Some Items of Interest to Process R&D Chemists and Engineers

A white paper authored by scientists from a cross-pharma impurities working group argue that current ICH guidelines do not adequately address how to handle small-molecule impurities arising from the payload, linker, or linker-payload of antibody–drug conjugates (Jones, M. T.; et al. AAPS PharmSciTech 2018, 10.1208/s12249-017-0943-6). A decision tree is presented that indicates a proposal for setting control limits on the basis of whether the impurity is conjugatable, contains the payload, or is unusually potent. A formula for daily impurity dose is presented, and the figures it generates indicate that small-molecule impurities should pose little risk to patient safety.

BMS scientists have provided more details about their Monte Carlo simulation-based methodology for using the historical yield and process mass intensity (PMI) of a class of reactions as a guide as to the probable efficiency, if a member of that reaction class is taken into development (Eastgate, M. D.; et al. ACS Sustainable Chem. Eng. 2018, 6, 1121). This initiative is driven by the opportunity to improve the chance of selecting a route during a route identification exercise, that when fully developed, offers the most efficiency versus alternative route options. Its use is demonstrated through a prediction of the efficiencies of different route options for the synthesis of fostemsavir, a potential HIV treatment. The authors stress the importance of not pushing the unfavorable environmental credentials of a reactant, reagent, and solvent outside of the consideration of the environmental footprint of the process, just because they feature outside of the GMP synthesis of the API. The article points out that the precompetitive sharing of PMI data between companies is required if this approach is to deliver an inflection point in its potential value to the wider community.

Merck scientists have shared the development and rationale behind the isolation process for boceprevir, an API isolated in an amorphous state, whose surface area was highly dependent on the processing conditions (Zarkadas, D.; et al. J. Pharm. Sci. 2018, 107, 217). Heating a suspension in a reactor and using FBRM to monitor the kinetics of agglomeration of the suspended particles is reported to generate glass transition temperature data that allow better predictions of surface area than the data produced by DSC measurements. The isolation process involved two discrete stages. The first used a continuous tee mixer precipitation process, and in-line chord length distribution measurements were used to help with its successful scale-up. The second stage involved a dynamic vacuum distillation step. The process has delivered control over the API surface area, a quality attribute critical for bioavailability and drug product manufacture, with a capability in excess of six sigma.

Chemists will be interested in a semiempirical quantum-mechanical method developed by the University of Copenhagen for predicting the preferred site of electrophilic aromatic substitution (EAS) (Jørgensen, M.; et al. Chem. Sci. 2018, 9, 660). The method, named RegioSQM, relies on identifying the aromatic carbon with highest proton affinity on the basis that such a protonated species resembles the intermediate of EAS. The calculations successfully predicted the observed regioselectivity (including examples of bisbromination) with a success rate of 81% when 525 brominations found in the literature were retrospectively analyzed. This figure rose to 96% after correction for cases in which the experimentally observed site of reaction was not where it was predicted to be computationally but was within 3.0 kcal/mol of the latter. While the method is presented as being generally applicable to electrophilic aromatic substitutions, the only validation data presented are for these bromination reactions. The method requires only the SMILES string of the reaction substrate and has been made freely available to the chemistry community.

Recent interest in couplings mediated by Pd(I) dimer complexes continues with the disclosure that they can be precatalysts for the amination of aryl halide electrophiles with both aliphatic and aromatic amines (Spokoyny, A. M.; et al. Dalton Trans. 2018, DOI: 10.1039/C8DT00119G; ChemRxiv preprint server, https://doi.org/10.26434/chemrxiv.5758551.v1). The Pd(I) dimer complexes examined all showed excellent stability toward storage in air. The article also provides an operationally straightforward synthesis of [Pd^I₂(MeCN)₆][BF₄]₂, a convenient precursor to air-stable Pd(I) dimer complexes bearing biaryl phosphine ligands.

