The m-CPBAâNH3(g) System: A Safe and Scalable Alternative for the

Article Cite This: Org. Process Res. Dev. XXXX, XXX, XXX−XXX

pubs.acs.org/OPRD

The m‑CPBA−NH3(g) System: A Safe and Scalable Alternative for the Manufacture of (Substituted) Pyridine and Quinoline N‑Oxides† Amey Palav,‡,§ Balu Misal,‡,§ Anilkumar Ernolla,‡ Vinod Parab,‡ Prashant Waske,∥ Dileep Khandekar,∥ Vinay Chaudhary,∥ and Ganesh Chaturbhuj*,§ ‡

Research and Development Center, Loba Chemie Pvt. Ltd., Tarapur, Thane 401 506, India Department of Pharmaceutical Sciences and Technology, Institute of Chemical Technology, Matunga, Mumbai 400 019, India ∥ Mettler-Toledo India Pvt. Ltd., Amar Hill, Saki Vihar Road, Powai, Mumbai 400 072, India Downloaded via UNIV OF NEW ENGLAND on January 16, 2019 at 07:01:58 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

§

S Supporting Information *

ABSTRACT: An improved, safe, and scalable isolation process for (substituted) pyridine and quinoline N-oxides in quantitative yields along with high purities using the m-CPBA−NH3(g) system is described. The safety was assessed by reaction calorimetry and differential scanning calorimetry studies for possible hazards during the conversion and isolation steps. Careful interpretation of the data substantiated the safety and scalability. The process flow is simplified to meet the industrial requirements of safety, cost-effectiveness, and utility minimization. The reaction was safely demonstrated at a 2.5 kg scale. KEYWORDS: pyridine, pyridine N-oxide, m-CPBA, ammonia

1. INTRODUCTION

differential scanning calorimetry (DSC) to identify the onset of degradation are very much required. Herein we report an industrial protocol for the preparation of N-oxides of (substituted) pyridines and quinolines using an organic oxidant. Among the many oxidants reported for Noxidation, m-chloroperoxybenzoic acid (m-CPBA) was chosen for use under controlled conditions of solvent compatibility, mode of addition, and temperature along with scavenging of excess peroxides in the reaction mixture.24,25

In recent years, the importance of heterocyclic N-oxides has been continuously growing in research and industrial applications. The N-oxide scaffold is an integral part of several drug molecules such as tirapazamine, an anticancer drug,1 and pyrithione zinc, a fungistatic and bacteriostatic drug.2 They have also been promising lead compounds in the new drug discovery for HIV,3 cancer,4 tuberculosis,5,6 and type II diabetes.7 Moreover, N-oxide derivatives of substituted pyridines serve as key raw materials for the synthesis of rabeprazole, an antiulcer drug,8 and lansoprazole, a proton pump inhibitor.9 The 2- and 4-substituted pyridine and quinoline N-oxides have a high demand in drug development10 and are synthesized by Polonovski rearrangement in the presence of acetic anhydride11 or trifluoroacetic acid.12,13 Several reagents are used for N-oxidation of heterocycles, such as peracetic acid/AcOH,14 perbenzoic acid,15 monoperphthalic acid,16 H2O2/AcOH,17 H2O2/[Mn(TDCPP)Cl],18 H2O2/MTO, BTSP/HOReO3,19,20 and dimethyl dioxirane.21 The aforesaid methods have restricted application at the industrial scale because of safety concerns like explosions22 at elevated temperatures with the decomposition of the oxidant23 and the product. The high water solubility of N-oxides poses difficulty in isolation and purification during biphasic extractive workup. To scale up the manufacture of N-oxides to higher volumes, there arises a need for the development of a simplified process that meets the industrial requirements of safety, cost-effectiveness, and minimization of utilities and waste. Safety in reactions at the manufacturing scale is of utmost importance. The reagents used and the products are of explosive nature, hence necessitating a careful assessment of heat changes and stability of the chemicals involved under these conditions. Safety studies using reaction calorimetry and © XXXX American Chemical Society

