Synthesis of polymacrocyclic compounds via Pd-catalyzed amination and evaluation of their derivatives as metal detectors

Introduction

Oxygen- and nitrogen-containing polymacrocyclic compounds which combine two or more macrocyclic moieties in one molecule have been always of great interest for researchers as they provide an immense scope for constructing versatile receptors of cations, anions and polar organic molecules. A special place in the panel of polymacrocycles is occupied by cryptands which possess two or more macrocyclic units sharing several atoms thus providing a sort of a cage which increases the selectivity and binding constants towards certain coordinated species. About three decades ago the first representatives of all important classes of polymacrocycles were already synthesized and their binding properties began to be explored. To note among them bis(crown) and bis(azacrown) compounds with isolated cavities [1], [2], spiro-conjugated crown ethers [3], bis(crown) ethers with aromatic spacer [4]. Macrobicycles and more complex macrotricycles (which actually are cryptands) can be organized as well without introducing an aromatic spacer [5], [6], they can comprise phenylene [7], [8] or binaphthalene [9] spacers, described are polymacrocycles of complex architecture like azacyclophanes [10] or catenanes [11]. As early as in 1991 many dozens of polymacrocyclic compounds were mentioned in a comprehensive review [12]. Since that time the number of such compounds increased dramatically, new approaches including catalytic methods were applied for their synthesis, and the main contemporary trend is the application of polymacrocycles in chemical sensing. The problems of the cryptands coordination with different metal cations for various purposes have been always of unceasing interest for researchers [13], [14], [15], [16], [17], with more demands every year for further research of the cryptands interaction with organic compounds [18], [19], [20], [21], [22]. Cryptands are widely studied as selective receptors for cations [23], [24], [25], anions [26], [27], [28], [29] and small organic molecules [30]. Creation of fluorescent chemosensors on the basis of cryptands is an important and prospective trend in the modern techniques of detection [31], [32], [33].

The majority of known cryptands have been synthesized via convenient synthetic procedure, and until now non-catalytic approaches have been dominant [34], [35]. However, metal-catalyzed synthesis has been applied in some cases to the synthesis of macropolycyclic compounds [36]. We were the first to use Pd-catalyzed amination reactions for the synthesis of nitrogen- and oxygen-containing macrocycles [37], [38]. Among them are macrocycles incorporating phenylene [39], naphthalene [40], anthracene and anthraquinone [41] endocyclic units, as well heterocycles like pyridine [42], [43], 2,2′-bipyridyl [44] and even cholane spacers [45]. We developed the synthesis of macrocyclic colorimetric chemosensors employing 1,8-diaminoanthraquinone and showed their usefulness in the optical detection of Pb(II) cations [46], [47]. A special attention has been paid recently to the synthesis of the macrocycles based on disubstituted quinoline [48] and 1,10-phenanthroline [49] for the synthesis of fluorescent chemosensors for metal cations. The synthesis of the chiral macrocycles with central endocyclic BINAM moiety was important for the creation of enantioselective fluorescent chemosensors for optically active amino alcohols [50], [51]. This mini-review is dedicated to synthesis of polymacrocyclic compounds using Pd(0)-catalyzed amination at the key step and to possibilities of the development of chemosensors for metal cations.

Bis- and trismacrocyclic compounds

Polymacrocyclic compounds with several isolated macrocyclic rings can be obtained by the reaction of dihaloarenes with polyazamacrocycles catalyzed by Pd(0) complexes. This approach was used by us in a series of works which employed the reactions of 1,3-dibromobenzenes with N,N′,N″-trimethylcyclen and N,N′,N″-trimethylcyclam [52]. The yields of the target bis(cyclen) and bis(cyclam) products 1 and 2 were 30 and 25%, while analogous reactions of 1,8-dichloroanthracene with N,N′,N″-trimethylcyclam resulted in lower yields of corresponding bismacrocycles 3 and 4 with face-to-face arranged macrorings (Scheme 1) [53], [54]. Another approach to bismacrocycles is based on the catalytic macrocyclization reaction of 3,5-dibromobenzyl-substituted azacrown ethers 7 and 8 with a variety of polyamines and oxadiamines 9 (Scheme 1) [55]. The yields of the corresponding bismacrocycles 10 comprising 1-aza-15-crown ether moiety reached 56%. BINAP or DavePhos ligands were most commonly employed for the catalytic macrocyclizations with polyamines to obtain the best possible yields. Some of bismacrocycles were tested in coordination with Zn(II) and Cd(II) cations using NMR titration method [56].

