Copper(I)–Phosphinite Complexes in Click Cycloadditions: Three‐Component Reactions and Preparation of 5‐Iodotriazoles

Abstract The remarkable activity displayed by copper(I)–phosphinite complexes of general formula [CuBr(L)] in two challenging cycloadditions is reported: a) the one‐pot azidonation/cycloaddition of boronic acids, NaN3, and terminal alkynes; b) the cycloaddition of azides and iodoalkynes. These air‐stable catalysts led to very good results in both cases and the expected triazoles could be isolated in pure form under ‘Click‐suitable’ conditions.


Introduction
Huisgen1 ,3-dipolarc ycloadditions represent ap owerful methodology for the preparation of aw ide range of five-membered-ring heterocycles. [1,2] These reactions have recently made as trong comeback owing to the huge interest attractedb y the copper(I)-catalyzed [3+ +2] cycloaddition of organic azides and alkynes, unarguably the best Click reaction to date. This transformation leads to the extremely efficient formationo f 1,4-disubstituted-[1,2,3]-triazolesa sasole regioisomer. [3] Barely ten years after the originalg roundbreaking reports by Sharpless and Meldal, [4] am yriado fl igandless as well as ligated coppers ystemsh ave been reported in the literature and have found applications in diverse fields. [5] Unsurprisingly,p hosphorous ligands, andP Ph 3 in particular, were among the first ligandsa ppliedt ot he cycloadditiono f azides and alkynes. [6,7] We recentlyr eported the excellent activity in copper-catalyzed azide-alkyne cycloaddition( CuAAC) reactions of novel copper(I) complexes bearing phosphinite or phosphonite ligands. [8] These complexes of general formula [CuBr(L)]a re stable andc an be handled with no particular precautionst oe xclude moisture or air.F urthermore, we showed that they outperformed related complexes with one phosphine or phosphitel igand in the synthesis of triazoles. This prompted us to further explore the potentialo ft his family of complexes in related copper-mediated cycloadditionr eactions. Herein, we report the applicationo fs uch complexes to two important re-actions:t he three-component preparation of 1,4-disubstituted triazoles from boronica cids, NaN 3 ,a nd terminal alkynes, as well as the cycloaddition of azides and iodoalkynes (Scheme 1).
Organic azides are generally stable towardsw ater and oxygen, and safe to use, [9] except those of low molecular weight, [10] which historically caused at angible azidophobiai n the chemical community. [4b] As ar esult, several approaches have been explored to avoid the handling and isolation of organic azides in [3+ +2] cycloadditions. These reactions are straightforward for aliphatic azides as these can be easily accessedf rom the corresponding halideso ra mines through as imple nucleophilics ubstitution with NaN 3 , [11] or TfN 3 , [12] respectively.H owever,r eactions involving aryl azides are significantly morel imited. Anilines can be reactedw ith tert-butyl nitrite and TMSN 3 to generate in situ the corresponding aryl azides, however,t he scope of this reaction is limited. [13] Alternatively, ap roline/copper(I) system has been reportedt om ediate the one-pot azidonation/cycloaddition reactions from aryl halides(iodide or bromide). [14,15] We turned our attention to the in situ preparation of aryl azides from the corresponding boronic acids because of their low toxicitya nd wide availability,s upported by the popularity of Suzuki-Miyaura cross-coupling reactions. To the best of our knowledge,n oh omogeneous, ligated copper(I) system has been reported in this context. To date, reports on this threecomponent copper-mediated transformationh ave focusedo n The remarkable activity displayed by copper(I)-phosphinite complexes of generalf ormula [CuBr(L)] in two challenging cycloadditionsi sr eported: a) the one-pota zidonation/cycloaddition of boronic acids, NaN 3 ,a nd terminal alkynes; b) the cy-cloaddition of azides and iodoalkynes.These air-stable catalysts led to very good resultsi nb oth cases and the expected triazoles could be isolated in pure form under 'Click-suitable' conditions. heterogeneous systems requiring high metal loadings (10 mol %[Cu]) and, often, alargeexcess of NaN 3 . [16,17] On the other hand, the preparation of 5-iodotriazoles from the corresponding organic azides andi odoalkynes remains achallenging transformation as only ahandfulofefficient catalysts has been described in the literature so far. [18,19] Halogenated heterocyclesa re particularly interesting from as ynthetic point of view,a nd iodotriazoles have indeed been used in several palladium-mediated cross-coupling reactions. [20] Despite the clear similarities, important differences in the reactionso f terminal and iodoalkynes have been reported. Whereas the ligandlessc ombination CuSO 4 /Na ascorbate is very popular for the cycloadditiono ft erminal alkynes, the presence of coordinating ligandsi sc rucial for any iodotriazole to be formed. In addition to ancillary ligands, most reported systemsa lso require an N-additive such as triethylamine or lutidine. These facts are alignedw ith the increasing evidencef or different mechanistic pathways in both reactions. Fokin first suggested that cleavage of the carbon-iodine bond is not required for the cycloaddition to take place. [21] Our DFT studies supported this and showedt hat either formationo facopper(III) metallacycle or direct activation of the iodoalkyne by p-coordination of the copperc atalyst accounted for the copper-acceleration effect and regioselectivityo ft his cycloadditionr eaction. [18e] Results and Discussion The three-componentr eaction In af irst step, we explored the three-component reaction between boronic acids, NaN 3 ,a nd terminala lkynes for the onepot preparation of 1,4-disubstituted triazoles. Both steps of this transformation (azidonation and cycloaddition reactions) are copper(I)-mediated, and therefore we chose {CuBr[PPh 2 -(OPh-2-OMe)]} A for the initial screening, as it had previously shown the highest activity in CuAAC reactions. [8] With paramethoxyphenylboronic acid and phenylacetylene as model startingm aterials, different solvents weres creened (Table 1). Reactions in acetone or acetonitrile only led to mixtures of various unknown byproducts. Reactions were cleaner in toluene but only 28 %o ft he expected triazole was observed as the startingb oronic acid and its corresponding boroxine derivative were the major products in the crude mixture. Satisfyingly, very cleanr eactions were obtained in am ixture of water/ MeOH (1:1). When water or MeOH wereu sed separately the overall conversion drastically droppedo wing to the presence of intermediate azide and boroxine.I ti si mportantt on ote that if all materials,p henylacetylene included,w ere added simultaneously,o nly 15 %c onversion to the expected triazole was observed under otherwise identicalreactionconditions.
The trimerization of aryl boronic acids to their anhydride trimers (boroxines) is well established in the literature (Scheme 2), [22] and it is broadly acknowledged that commercial an analytically pure form with no need for further purification. Different aryl boronic acids could be efficiently used in this preparation of triazoles 3,i ncluding heteroaromatic ones. Similarly,b oth functionalized alkynes as well as simple aliphatic ones were successfully used under these reaction conditions.

