A recent report (1) from workers at Chonnam National University (Gwangju, Korea)  describes a benzimidazole synthesis which:

•  produces good product yields (40-98%, for about 30 examples)
• and proceeds in one pot from three readily available components: sodium azide, an aldehyde, and 2-haloanilines 
• shows good functional group tolerance(nitro-, ester-, chloro-, and various heterocyclic functionalities on the aldehyde or haloaniline component).

The Benzimidazole Synthesis of Lee and coworkers (1)

 Naturally, there are many established ways to synthesize benzimidazoles, which are important substances used in the design of bioactive substances (2).  Recent work has sought to address specific drawbacks associated with these methods, which can include harsh reaction conditions and complicated product mixtures.

Further developments have focused on the use of 2-haloacetanilides, 2-haloarylamidines, arylamino oximes, and N-arylbenzimidamides (3).  This work notable due to the useful anthelmintic properties. Anthelmintic agents work to kill or repel intestinal worms. A review (3) discusses the synthesis of benzimidazoles, and cites the breakthrough discovery of thiabendazole by researchers at Merck in 1961.  Thiabendazole was found to have potent broad spectrum activity against gastrointestinal parasites. 

Early thiabendazole synthesis (3)

 The initial synthesis of thiabendazole occured via dehydrative cyclization of 1,2 diaminobenzenze in polyphosphoric acid (PPA). The commercialized process involved the conversion of N-arylamidines using hypochlorite (4). Although this process can be performed in ‘one-pot’ fashion it is more typically performed in two steps.

The ‘one-pot’ benzimidazole synthesis described by Lee et. Al. is showcased by its ability to produce thiabendazole in one step, from readily available starting materials (2-haloanilines, thiazole-4-carboxaldehyde) – in 97% yield.

Their work builds on the report of Driver and coworkers (5) that showed that benzimidazoles could be had from 2-azidoanilines in good yield. Indeed, Lee proposes a mechanism that produces an azidoaldimine intermediate, which foregoes the multistep preparation of 2-azidoaniline starting materials.

One proposed mechanistic pathway is shown, with the following steps:

• initial in situ formation of an aldimine, via addition of aniline to an aldehyde;
• Ar-X insertion of the copper catalyst;
• Cu-azide association, with transfer of azide to the aromatic ring;
• loss of nitrogen with concomitant ring formation, and catalyst regeneration
  

  One mechanistic explanation proposed by Lee and coworkers (1).

In developing their method, they investigated a number of factors:

• Solvent.  DMSO outperformed other polar solvents (NMP, DMF, DMAc).  Less polar solvents failed (toluene, diglyme).
• Source of Copper catalyst. The oxidation state of copper was not a factor, as Cu(I) and Cu(II) salts showed similar performance.
• Ligand Evaluation. Ligand selection was not a large factor. Several were tested; ultimately TMEDA was selected.
• Substituents on the aniline / pyridyl component. Base sensitive substituents were tolerated (benzoate ester) and 3-Cl groups were fine. The sensitivity to a broad range of substituents (the usual EWD- and ED-groups) was not rigorously determined
• Nature of the haloaniline. Although both bromo- and iodoaniline examples were given, the predominance of iodoaniline examples suggests it was prefered by the authors for unstated reasons.
• Reactivity of various aldehyde reactants. Aldehydes of varying classes were evaluated. Yields from aromatic substrates bearing ED groups(benzaldehyde, 4-Cl benzaldehyde, 4-methoxybenzaldehyde) produced the highest product yields.  Aliphatic aldehydes produced noticeably lower yields, with the curious exception of pivaldehyde. Several heterocyclic aldehydes (2- furyl- and 2-thionylaldehyde were tested and provided good results. 

A synopsis of the Lee Procedure follows:

Artie McKim.

