Mechanism of Acylative Oxidation–Reduction–Condensation

Feb 27, 2017 - We previously described a new organocatalytic oxidation–reduction–condensation for amide/peptide construction. The reaction system ...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/joc

Mechanism of Acylative Oxidation−Reduction−Condensation Reactions Using Benzoisothiazolones as Oxidant and Triethylphosphite as Stoichiometric Reductant Pavankumar Gangireddy, Vidyavathi Patro,† Leighann Lam,‡ Mariko Morimoto,§ and Lanny S. Liebeskind* Emory University, Department of Chemistry, 1515 Dickey Drive, Atlanta, Georgia 30322, United States S Supporting Information *

ABSTRACT: We previously described a new organocatalytic oxidation−reduction−condensation for amide/ peptide construction. The reaction system relies on triethylphosphite as the stoichiometric reductant and organocatalytic benzoisothiazolone/O2 in air as the oxidant. The reaction was assumed to generate catalytic quantities of S-acylthiosalicylamides as electrophiles, which are rapidly intercepted by amine reactants to generate amides/peptides and o-mercaptobenzamides. The latter are then gently reoxidized to the benzoisothiazolones under Cu-catalyzed aerobic conditions to complete the catalytic cycle. To gain a mechanistic understanding, we describe herein our studies of the stoichiometric generation of S-acylthiosalicylamides under oxidation−reduction− condensation conditions from a variety of benzoisothiazolones and carboxylic acids using triethylphosphite as the terminal reductant. These studies have revealed the presence of more than one reaction pathway when benzoisothiazolones react with triethylphosphite (including a rapid, direct deoxygenation of certain classes of benzoisothiazolones by triethylphosphite) and allow the identification of optimal reaction characteristics (benzoisothiazolone structure and solvent) for the generation of thioesters. These explorations will inform our efforts to develop highly effective and robust organocatalytic oxidation−reduction− condensation reactions that are based on the benzoisothiazolone and related motifs.



INTRODUCTION Oxidation−reduction−condensation is a powerful protocol for the dehydrative construction of C−C and Cheteroatom bonds from carboxylic acids (acylative) or from alcohols (alkylative) without recourse to strongly acidic or basic reagents.1−5 Its hallmark characteristic is pH neutral dehydration through the use of a gentle stoichiometric PIII reducing agent to accept an “O” atom coupled with a mild organic oxidizing agent to accept 2H to complete an overall redox dehydration. Throughout its long history as a synthetic method, a variety of oxidants have been explored (inter alia: azodiesters, quinones, disulfides, sulfenamides), and while the reductant has been largely restricted to triorganophosphines, occasionally phosphinites5 and triarylphosphites6 have been used. Recent efforts to render the reactions catalytic in the traditional hydrogen acceptors (azo compounds) have been described in the peer reviewed literature.7,8 The use of stoichiometric PIII reductants remains problematic. It suffers from a tedious chromatographic separation of the phosphine oxides, the oxidation byproducts, from the desired reaction products. This practical difficulty has led to the development of triorganophosphines covalently embedded within solid supports or to triorganophosphine variants adorned with water-solubilizing and other phase-separating functional groups.4 A different option would be to replace the higher molecular weight triorganophosphine and triarylphosphite reductants with inexpensive low molecular weight © XXXX American Chemical Society

trialkylphosphites such as (EtO)3P. Then, the oxidation− reduction−condensation reactions would produce water-soluble (EtO)3PO, which would be easier to separate from less polar reaction products by an aqueous wash, or by trituration when the desired products are solids. However, with one exception which focused on (i-PrO)3P as the reductant,9 concern over undesired Arbuzov-like side reactions at the alkoxy groups of the trialkylphosphite has contributed to the absence of practical uses of simple organophosphites in oxidation−reduction−condensation reactions. Our laboratory is exploring new protocols to render oxidation−reduction−condensation reactions catalytic in the organooxidant and practical and economical in both the terminal oxidant and the terminal reductant. Initial studies have focused on acylative oxidation−reduction−condensation reactions of carboxylic acids and amines in which organocatalytic benzoisothiazolones, 1, are linked (using a Cu catalyst) to stoichiometric O2 in dry air as the oxidant (Scheme 1).10 In this acylative reaction system, triethylphosphite functions as an effective stoichiometric reductant. Following precedent where S-acyl thiosalicylamides are directly generated from a carboxylic acid, a benzoisothiazolone and a stoichiometric triorganophosphine,11,12 the triethylphosphite-based catalytic Received: January 4, 2017 Published: February 27, 2017 A