2. RESULTS AND DISCUSSION In all of the initial experiments, the oxidation step showed quantitative conversion, so the only hurdle to be tackled was to separate m-chlorobenzoic acid generated during the reaction and the N-oxide from the reaction mixture. Attempts were made to precipitate m-chlorobenzoic acid as a salt using organic and inorganic bases such as triethylamine and sodium bicarbonate. The salts, which are water-soluble but insoluble in organic solvents, may be separated from the mixture either by extractive workup or filtration. Among the bases used, sodium salts resulted in poor yields because of the considerable water solubilities of the products, while the triethylammonium salts did not precipitate from the reaction solvent, which was expected to be filtered. To overcome these complexities, one more alternative was attempted to convert m-chlorobenzoic acid into its ammonium salt by sparging anhydrous ammonia gas into the reaction mass in an organic solvent (rather than by aqueous liquor ammonia). To our surprise, ammonium mchlorobenzoate precipitated from the reaction mixture and was separated quantitatively by filtration on a Celite bed. Hence, there was a necessity to optimize the aforesaid initial findings Received: October 31, 2018 Published: December 28, 2018 A

DOI: 10.1021/acs.oprd.8b00358 Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development

Article

to achieve maximum recovery of m-chlorobenzoic acid and the solvent. 2.1. Protocol Development. The experiments were planned to ascertain the effect of chlorinated hydrocarbon, hydrocarbon and ester solvents (Table 1). Although the Table 1. Solvent Optimization and Recovery Studya entry

solventb

class

yield (%)c

purity (%)d

solvent recovery (%)

1 2 3 4 5 6

DCE DCM hexanes cyclohexane EtOAc iPrOAc

1 2 2 2 3 3

90.0 89.0 98.0 98.1 98.0 98.0

95.0 95.6 99.0 99.2 96.0 99.8

85−90 55−60 62 60 82−85 90−92

a

Experimental conditions: pyridine (200.0 g, 2.5316 mol, 1.0 equiv). 10 vol, 75% m-CPBA (610 g, 3.53 mol, 1.4 equiv). cIsolated yields. d Purities of the isolated products as determined by GC. b

Figure 1. DSC graph (test range 30−400 °C) for 0.3 g/mL m-CPBA in iPrOAc.

Table 2. Time and Temperature Optimization Studya

oxidation preceded smoothly in hexanes and cyclohexane, it posed the difficulty of a nonfilterable cake of ammonium mchlorobenzoate that entrapped the product (Table 1, entries 3 and 4). In addition to this, the hydrocarbon solvents are highly flammable even with an electrostatic charge. The use of dichloromethane (DCM) gave a comparable yield with reasonable recovery of the solvent, while dichloroethane (DCE) gave similar results and better solvent recovery (Table 1, entries 1 and 2). However, it was suggested in the literature that DCM would be the solvent of choice for oxidation purposes.24 Nevertheless, there are certain issues at higher concentration of m-CPBA in DCM due to the intrinsic instability of m-CPBA, even though they are compatible.26 Because chlorinated solvents in ICH class 1 or 2 need to be replaced in order to address the issues of toxicity, high volatility, and recovery, the use of an environmentally benign ICH class 3 solvents such as ethyl acetate (EtOAc) or isopropyl acetate (iPrOAc) was desirable. These two solvents gave comparable yields and purities of the desired product with better recovery, and isopropyl acetate was the solvent of choice (Table 1, entries 5 and 6). We also studied the compatibility of m-CPBA with iPrOAc, and the onset temperature was found to be 50.07 °C (which is much lower than the decomposition temperature of 89 °C for m-CPBA in the solid state) with a heat release of 205.3 J/g26 (Figure 1). The temperature during the addition and quenching of m-CPBA is much lower than the onset temperature. Hence, this confirms that m-CPBA does not react or form a hazardous combination with iPrOAc. The oxidation is known to be an exothermic reaction. Avoiding a runaway reaction necessitates the determination of the heat of reaction using reaction calorimetry. Hence, initially, the temperature and time optimization studies were performed to access the ΔT during both the conversion and isolation steps (Table 2). As per the protocol of a 200.0 g input batch of pyridine, the oxidant m-CPBA was added in 10 portions while the temperature of the reaction mass was maintained between 10 and 15 °C. The temperature was optimized during the addition of m-CPBA and the course of the reaction until completion. The results revealed that it was necessary to maintain the temperature between 10 and 15 °C during the addition of m-CPBA and afterward to maintain it at 30−35 °C up to 4 h, which resulted in maximum conversion as well as purity of the product as assessed by GC. The aforesaid

entry

temperature (°C)b

time (h)c

conv. (%)d

1

10−15

2

20−25

3

30−35

2 4 6 2 4 6 2 4 6

78.2 82.3 85.6 85.7 88.7 90.5 98.5 98.5 98.5

a

Experimental conditions: pyridine (200.0 g, 2.53 mol, 1.0 equiv), iPrOAc (10 vol), 75% m-CPBA (610.0 g, 3.53 mol, 1.40 equiv). b During addition of m-CPBA. cAfter addition of m-CPBA. d Determined by GC.