Scheme 1:

The synthesis of bismacrocyclic compounds with isolated macrocycles.

Trismacrocyclic compounds were obtained by reacting N,N′-bis(bromobenzyl) substituted diazacrown ethers 12–15 with two equiv. of azacrown ethers 5 and 6, in all reactions the catalytic system Pd(dba)₂/DavePhos was used (Scheme 2). The yields of the target molecules 16–23 combining one diazacrown and two azacrown ethers were generally better than for bis(tetraazamacrocyclic) derivatives 1–4 and reached 44–46% [57]. It was equally possible to introduce N,N″-bis(bromobenzyl) compounds 24–27 in the similar reactions [58]. It was found out that in the majority of cases cyclen derivatives were obtained in much better yields (up to 45%) than corresponding cyclam derivatives (11–32%).

Scheme 2:

The synthesis of trismacrocyclic compounds with isolated macrocycles.

Macrobicyclic and macrotricyclic cryptands

As we gained more experience in the Pd-catalyzed synthesis of various N- and O-containing macrocycles, we decided to use this approach for constructing various cryptands of different architectures. For this purpose macrocycles on the basis of biphenyl and naphthalene 36 and 37 were modified with two 3-bromobenzyl substituents and introduced in the Pd-catalyzed macrocyclization reactions with various polyamines and oxadiamines (Scheme 3). As a result, macrobicycles with the central biphenyl moiety 40 were obtained in yields up to 30% [59], while corresponding cryptands with naphthalene moiety 41 were isolated in yields up to 35% [60]. Needless to say that the result of the macrocyclization processes in all cases strongly depended on the nature of polyamine (chain length, number of nitrogen and oxygen atoms), for this purpose the choice of the correct ligand (BINAP or DavePhos) is indispensable.

Scheme 3:

The synthesis of macrobicyclic compounds incorporating disubstituted biphenyl and naphthalene moieties.

Our earlier investigations revealed that the application of chiral phosphine ligands in the catalytic macrocyclization using 1,5-dichloroanthracene led to planar-chiral macrocycles 43 with ee up to 60% [61]. Further on some of these compounds with anthraquinone moieties were modified with 3-iodobenzyl substituents giving compounds 44 which were then introduced in the second macrocyclization reaction (Scheme 4). The yields of parent planar-chiral macrobicycles 45 were critically dependent on the length of the oxadiamine chain and reached 40% with the longest one [62].

Scheme 4:

The synthesis of macrobicyclic compounds based on 1,5-disubstittued anthraquinone with a planar-chiral fragment.

A great variety of cryptands incorporating diazacrown ether moieties was synthesized by Pd-catalyzed macrocyclization reactions using N,N-disubstituted diaza-15-crown and diaza-18-crown ethers bearing bromobenzyl (compounds 46–51), chloro- and bromopyridinylmethyl (compounds 52–55) groups (Scheme 5). It was found that BINAP ligand was much more efficient in the reactions employing bromobenzyl derivatives 56–61 while DavePhos was more helpful in the syntheses of pyridinyl-containing cryptands 62–65 [63], [64]. The best yields of the compounds with benzyl spacers were recorded for ortho-aminobenzyl derivatives 60, 61 (up to 58%) what is the best result among all processes of such type. As for macrobicycles with pyridinyl spacers, their yields were substantially lower, possibly, due to various concurrent catalytic processes with the participation of more active halogen atoms in halopyridine fragments. In accordance with this approach the derivatives with 12-membered diazacrown ether 66 and 67 were prepared as well as similar macrobicycles with diaminoadamantane linkers [65].

Scheme 5:

The synthesis of macrobicyclic compounds based on diazacrown ethers.