DOI: 10.1021/acs.joc.7b00020 J. Org. Chem. XXXX, XXX, XXX−XXX

Article

The Journal of Organic Chemistry

Table 1. Formation of Thioesters from BITs and Toluic Acida

Scheme 1. Aerobic, BIT-Catalyzed Amidation

BIT Series

reaction was assumed to proceed via an oxidation−reduction− condensation that generated catalytic quantities of electrophilic S-acylthiosalicylamides, 2, from the benzoisothiazolone, 1, the carboxylic acid, and, in the present case, the triethylphosphite. The in situ generated activated thioesters would then be intercepted by amine reactants to construct amides/peptides with expulsion of o-mercaptobenzamides, 3. The latter can be gently reoxidized to the benzoisothiazolone under Cu-catalyzed aerobic conditions to complete the catalytic cycle.13 Herein we describe our studies to probe details of the mechanism of the benzoisothiazolone catalyzed reaction beginning with the stoichiometric generation of S-acylthiosalicylamides from a series of benzoisothiazolones (Chart 1) and

%2

%(p-tolCO)2Ob

1 2 3 4 5 6 7

a b c dc ed fe g

N-Alkyl BITs 84 79 73 52 0 66 79 N-Aryl BITs

0 0 7 5 34 12 0

8 9 10 11 12 13 14

h if j kg l m n

0 59 0 20 0

complex mixture 9 10 24 0 22 no reaction

a

Conditions: BiT (1.0 mmol), p-toluic acid (1 mmol), (EtO)3P (1.2 mmol), anhydrous DMF (0.18 M) at 25 °C. bAnhydride maximum yield = 50%. cThionophosphate 5d forms in 14% yield. d Thioester 2e was sensitive to hydrolysis. Therefore, product yields shown were determined by 1H NMR analysis using 1,3,5-trimethoxybenzene as internal standard in DMF-d7. Thione 4e (separately purified and characterized) was also observed in 46% yield. eThioester 2f was very sensitive to hydrolysis on TLC and upon workup. Product yields shown were determined by 1H NMR analysis using 1,3,5-trimethoxybenzene as internal standard in DMF-d7. The BIT 1f disappeared over 12 h. fThe fate of the BIT 1i was not determinable. g 40% of a deoxygenation-dimer product was formed (see text for discussion).

Chart 1. Benzoisothiazolones Studied

and quantified (Table 1). Traces of 2-mercaptoethylbenzamides were also observed in some reactions. As seen in entry 1 the S-toluoylthiosalicylamide was generated in very good yield from 4-nitro-N-isopropylbenzoisothiazolone, 1a, and toluic acid in DMF. Thus, triethylphosphite is a practical replacement for the triorganophosphine reductants that were previously shown to react with 5-nitro-N-isopropylbenzoisothiazolone,1a, and a carboxylic acid via an oxidation−reduction−condensation reaction to provide thioesters.11,12 However, the remaining entries in Table 1 reveal dramatic trends in the yield of thioester obtained as the nature of the benzoisothiazolone was varied. All the N-alkyl BITs except 1e (bearing a 3-pyridyl core ring) reacted to form the corresponding S-toluoyl thioesters in good to excellent yields within 120 min. The 3-pyridyl BIT 1e is an outlier and did not produce the S-toluoylthiosalicylamide, but instead generated toluic anhydride and thione 4e. A reaction of 1e in DMF-d7 monitored in situ by 1H NMR showed the disappearance of 1e over 3 h concurrent with the appearance of p-toluic anhydride in 34% yield (50% max) and thione 4e in 46% yield. No intermediate generation of the thioester was detected. In comparison, a directly monitored reaction of the isomeric 5-pyridyl BIT 1f in DMF-d7 using 1,3,5-trimethoxybenzene as an internal standard showed the formation of thioester 2f and p-toluic anhydride in 66% and 12% yields,

carboxylic acids under oxidation−reduction−condensation conditions using triethylphosphite as the stoichiometric reductant. These studies have revealed the presence of more than one reaction pathway when a benzoisothiazolone reacts with triethylphosphite and have allowed the identification of optimal reaction characteristics (benzoisothiazolone structure and solvent) for the generation of thioesters. These studies will be important in guiding the choice of benzoisothiazolones as organocatalytic oxidants for the design of improved and practical oxidation−reduction−condensation reaction systems where triorganophosphites are used as reducing agents.



