Recent advances in controllable/divergent synthesis

Introduction

In the era of synthetic organic chemistry, divergence can produce stereodivergence (including diastereodivergence and enantiodivergence) [1-4] and regiodivergence [5,6]. In both cases, starting from the same substrate, different stereoisomers (diastereomers and enantiomers) or regioisomers can be obtained under different reaction conditions. Over the past two decades, researchers have found that by changing reaction conditions and modifying the substrate, two structurally distinct products that are neither stereoisomers nor regioisomers can be obtained from the same starting material (using the same reagents, if necessary), and significant progress has been made in recent years. Controllable/divergent synthetic strategies have increasingly attracted attention [5,7-14], for example, in 2024, Rana [15] and co-workers reported advances in solvent-controlled stereodivergent catalysis. Surprisingly, to our knowledge, there is currently no comprehensive review of studies on controllable/divergent synthesis. This review systematically examines, how these multidimensional control elements (including ligands, metal catalysts, solvents, time, temperature, acids/bases, and subtle modifications of substrates) synergize to achieve predictable product diversification. In addition, mechanistic insights are discussed providing illustrative examples across reaction classes, and emerging strategies for programming synthetic outcomes. The integration of these approaches promises to accelerate drug discovery and materials development through sustainable, atom-economic synthesis of complex molecular libraries.

Review

Ligand control

The precise regulation of product selectivity represents a fundamental challenge in transition-metal-catalyzed organic transformations, with significant implications for complex molecule synthesis. In this context, ligand-modulated divergent catalysis has emerged as a paradigm-shifting strategy, enabling programmable access to structurally distinct molecular architectures from identical substrate precursors through precise manipulation of metal coordination [16-18]. This sophisticated approach capitalizes on the stereoelectronic tunability of ancillary ligands to dictate reaction pathways, thereby offering unprecedented control over chemo-, regio-, and stereoselectivity parameters in catalytic manifolds. In 2015, the Jiang group developed a palladium-catalyzed regioselective three-component C1 insertion reaction (Scheme 1) [19]. In this reaction, an o-iodoaniline 1, phenylacetylene, and carbon monoxide were used as starting materials, and two natural product frameworks of phenanthridone and acridone alkaloids could be selectively obtained by controlling ligands. The reaction of o-iodoaniline with in situ-generated arynes under CO atmosphere under ligand-free conditions selectively afforded phenanthridinones. Intriguingly, switching to the electron-rich bidentate ligand bis(diphenylphosphino)methane (dppm) redirected the pathway to yield acridones. Time-dependent NMR studies revealed that the selectivity hinges on the aryne release kinetics from its precursor. Employing CsF, tetrabutylammonium iodide (TBAI), and water significantly accelerated aryne generation, thereby increasing its local concentration. This favored aryne coordination to the palladium center, followed by CO insertion and reductive elimination to furnish phenanthridinones. In contrast, when dppm was introduced, oxidative addition of the C–I bond to palladium formed the four-membered aryl–palladium complex Int-5. Steric hindrance from the bulky dppm ligand, combined with slower aryne release (using KF as the fluoride source), attenuated aryne coordination. Under these electron-deficient conditions, CO preferentially occupied the palladium coordination site. Sequential insertion of CO and aryne, followed by reductive elimination, culminated in acridone formation. This ligand-dependent mechanistic dichotomy underscores the critical interplay between aryne availability, steric modulation, and electronic effects in steering catalytic selectivity.

In 2016, the Jiang group achieved regioselective control in the gold-catalyzed intramolecular hydroarylation of alkynes by modulating the electronic and steric effects of ligands (Scheme 2) [20]. Mechanistically, the electron-deficient phosphite ligand L1 and the weakly coordinating OTf⁻ anion synergistically enhanced the electrophilicity of the gold center, enabling coordination with the amide group to form a three-coordinate Au(I)–π-alkyne intermediate Int-12. The umbrella-shaped steric shielding provided by the ligand-stabilized intermediate Int-9, followed by Friedel–Crafts-type addition and protonation to complete ortho-position cyclization. In contrast, para-position cyclization was exclusively achieved through π–π interactions between the electron-rich X-phos ligand and the substrate, compensating for the electron-deficient nature of the aromatic system and ensuring high regioselectivity.