protocol was subjected to reaction calorimetry assessment using a Mettler Toledo RC1 reaction calorimeter. The trends of heat changes indicated by the RC1 during the addition of mCPBA are depicted in Figure 2. These data were used to measure the heat of reaction, adiabatic temperature rise, and Stoessel criticality class (Figure 3). As depicted in Figures 2 and 3 the maximum heat rise during the addition of 10 portions of m-CPBA was 19.4 K, with a maximum temperature of synthetic reaction (MTSR) of 29.44 °C and an onset of decomposition temperature (TD24) of 101.5 °C as calculated by extrapolation using nth-order kinetic software with respect to the DSC data of reaction mixture, indicating that the reaction is safe even if the cooling fails, as the MTSR is much lower than the TD24 limit. The Stoessel criticality class is 1 for the m-CPBA addition operation, which is regarded as safe (Figure 3).27 2.2. Quenching of Excess Oxidant and Workup Procedure. The oxidation reaction was performed with 1.4 equiv of commercially available m-CPBA (75% w/w). There was a possibility that traces of peroxide could still be present at the end of the reaction time, which could lead to an explosion. To design a hazard-free workup procedure necessitated quenching of unreacted m-CPBA with suitable scavenger prior to ammonia sparging and evaporation of the solvent. To accomplish this, sodium sulfite, a cost-effective scavenger with no product contamination, was used. Sodium sulfite was added at the completion of the reaction until the test was negative on B



Article

Figure 2. RC1 trend of m-CPBA addition (red, Tr; blue, Tj; dark-pink, qr_rtc; green, Vr; dark-red, Tr − Tj; light-pink, Rt).

Figure 3. (a) Runaway graph based on current values of m-CPBA addition. (b) Stoessel criticality class for m-CPBA addition.

Figure 4. Trend of ammonia sparging (red, Tr; blue, Tj; dark-pink, qr_rtc; green, Vr; light-pink, Rt).

pH increased to 8−9. This neutralization protocol was studied using the RC1 reaction calorimeter to establish the safe handling temperature of a reaction mixture containing the explosive N-oxide. The RC1 trend of ammonia sparging is depicted in Figure 4. The data revealed that the maximum temperature rise during addition was 2.3 K with an MTSR of 12.3 °C and TD24 of 101.5 °C, which also reveals that the maximum temperature even after a cooling failure would reach 53.0 °C (Figure 5a). The

MN-Quantofix strip. No heat change was observed during the addition of sodium sulfite. This was followed by sparging with anhydrous ammonia gas to convert m-chlorobenzoic acid into ammonium m-chlorobenzoate. Formation of the ammonium salt is an exothermic process and hence needs to be performed at a lower temperature with control on the amount and rate of ammonia gas sparging. The rate of gas sparging was optimized to 180−200 mL/min while the reaction temperature was maintained between 10 and 15 °C until the reaction mixture C



Article

Figure 5. (a) Runaway graph based on current values of ammonia sparging. (b) Stoessel criticality class for ammonia sparging.

Figure 6. DSC graphs for (a) pure pyridine N-oxide and (b) pure pyridine N-oxide in iPrOAc (0.1 g/mL).

purity by HPLC). It is well-known in the literature that mchlorobenzoic acid can be recycled to m-CPBA.28 2.4. Kilogram-Scale Protocol Development. The RC1 trends revealed that the process can be scaled up in the equipment setup available for a general-purpose reactor. We scaled up the reaction to 2.5 kg of pyridine input in a 50 L glass reactor with a suitable jacketed cooling arrangement and outlet connected to the common scrubber. Initial addition of pyridine to the isopropyl acetate showed no exothermicity using a digital temperature indicator, substantiating the observation in the RC1 experiment. The time taken for the addition of the first portion of m-CPBA (762.0 g) was 15 min with a temperature rise from 10 to 13 °C. After addition, the reaction mixture was aged for a further 10 min to stabilize the temperature at 10 °C. The same exothermic pattern was observed for the subsequent five portions of m-CPBA. The seventh to 10th portions showed a slight temperature rise of 1 °C. The operational procedure of aging the reaction mixture for 10 min after the addition of each m-CPBA portion was followed even though the temperature stabilized quickly. The addition of sodium sulfite also showed no exotherm, indicating the complete consumption of m-CPBA. Even the equivalence ratio of m-CPBA used was 1.4 equiv (75% w/w), which corresponds to 1.0 equiv stoichiometrically. The crucial operation in the scale-up was ammonia sparging, which can

