A Telescoped Continuous Flow Enantioselective Process for Accessing Intermediates of 1-Aryl-1,3-diols as Chiral Building Blocks

A telescoped continuous flow process is reported for the enantioselective synthesis of chiral precursors of 1-aryl-1,3-diols, intermediates in the synthesis of ezetimibe, dapoxetine, duloxetine, and atomoxetine. The two-step sequence consists of an asymmetric allylboration of readily available aldehydes using a polymer-supported chiral phosphoric acid catalyst to introduce asymmetry, followed by selective epoxidation of the resulting alkene. The process is highly stable for at least 7 h and represents a transition-metal free enantioselective approach to valuable 1-aryl-1,3-diols.

1-Aryl-1,3-diols 1 are important synthetic building blocks for the pharmaceutical industry. 1They are key intermediates in the synthesis of numerous drugs, including ezetimibe (treatment of high blood cholesterol), 2 dapoxetine (premature ejaculation), 3 atomoxetine (attention deficit hyperactivity disorder), 4 and duloxetine (major depressive and anxiety disorders) 5 (Figure 1).Several synthetic routes have been developed to access optically active 1-aryl-1,3-diols using enantioselective reactions 5,6 and, most interestingly, organocatalysis. 7While asymmetric catalytic methods are more atom-efficient and produce less waste, the high cost of chiral ligands and organocatalysts often makes chiral auxiliaries the preferred option. 2,3,8To maximize the efficiency of existing catalytic enantioselective transformations, there has been a growing interest in the development of recyclable catalysts during the past decade. 9In particular, chiral phosphoric acids (CPAs) have seen widespread adoption due to their versatility. 10umerous applications of immobilized chiral CPAs have been reported to date, highlighting their significant potential to facilitate catalyst recovery. 11th regard to CPA-catalyzed enantioselective reactions with potential to synthesize optically active precursors of 1,3diols 1, Antilla and co-workers reported a highly enantioselective approach for allylboration of aldehydes using a 2,4,6tris-isopropyl-derived CPA, 12 known as TRIP 13 (Scheme 1A).A few years later, a copolymerization-based strategy was employed to immobilize TRIP onto a polystyrene resin, and the resulting supported catalyst (PS-TRIP) was successfully applied to enantioselective allylboration reactions as a highly recyclable organocatalyst. 11Even though some of the immobilized CPAs have been shown to be exceptionally active and robust, 11b,d they have not been widely utilized for the enantioselective synthesis of active pharmaceutical ingredients (APIs) and related compounds. 14ue to improved productivity, easier scalability, and waste reduction compared to more conventional batch procedures, telescoped continuous flow processes involving immobilized chiral catalysts have proven to be particularly useful for the multistep synthesis of optically active targets. 15Building on our previous efforts in flow synthesis of chiral APIs and their advanced intermediates, 16 we hypothesized that merging PS-TRIP-catalyzed asymmetric allylboration with selective epoxidation of the resulting chiral alkene in an uninterrupted flow process would open a simple and efficient entry to optically active 1,3-diols as key intermediates of atomoxetine, dapoxetine, duloxetine, and ezetimibe.The planned two-step process would produce enantioenriched epoxy alcohols 5 from readily available nonchiral aldehydes, which can then be easily transformed into the desired chiral diols 1 (Scheme 1B). 17By carefully selecting reaction conditions, we aimed to eliminate the need for any chromatographic purification, thereby facilitating larger-scale syntheses.
Our study began with optimizing the parameters of individual reaction steps.The activity of the PS-TRIP catalyst for asymmetric allylboration was explored in a flow setup consisting of two separate reagent feeds: solutions of benzaldehyde 2a (1.0 equiv) and allylboronic ester 3 (1.2equiv).The reagent streams were pumped at a flow rate of 100 μL/min each and were combined before entering a packed bed reactor containing 0.8 g of the supported catalyst (Scheme 1C).This corresponded to a residence time on the catalyst bed of ∼15 min.Several solvents were evaluated with the purpose of making the overall process greener. 18The effect of substrate concentration was also explored to maximize the productivity.The best results for obtaining alkene 4a were achieved in 97% yield and 90% enantiomeric excess (ee) using a substrate concentration of 0.15 M in toluene as the solvent (see Table S1 for details).