In 2018, the Jiang group developed a regiodivergent synthetic method for indolo[3,2-c]coumarins 10 and benzofuro[3,2-c]quinolinones 9 via controllable palladium(II)-catalyzed carbonylative cyclization (Scheme 3) [21]. When ligand L3 coordinates with the palladium center, the enhanced electrophilicity of palladium facilitates preferential coordination with the amino group and activates the alkyne to form the intermediate Int-14 instead of Int-14'. Subsequent nucleophilic cyclization generates intermediate Int-15. Following CO insertion, complex Int-16 is formed, and reductive elimination yields the benzofuro[3,2-c]quinolinone product 9 along with a Pd(0) species, which is reoxidized to Pd(II) by BQ (benzoquinone). When the ligand is switched to the sterically bulky and electron-rich dppm, the chemoselectivity is reversed: the palladium center now preferentially coordinates with the hydroxy group to form complex Int-17 and the amino group undergoes nucleophilic attack to generate Int-18. After CO insertion complex Int-19 is produced and reductive elimination ultimately affords the indolo[3,2-c]coumarin product 10.

In 2023, the Garg group achieved the first example of utilizing in situ-generated π-allylpalladium complexes to capture strained cyclic allene intermediates (Scheme 4) [22]. By modulating the ligands in the reaction system, two distinct polycyclic scaffolds, 13 or 14, could be synthesized with high selectivity. Mechanistically, the Pd(0) catalyst coordinates to substrate 11, followed by oxidative addition and release of carbon dioxide to form the zwitterionic π-allylpalladium intermediate Int-21. Under the reaction conditions, silyl triflate 12 undergoes a fluoride-mediated 1,2-elimination to generate the cyclic allene intermediate Int-22. Through a ligand-controlled regioselective migratory insertion process, reaction of Int-21 and Int-22 leads to the formation of π-allylpalladium intermediates Int-23 or Int-24, depending on the ligand employed. Finally, cyclization of Int-23 or Int-24 yields the tricyclic product 13 or the tetracyclic product 14, respectively.

In 2024, the Song group achieved a ligand-controlled regiodivergent and enantioselective ring expansion of benzosilacyclobutenes with internal naphthylalkynes by strategically modulating the ligand steric profiles (Scheme 5) [23]. Employing cavity-engineered phosphoramidite ligands, the reaction pathway bifurcated based on the steric demands of Si–C-bond activation. The methyl-substituted ligand (S)-8H-binaphthyl phosphoramidite L4, featuring a spacious cavity, favored sterically encumbered Si–C(sp³)-bond activation, selectively delivering axially chiral (S)-1-silacyclohexenyl arenes 17 with high enantiocontrol. Conversely, the bulky tert-butyl-decorated (R)-spirophosphoramidite L5 imposed a confined cavity, steering selectivity toward Si–C(sp²)-bond activation and predominantly afforded the regioisomeric (S)-2-silacyclohexenylarenes 18.

In 2025, Gong and co-workers reported a visible-light-mediated hydrogen atom transfer (HAT)/chiral copper dual catalytic system that achieved regiodivergent and enantioselective C(sp³)–C(sp³) and C(sp³)–N oxidative cross-couplings between N-arylglycine ester/amide derivatives and abundant hydrocarbon C(sp³)–H feedstocks (Scheme 6) [24]. This methodology also represents a highly challenging direct C(sp³)–H asymmetric amination. Mechanistic insights: When using a bulky, electron-rich chiral bisphosphine ligand L6, the glycine ester substrate coordinates with the copper catalyst to form a key intermediate complex Int-26. The sterically hindered and electron-rich environment around the copper center disfavors a direct interaction with nucleophilic alkyl radicals. Instead, the reaction proceeds via an outer-sphere mechanism, where the alkyl radical reacts with the copper-activated C=N unsaturated bond, enabling stereocontrolled C(sp³)–C(sp³) coupling. In contrast, with the anionic cyano-substituted bisoxazoline ligand L7, the glycine ester and copper catalyst form a distinct intermediate complex Int-28. The ligand’s reduced steric bulk and altered electronic properties facilitate direct interaction with alkyl radicals, forming a high-valent Cu(III) intermediate Int-29. This intermediate undergoes reductive elimination via an inner-sphere mechanism to generate the C(sp³)–N-coupled chiral product 22. Notably, benzoic acid acts as a critical additive, likely by stabilizing key intermediates and modulating the steric/electronic environment for enhanced enantiocontrol.

