showed modest yields on these complex substrates (e.g., Pt3–5 hydrosilylation to 9 at 30%; Pd2–3 Sonogashira provided low yield vs conventional catalysts despite higher TON). - Intermolecular C–O cross-coupling at a C–Br site (8) failed; reactivity was lower than C–I and steric hindrance impeded progress. - Several literature strategies were unsuccessful (Et2AlH/Ti(OiPr)4 hydrometallation/aldol variants; phosphazene-catalyzed coupling; Nozaki–Hiyama–Kishi approach; Pd2–3 clusters for Stille-type coupling), necessitating alternative tactics. - The carbonyl-olefin metathesis delivered moderate isolated yield (46%) and ~50% GC conversion under reported conditions, indicating room for optimization. - Catalyst species were not re-characterized post-reaction in this study (referenced prior characterizations), leaving potential in situ evolution untracked. - The NaBH4 alkene hydrogenation is presented with preliminary evidence; mechanistic origin remains to be established. - Use of stoichiometric reagents (LiAlH4, MnO2) and, in a confirming route, stoichiometric CrCl2 (Takai) impacts sustainability; heterogeneous analogues are suggested as alternatives.