Stoessel criticality class is 1 for the ammonia sparging operation, indicating that the reaction conditions are safe (Figure 5b). The temperature rise during neutralization is a function of the amount of ammonia that reacts, which is controllable by the rate of sparging. 2.3. Isolation of the Product and Regeneration of the Reagent Precursor. The aforesaid operation follows filtration of ammonium m-chlorobenzoate salt, leaving the N-oxide product in the filtrate. The N-oxide was isolated by distillation of the solvent under reduced pressure (20 mm Hg) at a temperature not exceeding 50 °C with 90% recovery of isopropyl acetate. The product purity was 98% by GC in comparison with the standard. To ascertain the safety during distillation of the solvent containing N-oxide, DSC of the isolated pure product as well as the product in iPrOAc at a concentration of 0.1 g/mL was studied. The onset points for the pure and solvent phases were 201.19 and 193.42 °C (Figure 6), which are very high limits compared with the operating temperatures during distillation. To recover m-chlorobenzoic acid from the salt, ammonium m-chlorobenzoate (452.4 g) was dissolved in 4.0 L of demineralized water followed by acidification with 580 mL of 20% HCl, filtration, and drying at 95 °C to constant weight to yield m-chlorobenzoic acid (398.0 g, 96% recovery, 99% D



Article

Table 3. N-Oxide Formation from Substituted Pyridines and N-Heterocyclic Substratesa

substituents entry

substrate

R1

R2

R3

N-oxide

yield (%)b

purity (%)c

MP/BPd (°C)

1 2 3 4 5 6 7 8 9 10 11 12 13 14

1a 2a 3a 4a 5a 6a 7a 8a 9a 10a 11a 12a 13a 14a

H 2-CH3 3-CH3 4-CH3 2-CH3 2-CH3 3-CH3 2-C6H5CH2 4-C6H5CH2 4-O2NC6H4CH2 2-pyridyl H 2-CH3 H

H H H H 3-CH3 6-CH3 5-CH3 H H H H H H H

− − − − − − − − − − − H H 8-OH

1b 2b 3b 4b 5b 6b 7b 8b 9b 10b 11b 12b 13b 14b

98 98.5 98.4 98.7 98.9 98.2 98.3 98.4 98.6 98.5 98.6 98.6 98.6 98.2

99.6 99.5 99.4 99.8 99.7 99.9 99.0 99.4 99.5 99.8 99.8 99.5 99.8 99.8

65−66 43−45 38−39 183−185 162−163d 108−110d 102−104 95−97 92−94 153−155 295−298 54−55 68−69 136−138

a

Experimental conditions: substrate (200.0 g, 2.53 to 1.86 mol, 1.0 equiv), 75% m-CPBA (610.0 to 450.0 g, 3.53 to 2.60 mol, 1.4 equiv), Na2SO3 (2g, 0.01 mol), and isopropyl acetate (10 vol). bIsolated yields. cDetermined by GC. dBoiling point.

Scheme 1. Process Flow for N-Oxidation of Pyridine

lead to a rapid exotherm (Figure 4) but can be controlled by the sparging rate. On the basis of RC1 data (Figure 4), a rapid exotherm was observed at a sparging rate of 180−200 mL/min. Hence, for the 2.5 kg input batch, initially sparging was begun at a low flow rate of 100−110 mL/min for 1 h. As no rapid exotherm was observed, the flow rate was increased to 150− 180 mL/min for a further 1 h and then to 180−200 mL/min where a similar exotherm pattern was observed. Filtration of the ammonium 3-chlorobenzoate salt proceeded with no hurdles such as caking or blocking of the filter pad. During solvent distillation, the thermic fluid bath temperature was maintained between 45 and 50 °C with a vacuum of 20 mm Hg, resulting in 90% solvent recovery and pyridine N-oxide (97% yield, 99.4% purity). 2.5. Scope of the Reaction. To investigate the scope and generality, the optimized protocol was extended to (sub-

stituted) pyridines and quinolines at the scale of 200.0 g (Table 3). The results revealed that the protocol was wellsuited for all of the substrates, which showed excellent yields and purities. The simplified schematic process flow is presented in Scheme 1. A DSC study of representative substituted pyridine N-oxides was also performed using pure samples as well as pure samples in iPrOAc in order to ensure the process safety for the scope of reaction. Table 4 shows the onset data for the pure products and corresponding heat releases. For the highly energetic substrate 4-(4-nitrobenzyl)pyridine N-oxide (10b), DSC in both the pure form and in iPrOAc was studied at higher concentration. In solution form, an endotherm was observed since the product was insoluble at a concentration of 0.5 g/mL (see the Supporting Information). The onsets for the pure sample and the sample in solvent were 173.38 and 176.11 °C, E