Olanzapine (2-methyl-4-(4-methyl-l-piperazinyl)-10H-thieno[2,3-b][1,5]benzodiazepine) is a successful antipsychotic drug first patented in 1971. However, because of its complex molecular structure, the mechanism of crystal growth of olanzapine is not well understood. Olanzapine exhibits approximately 60 solid forms, of which 56 are solvates and one is an amorphous compound. As a result, manufacturing of olanzapine is complex, and a fundamental understanding of its crystal growth mechanism can support the development of a robust crystallization process. A collaboration from Prof. Doherty’s group at the University of California, Santa Barbara, Eli Lilly Co., and Shanghai Jiao Tong University led to the successful prediction of olanzapine form I crystals (Sun, Y.; et al. Cryst. Growth Des. 2018, 18, 905; M. F. Doherty, corresponding author). Spiral crystal growth of stable olanzapine form I from five (nonsolvating) solvents (acetone, ethyl acetate, toluene, methyl isobutyl ketone, and n-butyl acetate) was modeled using a dimeric growth unit. These dimeric units are considered to be stabilized by multiple C–H···π contacts. The predicted morphologies align well the experimentally observed crystals.

Solvent selection for impurity purge in a (re)crystallization process is typically done empirically using either a traditional trial-and-error approach or a statistical design of experiments (DoE) approach, and either can be a time-consuming activity. Because of the recent progress with computational methods used for solid-state characterization, such tools could assist in a (re)crystallization solvent selection. A report from Pfizer (Abramov, Y. A. Cryst. Growth Des. 2018, 18, 1208) describes the success with such an approach. Three impurity purge factors are proposed on the basis of the impurity partition coefficient between the solvent and the crystalline API and the thermodynamic solubility of the API. Among others, the chemical structure of the impurity must be known in order to be able to calculate such purge factors. Compared with previously developed genotoxic impurity purge factors, which are semiquantitative risk factors, the computational purge factors proposed in this work are computed on the basis of thermodynamic considerations. In addition to several software platforms necessary in the computation of the purge factors, the main calculations were run in COSMOtherm (executing thermodynamic calculations of chemical potentials). A limited comparison of the proposed factors against experimental data for published compounds was conducted, showing that one of the three factors proposed exhibits superior predictive properties.

Growing crystals suitable for single-crystal X-ray analysis is challenging. Recently, Fujita developed a method for the absolute structure determination of chiral guests that involves the use of chiral crystalline sponges. A collaboration between DSM and Radboud University (de Poel, W.; et al. Cryst. Growth Des. 2018, 18, 126; R. de Gelder, corresponding author) applied this methodology using a porous metal–organic framework consisting of ZnI₂ and 2,4,6-tris(pyridin-4-yl)-1,3,5-triazine as a “sponge” for crystal structure determination of camphene and pinene. Because of the molecular freedom, camphene’s crystal structure cannot be resolved using classical X-ray diffraction, and pinene is a liquid at room temperature. XRPD was used to monitor the host–guest interaction and to define an end point for the formation of the molecular complex. Future work will evaluate the capability of this method to assist with chiral resolutions.

Cyclic amines are among the most commonly recurring motifs in bioactive natural product and drug structures. A new method for the C–H functionalizion of unprotected cyclic amines via their transiently generated imines, reported by Daniel Seidel and co-workers ( Nat. Chem. 2018, 10, 165), represents a novel approach to the modification of these key fragments. Although the addition of organometallic nucleophiles to cyclic imines is well-established, Siedel’s methodology couples this to a new protocol for in situ imine generation through deprotonation and treatment with a hydride acceptor such as benzophenone. The low-temperature in situ generation of the imine electrophiles sidesteps common side reactions of these species such as trimerization or aza-enolate formation, enabling the one-pot amine α-functionalization to be easily accomplished without the need for transition metals or amine protecting groups. A range of alkyl, vinyl, and aryl organolithium nucleophiles are demonstrated on a broad array of differently sized rings. Although the stereochemical outcome of the addition is controlled by the substrate, the reaction was found to be highly diastereoselective (for trans addition) where existing stereocenters are present. Moreover, existing stereocenters next to nitrogen do not suffer from erosion of enantiopurity under the reaction conditions. Although the use of ether as a solvent and the requirement for cryogenic temperatures are problematic in a process chemistry setting, this approach represents an original and inexpensive alternative to many existing methods for amine C–H functionalization.