Metal control

Over the past decade, the relentless pursuit of precision in natural products and pharmaceutical synthesis has driven remarkable advances in catalytic methodologies, particularly in the realm of catalyst-controlled chemoselective transformations [11,25-29]. In 2023, the Shu group developed a catalyst-controlled regioselective and enantioselective hydroamination reaction of electron-deficient alkenes (Scheme 7) [30]. By efficiently regulating the regioselectivity and enantioselectivity of alkene 23 hydrometallation through catalytic systems, they overcame the influence of steric and electronic effects during the hydrometallation process, simultaneously achieving the synthesis of chiral α-quaternary carbon amino acid derivatives 26 and α-chiral β-amino acid derivatives 27. Using a copper catalyst, the chiral α-quaternary carbon amino acid derivatives 26 were obtained with exclusive regioselectivity and excellent enantioselectivity. Employing a nickel catalyst, α-chiral β-amino acid derivatives 27 were synthesized with single regioselectivity and outstanding enantioselectivity.

In the same year, Rong and co-workers reported a highly efficient catalyst-controlled regio- and enantioselective hydroalkylation reaction, enabling the divergent synthesis of chiral C2- and C3-alkylated pyrrolidines through desymmetrization of readily available 3-pyrrolines (Scheme 8) [31]. The cobalt catalytic system (CoBr₂ with modified bisoxazoline ligands) achieved asymmetric C(sp³)–C(sp³) coupling via distal stereocontrol, efficiently producing C3-alkylated pyrrolidines, while the nickel catalytic system afforded C2-alkylated pyrrolidines through a tandem alkene isomerization/hydroalkylation process. This method utilized readily accessible catalysts, chiral BOX ligands L9, and reagents, delivering enantioenriched 2-/3-alkyl-substituted pyrrolidines with excellent regio- and enantioselectivity (up to 97% enantiomeric excess). Radical-clock experiments and deuterium-labeled silane studies revealed that cobalt catalysis proceeded via irreversible Co–H migratory insertion to achieve C3 selectivity, whereas nickel catalysis involved alkene isomerization to generate a (2,3-dihydropyrrolyl) intermediate Int-35, followed by C2-selective coupling.

In 2024, the Zheng group reported a catalyst-controlled cyclization reaction of bicyclo[1.1.0]butanes (BCBs) 32 with α-alkenylazides 33, achieving divergent synthesis of 2- and 3-azabicyclo[3.1.1]heptenes (aza-BCHepes) 35 or 36 (Scheme 9) [32]. This study developed a practical method for constructing novel 2- and 3-aza[3.1.1]heptene architectures from readily available α-alkenylazides and BCBs through catalyst-controlled (3 + 3) and (3 + 2) cyclization strategies. Two distinct pathways were established: (1) The titanium-catalyzed ring opening of bicyclobutane (BCB) 32 generates a γ-carbonyl radical intermediate Int-42, which undergoes trapping by vinylazide 33. Subsequent dinitrogen extrusion produces an iminyl radical species Int-44. This reactive intermediate then engages with a Ti(IV)-enolate complex through radical recombination, ultimately delivering 2-aza-bicyclo[3.1.1]heptene (BCHepe) while regenerating the Ti(III) catalyst to complete the catalytic cycle. (2) Scandium-catalyzed pathway: Activation of the donor–acceptor BCB via Sc(OTf)₃ coordination to its carbonyl group facilitates nucleophilic attack by vinylazide 33, forming an imino-diazonium intermediate Int-40 accompanied by a δ-carbanion. Transannular cyclization of this species affords 2-azidobicyclo[3.1.1]hexane (2-azidoBCHs). Subsequent thermal activation induces selective migration of the less sterically hindered secondary carbon center with concomitant dinitrogen elimination, yielding 3-aza-BCHepe as the final product.

