Electroreductive Radical Addition–Polar Cyclization Cascade to Access Cycloalkanes

Abstract

Compared with flat aromatic scaffolds, three-dimensional aliphatic ring systems feature high structural complexity and topological diversity, and thus have received increasing attention in drug discovery. Herein, we describe a mild and general electrochemical method for the modular synthesis of structurally distinct cyclic compounds including monocyclic alkanes, benzo-fused ring systems, and spirocycles from readily available alkenes and alkyl halides via a radical–polar crossover mechanism.

Graphical Abstract

Cyclic compounds are ubiquitous in natural products and drug molecules.¹ Approximately 95% of drug molecules contain at least one ring.^1a Due to the development of robust synthetic methodologies that allow for the convenient modification of aromatic rings, including cross-coupling reactions and nucleophilic aromatic substitution, flat aromatic scaffolds are overrepresented in medicinal chemistry compared with their saturated counterparts.² However, the incorporation of aliphatic rings into drug scaffolds could not only improve pharmacological properties such as solubility and target selectivity,³ but it also enables the sampling of a broader chemical space.¹ As such, three-dimensional aliphatic ring fragments have garnered increasing attention in drug discovery.

Among traditional ring-closure strategies, annulation reactions, such as pericyclic and formal cycloadditions, are often limited to the construction of ring systems of a specific size.⁴ For example, the Diels–Alder reaction leads to the generation of six-membered rings.^4c Radical cyclization⁵ and ring-closing metathesis⁶ usually do not work well for the formation of three- and four-membered rings due to the ring strain associated with the ring closure process. Therefore, developing a general and modular synthetic strategy to furnish cyclic products with different ring sizes remains highly desirable.

To this end, the radical addition–polar cyclization (RAPC) cascade represents a modular way to synthesize functionalized cycloalkanes, providing rapid access to complex organic structures (Scheme 1).⁷ Molander,^7a,d Fang,^7b,f and Aggarwal^7c independently developed photoredox-catalyzed RAPC to synthesize cyclopropanes using various nucleophiles (e.g., carboxylates and silicates) as alkyl radical precursors (Scheme 1A). Aggarwal further developed a more general approach using readily oxidizable arylboronate complexes as the radical precursor and electron-deficient alkenes bearing an alkyl iodide leaving group (X = I) as the radical acceptor for cycloalkane (n = 1–5) synthesis.^7e

Scheme 1. Radical Addition–Polar Cyclization (RAPC) Cascade.

Despite these advances, the development of novel RAPC systems will further enrich chemist’s toolbox toward rapidly accessing diverse saturated cyclic architectures. In particular, known methods for the RAPC nearly uniformly feature redox-neutral transformations by means of photoredox catalysis, wherein the radical precursors are limited to nucleophilic functional groups (alkylsilicates, -boronates, and -dihydropyridines). These coupling partners are often prepared from more readily available electrophilic precursors. Aggarwal^7c and Molander^7d expanded the scope to highly prevalent carboxylic acids, but these approaches are currently only demonstrated for the synthesis of three- and five-membered carbocycles. Notably, Luo and Fang reported a sole example of a cross-electrophile RAPC via reductive Ni catalysis, but the scope is currently limited to the coupling of N-hydroxyphthalimide-based redox-active esters with enynes bearing a tosylate leaving group.^7g

Here, we advance a distinct strategy to achieve RAPC via electroreductive cross-electrophile coupling (eXEC)⁸ using readily available alkyl halides (Scheme 1B). In this transformation, an alkyl bromide is first reduced to an alkyl radical, which undergoes intermolecular addition to an alkene to form a new C–C bond. The resultant carbon-centered radical is further reduced to a carbanion, which then attacks a pendant alkyl electrophile to forge the cycloalkane product. The electrophilic leaving group can be preinstalled on either the radical precursor or its alkene partner, giving rise to distinct bond disconnections. The overall reaction is reductive in nature (i.e., an overall two-electron reduction), thereby necessitating a new strategy to achieve the desired radical–polar crossover.⁹ We reasoned that electrochemistry¹⁰ represents an ideal means to execute this synthetic plan, as under a sufficient reductive potential, organic molecules can undergo efficient consecutive electron transfers at the cathode surface via an electrochemical–chemical–electrochemical–chemical (ECEC) mechanism.¹¹

In this regard, we recently demonstrated deep electroreductive chemistry as a general and versatile approach for a diverse range of cross-coupling and alkene functionalization reactions to construct C–C,¹² C–Si,¹³ C–B,¹⁴ and Si–Si bonds.¹⁵ Closely related to the objectives of this work, we developed eXEC of alkyl halides^12b and dialkylation of alkenes,^12c taking advantage of the distinct redox and steric properties of alkyl halides. We envisioned that an analogous strategy could be further expanded to an RAPC when a radical-acceptor alkene is introduced (Scheme 2). Specifically, when an alkyl bromide (B) is electrolyzed in the presence of an alkene (A) tethered with a primary alkyl chloride, the selective reduction of the more electrophilic B will take place (Scheme 2A; see Figure S1 for cyclic voltammetry data). The resultant radical C will add to alkene A to form a new carbon-centered radical D, which will undergo a second single-electron reduction to generate carbanion E. Finally, an intramolecular nucleophilic substitution will take place to afford product F.

