Multistep Continuous Flow Synthesis of Fine Chemicals with Heterogeneous Catalysts

Book Preface

Continuous-flow synthesis has many advantages over conventional batch synthesis from the viewpoint of efficiency, safety, environmental friendliness, and scalability. Among several types of continuous-flow methods, flow reactions using heteroge-neous catalysts are the most attractive and efficient system for multistep chemical transformations, because the use of activated reagents can be avoided and catalysts can be easily separated from products and used continuously. However, the appli-cation of heterogeneous catalysts to continuous-flow reactions is still limited for single-step reactions, and it has been regarded as a great challenge to synthesize fine chemicals under continuous-flow conditions using heterogeneous catalysts. In general, Active Pharmaceutical Ingredient (API) synthesis requires multistep chem-ical transformations. To realize continuous-flow synthesis of APIs with heteroge-neous catalysts, each reaction has to be “clean” without generating byproducts for the next reaction. For this reason, the precise design of whole synthetic routes and the development of catalysts to enable each transformation are essential and challenging points. Previously, our group reported a multistep continuous-flow synthesis of chiral API, Rolipram, without any quenching and purification operation. The whole process involves six chemical transformations through four kinds of heterogeneous catalysts. Although this example is the milestone of continuous-flow fine chemical synthesis, the scope of reactions and catalysts is still limited and needs to be expanded for the future development of this field.
In my Ph.D. thesis, I decided to focus on C–C bond formation and hydrogenation, because these types of reactions generally take place with high atom economy and generate water as a sole byproduct, which can be easily removed. To achieve a multi-step continuous-flow synthesis, my strategy is a construction of backbone of a target molecule at first by aldol-type reactions with substrates with high oxidation states, which can potentially act as a nucleophile, followed by conversion of the functional groups to final target by selective hydrogenation. I hypothesized that various kinds of fine chemicals can be synthesized by connecting these two types of reactions under continuous-flow conditions.
Synthesis of Nitro-Containing Compounds Through Multistep
Continuous Flow
Nitro alkenes are one of the most important, versatile, and frequently used inter-mediates in organic synthesis. In the first step, I investigated the effect of flow rates and concentrations on productivity and yield using nitromethane and benzalde-hyde as substrates using amine-functionalized silica with 2 CaClas a catalyst under continuous-flow conditions. The reactions were performed with various flow rates between 0.05 and 1.0 mL/min at 0.1 M concentration. With 0.05 ml/min flow rate, the yield was kept >90% to supply 36 mmol of nitromethane. The yield was kept with 0.1 ml/min flow rate, resulting in an increase of productivity by double. However, a further increase in flow rate from 0.25 to 1.0 ml/min resulted in a decrease in the yield. Next, concentrations were changed between 0.1 M and 1.0 M with 0.05 ml/min flow rate. Surprisingly, the yield was maintained >90% even with 1.0 M concentration, resulting in an increase of productivity by 10 times. These results indicated that longer residence time was the key to achieve high productivity. Under optimized reaction conditions, the scope of aldehyde was examined. With five kinds of aromatic alde-hydes, the yield was kept >80% to supply ~150 mmol of substrates. For the second step, I investigated various types of acid–base heterogeneous catalysts such as metal oxides, surface-functionalized SiO2, and polystyrene-immobilized catalysts. Metal oxides worked as heterogeneous base catalysts and promoted 1,4-addition of benzy-lamine and 1,3-ketoester under continuous-flow conditions. During the investigation,
I found that fresh preparation of catalysts was the key to achieve >80% yield and 48 h lifetime. DMAP-immobilized silica was employed as heterogeneous Lewis-base catalyst for Morita–Baylis–Hillman reaction. Although the catalyst was effec-tive under single continuous-flow conditions, the catalyst deactivation was observed after 6 h when combined in the first nitroolefin synthesis. Such deactivation problem was solved by changing the dehydrating agent in the first column from 2 to CaCl MS 4A, indicating that the deactivation was caused by leached Ca species. Finally, I could synthesize seven kinds of nitro-containing compounds in two steps under continuous-flow conditions without any workup and purification (Scheme1).