(1) Kim, Y.; Kumar, M.R.; Park, N.; Heo, Y.; Lee, S. J. Org. Chem. 2011, 76, 9577-9583.
(2) Tumulty, D.; Cao, K.; Homes, C.P. Org Lett. 2001, 3, 83.; Wu, Z. Rea. P.; Wickham, G.; Tetrahedron Lett. 2000, 41, 9871.;  Chari, M.A.; Shobha, P.S.D.;  Mukkanti, K. J. Heterocycl. Chem.  2010, 47, 153.
(3) Townsend, L.B.; Wise, D.S. Parasitology Today 6, 4 (1990) 107-112.
(4) Grenda, V. J.; Jones, R.E; Gal,G.; Sletzinger J. Org Chem. 30 (1965), 259-261.
(5) Shen, M.; Driver, T.G. Org Lett. 2008, 10, 3367.

 The work of Lee, Song and coworkers (1) provides an versatile route to symmetrical and unsymmetrical diarylalkynes. Their method is complementary to the powerful Sonagashira coupling of aryl halides and alkynes, and offers an alternative to the established ways to make diarylalkynes. 

The diarylalkyne synthesis of Park, Song and coworkers (1).

Conjugated alkynes have potential in nanoelectric materials and conductive organic films. The authors describe the disadvantages of some of the established ways to make this compound class:  

‘Protected’ acetylenes are easier to work with, but also have issues:

• Stille-type coupling of distannylalkynes: stoiciometric amounts of hazardous tin waste are generated
• bis-(trimethylsilyl)acetylenes require an excess of strong base.  

The authors have developed the use of propiolic acid (propynoic acid) as the alkyne

Another decarboxylative coupling: Becht et Al. (2)

component in the palladium-catalyzed coupling of various aryl halides.  Decarboxylativecoupling reactions are a relatively newdevelopment, with the advantage that innocuous CO2 is a byproduct. The Becht group (2) is active in this area and has published a decarboxylative coupling route to sterically hindered biaryl compounds.

Previous work in this area by the authors had produced a decarboxylative coupling reaction based on aryl- or alkyl alkynyl carboxylic acids (3). Advances toward an ecofriendly method were made, but they cite some limitations of this earlier work, including:

• excess of TBAF required (expensive, and TBAF is moisture sensitive)
• relatively high Pd loading (costly)
• ligand stability (P-tBu3) problematic.

Screening experiments using a model system (bromobenzene + propiolic acid) led them to these conditions:

• Pd source: Pd(PPh3)Cl2 was chosen; other sources evaluated were Pd(PPh3)2(OAc)2, Pd2(dba)3, Pd(OAc)2 and Pd(PPh3)4
• Base: DBU gave a quantitative yield with Pd(PPh3)Cl2.  Twelve other bases were tried, including the somewhat exotic 1,8-diaminonapthalene.
• Solvent: DMSO produced the highest product yields of solvents tested. Other solvents evaluated included toluene, diglyme, xylene, and NMP. Interetingly, although NMP has solvent properties which are similar to those of DMSO, its use only resulted in a 23% yield.
• Ligand: dppb outperformed PPh3, P t-Bu3, PCy3, dppf, and Xantphos.

Ultimately, their general procedure might paraphrased in this way:

Pd(PPh3)2Cl2 (105 mg, 0.15 mmol), dppb (128 mg, 0.30 mmol), aryl halides (6 mmol) and propiolic acid (212 mg, 3 mmol) were mixed with DBU (913 mg, 6 mmol) in a small RBF. DMSO (15 mL) was added and the the sealed flask was heated at 80C for 3 hr.

The reaction mixture was poured onto saturated NH4Cl (aq) and extracted with ether. Ether extracts were washed with brine, dried over magnesium sulfate, filtered, and concentrated in vacuo to provide a residue for purification by flash chromatography (ethyl acetate / hexane).                                                     

Lee, Song and coworkers also described the use of 2-butyndioic acid as a substitute for propiolic acid, and yields of symmetrical diarylalkynes are in most cases excellent.