Article

mixture was cooled to 10−15 °C, followed by the addition of m-CPBA (75% w/w) (610 g, 3.53 mol, 1.40 equiv) in 10 equal portions over a period of 45−50 min. The resulting mixture was stirred for 1 h at 10−15 °C and further stirred at 30−35 °C for 4 h, and the reaction was monitored by GC. After completion of the reaction, solid sodium sulfite (2.0 g, 0.01 mol) was added to the reaction mixture, which was stirred for 15 min. The peroxide content was determined using an MNQuantofix peroxide strip. The reaction mixture was cooled to 10−15 °C, and anhydrous ammonia gas was sparged from the commercially available cylinder at a rate of 180−200 mL/min until the pH was raised to 8−9; excess ammonia escaping was trapped in 5% aqueous oxalic acid. The resultant slurry was filtered through a Celite bed and washed with isopropyl acetate (2 × 200 mL). The filtrate was transferred to the rotary evaporator, and the product was isolated by evaporation of solvent with 90% solvent recovery at 20 mm Hg and a temperature not exceeding 50 °C. This afforded the viscous oily product, which eventually crystallized on standing to furnish 235.0 g of pyridine N-oxide, 1b (98% yield, 99.6% purity). The purity of the product was determined by GC. Melting point: 65−66 °C. 1H NMR (CDCl3, 400 MHz): δ 8.13 (dd, J = 5.4, 2.8 Hz, 2H), 7.21 (dd, J = 8.8, 6.4 Hz, 3H). FTIR (neat): 3361, 3108, 1733, 1716, 1653, 1605, 1558, 1540, 1506, 1460, 1223, 1171, 1070, 1015, 914, 832, 765, 672 cm−1. HRMS elemental analysis calculated for C5H5NO (MH+), 95.0371; found, 95.0371. RC1 Reaction Procedure. The experiment was performed by isothermal calorimetry in the RC1e calorimeter. Initially, isopropyl acetate (350.0 mL, 17.5 vol) was added, followed by pyridine (20.0 g, 0.252 mol) under stirring at 300 rpm. The homogeneous solution was cooled to 10 °C, and m-CPBA (61.0 g, 0.353 mol, 1.40 equiv) was added in 10 equal portions. After each addition, the heat of reaction was recorded. Each portion of m-CPBA required approximately 2 min for addition, and we followed a trend of waiting for 5 min after the addition of each portion. The reaction mixture was maintained at 10 °C for 90 min. The temperature of the reaction mixture was raised to 30 °C and maintained for 1 h. The excess peroxides were quenched by addition of sodium sulfite (0.2 g, 0.001 mol), and the heat of peroxide quenching was measured and found to be negligible. The reaction mixture was cooled to 10 °C, and ammonia gas was sparged at a rate of 180−200 mL/min. During sparging at this rate, a rapid exotherm was recorded, and the sparging was stopped to stabilize the reaction mixture temperature at 10 °C. The sparging of ammonia gas was continued at a low flow rate of 100−110 mL/min until the pH of the reaction mixture increased to pH 8−9. Excess ammonia escaping was trapped in 5% aqueous oxalic acid. The slurry was filtered through a Celite bed and washed with isopropyl acetate (2 × 20 mL). The filtrate was transferred to the rotary evaporator, and the product was isolated by evaporation of the solvent. This afforded in the viscous oily product, which eventually crystallized on standing to furnish 23.3 g of pyridine N-oxide (97% yield, 99.4% purity by GC). The melting point was 65−66 °C. Kilogram-Scale Procedure. A 50 L glass reactor equipped with a Teflon blade stirrer in a circulation bath with an external chiller with a 14−16 L/min thermal fluid flow rate was charged with isopropyl acetate (43.7 L, 17.5 vol) followed by pyridine (2.5 kg, 31.64 mol) under stirring at 150 rpm. The reaction mixture was cooled to 10−15 °C. While the temperature was maintained at 10−15 °C, m-CPBA (75% w/w) (7.62 kg, 44.15

Table 4. DSC Data for Substituted N-Oxides entry

substrate

melting range (°C)

onset for the pure form (°C)

heat release (J/g)

1 2 3 4 5 6 7 8 9

1b 3b 4b 5b 9b 10b 12b 13b 14b

65−66 38−39 183−185 162−163a 92−94 153−155 54−55 68−69 136−138

201.19 152.01 202.16 140.34 130.12 173.38 208.72 125.64 159.24

1302.77 971.33 1270.88 1087.53 832.35 457.34 764.86 856.86 882.62

a

Boiling point.

respectively, from which it can be concluded that the isolation step during the solvent distillation is safe for this molecule.