Next, various strategies were explored for the subsequent epoxidation, initially under batch conditions (Table 1).We found that hydrogen peroxide as an oxidant resulted in overoxidation of the desired chiral alcohol (5a) to the corresponding ketone 6a, making the process unsuitable for further development (Table 1, entry 1).Dimethyldioxirane (DMDO), generated from acetone and Oxone (2KHSO 5 • KHSO 4 •K 2 SO 4 ) in a buffered aqueous solution, 19 showed high conversion and selectivity (Table 1, entry 2) but involved miscibility issues with toluene.To avoid solubility problems that could affect the reactivity in flow, we next evaluated organic peracids.Commercially available solutions of peracetic acid (PAA) showed high selectivity but only poor conversion (Table 1, entries 3 and 4).
Although the in situ generation of peracids under continuous flow conditions is well-known, 20 preliminary tests showed significant overoxidation to ketone 6a, probably due to the large excess of H 2 O 2 required in these reactions.Therefore, we finally tested m-chloroperbenzoic acid (mCPBA) as the epoxidation agent.Gratifyingly, excellent conversion and selective epoxidation were achieved in the presence of 4.0 equiv of mCPBA, making it the preferred oxidant for further development (Table 1, entries 5−7).Although the diastereoselectivity of the epoxidation process is minimal, the late removal of the chiral center on the epoxide makes it not relevant for synthesis of diols 1.The mCPBA-mediated selective epoxidation was then transferred to continuous flow using a simple coil reactor, ensuring conversions of ≥90% within residence times of ∼10 min at 85 °C (see Table S2 for details).
Following step-by-step optimization, we combined the PS-TRIP-catalyzed asymmetric allylboration of benzaldehyde (2a) and the subsequent epoxidation in a telescoped flow sequence to access epoxy alcohol 5a, a chiral intermediate of atomoxetine and dapoxetine (Scheme 2A).Downstream to the packed bed reactor, the mCPBA feed served a double role.Apart from functioning as an epoxidation agent, it also quenched any unreacted allyl pinacol ester, thereby preventing racemic background reactions in the case of uncompleted allylboration.To safely quench any excess oxidant, the outlet of the reactor was directed into a stirred solution of Na 2 S 2 O 5 .With the optimized setup in hand, we performed a continuous long run for 7 h.The overall process was followed by off-line HPLC with samples taken and analyzed every hour.We were pleased to find no decrease in either the conversion or enantioselectivity, showing the robustness of the process (Scheme 2B).Contrary to previous reports on enantioselective allylboration reactions, 11d,12,21 the process presented here did not require any chromatographic purification but a simple acid/base extractive workup to isolate the desired chiral adduct in sufficiently pure form.
To obtain potential precursors of ezetimibe and duloxetine, the two-step flow synthesis was next attempted using 4fluorobenzaldehyde (2b) and 2-thiophenecarboxaldehyde (2c)  The Journal of Organic Chemistry as the substrate, respectively (Scheme 2C).Epoxy alcohol 5b was smoothly produced from aldehyde 2b during a continuous 3 h run (90% yield, 92% ee) under conditions identical to those applied in the synthesis of 3a.In the targeted synthesis of oxirane 5c from aldehyde 2c, the epoxidation step resulted in a complex mixture, probably due to the polymerization of the thiophene ring. 22In this case, the process was stopped after the allylboration step (performed using the setup shown in Scheme 1C; see also the Supporting Information for details) to afford alkene 4c in 99% yield and 66% ee.