Scheme 2. Electrochemical Strategies for Cross-Electrophile RAPC.

In another scenario, when a primary alkyl bromide leaving group is attached to a tertiary alkyl bromide (G), a regioselective (n + 2) annulation (n = number of main chain atoms between the two bromine atoms in G) will occur (Scheme 2B).¹⁶ The regioselectivity would arise from substantially different reduction potentials of primary and tertiary alkyl bromides (as an example, E_red for ⁿBuBr and ^tBuBr is −2.69 and −2.13 V versus Ag wire, respectively).¹⁷ As such, the more substituted alkyl bromide moiety would be preferentially reduced. The resultant tertiary alkyl radical H will trigger an analogous RAPC in the presence of an alkene J to generate radical K and then carbanion L enroute to producing carbocycle M via intramolecular substitution of the pendant primary alkyl bromide.

To test our hypothesis, we utilized α-(2-chloroethyl)styrene (1) as the model substrate and tertiary alkyl bromide 2 as the radical precursor (Scheme 3; see Table S1 for optimization data). The desired cyclopropane product 3 was obtained in 80% yield under constant current electrolysis in an undivided cell equipped with a graphite cathode and a sacrificial Mg anode, employing tetrabutylammonium perchlorate and 1,2-dimethoxyethane (DME) as electrolyte and solvent, respectively. Two C(sp³)–C(sp³) bonds and two quaternary carbon centers are constructed in a single transformation. Control experiments highlighted the vital role of electrochemical setup in enabling both reactivity and chemoselectivity, showing that (i) the desired product 3 was not formed in the absence of an electric current, and both starting materials were largely recovered; and (ii) using magnesium powder as the reductant in lieu of electroreduction led to the full conversion of alkyl bromide 2, but only 19% yield of the desired product 3.

Scheme 3. Proof of Concept and Control Experiments.

^aReaction conditions: 1 (1 mmol, 1 equiv), 2 (2 mmol, 2 equiv), ⁿBu₄NClO₄ (2 mmol, 2 equiv), DME (4 mL), graphite cathode, Mg anode, undivided cell, 2.5 mA of constant current (corresponding to a current density of 0.4 mA cm⁻²), 3.0 F mol⁻¹ of charge, 22 °C, 32 h. Isolated yield. ^bSame as footnote a except that no current was passed. ^cReaction conditions: 1 (1 mmol, 1 equiv), 2 (2 mmol, 2 equiv), ⁿBu₄NClO₄ (2 mmol, 2 equiv), Mg powder (4 mmol, 4 equiv), DME (4 mL), 22 °C, 32 h. Yield was determined by 1H NMR analysis.

Under these electrolysis conditions, we next explored the scope of the electroreductive cyclization (Scheme 4). Our protocol provided modular access to a panel of functionalized cyclopropanes in good yields (3–11). Cyclopropanes frequently appear in drug molecules and natural products¹⁸ and are useful synthetic building blocks in organic synthesis.¹⁹ In addition to styrene-derived substrates (3–8), vinylboronates (9), acrylates (10), and enynes (11) were all found to be viable substrates. Functional groups that are sensitive to reductive conditions such as aryl halides (4 and 5), pyridines (7), and esters (10) are tolerated. With respect to the radical precursor, various alkyl bromides with different degrees of substitution (tertiary and secondary), structures (acyclic, monocyclic, and bicyclic), and functional groups (ether, thioether, and carbamate) were compatible. To expand the generality of the method, we applied our method to the synthesis of larger ring systems, such as cyclobutanes (12) and cyclopentanes (13 and 14). Though the yields are modest, to the best of our knowledge, the RAPC strategy with alkyl chloride electrophiles has not been previously demonstrated in the synthesis of cyclobutanes and is inefficient in cyclopentane synthesis (10% yield).^7c,e Finally, indane 16 and chromane 18 were synthesized in good yields from styrenes bearing an ortho-substituent with a leaving group.

Scheme 4. Scope of Electroreductive Cyclization^a.