■

EXPERIMENTAL SECTION General Information. All of the reactions were carried out under the fume hood. All of the chemicals and reagents were obtained from Loba Chemie with a minimum purity of 99% (m-CPBA 75% w/w) and used as received. Melting points were recorded on a Veego VMP-CM apparatus. Thin-layer chromatography was done on precoated silica gel plates (Kiesegel 60 F254; Macherey-Nagel). Gas chromatography was performed on an Agilent 7890B gas chromatograph with a BP5 column (30 m × 0.32 μm × 0.25 μm) at an injection temperature of 240 °C, a detector temperature of 250 °C, an initial temperature of 100 °C for a 2 min hold, a final temperature of 250 °C for a 2 min hold, a flow rate of 1.2 mL/ min, a split ratio of 83:1, and a run time of 19 min. The peroxide content was checked on MN-Quantofix peroxide strips (Macherey-Nagel). The flow rate for ammonia gas sparging was measured using an Agilent ADM flow meter (G6691-40500). FTIR spectra were recorded on a Thermo Scientific Nicolet IS 5 spectrometer. NMR spectra were recorded on an Agilent MR-400 spectrometer using CDCl3 or D2O as the solvent. Chemical shifts are reported in parts per million downfield from tetramethylsilane. High-resolution mass spectrometry (HRMS) data for the representative compounds were recorded using a Thermo Fisher Scientific Q-Exactive plus Biopharma high-resolution Orbitrap by the direct infusion method. The isothermal study was carried on a RC1e calorimeter with an AP01-0.5-RTC-3w reactor, an AP01-0.5 PTFE cover, and a 3-down pitched-blade stirrer. DSC measurements were performed with a Mettler Toledo DSC 1 module. About 4 mg of sample or 0.1 g/mL solution (unless otherwise stated) was placed in a 40 μL high-pressure gold-plated crucible and measured with dynamic heating run from 25 to 400 °C. The nth order kinetic software was used to calculate the activation energy and TMR. Cautionary note: Reactions and subsequent operations involving peracids should be performed behind a safety shield on a small scale. For relatively fast reactions, the rate of addition of oxidant should be controlled. The excess peroxide should be quenched with a peroxide scavenger before distillation. General Procedure for 200 g Scale. A 3 L four-neck round-bottom flask arranged with an overhead stirrer and outlet connected with a trap containing 5% aqueous oxalic acid was charged with pyridine (200.0 g, 2.52 mol) and isopropyl acetate (2.0 L, 10 vol) under stirring at 125 rpm. The reaction F



■

mol, 1.40 equiv) was added in 10 equal portions. Each portion of m-CPBA was added in 15 min, and the mixture was aged for a further 10 min before the next addition. After complete addition, the reaction mass was maintained for a further 1 h at 10−15 °C. The temperature of the reaction mixture was then raised to 30−35 °C and maintained for a further 4 h, and the reaction was monitored by GC. After completion of the reaction, solid sodium sulfite (25.0 g, 0.19 mol) was added to the reaction mixture, followed by stirring for 15 min. The peroxide content was determined using an MN-Quantofix peroxide strip. The peroxide test was negative. The reaction mixture was cooled to 10−15 °C, and anhydrous ammonia gas was sparged from the commercially available cylinder at a rate of 100−110 mL/min for 1 h. Then the rate was increased to 150−180 mL/min for a further 1 h and to 180−200 mL/min for 2 h while the temperature of the reaction mass was maintained between 10 and 15 °C in order to increase the pH of the reaction mass to pH 8−9. The excess ammonia was scrubbed using a common scrubber facility. The resultant slurry was filtered through a Celite bed on an SS 316L Nutsche filter, and the cake was washed with isopropyl acetate (2 × 2.0 L). The filtrate was transferred to another 50 L glass reactor, and the solvent was distilled by maintaining the thermal fluid bath temperature at 40−45 °C under vacuum (20 mm Hg). After complete distillation of the solvent, the residue was degassed for 15 min to ensure complete removal of solvent traces. The viscous oily product was unloaded hot into an HDPE container and eventually crystallized on standing to furnish 2.92 kg of pyridine N-oxide (97% yield, 99.4% purity). The GC graph of the kilogram-scale batch was concordant with that of standard pyridine N-oxide (see the Supporting Information). The melting point was 65−66 °C.