The coordination of some of these cryptands with different metal cations was studied using NMR titration. Stoichiometry and binding constants were established for the complexes with Zn(II), Cd(II), Pb(II), Ag(I), Ga(III), In(III), Y(III), and Sr(III) of several cryptands 58–61 while bicycles 66 and 67 were tested in the coordination with Li(I). In order to improve sensing properties of macrobocycles two alternative approaches for introducing fluorophore groups were proposed: the first one is a non-catalytic reaction with the popular dansyl chloride which afforded corresponding didansyl-substituted cryptands 68 and 69 in high yields; the second employed the catalytic arylation with bromosubstituted pyrene and quinoline which gave corresponding derivatives 70–72 (Scheme 6).

Scheme 6:

Modifications of macrobicycles with fluorophore groups.

Thus modified macrocycles were studied as potential fluorescent chemosensors for metal cations using a panel of perchlorates (Li(I), Na(I), K(I), Mg(II), Ca(II), Ba(II), Al(III), Mn(II), Fe(II), Co(II), Ni(II), Cu(II), Zn(II), Cd(II), Hg(II), Ag(I), Pb(II)) and nitrates (Ga(III), In(III), Y(III)). The addition of metal salts caused different changes in UV and fluorescence spectra depending on the nature of the central diazacrown moiety and oxadiamine linker. Fluorescence at 540 nm of dansyl-substituted compounds generally drastically diminished in the presence of Cu(II) and Al(III) perchlorates (Fig. 1) with hypsochromic shift of residual emission maxima (465 nm), while other cations either did not change emission intensity notably, or diminished it partially, or with certain cryptands In(III) (5 equiv.) caused quenching of the initial emission by half and a new emission band with a notable hypsochromic shift (415 nm) emerged. Such behavior of the fluorescent properties of compounds 68, 69 makes them promising fluorescent molecular probes for Cu(II), Al(III) and, in some cases, for In(III).

Fig. 1:

Spectra of fluorescence of selected macrobicycles in the presence of different metals cations (5 equiv.).

A vast series of macrobicyclic compounds was obtained by employing N,N″-trans-disubstituted cyclen and cyclam bearing bromobenzyl (compounds 73–77) and halogenopyridinylmethyl (compounds 87–90, 93, 94) groups (Scheme 7). The yields of the bicyclic benzyl-containing cryptands 78–82 ranged from humble 8% to quite satisfactory 44% depending on the nature of polyamines used and the type of the starting tetraazamacrocycle; in the case of cyclen yields were generally notably higher. Interesting macrotricyclic by-products were also obtained and in some cases isolated in individual state [66]. A detailed study was carried out to determine the best route of formation of these macrotricycles and using in situ obtained bis(polyamine) derivatives 83 and 84 several cylindrically-shaped cryptands 85 and 86 were synthesized. As in the case with diazacrown derivatives, the cryptands possessing disubstituted pyridine spacers (compounds 91, 92, 95, 96) were obtained in lower yields [67]. Cyclen-based macrobicycles 99 with naphthalene spacers were synthesized from corresponding bromonaphthalenylmethyl derivatives 98, they are valuable as potential fluorescent detectors for metal cations [68]. The same approach was applied to the synthesis of unusual cryptands combining cyclen or cyclam central moiety and diaminoadamantane fragments [69].

Scheme 7:

The synthesis of macrobicyclic compounds based on disubstituted cyclen and cyclam.

Some of the cyclen/cyclam-based macrobicycles were studied in the complexation with metal cations using NMR titration technique. 1:1 Stoichiometry was established for all complexes with Zn(II), Cd(II), Pb(II), Ag(I) and Hg(II) and the values of binding constants were shown to be strongly dependent on the nature of the cryptand and metals under investigation.

