2-Aminobenzothiazole-Containing Copper(II) Complex as Catalyst in Click Chemistry: An Experimental and Theoretical Study

: The reaction of copper(II) acetate with the 2-aminobenzothiazole (abt) heterocycle a ﬀ ords the new copper(II) complex of formula [Cu(abt) 2 (OOCCH 3 ) 2 ] ( 1 ) in a straightforward manner. Compound 1 served as a precatalyst for azide / alkyne cycloaddition reactions (CuAAC) in water, leading to 1,4-disubstituted-1,2,3-triazole derivatives in a regioselective manner and with excellent yields at room temperature. The main advantages of the coordination of such a heterocyclic ligand in 1 are its strong σ -donating ability (N-Cu), nontoxicity and biological properties. In addition, the click chemistry reaction conditions using 1 allow the formation of a great variety of 1,2,3-triazole-based heterocyclic compounds that make this protocol potentially relevant from biological and sustainable viewpoints. A molecular electron density theory (MEDT) study was performed by using density functional theory (DFT) calculations at the B3LYP / 6-31G(d,p) (LANL2DZ for Cu) level to understand the observed regioselectivity in the CuAAC reaction. The intramolecular nature of this reaction accounts for the regioselective formation of the 1,4-regioisomeric triazole derivatives. The ionic nature of the starting copper-acetylide precludes any type of covalent interaction throughout the reaction, as supported by the electron localization function (ELF) topological analysis, rea ﬃ rming the zwitterionic-type ( zw -type) mechanism of the copper(I) / aminobenzothiazole-catalysed azide-alkyne cycloaddition reactions.


Introduction
Triazole-containing heterocycles have found many applications in several disciplines, such as medicine, biology and materials science [1]. In general, these compounds are prepared through azide-alkyne cycloaddition (AAC) reactions [2,3]. Indeed, the non-catalyzed [3 + 2] cycloaddition (32CA) reaction, also known as "Huisgen's 1,3-dipolar cycloaddition", between azides and alkynes is characterized by low rates and small yields at room temperature, due to the high kinetic barrier of the reaction [4,5], as well as the formation of two isomers, that is, 1,4-and 1,5-disubstituted 1,2,3-triazoles. Later, the discovery by the groups of Meldal [6] and Sharpless [7] that copper(I) is capable of accelerating the 32CA reaction of azides and alkynes with high regioselectivity, to afford the 1,4-disubstituted 1,2,3-triazole isomer under mild conditions, led to a breakthrough of such disubstituted 1,2,3-triazole isomer under mild conditions, led to a breakthrough of such a cycloaddition process. Since then, the copper-catalyzed azide-alkyne cycloaddition (CuAAC) reaction has become the most representative click reaction in the panorama of click chemistry due to its mild conditions and vast applications, in particular in life sciences [8][9][10][11]. However, the thermodynamic instability of copper(I), as well as its easy oxidation to copper(II), have led to the search for suitable stabilizing ligands for Cu(I) ion. Consequently, the formation of the active Lewisacid copper(I) species is becoming crucial in developing CuAAC reactions in terms of stability and activity [12]. A variety of homogeneous catalytic systems were reported for the click of 1,4disubstituted 1,2,3-triazoles [13][14][15][16][17]. In this regard, several polydentate ligands containing nitrogen atoms have been used to thermodynamically stabilize copper ions against destabilizing reactants found in the reaction medium. Some of these interesting ligands include the tris-(benzyltriazolylmethyl)amine (TBTA) [18], dipicolinate [19], tris(triazolyl)methanol (TBTM) [20], 2pyrrolecarbaldiminato [21] and crowded tetradentate tris(2-dioctadecylaminoethyl)amine [22].
Inspired by Nature's strategy, the use of ligands with biological activities can broaden the CuAAC applications. Many heterocyclic nuclei, such as benzothiazole, have been recently reviewed as antimicrobial agents [23,24] and 2-aminobenzothiazole derivatives were reported with a variety of biological applications [25,26]. The 2-aminobezothiazole ligand (abt) contains sulfur and nitrogen in a ring, an exocyclic nitrogen and π-electrons, giving the ligand a high coordination ability. Its coordination to the metal center via the nitrogen atom of the ring occurs in most of the reported crystal structures [27,28]. It is worth noting that the type of complexes have been widely studied as they exhibit interesting biological and photophysical properties [29][30][31].
Having in mind the above facts, we present here the synthesis of a Cu(II)-abt complex in ethanol/water and the study of its catalytic potential for the click of 1,4-disubstituted-1,2,3-triazole compounds via azide-alkyne cycloaddition reactions in water at room temperature. A theoretical study by the MEDT method and ELF topological analysis is also carried out to explain the mechanism and intermediates species that participate in this chemical process. It deserves to be noted that the regioselective click of 1,2,3-triazoles is performed in absence of any reducing agent or additive bases, while the catalyst is found to be highly active at lower loadings.