Despite the advances in asymmetric catalysis that recent decades have brought, covalently linked chiral auxiliaries are still a relatively common sight in organic synthesis, even in the pharmaceutical industry. Although many popular auxiliaries such as Oppolzer’s sultam can be recycled, their reuse typically adds several operations to a batch process, leading to increased cycle time and cost. In contrast, a recent paper by Sullivan and Newman ( Chem. Sci. 2018, DOI: 10.1039/c7sc05192a) describes the design of a continuous flow process that allows for the recycling of a chiral auxiliary without isolation, rendering the reaction “pseudocatalytic”, as multiple equivalents of product can now be produced for a single equivalent of chiral auxiliary. Although the reaction studied—the asymmetric reduction of α,β-unsaturated carbonyl compounds—has been reported using many varied catalytic systems, certain substrate classes remain problematic even for modern transition metal catalysis. For the system studied, the entire process could be accomplished with a total residence time of 30 min in flow, compared with around 22 h required to run the three steps in batch. Moreover, both the yield and diastereoselectivity were also improved in flow. Beyond this reaction, the design and engineering principles of this system may be applicable to other cases where no acceptable catalytic alternative is available or where increased efficiency is sought.

Chiral amino acids are of huge importance to the field of organic synthesis in catalysis, for the preparation of chiral auxiliaries, and as building blocks themselves. This broad utility has driven the development of numerous approaches to their synthesis. However, with the ever-increasing accessibility of the tools of chemical biology, biocatalytic and enzymatic methods have moved to the fore. A recent review by Yu-Guo Zheng and co-workers ( Chem. Soc. Rev. 2018, DOI: 10.1039/c7cs00253j) provides a useful overview of the numerous enzymatic reactions that can be used to synthesize nonracemic chiral amino acids, with a focus on unnatural substrates. The review focuses on four main categories of enzymatic reactions: asymmetric reductive amination of keto acids catalyzed by amino acid dehydrogenases (AADHs), asymmetric transfer of an amino group to keto acids by aminotransferases, enantioselective addition of ammonia to α,β-unsaturated acids by ammonia lyases, and aldol-type condensation of amino acids with aldehydes enabled by aldolases and hydroxymethyltransferases.

Solvents constitute a major part of waste streams from the execution of chemical processes and therefore have the potential to have a major impact on the environmental burden of a chemical reaction. Thus, judicious choice of the solvent for a reaction can be as important in minimizing cost, energy usage, and waste generated with a particular chemical step as the selection of the reagents and conditions themselves. New developments in the availability of renewable organic solvents as well as more specialized alternatives such as ionic liquids, liquid polymers, deep eutectic solvents, and supercritical fluids are summarized in a recent review by Jason Hallett and co-workers ( Chem. Rev. 2018, 118, 747). The performance of each class of solvent is outlined with respect to the chemistry for which it is suitable, its extraction properties, its environmental impact, and the wider context of the types of processes where it may be useful. The review contains 498 references and serves as an excellent introduction to solvents beyond those commonly found in a standard organic chemistry laboratory.

Iodine(III) reagents such as diacetoxyiodobenzene, iodosobenzene, and Koser’s reagent are capable of performing unique and powerful transformations in organic synthesis. However, the use of these reagents on a large scale is limited by their high cost, a lack of large-scale suppliers, and the high PMI associated with their use. Although catalytic systems based around these reagents have been reported for some time, to date no system that make uses of the simplest oxidant—molecular oxygen—has been reported. Now, David Powers and co-workers have disclosed a system that harnesses intermediates in the oxygen-driven autoxidation of simple aldehydes to achieve both the synthesis of iodine(III) reagents and catalytic versions of some of their best known reactions ( Nat. Chem. 2018, 10, 200). With acetaldehyde as the aldehyde reagent, combined with 1 mol % CoCl₂ as an initiator, a number of popular hypervalent iodine reagents can be prepared in high yield using only oxygen as the terminal oxidant. Furthermore, these reagents can also be generated and used in situ to accomplish the oxygen-driven oxidation of a number of classes of organic molecules. This provides an interesting and novel oxidase catalysis platform with fairly broad applicability. Clearly, the use of pure oxygen as an oxidant is impractical in a process setting, but the authors show that air can also be used, albeit in lower yield. Although the majority of reactions are reported in DCE, which is unappealing because of its toxicity, other solvents such as MeCN and AcOH were also demonstrated to be competent in the reaction.