To illustrate the applicability of epoxides 5 in the synthesis of 1-aryl-1,3-diols 1, we performed the ring opening of epoxide 5a in acidic media, affording triol 7a in high yield (Scheme 3).Further transformations of triols 7 to the corresponding diols 1 are known in the literature. 17 summary, we have developed a telescoped continuous flow process using an immobilized CPA-mediated enantioselective allylboration as the key step followed by mCPBAmediated selective alkene epoxidation.Our strategy consists of a transition-metal free catalytic method to access triols 7 and diols 1 in high yield and enantiocontrol by using a robust immobilized organocatalyst.By exploiting an uninterrupted flow process, we obtained chiral epoxides 5 in a simple and efficient manner, without the need for any chromatographic purification.With a cumulative residence time of <30 min, the protocol enabled a notable chemical intensification compared to earlier methodologies.
■ EXPERIMENTAL SECTION General Information.All solvents and chemicals were obtained from typical commercial vendors and used as received, without any further purification. 1H, 19 F, and 13 C NMR spectra were recorded on a Bruker Avance III 300 MHz instrument at room temperature, in CDCl 3 as the solvent, at 300 and 75 MHz.Chemical shifts (δ) are reported in parts per million relative to the residual solvent peak (CDCl 3 , 1 H, 7.26 ppm; 13 C, 77.16 ppm).Coupling constants are reported in hertz.Multiplicity is reported with the usual abbreviations.
When required, column chromatographic purification was performed by using a Biotage Isolera automated flash chromatography system with cartridges packed with KP-SIL, 60 Å (32−63 μm particle size).Analytical thin-layer chromatography (TLC) was carried out using Merck silica gel 60 GF254 plates.Compounds were visualized by means of ultraviolet (UV) or KMnO 4 .
Analytical HPLC analysis was carried out on a C18 reversed-phase (RP) analytical column (150 mm × 4.6 mm, particle size of 5 mm) at 37 °C by using mobile phases A [90:10 (v/v) water/acetonitrile with 0.1% TFA] and B (acetonitrile with 0.1% TFA) at a flow rate of 1.5 mL/min.The following gradient was applied: linear increase from 3% to 5% B over 3 min, linear increase from 5% to 30% B over 4 min, linear increase from 30% to 100% B over 3 min, hold at 100% B for 2 min, linear decrease from 100% to 3% B over 0.5 min, and hold at 3% B for 2.5 min.
The ee of the compounds was determined by chiral HPLC or chiral GC.Chiral HPLC analysis was performed on a Shimadzu HPLC system (DGU-403 degassing unit, CTO-40S column oven, CBM20 system controller, SPD-40 UV−visible detector, LC-20AT pumps).Chiral GC analysis was performed on a Trace-GC (ThermoFisher) GC system equipped with a flame ionization detector (FID), using an Rt-BDEXse column [30 m × 0.32 mm (inside diameter) × 0.25 μm df] (Restek GmbH) and helium as a carrier gas (linear velocity of 0.5 mL min −1 ).FID was used for detection, and the detector gases used for flame ionization were hydrogen and synthetic air (5.0 quality).
Optical rotation was measured in CHCl 3 (HPLC-grade) at 25 °C against the sodium D line (λ = 589 nm) on a PerkinElmer Polarimeter 341 using a 10 cm path length cell.The specific rotation was calculated with the following equation where T is the temperature in degrees Celsius, D is the sodium D line emission, α is the angle of rotation, c is the concentration of the solution in grams per 100 mL, and d is the length of the polarimeter tube in decimeters (here 1 dm).The given data were calculated as the average of three measurements.The absolute configuration was determined by comparison of the optical rotation for compound 4c, and the absolute configurations of other compounds were assigned by analogy. 12The Journal of Organic Chemistry High-resolution mass spectra were recorded in either negative or positive mode on an Agilent 6230 TOF LC/MS instrument (G6230B) by flow injections on an Agilent 1260 Infinity Series HPLC instrument (HiP degasser G4225A, binary pump G1312B, ALS autosampler G1329B, TCC column thermostat G1316A, and DAD detector G4212B).
Equipment for the continuous flow reactions was assembled using commercially available components.Liquid streams were pumped by using Syrris Asia syringe pumps.Flow systems were pressurized by using an adjustable backpressure regulator (BPR) from Zaiput and/or by using a fixed-pressure BPR from IDEX.Reaction coils were heated by means of a conventional oil bath.Reagent feeds were streamed directly or by using injection valves and sample loops.Sample loops and reactor coils were made by using perfluoroalkoxy alkane (PFA) tubings (1/16 in.outside diameter, 0.80 mm inside diameter or 1/8 in.outside diameter, 1.58 mm inside diameter).Details of reaction setups as well as general procedures can be found in the following sections.