Results and Discussion
The copper(II)-abt complex of formula [Cu(abt)2(OOCCH3)2] (1) was simply prepared by the addition of an aqueous solution of copper(II) acetate to a methanolic solution of abt at room temperature (Scheme 1). Green crystals of 1 were grown from the resulting solution. The optimized structure of 1 with the atom numbering of the non-carbon atoms is shown in Figure 1. Bond distances and angles of the complex 1 are summarized in Table 1. The copper(II) ion in 1 is six-coordinate with two trans-positioned nitrogen atoms from to abt molecules and four oxygen atoms of two chelating acetate ligands building an elongated octahedral CuN2O4 environment. Each acetate ligand in 1 adopts an asymmetrical didentate coordination, with one short (1.981 and 1.980 Å for Cu1-O1 and Cu1-O4, respectively) and one long (2.554 Å for Cu1-O2 and Cu1-O3) copper-oxygen bond lengths. The Cu-Nabt bond distances [2.130 Å for Cu-N5 and Cu-N6, respectively] are somewhat shorter than the longest copper to oxygen bonds. The value of the N5-Cu-N6 bond angle (179.9°) is Scheme 1. Schematic illustration of the synthesis of 1.
The optimized structure of 1 with the atom numbering of the non-carbon atoms is shown in Figure 1. Bond distances and angles of the complex 1 are summarized in Table 1. The copper(II) ion in 1 is six-coordinate with two trans-positioned nitrogen atoms from to abt molecules and four oxygen atoms of two chelating acetate ligands building an elongated octahedral CuN 2 O 4 environment. Each acetate ligand in 1 adopts an asymmetrical didentate coordination, with one short (1.981 and 1.980 Å for Cu1-O1 and Cu1-O4, respectively) and one long (2.554 Å for Cu1-O2 and Cu1-O3) copper-oxygen bond lengths. The Cu-N abt bond distances [2.130 Å for Cu-N5 and Cu-N6, respectively] are somewhat shorter than the longest copper to oxygen bonds. The value of the N5-Cu-N6 bond angle (179.9 • ) is very close to linearity. These findings are in agreement with the crystallographic data of this compound, whose single-crystal X-ray structure was reported by Dojer et al. in 2019 [28].
Catalysts 2020, 10, x FOR PEER REVIEW  3 of 19 very close to linearity. These findings are in agreement with the crystallographic data of this compound, whose single-crystal X-ray structure was reported by Dojer et al. in 2019 [28].   [28].

Catalytic Studies
The as-prepared copper(II) complex was then tested in 32CA reactions between a variety of azides and terminal alkyne derivatives. The reaction between benzyl azide (1a) and phenylacetylene (2a) in the water at room temperature was chosen as the model reaction and it was performed under a variety of conditions (see Scheme 2 and Table 2). The controlled experiments showed that the 32CA reaction does not take place in the absence of the catalyst (entry 1). It was observed that the use of copper(II) acetate only in water as a reaction medium led to poor yields of the corresponding 1,2,3-