Anion Exchange Resins as Catalysts for Direct Aldol-Type
Reactions
To establish a general method of aldol condensations under continuous-flow condi-tions, I started the investigation of heterogeneous base catalysts. As a model reaction, α-tetralone and benzaldehyde were used as substrates and various heterogeneous catalysts were evaluated under batch conditions. Amine-functionalized silica cata-lysts did not give any product, although they were effective for nitroalkene synthesis. Common solid bases such as metal oxides, hydrotalcite, and 2OKF/Al3 resulted in
low conversion and yield presumably due to low basicity and deactivation by gener-ated water. On the other hand, a basic resin having ammonium hydroxide afforded the desired α,β-unsaturated ketone in 90% yield. Because the physical properties of these resins differ significantly depending on the solvent, I decided to investigate the effect of solvent to catalyst activity. As a result, EtOH and Toluene gave supe-rior results compared to THF. However, no conversion was observed using EtOAc or DCM as a solvent. Although the origin of the solvent effect was still unclear, swelling of resins seemed to be the one key factor to achieve high catalyst activity. Interest-ingly, the aldol product could be obtained in >99% chemoselectivity by decreasing reaction temperature from rt to −o40 C using the same catalyst in moderate yield. The catalyst efficiently worked even under continuous-flow conditions to obtain the desire α,β-unsaturated compounds in >80% yield with >99% selectivity for >48 h. Throughout the investigation of substrate scope, the catalyst was found to be effective for other kinds of nucleophiles such as benzyl nitriles, and even acetonitrile could be employed as nucleophile when using a solvent amount. Interestingly, a similar solvent effect was observed for other nucleophiles (Scheme2).

Polysilane-Supported Pd Catalysts for Continuous-Flow
Hydrogenations
Previously, our laboratory reported polysilane-supported Pd/Al2O3 catalysts for hydrogenation of alkenes and alkynes under continuous-flow conditions. I thought that changing support materials would affect the property of the catalysts and could be applicable for hydrogenation of nitriles. I prepared polysilane-supported Pd cata-lysts with different inorganic supports and evaluated using decanenitrile as a model substrate under continuous-flow conditions. Using 2AlO3 as support, the reaction proceeded to achieve 70% conversion under 1.5 bar 2of . HHowever, secondary
amine was obtained as a side product. It was found that the addition of 1.5 eq. of HCl improved conversion to 76% and selectivity to 96%. On the other hand, 100% selectivity was achieved using S2 iOas catalyst support, although the lower conver-sion was observed. By increasing the amount of catalyst and reaction temperature, 99% selectivity was maintained to achieve 89% yield. On the other hand, the selec-tivity was affected by the concentration of the substrate. When the concentration was decreased from 0.2 M to 0.1 M, the selectivity was decreased to 89%. Decreasing the flow rate from 0.1 ml/min to 0.05 ml/min diminished selectivity to 88% as well. These results indicate that high concentration and short residence time are the keys to prevent a non-catalyzed undesired side reaction. Very interestingly, when the reac-tion was performed under batch conditions, a complex mixture was obtained in low conversion. This result indicates the advantage and uniqueness of the continuous-flow method. This method could be applicable for various kinds of aromatic, heteroaro-matic, and aliphatic nitriles. This catalyst kept its activity even after 120 h continuous-flow reaction and no Pd leaching was detected in the resulting solution. I also inves-tigated the support effect in the hydrogenation of aliphatic nitrocompounds. For this reaction, Bone charcoal was found to be optimal support to give the desired compounds in excellent yields with selectivities. The catalyst could be applicable for various kinds of aliphatic nitrocompounds including secondary and tertiary nitrocom-pounds. In conclusion, I developed general methods to prepare primary amines from easily available starting materials under continuous-flow conditions (Scheme3).