Although the synthesis of enantioenriched trifluoroisopropanol via the Noyori-type Ru-mediated reduction of trifluoroacetone has been reported, the high hydrogen pressure required and the potential for runaway decomposition of the thermally unstable ketone starting material cause serious safety concerns for the use of this process on scale. In order to enable the efficient kilogram-scale preparation of chiral nonracemic trifluoroisopropanol, Chris Senanayake and co-workers at Boehringer Ingelheim have studied the asymmetric hydrogenation of the corresponding enol acetate under rhodium catalysis ( J. Org. Chem. 2018, 83, 1448). A highly efficient and enantioselective process was developed using the C₂-symmetric P-chiral diphosphine ligand t-Bu-SMS-Phos, but it was difficult to prepare the quantities needed using the published chemistry—despite the low loading and ligand recycling used. Thus, the authors report a much improved chiral-auxiliary-based approach that was used to prepare over 4 kg of the ligand, eliminating the numerous chromatographic purifications required in the original route. The hydrogenation itself was demonstrated on a 1 kg scale, and 97% of the ligand could be recovered and recycled.

The synthesis of ketones through carbonylation of organometallic reagents as the C-nucleophiles is well-established, but it suffers from the drawback of generating stoichiometric amounts of metal salts as waste. Beller and co-workers have reported a Pd-mediated, directing-group-free C–H carbonylation of a range of (hetero)arenes with readily available olefins ( ACS Cent. Sci. 2018, 4, 30). Model studies on the reaction of N-methylindole with 1-octene demonstrated the requirement to add p-TsOH as a cocatalyst to ensure sufficient reactivity. A range of bidentate phosphines were evaluated because of their known propensity to form linear over branched products, with Xantphos emerging as the best, providing a 92% yield with excellent selectivity for C-3 (>20:1) of the heteroarene and a good linear/branched ratio of products (88:12). From the olefin perspective, steric bulk increased the linear selectivity, while 1,1-disubstituted, internal (leading to branched products), tetrasubstituted, and cyclic olefins were all well-tolerated. In the case of (−)-β-citronellene, the terminal olefin was selectively carbonylated to give the linear ketone. Indoles, pyrroles, furans, thiophenes, and electron-rich arenes were all successful substrates. If C-3 of the indole is blocked, the reaction takes place at C-2, and the C2/C3 selectivity for pyrrole could be controlled by changing the steric bulk of the N substituent. Mechanistic and DFT studies suggest that the reaction proceeds through a Pd–hydride complex, with the regioselectivity determined by the strain energy in the transition state.

There are a number of challenges inherent in the development of a direct asymmetric Michael reaction between an α,β-unsaturated aldehyde and an unactivated ketone, including control of 1,2- versus 1,4-selectivity and generation of a suitable intermediate chiral species to control the enantioselectivity. Hayashi and Umekubo have reported on this reaction mediated by pyrrolidine accompanied by a chiral silyl ether to provide the product in 74% yield with 91% ee (syn/anti = 5:1) ( Angew. Chem. Int. Ed. 2018, 57, 1958). To explain the success of the reaction, labeling studies were carried out to determine the relative rates of formation of both the intermediate iminium ion and the enamine. These showed that both possible iminium species were formed from the aldehyde, with pyrrolidine reacting slightly faster, though only pyrrolidine reacts with the ketone to form the enamine. In order to rationalize the high levels of enantioselectivity observed, the chiral iminium ion must react preferentially with the racemic enamine, and this is postulated to occur because of its higher electrophilicity. Replacing pyrrolidine with l-trans-4-hydroxyproline led to higher levels of diastereoselectivity but had no impact on the enantioselectivity, while the addition of both water and p-nitrophenol is critical for reactivity. A range of aromatic and heteroaromatic moieties are tolerated at the β-position of the aldehyde, though alkyl groups are not. In addition, six- and seven-membered cyclic ketones give high yields and enantioselectivities, though both suffer in the case of the analogous five-membered rings.