RESULTS AND DISCUSSION The study commenced with an exploration of the reaction of the N-alkyl and N-aryl benzoisothiazolones shown in Chart 1, prepared according to known procedures,12−15 with triethylphosphite and p-toluic acid in DMF. The BITs 1a−n were treated with 1.0 equiv of p-toluic acid and 1.2 equiv of P(OEt)3 in DMF (0.18M) at 25 °C. Reactions were monitored by TLC for the formation of the corresponding S-toluoyl 2-mercaptosalicylamides, 2a−n. These products as well as the side product p-toluic acid anhydride were isolated B

DOI: 10.1021/acs.joc.7b00020 J. Org. Chem. XXXX, XXX, XXX−XXX

Article

The Journal of Organic Chemistry respectively. However, the S-toluoyl thioester 2f derived from 1f was not sufficiently stable to survive workup and isolation. In contrast to the slower reactions of the N-alkyl BITs, all N-aryl BITs rapidly disappeared within minutes in the presence of triethylphosphite at room temperature in DMF. However, with two exceptions, the N-aryl BITs proved to be uniformly poor substrates for generating stable thioesters in DMF as solvent. The exceptions are the two BITs 1j and 1l bearing N-2-pyridyl substituents: 1j gave a stable S-toluoyl thioester in 59% yield, while 1l produced its S-toluoyl thioester, but in only 20% yield. As is described below, the inability of the N-aryl BITS to systematically provide the corresponding thioesters under oxidation−reduction−condensation conditions is caused by an unusual BIT deoxygenation pathway that is favored when the BIT N-substituent is an aromatic group. To gain insight into the differential reactivity of variously substituted benzoisothiazolones with a carboxylic acid in the presence of triethylphosphite, a 31P NMR study of the interaction of benzoisothiazolones with triethylphosphite in the absence of the carboxylic acid was undertaken in three different solvents (DMF, EtOAc, and toluene) (Table 2). In an NMR

to (EtO)3P with facility, generating (EtO)3PO rapidly in all three solvents, with the fastest reaction taking place in DMF. In contrast, all the N-aryl substituted BITs, except for 1n, were uniformly very reactive to deoxygenation by triethylphosphite in DMF. Deoxygenation of the BIT was complete in DMF within 2 h. Slower deoxygenation reactions of the same set of N-aryl BITs were observed in EtOAc, while very slow reactions were observed in toluene. The BIT 1n is an outlier among the N-aryl substrates as explained below. Knowing that triethylphosphite can rapidly extract an oxygen atom from an N-aryl benzoisothiazolone in polar solvents, preparative scale deoxygenation reactions were carried out to determine the fate of the deoxygenated benzoisothiazolone. Initial evidence for the fate of the benzoisothiazolone was obtained from the attempt of entry 11 in Table 1 to convert BIT 1k into a p-toluoyl thioester using p-toluic acid and triethylphosphite. As reported, only a low yield (24%) of p-toluic anhydride was generated in this reaction. The major product isolated in 80% yield was, 6, derived from deoxygenation of the benzoisothiazolone followed by dimerization (Scheme 2). In a separately run reaction the deoxygenation

Table 2. 31P NMR Studies of the Direct Deoxygenation of Benzoisothiazolones by (EtO)3P

Scheme 2. Deoxygenation−Dimerization of Benzoisothiazolone 1k

% OP(OEt)3b a

entry

BIT series

N-Alkyl BITs 1 a 2 b 3 c 4 d 5 e 6 f 7 g N-Aryl BITs 8 h 9 i 10 j 11 k 12 l 13 m 14 n

reaction time

DMF

EtOAc

toluene

h h h h h h h

0,