Catalytic Studies
The as-prepared copper(II) complex was then tested in 32CA reactions between a variety of azides and terminal alkyne derivatives. The reaction between benzyl azide (1a) and phenylacetylene (2a) in the water at room temperature was chosen as the model reaction and it was performed under a variety of conditions (see Scheme 2 and Table 2). The controlled experiments showed that the 32CA reaction does not take place in the absence of the catalyst (entry 1). It was observed that the use of copper(II) acetate only in water as a reaction medium led to poor yields of the corresponding 1,2,3-triazole ca. 40% after stirring for 24 h at room temperature (entry 2). Interestingly, the corresponding 1,4-triazole was formed in high yields (98%) within 12 h by using 5 mol% of 1 and water as solvent (entry 5) compared with copper(II) sulphate in the presence of 2-abt as a ligand (entry 4). The catalyst loading for 1 could be further reduced to 2 mol%, affording an excellent yield of the corresponding 1,4-disubstituted 1,2,3-triazole (92%) in water within 4 h at room temperature ( Table 2, entry 7). Further, several solvents were investigated for the model reaction ( Table 2, entries 9-14). Hexane, ethanol, acetonitrile, acetone and toluene as solvents gave poor to moderate reaction yields (40-67%, respectively). The ethanol/water solvent mixture also gave a moderate yield ca. 62% with 2 mol% of the catalyst. As can be seen in Table 2, the maximum yield of the triazole product was obtained with 2 mol% of the catalyst in water medium at room temperature.
Catalysts 2020, 10, x FOR PEER REVIEW  4 of 19   triazole ca. 40% after stirring for 24 h at room temperature (entry 2). Interestingly, the corresponding 1,4-triazole was formed in high yields (98%) within 12 h by using 5 mol% of 1 and water as solvent (entry 5) compared with copper(II) sulphate in the presence of 2-abt as a ligand (entry 4). The catalyst loading for 1 could be further reduced to 2 mol%, affording an excellent yield of the corresponding 1,4-disubstituted 1,2,3-triazole (92%) in water within 4 h at room temperature ( Table 2, entry 7). Further, several solvents were investigated for the model reaction (Table 2, entries 9-14). Hexane, ethanol, acetonitrile, acetone and toluene as solvents gave poor to moderate reaction yields (40-67%, respectively). The ethanol/water solvent mixture also gave a moderate yield ca. 62% with 2 mol% of the catalyst. As can be seen in Table 2, the maximum yield of the triazole product was obtained with 2 mol% of the catalyst in water medium at room temperature.

Scheme 2.
Azide-alkyne cycloaddition in the presence of complex 1. Based on the above-optimized reaction conditions, the scope of our method was explored by using different para-substituted phenyl acetylenes as well as different benzyl and alkyl azides. The 32CA reactions progressed effectively to afford the corresponding triazole derivatives in a short time for all cases and the corresponding 1,2,3-triazoles were isolated in good to excellent yields (see Table  3). Interestingly, the clicking reaction between alkyne-containing polycyclic and benzyl or alkyl azides was conducted and the corresponding 1,2,3-triazoles were obtained in excellent yields by precipitation in water and without need for any further purification by conventional methods (see Table 3, entries 11 and 12). The products were fully characterized by 1 H and 13 C NMR spectroscopy and mass spectrometry (see Supplementary Materials for details).  Based on the above-optimized reaction conditions, the scope of our method was explored by using different para-substituted phenyl acetylenes as well as different benzyl and alkyl azides. The 32CA reactions progressed effectively to afford the corresponding triazole derivatives in a short time for all cases and the corresponding 1,2,3-triazoles were isolated in good to excellent yields (see Table 3). Interestingly, the clicking reaction between alkyne-containing polycyclic and benzyl or alkyl azides was conducted and the corresponding 1,2,3-triazoles were obtained in excellent yields by precipitation in water and without need for any further purification by conventional methods (see Table 3, entries 11 and 12). The products were fully characterized by 1 H and 13 C NMR spectroscopy and mass spectrometry (see Supplementary Materials for details).