The development of the approach included the synthesis of macrobicyclic compounds based on tetrasubstituted cyclen and cyclam bearing additional groups at nitrogen atoms. The reactions of dibromotetrabenzyl substituted cyclen and cyclam 100–103 with polyamines afforded corresponding macrocycles 106–109 in yields up to 31%, in many cases macrotricyclic cryptands were obtained in comparable yields (Scheme 8) [70]. The reactions with di(2-pyrimidyl) derivatives of cyclen 104 and 105 gave macrobicycles 110 and 111 also in reasonable yields; these compounds are of interest as they possess additional coordination sites [71]. The combination of appropriate substituents at the nitrogen atoms of cyclam and enough short diamine linkers opens a way to the synthesis of planar-chiral macrobicycles due to restricted rotation of the diamine chain around the tetraazamacrocyclic moiety. Thus, using tetrasubstituted dibromo derivatives of cyclam 102, 112 and 113 we managed to synthesize a series of such macrobicycles 114–116 with diaminoethane and diaminopropane linkers in quite good yields up to 50% (Scheme 8). The search for the most efficient chiral phosphine ligand led to a series of Josiphos-type ligands, among which L1–L3 were demonstrated to be more efficient, however, they provided only very modest enantioinduction [72].

Scheme 8:

The synthesis of macrobicyclic compounds based on tetrasubstituted cyclen and cyclam.

The elaborated synthesis of macrobicycles comprising tetrasubstituted cyclen moiety allowed the preparation of the macrobicycle 117 with two dansyl fluorophore groups attached to the nitrogen atoms of the trioxadiamine linker. On the other hand, the formation of the didansyl-containing tetrasubstituted cyclen 118 allowed the synthesis of another macrobicycle 119 which bears dansyl groups at the nitrogen atoms of the central cyclen fragment (Scheme 9). The fluorescent titration of compounds 117 and 119 with a variety of metal salts (Fig. 2) showed that compound 119 was more selective towards Cu(II) ions as only this metal quenched emission almost totally while other cations changed it only to some extent. The ligand 117 was less efficient as the addition of Cu(II), Al(III) and Cr(III) ions led only to partial quenching of fluorescence.

Scheme 9:

The synthesis of cyclen-based macrobicyclic compounds with dansyl fluorophore groups.

Fig. 2:

Fluorescence spectra of macrobicycles 119 (a) and 117 (b) in the presence of different metals cations (5 equiv.).

Another approach was elaborated for the synthesis of bicyclic ligands based on triazacyclononane (TACN) and triazacyclododecane (TACD) possessing dansyl fluorophore groups. For this purpose one dansyl group was introduced into free triazacycle 120 or 121, then the compounds 122 and 123 were modified with bromobenzyl substituents to give derivatives 124–126, the latter compounds were introduced in the Pd-catalyzed macrocyclization reactions with oxadiamines in which DavePhos ligand was found to be advantageous over BINAP (Scheme 10). The resulting bicyclic compounds 127–129 were obtained in rather small yields due to formation of various oligomeric by-products [73].

Scheme 10:

The synthesis of bicyclic compounds based on triazacyclononanes with dansyl fluorophore groups.

The sensing properties of the synthesized bicycles were tested using a panel of 20 metal cations and it was revealed that almost all the ligands independent of the oxadiamine linker nature fully quenched their fluorescence in the presence of Cu(II) and Al(III) ions (Fig. 3). However, the selectivity of TACN- and TACD-based bicycles was different as in the latter case other ions (like In, Ga, Hg) also substantially diminished the emission intensity; it means that these cations may mask the action of copper and aluminum ions. On the other hand, the fluorescence of the ligand 129 in the presence of K(I) showed by a notable hypsochromic shift of the emission maximum.

Fig. 3:

Fluorescence spectra of TACN-based bicycles in the presence of different metals cations (5 equiv.).

The introduction of another fluorophores (naphthalene, acridine) in the molecule of TACN was achieved via the aminal derivative 130 (Scheme 11). The precursor 131 bearing naphthalene moiety and two bromobenzyl substituents was obtained in total 45% yield via a 4-step procedure while similar acridine-containing derivative 132 was synthesized in 37% yield after two steps (deprotection step occurred simultaneously together with the alkylation). The macrocyclization with the longest trioxadiamine led to corresponding bicycles 133 and 134 in low yields [74]. It is obvious that the introduction of additional substituents in the parent macrocycles increases the steric hindrances what causes the decrease in the efficiency of the intramolecular cyclization and leads to excessive formation of oligomeric by-products.

Scheme 11:

The synthesis of bicyclic compounds based on TACN with naphthalene and acridine fluorophore groups.