Coumarins and dihydrocoumarins are prevalent scaffolds in both natural products and bioactive compounds and can be accessed through the hydroarylation of α,β-unsaturated acids with substituted phenols. Currently reported conditions utilize TFA to achieve this transformation, with electron-rich phenols being significantly preferred. Aubé and co-workers have developed a mild alternative protocol for this transformation using HFIP as the solvent with a catalytic amount of acetyl chloride added in order to generate HCl in situ ( Eur. J. Org. Chem. 2018, 306). Control studies showed no conversion in either neat acetyl chloride or HFIP alone, and although other fluorinated solvents were somewhat effective, HFIP proved to be superior, in some part because of its ability to solubilize the starting materials. A range of examples are provided, including electron-rich phenols and a number of phenol-containing natural products. Electron-rich cinnamic acids performed in a superior manner to electronically neutral derivatives, and a number of heterocycle-containing cinnamic acids are also reported. Mechanistic studies by NMR clearly demonstrate the development of a hydrogen bond between HFIP and the cinnamic acid, and the reaction is believed to proceed via 1,4-addition of the ortho carbon of the phenol followed by cyclization.

Aryl and heteroaryl halides are important building blocks in medicinal chemistry because of their propensity to participate in a variety of metal-mediated coupling reactions. However, their preparation is often complicated by the need to use either corrosive reagents or the requirement to add Lewis or Brønsted acids as activators of NXS-based reagents. Crousse and co-workers have reported a mild regioselective halogenation using HFIP as the solvent with the rationale that an activator would not be required because NXS would be activated in an electrophilic manner through hydrogen bonding to the solvent ( J. Org. Chem. 2018, 83, 930). A range of electron-rich arenes and heteroarenes were successfully brominated using NBS, primarily at the para position, with the reactions generally occurring at room temperature in an hour. Less electron-rich arenes than toluene were not successful substrates, and no benzylic bromination was observed. The methodology was extended to the use of both NIS and NCS, though the latter had a slightly narrower substrate scope. The reaction was also demonstrated on a gram scale and applied to one-pot sequential halogenations and halogenation/Suzuki cross-couplings.

The application of high-throughput experimentation for medicinal chemistry reaction optimization is to some degree limited by material availability, particularly for complex advanced intermediates. Whereas employing high-throughput screening-based instrumentation has enabled reactions to be performed on the submilligram scale in well plates, there are limitations in the range of solvents that can be used as well as the real-time resolution of the analysis. Workers at Pfizer have disclosed a flow-based screening platform that enables >1500 reactions to be evaluated in 24 h using <100 mg of material (Perera, D.; et al. Science 2018, 359, 429). The system, which is assembled from commercially available instrumentation, combines the reaction components using an autosampler and injects them into a flowing stream of what becomes the reaction solvent as a result of dilution. Control experiments indicate both the efficiency of segment mixing and the dispersion into the solvent. The dispersed segments are passed through a heated coil before being analyzed by high-resolution real time LC–MS. The residence time in the coil for the reported Suzuki–Miyaura coupling is 1 min at 100 °C, and there are two LC–MS systems for analysis in order to continually analyze the reactions as they emerge from the coil. Variation of the natures of both the nucleophile and the electrophile led to evaluation of 5760 reactions, with both positive and negative controls scaled up in a batch manner, to confirm the observed reactivities. In addition, a screen was executed to identify reaction conditions for a Pd-mediated coupling of 5-bromooxindole, which had previously failed in a number of Pfizer combinatorial libraries.

The classical Appel reaction represents an efficient method for the dehydration of primary amides to nitriles, but it suffers from disadvantages such as the use of toxic CCl₄ and the requirement for a stoichiometric amount of Ph₃P, which can lead to difficulties with purification. Malkov, Rubstov, and co-workers have reported a variation of the Appel reaction that avoids these issues and utilizes Ph₃P═O as a catalyst ( Org. Lett. 2018, 20, 728). Model studies using benzamide as the substrate indicated that as little as 1 mol % Ph₃P═O was effective, though the stoichiometries employed for oxalyl chloride (2 equiv) and Et₃N (3 equiv) were critical for complete conversion. Several solvents were able to mediate the reaction, with MeCN shown to be the best. From a scope perspective, aryl, heteroaryl, and aliphatic amides were all successful substrates, though aliphatic amides displayed lower reactivity. Phthalimide-protected amino acid amides were also successfully dehydrated, though the corresponding Boc-protected substrates led to multiple side products. Two potential mechanistic pathways were proposed, and the one featuring the intermediacy of an N-acyltriphenylphosphine imide (originally rejected by Appel) was shown to be highly probable through control studies.