Synthesis of the Catalysts.The synthesis of PS-TRIP catalysts was performed following Pericas's procedure.11d,g The catalyst loading of the resin was calculated on the basis of the P elemental analysis by using the following formula: % 1000 number of P atoms MW( ) 100 Anal.P, 0.48%; f = 0.15 mmol/g.
General Procedure for the Batch Synthesis of Racemic 4. The corresponding aldehyde (1.0 mmol, 1.0 equiv) was dissolved in 5 mL of toluene, and AllylBpin (225.1 μL, 1.2 mmol, 1.2 equiv) was added dropwise at room temperature.The reaction mixture was stirred overnight and then concentrated under vacuum.The reaction crude was purified by column chromatography on silica gel (1:0 to 7:3 hexanes/Et 2 O, followed at 210 nm).The reported data match the literature.
Experimental Procedure for the Telescoped Flow Synthesis of Oxiranes 5. First, 0.8 g of the PS-TRIP catalyst was loaded into an adjustable Omnifit glass column [10 mm (inside diameter)].Prior to the reactions, the catalyst bed was swollen by pumping toluene at a rate of 200 μL/min for 30 min.Stock solutions (in toluene) of aldehyde 2a (0.30 M, 100 μL/min, 1.0 equiv) and 3 (0.36 M, 100 μL/min, 1.2 equiv) were pumped independently (overall flow rate of 200 μL/min) and combined at room temperature just before the catalyst-containing Omnifit column by using a Syrris Asia syringe pump.A check-valve was added to the exit of the packed bed reactor to avoid back flow.Then, a mCPBA a solution in toluene (0.30 M, 400 μL/min, 4.0 equiv) was combined in the reaction coil heated to 85 °C in an oil bath.b The system was pressurized at 5 bar by using a Zaiput BPR.The reaction outcome was quenched by collecting the mixture directly into a stirred aqueous solution of 1.0 M Na 2 S 2 O 3 .More detailed information about the flow setup is shown in section S3 of the Supporting Information.
Workup for 1 h of the Reaction in Continuous Flow.The reaction mixture was collected over 30 mL of a 1.0 M solution of Na 2 S 2 O 3 .The organic phase was then separated and washed with 1.0 M NaOH (3 × 50 mL), saturated NaHCO 3 (1 × 50 mL), and brine (1 × 50 mL).The organic phase was then dried over MgSO 4 , filtered, and concentrated under vacuum.
Workup for 3 h of the Reaction in Continuous Flow.The reaction mixture was collected over 90 mL of a 1.0 M solution of Na 2 S 2 O 3 .The organic phase was then separated and washed with 1.0 M NaOH (3 × 150 mL), saturated NaHCO 3 (1 × 150 mL), and brine (1 × 150 mL).The organic phase was then dried over MgSO 4 , filtered, and concentrated under vacuum.
Note that the epoxidation step was not suitable for the synthesis of compound 5c due to the undesired polymerization of the thiophene in the reaction. 22In contrast to compounds 5a and 5b, in which colorless or pale yellow solutions were observed, during the synthesis of 5c, a black precipitate is formed in the case of 4c and mCPBA, leading to the clogging of the system and a complex mixture of byproducts.
Therefore, the 3 h run for the synthesis of 4c was performed by collecting the allylation product of 2a and 3 just after the packed bed reactor, using the optimal reaction conditions for the allylboration reaction (Table 1, entry 6).The reaction outcome was quenched by collecting directly into a stirred 1.0 M Na 2 S 2 O 3 (90 mL), and the organic phase was extracted in toluene, dried over MgSO 4 , filtered, and concentrated.The reaction crude was purified by column chromatography on silica gel (1:0 to 7:3 hexanes/Et 2 O, followed at 210 nm).