Comparison of Complex 1 with other Catalysts
Catalysts 2020, 10, x FOR PEER REVIEW 5 of 19

Comparison of Complex 1 with other Catalysts
Catalysts 2020, 10, x FOR PEER REVIEW 5 of 19

Comparison of Complex 1 with other Catalysts
Catalysts 2020, 10, x FOR PEER REVIEW 5 of 19

Comparison of Complex 1 with other Catalysts
Catalysts 2020, 10, x FOR PEER REVIEW 5 of 19

Comparison of Complex 1 with other Catalysts
Catalysts 2020, 10, x FOR PEER REVIEW 5 of 19

Comparison of Complex 1 with other Catalysts
Catalysts 2020, 10, x FOR PEER REVIEW 5 of 19

Comparison of Complex 1 with other Catalysts
Catalysts 2020, 10, x FOR PEER REVIEW 5 of 19

Comparison of Complex 1 with other Catalysts
In order to prove the efficiency of the complex 1-mediated azide-alkyne cycloaddition reaction, a comparison of this catalyst with other copper(II) complexes for the reaction of phenylacetylene (1a) and benzyl azide (2a) to afford 1,4-disubstituted 1,2,3-triazoles was done (see Scheme 2 and Table 4). It was observed that the presence of a reducing agent, high temperatures and inert atmosphere are requisites to operate some catalytic systems. Importantly, complex 1 occurs in an effective manner for the synthesis of 1,2,3-triazole, working in air and at room temperature in water, followed by an easy work-up protocol for the final separation of the product.

Mechanistic Studies
Based on the reported mechanism [38][39][40][41][42][43][44], the reaction mechanism for the synthesis of 1,4-disubstituted 1,2,3-triazoles in the presence of 1 is shown in Scheme 3. Firstly, the catalytic copper(I) species arises through the reduction of the copper(II) complex by the terminal alkyne via the oxidative alkyne homocoupling process [45]. Then, the in-situ coordination of the terminal alkyne to the copper(I) species leads to the dinuclear acetylide-copper intermediate. The reaction of this dinuclear complex and organic azide conducts to a six-membered ring, forming the triazolide ring intermediate.  Scheme 3. Generally accepted mechanism of CuAAC catalyzed by complex 1 (L = abt).

MEDT Study
Recently, the molecular electron density theory [46] (MEDT) was proved to be a powerful tool for the theoretical understanding of 32CA reactions allowing the explanation of the reactivity of the simplest three-atom-components (TACs) towards ethylene, based on their respective electronic structure [47]. Accordingly, the CuAAC reaction catalyzed by the Cu(I)-abt complex was investigated through the MEDT study (Scheme 4). The terminal alkyne is firstly metallated with copper(I) via an easy deprotonation of the alkyne, yielding a binuclear copper(I)-acetylide species (see Scheme 4). After coordination of the azide to the copper(I)-acetylide, the cycloaddition subsequently affords the corresponding 1,2,3-triazole compounds. Herein, the reaction path associated to the CuAAC reaction of propyne (4) with methyl azide (5) yielding 1,4-dimethyl-1,2,3-triazole (6) is investigated within MEDT through DFT methods at level B3LYP/6-31G(d,p) (LANL2DZ for Cu) (see Scheme 4).