The decoration of diazacrown ethers with two 3,5-dibromobenzyl substituents (compounds 135 and 136) gave a possibility to synthesize trismacrocyclic compounds 137 and 138 (Scheme 12). This double macrocyclization process was possible in the presence of RuPhos catalyst. The presence of four reaction centers in the starting compounds made this synthesis a real challenge as only one reaction route out of many possible others was desirable. The modification of one trismacrocycle with four dansyl fluorophore groups afforded compound 139 in high yield, the latter was tested as a fluorescent chemosensor for metal cations [75]. The addition of Cu(II) and Al(III) ions led to full quenching of emission (Fig. 4), however, this ligand was less selective compared to macrobicycles described above which contained 1 or 2 dansyl groups: some cations like Fe(III), Ga(III), Ni(II), Co(II), Hg(II) caused notable emission quenching as well. As for UV spectrum of the ligand, only the addition of Cu(II) and Al(III) led to principal spectral changes, thus tricycle 139 can be regarded as a dual-channel molecular probe for these cations.

Scheme 12:

The synthesis of trismacrocyclic derivative of diazacrown ether and its modification with four dansyl fluorophore groups.

Fig. 4:

UV (a) and fluorescence (b) spectra of compound 139 in the presence of different metals cations (5 equiv.).

The possibility of the synthesis of more sophisticated cryptand-type macrotricycle using tetrabromide 135 was demonstrated by the two-step synthesis of compound 141 which employed two Pd-catalyzed macrocyclization reactions (Scheme 13). The intermediate macrobicycle 140 was also used for the modifications with fluorophore groups (pyrene and quinoline) by Pd-catalyzed amination using two resting bromine atoms.

Scheme 13:

The synthesis of macrotricyclic derivative of diaza-15-crown-5 ether, the catalytic introduction of fluorophore substituents.

Another approach to macrotricyclic compounds based on diazacrown ether was shown by the two-step modification of the macrobicycles 58, 59 (Scheme 14). They were provided with two bromobenzyl groups to make corresponding compounds 143 and 144 in high yields, but the macrocyclization reaction was efficient only in one case to give the target cryptand 145 while other intermediates afforded mainly mixtures of oligomers [75].

Scheme 14:

The synthesis of macrotricyclic derivative of diaza-18-crown-6 ether.

Three different routes to macrotricyclic compounds containing the central cyclen or cyclam moieties have been proposed. The first one employed accurate modification of macrobicycles with two bromobenzyl substituents by playing with different reactivity of dialkyl and alkylaryl amino groups (Scheme 15). These intermediates were introduced in the Pd-catalyzed macrocyclization to give unusual spherically-shaped cryptands 147–149, surprisingly, the yield in the case of compound 148 attained 44% [76].

Scheme 15:

The synthesis of spherically-type macrotricyclic derivatives of cyclen.

To change the geometry of the macrotricyclic cryptands, two secondary amino groups in N,N′-bis(3-bromobenzyl)cyclen 73 were protected with Boc groups giving derivative 150 which was then transformed via three-step sequence of Pd-catalyzed macrocyclization, bromobenzylation and the second Pd-catalyzed macrocyclization to produce macrotricycles 151–153 in reasonable yields (Scheme 16) [76]. Cylindrically-shaped macrotricyclic compounds were successfully obtained by the catalytic amination of N,N′-bis(bromobenzyl) substituted diazacrown ethers 58, 59 and cyclen or cyclam 73–76 with free diazacrown ethers (Scheme 17). The outcome of these reactions depended strongly on the mutual orientation of bromine and nitrogen reaction centers what is important for such more sterically rigid reactants, and in some cases they yields reached 20–30% [77].

Scheme 16:

The synthesis of Boc-containing macrotricyclic derivatives of cyclen.

Scheme 17:

The synthesis of cylindrically-shaped macrotricyclic derivatives of diazacrown ethers and tetraazamacrocycles.

Polymacrocyclic conjugates with porphyrins

Porphyrins are extremely important and demanded macrocyclic compounds possessing numerous useful and interesting properties among which is their ability to form various metal complexes and unique photophysical characteristics. Taking these facts into consideration, we decided to combine macrocyclic compounds with porphyrins in polymacrocyclic conjugates in view of creating new fluorescent chemosensors for metal cations. Using Pd-catalyzed amination of meso-bromophenyl porphyrins 164, 167 and 168 with diamino derivative of azacrown ethers 163 and 166 we managed to obtain corresponding bis- and trismacrocyclic conjugates 165, 169 and 170 (Scheme 18) which were tested as potential chemosensors with 17 different metal cations.