Cycloaddition of azidesa nd iodoalkynes
We started screeningd ifferent phosphinite and phosphonite complexes with benzyl azide and iodoethynylbenzene as the model substrates. Reactions were carriedo ut neat at room temperature (Scheme 5).
No reactivity trends could be drawn from the obtained results. Average conversionsw ere observed with complexes B and C,w hereas complexes with no functionalized ligandsg ave even poorer results. No reactionw as observed in different solvents including MeCN, EtOH, toluene, and water,a nd only 12 %c onversion was found in THF.
Severalr eports in the literaturef eature the importance of nitrogen additives in this reaction, including our own work with the related [CuI(PPh 3 ) 3 ]c omplex. [18e] No improvement, however,w as observed with 5mol %o f N,N-diisopropylethylamine, 2,6-lutidine, or triethylamine. In the presence of 1,10-phenanthroline,t he reactionconversion dropped to 25 %. Nevertheless, during the catalyst screening we noticedt hat complexes B and C remained active overs everal days [27] and by simply raising the reactiont emperature to 40 8C, full conversion into triazole 6a was obtained overnight with C.Bycontrast, only 20 %conversion into 6a was observed with B,w hich indicatest hat this complex is significantly more heat sensitive.
The optimized reactionc onditions were next applied to different pairs of azides and iodoalkynes (Scheme 6). In all cases, either high or complete conversionsi nto 5-iodotriazoles were obtained. Interestingly,w hereas dehalogenated triazoles were sometimes observed in the reactions with [CuI(PPh 3 ) 3 ], [18e] this was never the case with phosphinite catalyst C.T his allowed us to use iodoalkynes with R' = cyclopropyl, N,N-dimethylaminomethyl, andb utyl, which with our previousp hosphine system produced % 5-10 %o f5 -H-triazoles (Table 2, entries 1-2). The formation of such byproducts is problematic not only because it is undesired, but also because 5-H-and 5-I-triazoles are inseparable even by using chromatographic techniques. In this work, all formed iodotriazoles 6 could be easily isolated after as imple extraction andw ashing with pentane.
Catalyst C was also efficient with aryl azides,e ven sterically hindered ones such as mesityl azide ( Table 2, entry 3). Another important advantage of this catalytic system is that it shows aw ider functional group tolerance. Unsaturated C=Cb onds, or pyridines did not prevent the reaction from taking place and the respective iodotriazoles could be easily isolated, whereas no reactionw hatsoever had previously been observed with ap hosphine catalyst ( Table 2,