Solvent control

The solvent microenvironment emerged as a critical determinant in governing stereochemical outcomes, exerting profound influence through multifaceted solute–solvent interactions [5]. Solvent polarity, hydrogen-bonding propensity, and dielectric characteristics collectively orchestrate stereodivergent pathways through dynamic coordination effects and differential stabilization of transition states. Notably, these solvent-mediated electronic and steric modulations frequently dictate reaction stereoselectivity, with even subtle solvent permutations potentially inducing complete stereochemical inversion in sensitive systems [33-38]. In 2023, He and Sessler disclosed a versatile one-pot synthesis of structurally diverse macrocycles through the dynamic self-assembly of α,α’-linked oligopyrrolic dialdehydes and alkyldiamines (Scheme 10) [39]. Their investigation revealed distinct solvent-mediated selectivity in product formation. Condensation of the pyridine-bridged oligopyrrolic dialdehyde 37 with simple alkyldiamines proceeded with solvent-independent regioselectivity, exclusively furnished [2 + 2] macrocyclic adducts. Strikingly, when 37 was combined with 2,2’-oxybis(ethylamine) (38), the reaction pathway exhibited pronounced solvent dependency. Reactions in methanol, ethanol, or chloroform selectively generated the [1 + 1] macrocycle 39 as the sole product. In contrast, polar aprotic solvents such as dimethyl sulfoxide (DMSO), N,N-dimethylformamide (DMF), or acetonitrile (MeCN) favored precipitation of the [2 + 2] macrocycle 40. Notably, the macrocycle 40 underwent spontaneous structural reorganization in chloroform or dichloromethane (DCM), converting entirely into the thermodynamically stable [1 + 1] isomer 39. This work demonstrates a solvent-driven approach for dynamically interconverting macrocycle sizes, governed by thermodynamic stability and solubility differences.

In the same year, Chauhan, Koenigs, and co-workers demonstrated solvent-controlled bifurcation in the light-driven reactivity of cyclic diazo imides 41 with thiols 42, unveiling two mechanistically distinct pathways (Scheme 11) [40]. In dichloromethane (DCM), the reaction proceeds via a carbene intermediate, enabling cascade C(sp²)–H functionalization/thiolation to deliver indane-fused pyrrolidines 43 in excellent yields (up to 92%). Strikingly, switching the solvent to acetonitrile completely suppresses carbene formation under identical conditions, redirecting the pathway toward an unconventional diazo reduction wherein aryl thiols act as stoichiometric reductants. Mechanistic insights, elucidated through control experiments and DFT calculations, revealed that photoexcitation of diazo imide 41 triggers nitrogen extrusion (ΔG^‡ = +10.0 kcal·mol⁻¹), generating the triplet carbene intermediate Int-45. In DCM, this species undergoes intramolecular cyclization into a proximal C(sp²)–H bond (ΔG^‡ = +19.7 kcal·mol⁻¹) to form Int-46, which reacts with 4-MePhSH (ΔG^‡ = +14.9 kcal·mol⁻¹) to yield radical intermediate Int-47 and a thiyl radical (4-MePhS^·). Sequential thiol-assisted hydrogen shifts produce Int-48, followed by barrierless thiyl radical addition and intersystem crossing (ISC) to furnish the final product. In contrast, acetonitrile’s polar aprotic environment destabilizes the carbene pathway, favoring direct reduction of the diazo moiety via electron transfer from the thiol. This solvent-gated selectivity underscores the critical role of reaction medium polarity in modulating reactive intermediates, offering a strategic lever to toggle between C–H functionalization and reductive manifolds in photochemical transformations.

In 2024, the Cheng group developed a palladium/chiral norbornene (NBE)-catalyzed cyclization reaction between aryl iodides 45 and phosphoramides 46 under varying solvent conditions of toluene (PhMe) and acetonitrile (MeCN), based on their studies of the Catellani reaction (Scheme 12) [41]. This method exhibited a broad substrate scope for both aryl iodides and phosphoramides, and enabled enantioselective access to both enantiomers of chiral P(V) molecules 47 or 48 using a single chiral NBE catalyst.