A recent report by Tomkinson and co-workers from the University of Strathclyde ( J. Org. Chem. 2018, 83, 1510) describes the serendipitous discovery of a simple and effective protocol for the regioselective functionalization of electron-deficient heterocyclic N-oxides via treatment with 4-nitrobenzoyl chloride and a cyclic thioether under basic conditions. The transformation gave high yields of the corresponding thioether products and was tolerant of a wide range of substrates. The team circumvented the need for excess sulfide by the use of diisopropyl ether as a suitable solvent for the reaction. In addition, the products from these transformations provide an effective synthetic handle for further functionalization.

Sulfur fluoride exchange (SuFEx) click chemistry has a wide range of applications that are limited to S–O and S–N links. To broaden the scope of the applications, Sharpless and co-workers at the Scripps Research Institute in California have shown the extension of SuFEx click chemistry to include S–C bonds via the reaction of organolithium reagents with thionyl tetrafluorides ( Angew. Chem., Int. Ed. 2018, 57, 1939). The team prepared a variety of arylithium reagents by either halogen–lithium exchange or direct deprotonation, which under low-temperature conditions afforded monosubstituted products in good to excellent yields. In addition, the team synthesized a variety of sulfonimidamides, sulfoxides, and sulfoximines from a tunable S–F synthetic handle. It is worth nothing that the reported protocol is complementary to the established S–O and S–N SuFEx click chemistry.

There have been a variety of synthetic approaches to access allylic alcohols because of their wide applicability in synthetically and biologically relevant molecules. Mankad and co-workers at the University of Illinois at Chicago have reported a new approach to synthesize allylic alcohols via Cu-catalyzed hydrocarbonylative coupling of alkynes and alkyl halides ( J. Am. Chem. Soc. 2018, 140, 1159). After extensive experimentation, the team found the optimal conditions to employ a copper catalyst (^CliPrCuCl) with an N-heterocyclic carbene (^CliPr) ligand and polymethylhydrosiloxane as the reducing agent under CO at 6 atm pressure in the presence of KOMe as the base. The reaction tolerated a variety of functional groups, including benzyl ether, chloroallyl, and internal and terminal alkynes. Furthermore, the team found that tertiary alkyl halides may have a different radical mechanistic pathway compared with the primary and secondary allyl halide pathways.

Oxygen transfer reactions are a versatile tool in synthetic organic chemistry. Doyle and co-workers have reported the efficient generation of anhydrous vicinal di- and tricarbonyl compounds catalyzed by dirhodium(II) tetraoctanoate with dinitrogen extrusion ( Org. Lett. 2018, 20, 776). The team found that the treatment of diazoacetoacetates with 2,6-dichloropyridine N-oxide in chloromethane and 4 Å molecular sieves afforded the corresponding carbonyl compounds in high yields. In addition, the products are sufficiently reactive for further carbon–carbon bond transformations.

The synthesis of all possible stereoisomers continues to be of great importance to the pharmaceutical and agrochemical industries. Hartwig and co-workers at the University of California at Berkeley have shown the stereodivergent allylation of azaaryl acetamides and acetates to form all four individual stereoisomers of products containing vicinal stereogenic centers ( J. Am. Chem. Soc. 2018, 140, 1239). The team developed a combination of two catalysts, a chiral metallacyclic iridium complex and a chiral bisphosphine copper(I) complex, that was tolerant of a variety of functionalities including pyridyl, benzoxazolyl, and isoquinolinyl. Furthermore, the carbamate tolerated groups such as alkyl, heteroaryl, and alkenyl substituents, which afforded single diastereomers in most cases. The reported protocol may enable the rapid synthesis of all stereoisomers of chiral compounds for studies in understanding structure–activity relationships in drug discovery, for example.

Diversity-oriented synthesis (DOS) is a very attractive methodology because it provides rapid access to a number of important structural scaffolds. However, there are few reports of DOS involving heterocycles. Wang and co-workers have reported an efficient and facile cascade deprotection, cyclization, and ionic hydrogenation that affords diversified synthesis of heterocycles ( J. Org. Chem. 2018, 83, 1387). The team found that refluxing substrates such as hexahydrophenoxazine in the presence of Al(OTf)₃ and Et₃SiH in acetonitrile resulted in excellent yields of cyclized products. Various functional groups were tolerated, including electron-rich, electron-poor, and halogen substituents. However, there is still room for wider applicability of the reported protocol.