■

CONCLUSIONS

■

ASSOCIATED CONTENT

Article

AUTHOR INFORMATION

Corresponding Author

*Tel.: 91-22-33612212. Fax: 91-22-33611020. E-mail: gu. [email protected]. ORCID

Ganesh Chaturbhuj: 0000-0002-6263-8777 Notes

The authors declare no competing financial interest. † Provisional Indian patent no. TEMP/E-1/29261/2018MUM.

■

ACKNOWLEDGMENTS The authors thank Loba Chemie Pvt. Ltd. for supporting this work, Mr. Parag Chaudhary for preparing the DSC samples, Ms. Aprajita Singh for performing calorimetric experiments on the RC1, and Mr. Vijay Bhujle and Mr. Ganesh Kumar Janardhan (GVS Cibatech Pvt. Ltd.) for providing insight for interpretation of thermal analysis.

■

REFERENCES

(1) Wardman, P.; Priyadarsini, K. I.; Dennis, M. F.; Everett, S. A.; Naylor, M. A.; Patel, K. B.; Stratford, I. J.; Stratford, M. R.; Tracy, M. Chemical properties which control selectivity and efficacy of aromatic N-oxide bioreductive drugs. Br. J. Cancer. Suppl. 1996, 27, S70−S74. (2) Chandler, C. J.; Segel, I. H. Mechanism of the antimicrobial action of pyrithione: effects on membrane transport, ATP levels, and protein synthesis. Antimicrob. Agents Chemother. 1978, 14, 60−68. (3) Balzarini, J.; Stevens, M.; De Clercq, E.; Schols, D.; Pannecouque, C. Pyridine N-oxide derivatives: unusual anti-HIV compounds with multiple mechanisms of antiviral action. J. Antimicrob. Chemother. 2005, 55, 135−138. (4) Ghattass, K.; El-Sitt, S.; Zibara, K.; Rayes, S.; Haddadin, M. J.; El-Sabban, M.; Gali-Muhtasib, H. The quinoxaline di-N-oxide DCQ blocks breast cancer metastasis in vitro and in vivo by targeting the hypoxia inducible factor-1 pathway. Mol. Cancer 2014, 13, 12. (5) Sainz, Y.; Montoya, M. E.; Martínez-Crespo, F. J.; Ortega, M. A.; López de Ceráin, A.; Monge, A. New quinoxaline 1,4-di-N-oxides for treatment of tuberculosis. Arzneim. Forsch. 1999, 49, 55−59. (6) Vicente, E.; Villar, R.; Perez-Silanes, S.; Aldana, I.; Goldman, R. C.; Monge, A. Quinoxaline 1,4-di-N-oxide and the potential for treating tuberculosis. Infect. Disord.: Drug Targets 2011, 11, 196−204. (7) Maddileti, D.; Mannava, C.; Swapna, B.; et al. Blonanserin Noxide lowers glucose levels in animal models. J. Pharm. Drug Dev. 2017, 4, 102. (8) Ramchandra Reddy, P.; Himabindu, V.; Jaydeepkumar, L.; Madhusudhan Reddy, G.; Vijaya Kumar, J.; Mahesh Reddy, G. An improved process for the production of rabeprazole sodium substantially free from the impurities. Org. Process Res. Dev. 2009, 13, 896−899. (9) Ahn, K.-H.; Kim, H.-W.; Kim, J.-R.; Jeong, S.-C.; Kang, T.-S.; Shin, H.-T.; Lim, G.-J. A New Synthetic Process of Lansoprazole. Bull. Korean Chem. Soc. 2002, 23, 626−628. (10) Ochiai, E. Recent Japanese work on the chemistry of pyridine 1-oxide and related compounds. J. Org. Chem. 1953, 18, 534−551. (11) Nagano, H.; Nawata, Y.; Hamana, M. The Mechanism of the reaction of nicotinic acid 1-oxide with acetic anhydride. Chem. Pharm. Bull. 1987, 35, 4068−4077. (12) Grierson, D. The Polonovski reaction. Org. React. 1990, 39, 85−295. (13) Cave, A.; Kanfan, C.; Potier, P.; Le Men, J. Modification de la reaction de polonovsky. Tetrahedron 1967, 23, 4681−4689. (14) Mosher, H. S.; Turner, L.; Carlsmith, A. Pyridine N-oxide. Org. Synth. 1953, 33, 79. (15) Meisenheimer, J. Ü ber Pyridin-, Chinolin- und Isochinolin-Noxyd. Ber. Dtsch. Chem. Ges. B 1926, 59, 1848−1853.