Time control

Time control of chemical reactivity offers an inherent strategy to program synthetic pathways through kinetic discrimination of transient intermediates. Diverging from additive-dependent or stimulus-responsive approaches, this paradigm capitalizes on the chronoselective evolution of reactive species to unlock sequence-controlled transformations. In 2020, the You group reported a reaction-time-dependent enantiodivergent synthesis method. Under the same chiral catalytic system, they achieved selective synthesis of either enantiomer of a target product by controlling the reaction duration (Scheme 13) [42]. When performing the asymmetric intermolecular allylic amination of 6-hydroxyisoquinoline (49) with tert-butyl(1-phenylallyl)carbonate ((rac)-50) using an Ir catalyst derived from [Ir(cod)Cl]₂ and the Carreira chiral phosphoramidite ligand (S)-L10, along with the addition of 3,5-dichlorobenzoic acid as an additive in MeOH at room temperature, the reaction proceeded smoothly for 10 hours to yield the aminated product 51. Interestingly, when the reaction was quenched after 6 minutes in the absence of a Brønsted acid additive, the opposite enantiomer 52 was obtained. Mechanistically, an initial kinetic resolution (KR) of (rac)-50 occurs via an Ir-catalyzed asymmetric allylic amination. Due to the higher reactivity of (S)-50, it reacts with 6-hydroxyisoquinoline within 6 minutes to generate 52, while (R)-50 remains largely unreacted during this period (k₁^R << k₁^S). However, as the reaction progresses, 52 undergoes further reaction with MeOH under the catalytic system to form 52'. Meanwhile, the less reactive (R)-50 gradually reacts with 6-hydroxyisoquinoline, leading to the accumulation of 51. Since 51 is highly stable and resistant to reaction with MeOH (k₂^R << k₂^S ≈ k₁^R), it can be obtained with high optical purity after an extended reaction time (10 hours)

In 2023, Yang and Liang jointly reported a tetrasilane (ODCS)-based method for time-controlled, palladium-catalyzed C–H activation in the divergent synthesis of silacyclic compounds (Scheme 14) [43]. This reaction employs the ODCS reagent to capture a five-membered C,C-palladacycle species, using reaction time as a control switch to enable transformations of three distinct substrates – acrylamides, 2-halo-N-methylacryloylbenzamides, and 2-iodobiphenyls – thereby selectively synthesizing silacyclic compounds with varying ring sizes, including ten-membered, seven-membered, and five-membered rings. Mechanism (Scheme 15): Substrate 53 undergoes oxidative addition with Pd(0), followed by intramolecular carbopalladation to form the σ-alkylpalladium intermediate Int-50. The intermediate Int-50 undergoes C–H activation to generate the spiro-palladacycle Int-51, which proceeds via two possible pathways: 1) Path a: oxidative addition/reductive elimination or 2) path b: transmetalation/reductive elimination giving rise to intermediates Int-53 or Int-53'. Reductive elimination of Int-53 or Int-53' regenerates Pd(0) and produces intermediate 55. With the assistance of the base K₂CO₃, the ten-membered silacycle 55 undergoes rapid ring contraction via cleavage of two Si–O bonds and formation of one Si–O bond, leading to 56 and Int-55. Concurrently, Int-55 dimerizes to form 54, which is further transformed into cyclosiloxanes under K₂CO₃ and DMA conditions. Intermediate 56 undergoes additional ring contraction through cleavage of Si–O/Si–C bonds and formation of a Si–C bond, yielding 57 and Int-56, with Int-56 polymerizing to generate cyclosiloxanes. An alternative pathway involving cleavage of another Si–O bond during the conversion from 55 to 56 and subsequently to 57 cannot be excluded.

Temperature control

Temperature, as a readily adjustable physical parameter in organic synthesis, offers a simple and versatile approach to control regioselectivity. It profoundly influences reaction kinetics, stability of intermediates, and reaction equilibria. Through precise temperature modulation, chemists can effectively steer the formation of regioisomers, often achieving desired selectivity with minimal alterations to other reaction components. For instance, Rana and colleagues developed a temperature-dependent regiodivergent strategy to access functionalized maleimides and itaconimides [44]. This thermochemical strategy provides a robust platform for controlling reaction pathways while maintaining synthetic simplicity. In 2022, García-García and Fernández-Rodríguez reported on the practicality of metal-free BCl₃-catalyzed borylation cyclization reactions in synthesis (Scheme 16) [45]. Biphenyl-embedded 1,3,5-trienes-7-yne compounds 58 react with BCl₃ under catalyst-free and additive-free conditions to form novel polycyclic boronated structural units. By adjusting the temperature of the reaction medium, it is possible to precisely control the reaction pathway, thereby obtaining two different boronated frameworks from the same starting material: boronated phenanthrene derivatives 59 at 60 °C and phenanthrene-fused boronated cyclobutane 60 at 0 °C.