^aReaction conditions: chloroalkyl alkene (1 mmol, 1 equiv), alkyl bromide (2 mmol, 2 equiv), ⁿBu₄NClO₄ (2 mmol, 2 equiv), DME (4 mL), graphite cathode, Mg anode, undivided cell, 2.5 mA of constant current (corresponding to a current density of 0.4 mA cm⁻²), 3.0 F mol⁻¹ of charge, 22 °C, 32 h. Isolated yields. ^bWith 3 equivalents of alkyl bromide. ^cYield was determined by 1H NMR analysis.

We then evaluated the scope of electroreductive (n + 2) annulation of alkenes with alkyl dihalides (Scheme 5). We first applied this strategy to the synthesis of cyclopentane 19 from commercially available tert-butyl methacrylate and 1,3-dibromo-3-methylbutane. Two quaternary carbon centers were installed in an efficient and regioselective manner. This transformation thus provides a modular approach to rapidly accessing polysubstituted cyclopentanes from simple reactants. Other classes of alkenes including vinylboronates (20 and 21), styrenes (22–24), vinylsilanes (25), enynes (26), and dienes (27–30) all underwent the desired (3 + 2) annulation smoothly; however, N-vinylphthalimide and diphenylacetylene were currently not tolerated (Figure S4). When a secondary alkyl bromide was used as the radical precursor, the yield was reduced due to the more sluggish cathodic reduction of secondary alkyl halides (24).¹⁷ Notably, cyclic tertiary alkyl bromides and exocyclic alkenes (e.g., 1-methylenetetralin) can be converted into the corresponding spirocyclic products (21 and 22). To demonstrate the practicality of our method, we scaled up the reaction of 1-phenylvinylboronate to afford 1.02 g of 20 in 85% yield. Further, we achieved a concise synthesis of sesquiterpene ar-macrocarpene²⁰ (33) via (4 + 2) annulation from readily available starting materials (31 and 32).

Scheme 5. Scope of Electroreductive (n + 2) Annulation^a.

^aReaction conditions: alkene (1 mmol, 1 equiv), alkyl dibromide (2 mmol, 2 equiv), ⁿBu₄NClO₄ (2 mmol, 2 equiv), DME (4 mL), graphite cathode, Mg anode, undivided cell, 2.5 mA of constant current (corresponding to a current density of 0.4 mA cm⁻²), 3.0 F mol⁻¹ of charge, 22 °C, 32 h. Isolated yields. ^bPerformed on 4-mmol scale, see the Supporting Information for details. ^cYield was determined by 1H NMR analysis. ^dWith 3 equivalents of alkyl dibromide and 3.5 F mol⁻¹ of charge.

Finally, we conducted density functional theory (DFT) calculations to support the proposed radical–polar crossover mechanism (Table 1). An alternative radical cyclization pathway is unlikely due to its endergonic nature and very high Gibbs free energy of activation (ΔG^‡ = 24.6, 36.1, and 30.8 kcal/mol for 3-, 4-, and 5-membered ring formation, respectively).^7a In contrast, the anionic x-exo-tet cyclization is highly exergonic and kinetically feasible (ΔG^‡ = 3.0, 12.4, and 8.5 kcal/mol for x = 3, 4, and 5, respectively). Besides, the presence of an anion-stabilizing substituent on the alkene (e.g., aryl, boryl, silyl, ester, vinyl, and alkynyl) is important to ensure a rapid reduction of radical 34 to anion 35 prior to the cyclization.

Table 1.

Computed Gibbs Energies for Radical and Anionic Cyclizations^a

n	radical		anionic
	ΔG‡	ΔG	ΔG‡	ΔG
1	24.6	28.5b	3.0	−49.8
2	36.1	31.9	12.4	−49.3
3	30.8	16.8	8.5	−64.9

^aComputed at the SMD(DME)/M06-2X/6-311+G(d,p)//M06-2X/6-311+G(d,p) level.²¹ Relative Gibbs energies are reported in kcal mol⁻¹.

^bProduct 36 forms a complex with chlorine atom.

In conclusion, we have developed a selective electrochemical approach to access functionalized cycloalkanes. Various ring systems, including cyclopropanes, cyclobutanes, cyclopentanes, cyclohexanes, indanes, chromanes, and spirocycles, are constructed from alkenes and alkyl halides via cyclization or annulation. The reaction proceeds through the addition of a cathodically generated alkyl radical to an alkene, followed by electrochemical radical–polar crossover and anionic cyclization. We will continue to explore this simple yet general mechanistic paradigm of electroreductive radical–polar crossover in other synthetically useful transformations.

Data Availability Statement

The data underlying this study are available in the published article and its Supporting Information.