MEDT Study
Recently, the molecular electron density theory [46] (MEDT) was proved to be a powerful tool for the theoretical understanding of 32CA reactions allowing the explanation of the reactivity of the simplest three-atom-components (TACs) towards ethylene, based on their respective electronic structure [47]. Accordingly, the CuAAC reaction catalyzed by the Cu(I)-abt complex was investigated through the MEDT study (Scheme 4). The terminal alkyne is firstly metallated with copper(I) via an easy deprotonation of the alkyne, yielding a binuclear copper(I)-acetylide species (see Scheme 4).
After coordination of the azide to the copper(I)-acetylide, the cycloaddition subsequently affords the corresponding 1,2,3-triazole compounds. Herein, the reaction path associated to the CuAAC reaction of propyne (4) with methyl azide (5) yielding 1,4-dimethyl-1,2,3-triazole (6) is investigated within MEDT through DFT methods at level B3LYP/6-31G(d,p) (LANL2DZ for Cu) (see Scheme 4). Azide components (TACs) participate in the zw-type 32CA with high activation energies [48], which can be favored through a more polar process by the increase of the electrophilic and nucleophilic character of the reagents. Consequently, the 32CA reaction catalyzed by Cu(I)-abt complex was analyzed using the global reactivity indices defined within Conceptual DFT (CDFT) [49,50] and the results are shown in Table 5. Compound 4 was found as a marginal electrophile and a marginal nucleophile within the electrophilicity [51] and nucleophilicity scales [52], a feature that explains the non-participation of 4 in polar cycloaddition reactions. In contrast, compound 5 is a strong electrophile (ω = 1.22 eV) and a moderate nucleophile (N = 2.15 eV). However, the coordination of 4 to copper(I)-abt increases the values of the ω and N indexes to 1.14 and 3.11 eV being thus classified as both strong electrophile and nucleophile. Therefore, the CuAAC reactions of 5 with Ac are polar reactions due to the strong nucleophilic (Ac) and the strong electrophilic (5) characters (see Table 5). Recently, a lot of studies of polar 32CA reactions have shown that the analysis of the electrophilic and nucleophilic Parr functions allows us to explain the experimentally observed regioselectivity [53]. Thus, the electrophilic Pk + and nucleophilic Pk -Parr functions at 4, 5 and Ac were analyzed (see Figure 2). The analysis of the nucleophilic Pk -Parr functions at 4 indicates that the terminal C5 carbon is the most nucleophilic center of this molecule (Pk -= 0.54), although the C4 carbon is also nucleophilically activated (Pk -= 0.36). On the other hand, the analysis of the electrophilic Pk + Parr functions at 5 reveals that the terminal N1 nitrogen is the most electrophilic center of this TAC (Pk + = 0.56). These results suggest that along with a polar 32CA reaction, the 1,5-regioisomer will be the main reaction product. However, the very low polar character of this non-catalyzed AAC reaction accounts for its low regioselectivity. Azide components (TACs) participate in the zw-type 32CA with high activation energies [48], which can be favored through a more polar process by the increase of the electrophilic and nucleophilic character of the reagents. Consequently, the 32CA reaction catalyzed by Cu(I)-abt complex was analyzed using the global reactivity indices defined within Conceptual DFT (CDFT) [49,50] and the results are shown in Table 5. Compound 4 was found as a marginal electrophile and a marginal nucleophile within the electrophilicity [51] and nucleophilicity scales [52], a feature that explains the non-participation of 4 in polar cycloaddition reactions. In contrast, compound 5 is a strong electrophile (ω = 1.22 eV) and a moderate nucleophile (N = 2.15 eV). However, the coordination of 4 to copper(I)-abt increases the values of the ω and N indexes to 1.14 and 3.11 eV being thus classified as both strong electrophile and nucleophile. Therefore, the CuAAC reactions of 5 with Ac are polar reactions due to the strong nucleophilic (Ac) and the strong electrophilic (5) characters (see Table 5). Recently, a lot of studies of polar 32CA reactions have shown that the analysis of the electrophilic and nucleophilic Parr functions allows us to explain the experimentally observed regioselectivity [53]. Thus, the electrophilic P k + and nucleophilic P k − Parr functions at 4, 5 and Ac were analyzed (see Figure 2). The analysis of the nucleophilic P k − Parr functions at 4 indicates that the terminal C5 carbon is the most nucleophilic center of this molecule (P k − = 0.54), although the C4 carbon is also nucleophilically activated (P k − = 0.36). On the other hand, the analysis of the electrophilic P k + Parr functions at 5 reveals that the terminal N1 nitrogen is the most electrophilic center of this TAC (P k + = 0.56). These results suggest that along with a polar 32CA reaction, the 1,5-regioisomer will be the main reaction product. However, the very low polar character of this non-catalyzed AAC reaction accounts for its low regioselectivity.  