In the same year, Lu's research group reported a temperature-controlled site-selective olefin hydroalkylation reaction (Scheme 17) [46]. By adjusting only the reaction temperature, different skeletal structures of nitrogen α- and β-alkylated products could be obtained from the same olefin substrates 61. At 10 °C, the catalytic system consisting of NiBr₂(diglyme), oxazoline ligand, (EtO)₃SiH, and K₃PO₄(H₂O) achieved β-selective hydroalkylation. When the temperature was raised to 100 °C, the reaction selectively produced α-branched products. DFT calculations showed that at low temperatures, the six-membered nickel ring captures radicals and undergoes reductive elimination to form β-products (kinetic control); at high temperatures, the formation of a five-membered nickel ring leads to α-products (thermodynamic control). Therefore, the formation of the more stable nickel ring drives migration, while the thermodynamic and kinetic properties of different reductive elimination intermediates jointly determine the switchable site selectivity.

Acid–base control

The strategic modulation of acid–base interactions has emerged as a powerful paradigm in organic synthesis, enabling precise control over reaction pathways, selectivity, and catalytic efficiency. By exploiting dynamic acid–base equilibria or stimuli-responsive systems, chemists can manipulate substrate activation, stabilize reactive intermediates, and orchestrate complex multistep transformations under mild conditions [47]. In 2016, Lu's team designed a new class of acetylene carbonate reagents and successfully applied them to copper-catalyzed decarboxylative amination/hydroamination sequences (Scheme 18) [48]. By controlling acidic and basic reaction conditions, the authors achieved the controllable synthesis of two types of functionalized indoles. When treated with acid (BF₃·E₂O), the intermediate 2-methylene-3-aminoindoline 69 undergoes an aza-Cope rearrangement to form 2-benzylindole 70; when treated with a base (Cs₂CO₃), this intermediate undergoes a 1,3-proton migration process to convert back to 3-aminoindole 71. The possible mechanism for the formation of the key intermediate 69 is outlined in Scheme 19: first, substrate 67, under the action of a copper catalyst and diisopropylethylamine, undergoes a decarboxylation process to generate the allylidenecopper intermediate Int-63 and its resonance form Int-64. Subsequently, these intermediates undergo a propargylation process (Int-63, Int-64 to Int-65) followed by a proton elimination process to generate Int-66 (Int-5 to Int-66). Then, Int-66 undergoes an intramolecular amination through copper-catalyzed activation to form Int-68, and finally, 2-methylene-3-aminoindoline 69 is generated via a proton transfer promoted by diisopropylethylamine.

In 2023, the Jiang research group achieved a chemically divergent photocatalytic asymmetric synthesis using a dual catalytic system consisting of a chiral phosphoric acid and dicyanopyrazine (DPZ) as the photosensitizer (Scheme 20) [49]. By regulating the chemical selectivity of a three-component radical cascade reaction involving α-brominated aryl ketones 72, olefins 73, and 1-methylquinoxalin-2(1H)-one (74) with an inorganic base, they were able to obtain two important types of products with high yield and enantioselectivity. Through mechanistic experiments and DFT calculations, the authors proposed a possible mechanism for the reaction: first, DPZ is excited by light to form the excited state DPZ*, which then oxidizes bromide ions through single-electron transfer to generate corresponding radical anions. These radical anions undergo single-electron transfer with substrate 72 to form radical intermediate Int-70, completing the DPZ catalytic cycle. Intermediate Int-70 adds to substrate 73 to form radical intermediate Int-71, which further adds to hydrogen-bond-activated substrate 74 to form hydrogen-bonded complex Int-72. When Na₃PO₄ is used as the inorganic base, bromine radicals abstract hydrogen to form product 75; whereas when Na₂HPO₄ is used as the inorganic base, its weaker basicity leads to protonation of complex Int-72 to form intermediate Int-73, which then preferentially undergoes single-electron transfer with the DPZ radical anion, followed by cyclization and dehydration to yield bicyclic product 76.