Scheme 18:

The synthesis of bis- and trismacrocyclic conjugates of porphyrins with azacrown ethers.

It was established that compounds 165 and 169 could act as selective detectors of Cu(II) ions by full quenching of fluorescence (Fig. 5). Other metals did not intervene in Cu(II) detection. It is worth noting that the conjugate 165 with one azacrown moiety was less sensitive to this metal than the conjugate 169 with two azacrowns attached to 5,15-diphenylporphyrin central unit [78], [79].

Fig. 5:

Fluorescence spectra of porphyrin conjugates 165 (a, b) and 169 (c, d) in the presence of 2.5 equiv. metal cations (a, c) and different amounts of Cu(II) perchlorate (b, d).

Other porhyrin conjugates were synthesized using previously elaborated trioxadiazacrown ethers on the basis of biphenyl 36 and naphthalene 37. The arylation of these macrocycles was conducted with Zn(II) mono- and bis(bromophenyl)porphyrinates. As a result, in the reactions with Zn(II) 5-(4-bromophenyl)porphyrinate 171 tris- and bismacrocyclic conjugates 172–175 were obtained: compounds 172 and 174 with two porphyrin units and compounds 173, 175 with one porphyrin unit (Scheme 19) [80]. The reaction with Zn(II) 5,15-bis(4-bromophenyl)porphyrinate 168 produced trismacrocyclic structure 176 with two trioxadiazamacrocycles and bismacrocycle 177 with one trioxadiazacrown ether unit [81].

Scheme 19:

The synthesis of bis- and trismacrocyclic conjugates of porphyrins with biphenyl- and naphthalene-based diazacrown ethers.

The experiments were conducted to find out the most selective chemosensors for metal cations. The emission of trismacrocycle 172 diminished substantially in the presence of Cr(III) cations, while full quenching was observed with Cu(II) and Al(III) (Fig. 6a), bismacrocycle 173 was not so sensitive to the presence of the metal ions and its emission diminished notably with Cu(II) and Pb(II) ions (Fig. 6b). On the other hand, another trismacrocyclic ligand 174 on the basis of the naphthalene-containing trioxadiazacrown ether showed much better selectivity towards Cu(II) cations as only in the presence of this metal its emission quenched totally while other 17 cations altered it insignificantly (Fig. 6c).

Fig. 6:

Fluorescence spectra of porphyrin conjugates 172 (a), 173 (b), 174 (c), and 179 (d) in the presence of 5 equiv. metal cations.

More sophisticated polymacrocyclic systems were obtained by the Pd-catalyzed arylation of diazacrown-based macrobicycles with Zn(II) porphyrinates 164 and 171 (Scheme 20) [80], 81]. As in all processes dealing with the arylation of the secondary amino groups in macrocycles with bromophenylporphyrins, the application of Pd(dba)₂/DavePhos catalytic system was crucial for successful accomplishment of the process. Compound 179 was shown to be rather selective fluorescent chemosensor for Cu(II) cations as the addition of only this metal led to substantial emission quenching (Fig. 6d).

Scheme 20:

The synthesis of polymacrocyclic conjugates of porphyrins with diazacrown-based macrobicycles.

Conclusions

This mini-review demonstrates that on the basis of Pd-catalyzed amination we have successfully developed several approaches to polymacrocyclic compounds with isolated macrocycles and to macropolycyclic cryptands of different architecture. The dependence of the yields of the compounds on the structural types of starting compounds and on the nature of polyamine linkers introduced for the formation of new macrorings was shown. Many compounds thus synthesized were modified with fluorophore groups and tested for the possibility of their application as fluorescent chemosensors of metal cations. The majority of dansyl-containing polymacrocycles proved to be selective towards Cu(II) and Al(III) ions while porphyrin-based conjugates with polyazapolyoxamacrocyclic units can be regarded as prospective fluorescent chemosensors of Cu(II) ions.