An improved, safe, and scalable process for the preparation and isolation of N-oxides of (substituted) pyridines and quinolines in 98−99% yield with purity in the range of 98−99% using the m-CPBA−NH3(g) system was developed. The safety of the reaction during the conversion and isolation steps was substantiated by calorimetric data for the reaction and isolated product. The proposed protocol is not only economical but also eco-friendly because of efficient recovery and reusability of the solvent as well as the reagent precursor m-chlorobenzoic acid. The simplified process flow, minimized utility use and waste, and high-purity products are the key features of this protocol for scale-up, which meets the industrial requirements of safety and cost-effectiveness. The improved heat capacity due to the solvent:reactant ratio of 17.5, the controlled addition of m-CPBA and ammonia, the scavenging of excess mCPBA, and cautious solvent recovery allow the protocol to satisfy most of the safety criteria. Hence, it was feasible for us to demonstrate the safe operation at a scale as high as a 2.5 kg batch.

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.oprd.8b00358. Spectral data and calorimetric analysis (PDF) G



Article

(16) Bobranski, B.; Kochanska, L.; Kowalewska, A. Ü ber die Einwirkung von Sulfurylchlorid auf Pyridinoxyd. Ber. Dtsch. Chem. Ges. B 1938, 71, 2385. (17) Epsztajn, J.; Bieniek, A.; Kowalska, J. A. Application of organolithium and related reagents in synthesis. Part 9. Synthesis and metallation of 4-chloropicolin- and 2-chloroisonicotinanilides. A useful method for preparation of 2,3,4-trisubstituted pyridines. Tetrahedron 1991, 47, 1697−1706. (18) Thellend, A.; Battioni, P.; Sanderson, W.; Mansuy, D. Oxidation of N-heterocycles by H2O2 catalyzed by a Mn-Porphyrin: An easy access to N-oxides under mild conditions. Synthesis 1997, 1997, 1387−1388. (19) Copéret, C.; Adolfsson, H.; Khuong, T.-A. V.; Yudin, A. K.; Sharpless, K. B. A simple and efficient method for the preparation of pyridine N-oxides. J. Org. Chem. 1998, 63, 1740−1741. (20) Rudolph, J.; Reddy, K. L.; Chiang, J. P.; Sharpless, K. B. Highly efficient epoxidation of olefins using aqueous H2O2 and catalytic methyltrioxorhenium/pyridine: pyridine-mediated ligand acceleration. J. Am. Chem. Soc. 1997, 119, 6189−6190. (21) Copéret, C.; Adolfsson, H.; Chiang, J. P.; Yudin, A. K.; Sharpless, K. B. A simple and efficient method for the preparation of pyridine N-oxides II. Tetrahedron Lett. 1998, 39, 761−764. (22) Brown, D. B.; Ironside, M. D.; Shaw, S. M. Safety Notables: Information from the literature. Org. Process Res. Dev. 2016, 20, 575− 582. (23) Islam, A. N.; Tofik, N. I.; Tofik, B. B.; Mamedali, M. A. Oxidation of pyridine bases by hydrogen peroxide. J. Chem. Chem. Eng. 2011, 5, 82−88. (24) Kubota, A.; Takeuchi, H. An Unexpected Incident with mCPBA. Org. Process Res. Dev. 2004, 8, 1076−1078. (25) Caron, S.; Dugger, R. W.; Ruggeri, S. G.; Ragan, J. A.; Ripin, D. H. Large-Scale Oxidations in the Pharmaceutical Industry. Chem. Rev. 2006, 106, 2943−2989. (26) Zhang, X.; Hu, A.; Pan, C.; Zhao, Q.; Wang, X.; Lu, J. Safer preparation of m-CPBA/DMF Solution in Pilot Plant. Org. Process Res. Dev. 2013, 17, 1591−1596. (27) Stoessel, F. Thermal Safety of Chemical Processes: Risk Assessment and Process Design; Wiley: New York, 2008. (28) McDonald, N.; Steppel, R. N.; Dorsey, J. E. m-Chloroperbenzoic acid. Org. Synth. 1970, 50, 15.

H


The m-CPBAâNH3(g) System: A Safe and Scalable Alternative for the

Recommend Documents

The m-CPBAâNH3(g) System: A Safe and Scalable Alternative for the