Substrate control

Substrate control has emerged as a powerful strategy in organic synthesis, enabling precise manipulation of reaction pathways and stereochemical outcomes through the intrinsic structural and electronic features of the starting material. By exploiting preorganization, steric effects, or directing groups within the substrate, chemists can achieve high levels of regioselectivity, diastereoselectivity, and enantioselectivity without relying on external catalysts or additives. This approach has been successfully applied in the synthesis of complex natural products, pharmaceuticals, and functional materials, often streamlining multistep sequences and minimizing protecting-group strategies [50,51]. In 2016, Li and co-workers developed divergent coupling conditions for iminamides 77 with receptor-type diazo compounds 78 or 79 under ruthenium catalysis, generating indoles 81 and 3H-indoles 80, respectively (Scheme 21) [52]. α-Diazo-β-ketoesters form indoles by cleaving the C(N₂)–C(acyl) bond, while diazomalonates form 3H-indoles through C–N-bond cleavage. Mechanistically, the cyclometalation of iminamides follows a concerted metalation–deprotonation (CMD) mechanism to generate ruthenium intermediate Int-75. Subsequently, diazo compound 78 or 79 coordinates with intermediate Int-75, followed by deazidation to form the ruthenium carbenoid species Int-76. The ruthenium–aryl bond in this intermediate migrates into the carbenoid unit, providing heptacyclic ruthenium ring intermediate Int-77. Intermediate Int-78 is then formed via ruthenium migration insertion into the C=N bond from Ru–C(alkyl). For diazoketoester substrates, the final product 81 is released from Int-78 through protonation, intramolecular nucleophilic addition, and subsequent release of one molecule of amide, reactivating the active ruthenium(II) catalyst. In contrast, for diazomalonates, intermediate Int-78 releases ammonia with the help of Ru(II) or acetic acid, ultimately yielding 3H-indole 80. This change in selectivity may be due to the reduced electrophilicity of the ester carbonyl.

In 2021, Dong and Xie reported the development of an azido Matteson reaction, which achieves carbene insertion into an N–B bond of aminoboranes 84 or 86 (Scheme 22) [53]. In this methodology, by controlling the carbene leaving group (alkyl chlorides/alkyl bromides) and the Lewis acid activator, a selective mono- or di-methylene insertion reaction can be carried out, generating α-/β-boryl-substituted tertiary organic amines 83 from simple secondary organic amines. Using N-alkyl-N-arylaminoboranes as the reactant, the reaction proceeds at −78 °C with CH₂Br₂ and n-BuLi, followed by a reaction with ZnCl₂ at room temperature. The product is then hydrolyzed with a NaOH solution of H₂O₂ to yield amino alcohols. The mechanism involves the formation of borate intermediate Int-79 from substrate 83 under the action of CH₂BrLi. This is followed by an N-1,2-migration to form borate ester 86, which then reacts with another molecule of CH₂BrLi to form the more stable borate Int-80. Subsequently, a C-1,2-migration leads to the formation of the double-insertion product 84. If the amine portion is more electron-deficient or has more delocalized nitrogen electrons (such as indole substrates), Int-79 is more stable at −78 °C, favoring the formation of the mono-insertion product 86.

In 2022, Wu and colleagues reported a novel methodology for constructing α-ketoamides 90 or 92 and amides 91 through copper-catalyzed dicarbonylation and monocarbonylation reactions involving alkyl halides 88 (Scheme 23) [54]. Using alkyl bromides, CuBr as the catalyst, bpy as the ligand, Co₂(CO)₈ as the additive, Cs₂CO₃ as the base, and 1,4-dioxane as the solvent under 40 bar CO pressure at 80 °C, they successfully synthesized α-ketoamides 90. When alkyl iodides were used as substrates, both dicarbonylation and monocarbonylation processes occurred simultaneously with Cu(OAc)₂, favoring the dicarbonylation process. In contrast, using CuBr(Me₂S) the monocarbonylation process was favored. Possible reaction mechanisms: First, CO coordinates with copper salts to form (carbonyl)copper species Int-83. Subsequently, in the presence of a base, the amine undergoes nucleophilic attack on the coordinated CO, generating (carbamoyl)copper complex Int-84. Then, alkyl bromide undergoes a single-electron-transfer (SET) process with Int-84, forming intermediate Int-85 and an alkyl radical, which is captured by CO to yield an acyl radical. Alternatively, under the action of a base, the amine can undergo anionic ligand exchange with (carbonyl)copper species Int-83, generating an electron-rich amino copper(I) species Int-84', which activates alkyl bromide through an SET process, followed by immediate insertion of CO to form complex Int-85. Nucleophilic activation of the acyl radical initiates through its reaction with intermediate Int-85, generating the critical acyl(aminoacyl)copper species Int-86. Subsequent reductive elimination from this intermediate liberates the α-ketoamide product 92 while regenerating the catalytic species Int-82. Comparative kinetic analysis revealed a marked preference for alkyl iodide activation, as demonstrated by its substantially lower activation energy barrier compared to alkyl bromide analogs (path b). This energetic advantage facilitates preferential formation of intermediate Int-87 via oxidative addition. Rapid coupling with the in situ-generated acyl radical produces copper-bound intermediate Int-88. Base-mediated anionic exchange then displaces the halide ligand with amine, yielding intermediate Int-89. Final reductive elimination from this species affords amide product 91 with concurrent regeneration of the catalyst Int-83. Notably, a competitive pathway emerges through alternative reactivity of Int-88 (path c). The coordinated CO ligand undergoes nucleophilic attack by the amine, bypassing halide exchange to instead generate Int-86. This mechanistic crossover establishes a product dichotomy between α-ketoamide 92 and amide 91, with the branching ratio governed by relative rates of base-mediated exchange versus CO activation at the copper center.