Conclusions
The utility of phosphinite-copper(I) complexesi nc ycloaddition reactionsh as been explored.I np articular,w eh ave developed two competent catalytic systems for the 1,3-dipolar cycloaddition of terminal alkynes with aryl azides generated in situ from the corresponding boronic acids as well as azides and iodoalkynes.T hese rely on robust air-stable copper complexes that can be handledi na ir with no particular precautions. Importantly,b oth the reaction conditions as well as the isolation of the desired product are 'Click-friendly'.
Phosphinite-copper complex A represents the first homogenous copper(I) complex for the one-pot azidonation/cycloaddition of boronic acids and allowst he clean formation of the desired 5-H-triazoles with al ower metal loadings than reported for heterogeneous alternatives. On the other hand, our results in the synthesis of 5-iodotriazoles clearly show that phosphinite-copper complex C is not am erea nalog of [CuI(PPh 3 ) 3 ], with somewhat improved reactivity.E ven if slightly higher metal loadings (5 insteado f1mol %) and mild heatinga re required, the use of ap hosphinite ligand has ap rofoundi mpact on the scopeo ft he catalytic system,b othi nt erms of steric hindrance and functionalgroup tolerance.
Compared with ubiquitous phosphine ligands, phosphinites (or phosphonite) ligands have only been used in ah andfulo f copper-catalyzed transformations. [28] Moreover,t ot he best of our knowledge,n oo ther pre-formed copperc omplexes with these ligandshave found applications in catalysis to date.
However,t he remarkable improvement in reactivity and applicability maket hese system attractive candidates for other copper-mediated processes. Effortsi nt his direction are currently ongoing in our laboratory and will be reported in due course.

Experimental Section
Catalytic reactions were carried out in air and by using technical solvents without any particular precautions to exclude moisture or oxygen. The reported isolated yields for the catalytic studies are an average of two independent reactions.
CAUTION: Although we did not experience any problems, the cycloaddition of azides and (iodo)alkynes is highly exothermic and, as ac onsequence, adequate cooling should always be available when performing these reactions in the absence of solvent.
A. Three-componentr eaction( boronic acid, NaN 3 ,alkyne) In av ial fitted with as crew cap, {CuBr[PPh 2 (OPh-2-OMe)]} A (11mg, 5mol %), ab oronic acid (0.5 mmol), sodium azide (0.5 mmol), water (1.5 mL), and MeOH (1.5 mL) were added and stirred for 18 h. Then, at erminal alkyne (0.5 mmol) was added and the solution was stirred for 18 h. The precipitate was extracted with ethyl acetate, stirred vigorously in aqueous saturated ammonium chloride solution (10 mL) for 3h.A fter separation, the organic layer was concentrated under reduced pressure and the resulting solid residue was washed with water,d iethyl ether and pentane, then dried under reduced pressure. In all examples, the Scheme6.Scope of the azido-iodoalkyne cycloaddition. Isolated yields are the average of at least two independent experiments. 1 HNMR conversions are showninb rackets for reactionsthatdid not reach completion.

B. [3+ +2] Cycloaddition of azidesand iodoalkynes
{CuBr[PPh 2 (OPh-4-OMe)]} C (11mg, 5mol %), azide (0.5 mmol), and iodoalkyne (0.5 mmol) were loaded into av ial fitted with as crew cap. The reaction was allowed to proceed at 40 8Cf or 18 h. Then, saturated aqueous ammonium chloride solution (10 mL) was added and the resulting mixture was stirred vigorously for 3h.T he resulting precipitate was filtered and washed with water and pentane, then dried under reduced pressure. In all examples, the crude products were estimated to be > 95 %p ure by 1 HNMR spectroscopy.