Malcolmson and co-workers at Duke University have shown that with a copper catalyst, α-aminoalkyl nucleophiles can be generated from infrequently used 2-azadienes, which add to ketones to furnish 1,2-amino tertiary alcohols with excellent enantioselectivities and diasteroselectivities ( J. Am. Chem. Soc. 2018, 40, 598). 2-Azadienes can be synthesized in two steps from benzophenone imines and undergo migratory insertion at their least-hindered π bond with a chiral Cu–H catalyst to generate the nucleophile. After addition to ketones, the resulting diphenyl imines can be easily hydrolyzed to the free amines. Ph-BPE was found to be a uniquely effective ligand with Cu(OAc)₂ for both the desired reactivity and enantioselectivity. Dimethoxymethylsilane is used as the reducing agent, resulting in a silyl ether after ketone addition, which can be desilyated with ammonium fluoride. 4-Substituted 2-azadienes provide the corresponding products in useful yields, but the added steric hindrance leads to competitive ketone reduction.

The direct asymmetric alkylation of ketones remains a challenging reaction for the synthesis of α,α-disubstituted ketones, making alternative routes that use readily available starting materials attractive. Studer and co-workers at Westfälische Wilhelms-Universität have described a novel route to α-chiral ketones using a transition-metal-free, radical-induced, stereospecific 1,2-migration of vinyl boron ate complexes with alkyl iodides ( Angew. Chem., Int. Ed. 2018, 57, 2441). The vinyl boron ate complexes are generated from enantioenriched boronic esters and vinyllithium at low temperature. Upon light irradiation, alkyl iodides add to the vinyl moiety, and after a 1,2-rearrangement, boronic ester diastereomers are formed (∼1:1 diastereomer ratio) with preservation of chirality at the alkyl center. The boronic ester moiety can be oxidized to the ketone using traditional methods or cleaved with TBAF/Mn(OAc)₃ to form an enantioenriched homologated alkyl chain. Several commercially available alkyl iodides were employed, including perfluoryoalkyl iodides. With perfluoryoalkyl iodides, the resulting α-perfluorylalkylated ketones undergo dehydrofluorination upon treatment with silica gel, forming useful fluorinated enones.

Meek and co-workers at the University of North Carolina have developed a practical method for the synthesis of useful allylic-alcohol-containing alkenyl boronates from epoxides and di-[B(pin)]-methane ( Org. Lett. 2018, 20, 469). The simple strategy involves the addition of lithiated 1,1-diborylmethane to an epoxide followed by stereoselective Pd-catalyzed dehydroboration to obtain an (E)-alkenylboron. A variety of epoxides can be opened, including 1,1- and 1,2-disubstituted epoxides. No loss in enantiopurity was observed when an enantiomerically enriched epoxide was used. Substitution on the 1,1-diboryl fragment is not tolerated; however, a borylsilylmethane was used to obtain an (E)-alkenylsilane, albeit in lower yield.

Only a handful of asymmetric reactions have been developed to transform simple unactivated α-olefins into highly functionalized compounds, with no examples of efficient enantioselective Markovnikov hydro- or protoboration of unactivated alkenes. Hong, Shi, and co-workers at Shanghai Institute of Organic Chemistry have developed a mild and general copper-catalyzed asymmetric protoboration of α-olefins to provide a diverse array of chiral secondary alkylboronic esters with high enantioselectivities ( Angew. Chem., Int. Ed. 2018, 57, 1376). The impressive Markovnikov selectivity is due to a novel NHC ligand based on the acenaphthoimidazolylidene framework (ANIPE). The ligand precursor is stable toward air and moisture and available in two steps from a known chiral amine. An X-ray crystal structure of the metal–ligand complex shows a very large buried volume for a monodentate ligand, providing the steric bulk required to deliver the Markovnikov product with good to high regioisomer ratio (rr). Also crucial for the selectivity is the use of B₂dmpd₂ as the boron source. B₂dmpd₂ is commercially available, and the derived products provide the same synthetic utility as more common boronic esters.