In 2022, the Jiang research team developed regulated SuFEx click chemistry between fluorosulfonyl imides and TMS-alkynes, enabling the rapid construction of S(VI)–C(sp²) or S(VI)–C(sp) bonds efficiently (Scheme 24) [55]. This linkage utilizes the high bond dissociation energy (BDE = 135 kcal/mol) of silicon–fluorine bonds, employing trifluoroborate as a fluorine transfer reagent to simultaneously cleave the S(VI)–F bond and activate the Si–C bond. DFT calculations indicate that the reaction proceeds via the formation of a difluoroborate phenylacetylene intermediate 94’’ by in situ generation from boron trifluoride etherate and silicon-protected phenylacetylene 94, which activates the S–F bond of the fluorosulfonyl imide to form sulfonyliminium cations Int-95. These then add to the activated phenylacetylene to construct the S–C bond, followed by intramolecular 1,5-hydrogen migration and aqueous workup to remove benzaldehyde, yielding the target sulfonylimine products.

Both original protoberberine and protonitidine alkaloids are characterized by an isoquinoline ring skeleton. An analysis of their molecular structures revealed that the two alkaloids share a basic structure, differing only in the junction of the B-ring. In 2021, Liu and Jiang designed new pyridyne precursors, which underwent cycloaddition reactions with substituted furans as diene component to produce the corresponding epoxy-cycloaddition adducts. The authors developed an Ir/Sc tandem catalytic reaction to convert these adducts into polysubstituted 3-haloisoquinolines 99 in one pot. After obtaining isoquinoline compounds 99 with different substituents and polysubstituted annular boronic acids 98, a Suzuki coupling was employed to synthesize advanced isoquinoline intermediates 100. Following this, a 6π electrocyclization reaction and nucleophilic reaction were developed to achieve C–C and C–N bond constructions, respectively, leading to the synthesis of differently substituted protonitidine alkaloids and protoberberine alkaloids (Scheme 25) [56].

Conclusion

Developing streamlined and versatile approaches for the rapid assembly of structurally diverse organic molecules represents a pivotal challenge in organic synthesis, pharmaceutical research, and advanced materials development. Recent advances in controllable/divergent synthesis methodologies, which enable the construction of variously functionalized architectures from common precursors, have emerged as particularly promising due to their inherent efficiency. Contemporary strategies for controlling reaction pathways and selectivity predominantly involve precise manipulation of catalytic systems (metal catalysts/ligands), reaction parameters (solvent, temperature, time), acid/base mediation, and strategic substrate engineering. This review systematically organizes recent breakthroughs according to critical control elements governing product divergence. Through mechanistic investigations of pivotal bond-forming steps and comparative analysis of representative case studies, we provide fundamental insights into the origin of selectivity variations and reaction pathway control. The discussion emphasizes structure–reactivity relationships and catalytic design principles that enable predictable access to distinct molecular architectures from shared synthetic intermediates. This review serves as a conceptualized platform for controllable/divergent synthesis, arousing more state-of-the-art tactics in chemical synthesis.