Yields of unsymmetrial diarylalkynes from propiolic acid suffer somewhat, but yields of 65-89% were reported. An interesting feature of this reaction is that it apparently works best when a) one aryl halide is a bromoarene and the other an iodoarene and b) both arenes are added in the beginning of the reaction.

The authors also describe mono- vs. di-coupling via temperature control, and state that this work represents the first successful report of double decarboxylative coupling using 2-butyndioic acid.

Artie McKim.

(1) Park, K.; Bae, G.; Moon, J.;Choe, J.; Song, K.H.; Lee, S. J. Org. Chem.  2010, 75, 6244-6251.

(2) Becht, J.-M.; Catala, C.; Drian Claude, L.; Wagner, A. Org Lett 2007, 9, 1781-3.

(3) Moon, J.; Jang, M.; Lee, S. J. Org. Chem. 2009, 74, 1404-1406  b) Kim, H. ; Lee, P.H. Adv. Synth. Catal. 2009, 351, 2827-2832.

The sulfide oxidation procedure of Page and coworkers (1)

Our first trial was based on the work of  Page and coworkers at the University of Liverpool (1). 

 A first look at this procedure suggested it would be simple to implement in the lab and could be used to produce preparative quantities of the sulfoxide we required. Features that made this reaction attractive include:

• relatively brief reaction times (under 2 hours)
• good to excellent product yields (69-91%)
• No exotic reagents are required.

The mechanism of peroxide activation involves a peroxyimidic acid formed in situ by

peroxyimidic acid intermediate formation.

reaction between acetonitrile and hydrogen peroxide. The authors mention the need to minimize the concentration of this intermediate as it can decompose to acetamide if allowed to react again with peroxide (2). This is achieved by using a large excess of acetonitrile and through the slow introduction of peroxide. As previously mentioned, this is essentially a modification of the Payne epoxidation of olefins; Page and coworkers have extended this to sulfoxide synthesis.

Based on their experimental procedure and some preliminary tests, this is the procedure we put together for our purposes: 

 The progress of the reaction was followed by FID-GC. When the mixture was worked up we found the following:

•  all of the sulfide starting material had been consumed.
• Although the reaction is selective to sulfoxide formation, a considerable amount of sulfone byproduct was formed also.

In our hands, a contained yield of about 70% was the best we could do. This was after several small-scale trials wherein we varied the following:

• reaction time: we saw that little reaction had occurred after two hours, and prolonged the reaction time realtive to the generic procedure given in the paper.
• peroxide addition time and dilution: clearly overoxidation can be minimized by controlling this variable, and we worked to add the peroxide as slowly as possible. The Page paper does not prescribe a dilution volume for the aqueous peroxide (‘added dropwise as a dilute methanol solution’) and this could be a worthwhile factor to optimize.

Our conclusion: this is a serviceable reaction which is easy to perform. We suspected that a more selective method was available, however, and did not scale this reaction to produce a larger batch for purification. If we wanted to improve the selectivity of the reaction, some things we would look into would include:

• The authors stated that a large excess of acetonitrile can minimize overoxidation. The generic experimental procedure in the Page paper recommended a slight excess (1:1.5 molar ratio relative to starting sulfide). Would a larger amount of acetonitrile make a difference?
• It appears that the paper we followed used a substoichiometric amount of potassium carbonate. We thought a larger amount of base might make a difference but didn’t examine this.
• Further consideration of temperature / time conditions might improve product yields.

Artie McKim.

(1) Page, P.C.B.; Graham, A.E.; Bethell, D.; Park, B.K. Synthetic Communications 23, 11  (1993) 1507-1514.   

 (2) The Radziszewski Reaction was a new one to me. It is presumably a useful way to convert nitriles to amides, when other functional groups present are resistant to oxidation conditions. Page and coworkers provide this reference: Ogata, Y. ; Sawaki, Y. Tetrahedron 20 (1964), 2065.