An elegant synthesis of pharmaceutically relevant 4-aryl-2-quinolinones has very recently been reported by researchers from China (Chu, W.; et al. J. Org. Chem. 2018, 83, 1422). The authors showcase a one pot protocol in which a visible-light-induced decarboxylation coupling/intramolecular cyclization method is employed on N-acetyltetrahydroquinolines and α-oxophenylacetic acids; over 24 examples are presented, with yields as high as 92% reported. Under the optimized reaction conditions, the reaction tolerates modification of the tetrahydroquinoline (modification at the 2-, 5-, or 6-position on the quinoline backbone) along with the incorporation of a variety of α-oxophenylacetic acids containing either electron-rich or electron-deficient groups. The authors utilized the methodology to synthesize an HBV inhibitor in 85% yield when the tetrahydroquinoline fragment was substituted for 4-chloroacetanilide. From a synthetic point of view, the method has numerous benefits, including atom economy, limited waste, and the use of readily available starting materials. The postulated mechanism was supported by control experiments involving TEMPO trapping of the key benzoyl radical generated by light irradiation of the keto acid.

The Peng group at Zhejiang Normal University in China have recently reported an iodine(III)-mediated oxidative ring contraction of aryl cyclobutanols to aryl cyclopropyl ketones (Peng, B.; et al. Adv. Synth. Catal. 2018, DOI: 10.1002/adsc.201701237). The authors show 26 successful examples of this particular transformation, with yields of up to 93% achievable. While known ring contractions of cyclobutanols have previously utilized acid/base or palladium catalysis, this new iodine(III)-promoted ring contraction protocol has a number of advantages: the method is mild, avoids the use of metals, and allows a broad range of functional group incorporation (an identified problem with traditional approaches). The authors demonstrated that the reaction can be scaled to a gram scale without significant loss in yield (cyclopropyl phenyl ketone example, 80% yield on a 1 g scale). In cases where more electron-rich arene substrates are used, the addition of 2,6-dichloropyridine facilitates the transformation that failed to proceed or was poor-yielding in its absence. The authors postulate a reaction mechanism based on their own experimental observations and literature precedents showing that cyclobutanol reacts with a molecule of PhI(OTf)₂ (formed from PhI(OAc)₂ and two molecules of TMSOTf) to initiate the ring contraction sequence.

The Hilinski group at the University of Virginia have recently disclosed their methodology for the site-selective C(sp³)–H amination of a range of substrates (Hilinski, M. K.; et al. Chem. Sci. 2018, 9, 935). The organocatalyst that facilitates the transformation is a tetrahydroisoquinoline iminium salt bearing a CF₃ substituent and gem-dimethyl moiety; at 20 mol % catalyst loading with iminoiodane (2 equiv) as the nitrogen source, a range of benzylic and aliphatic compounds could be aminated in yields of up to 87%. The methodology is mild and was shown to be site-selective, a point highlighted by late-stage aminations of ambroxide and a derivative of dehydroabietylamine with moderate diastereoselectivities. The authors postulate a mechanism that involves a diaziridinium intermediate as the nitrogen transfer agent, and this was supported by various control experiments. This organocatalytic method provides a real alternative to the use of rare or toxic metals to bring about C(sp³)–H aminations.

The Shipman group at the University of Warwick have recently disclosed their findings on the synthesis of hexahydropyridazine analogues from 3-alkylidene-1,2-diazetidines (Shipman, M.; et al. J. Org. Chem. 2018, 18, 491). The group utilized addition of various carbenes, such as difluorocarbene (generated from TMSCF₃ and NaI) or dichlorocarbene (generated from CHCl₃, TEBAC, and NaOH), across the exocyclic double bond. Yields obtained for the synthesis of 1,1-dichloro-4,5-diazaspiro[2.3]hexanes reached 64%, while the difluoro derivatives could be accessed in yields of up to 97% yield. The group subsequently reported the synthesis of larger 1,2-diazaspiro[3.3]heptanes in yields of up to 99%; tetracyanoethylene (TCNE), a favored reagent for [2 + 2] cycloadditions, was employed to react with electron-rich alkenes (six examples were shown to react smoothly under the selected conditions). The new spirocycles reported show potential to be incorporated into medicinal chemistry programs as hexahydropyridazine analogues.