Figure 3 shows electrophilic Pk + and nucleophilic Pk -Parr functions of 4, 5 and Ac. The analysis of the nucleophilic Pk -Parr functions at 4 indicates that the terminal carbon (C5) is the most nucleophilic center of this molecule (Pk -= 0.57), while the most electrophilic center of 5 is the terminal N1 nitrogen (Pk + = 0.57). Thus, the very low polar character of this non-catalyzed azide-alkyne  In the first step of the CuAAC mechanism, 4 binds to a copper(I) cation as a π-ligand. This coordination increases the acidity of the terminal alkyne, favoring the formation of mononuclear Cu(I)-acetylide upon deprotonation. Then, the coordination of the C5 carbon atom to a second copper(I) center leads to the dinuclear Ac species. Subsequently, the coordination of 5 to a copper(I) cation in Ac via its N3 azide nitrogen allows the formation of the reactive complex (RC). This intermediate is the starting point for the stepwise sequences represented in Figure 3. This step is highly exothermic at 17.31 kcal/mol. The computed barrier for the formation of the first N1-C4 single bond via TS1 is 2.66 kcal/mol, a value which is smaller than the barrier for the non-catalyzed reaction (ca. 30.3 kcal/mol) [54]. Subsequently, the formation of the intermediate complex (IC) is exothermic by ca. 13.74 kcal/mol. From this intermediate, the barrier for the ring contraction leading to the Cu(I)triazolyl derivative is 11.17 kcal/mol (TS2). Finally, the formation of the AT derivative is exothermic by 112.14 kcal/mol. These results allow us to explain the easy formation of the corresponding triazoles in water. Moreover, the use of abt as ligand for copper(I) in the CuAAC process shows that this heterocyclic ligand causes a decrease of the energy barrier of about 3-6 kcal/mol in the aqueous phase, by comparison with our previously reported catalytic system using the water molecule as the only ligand [55]. In the first step of the CuAAC mechanism, 4 binds to a copper(I) cation as a π-ligand. This coordination increases the acidity of the terminal alkyne, favoring the formation of mononuclear Cu(I)-acetylide upon deprotonation. Then, the coordination of the C5 carbon atom to a second copper(I) center leads to the dinuclear Ac species. Subsequently, the coordination of 5 to a copper(I) cation in Ac via its N3 azide nitrogen allows the formation of the reactive complex (RC). This intermediate is the starting point for the stepwise sequences represented in Figure 3. This step is highly exothermic at 17.31 kcal/mol. The computed barrier for the formation of the first N1-C4 single bond via TS1 is 2.66 kcal/mol, a value which is smaller than the barrier for the non-catalyzed reaction (ca. 30.3 kcal/mol) [54]. Subsequently, the formation of the intermediate complex (IC) is exothermic by ca. 13.74 kcal/mol. From this intermediate, the barrier for the ring contraction leading to the Cu(I)-triazolyl derivative is 11.17 kcal/mol (TS2). Finally, the formation of the AT derivative is exothermic by 112.14 kcal/mol. These results allow us to explain the easy formation of the corresponding triazoles in water. Moreover, the use of abt as ligand for copper(I) in the CuAAC process shows that this heterocyclic ligand causes a decrease of the energy barrier of about 3-6 kcal/mol in the aqueous phase, by comparison with our previously reported catalytic system using the water molecule as the only ligand [55].

ELF Topological Analysis
In order to get more understanding of the CuAAC reactions catalyzed by Cu(I)-abt, all intermediates in the reaction between Ac and 5 were characterized by a topological analysis of the ELF [56]. The results are given in Table 6 and the ELF-based Lewis representations are illustrated in Figure 4. ELF [56]. The results are given in Table 6 and the ELF-based Lewis representations are illustrated in Figure 4.  The topological analysis of the ELF of 5 is characterized by the presence of two V(N1,N2) and V'(N1,N2) disynaptic basins integrating a total population of 4.00 e, two V(N1) and V(N3) monosynaptic basins integrating 3.94 and 3.53 e and one V(N2,N3) disynaptic basin integrating 2.51 e. Accordingly, Lewis' bonding model for 5 is identified by one N1-N2 double bond, a N2-N3 single bond and two lone pairs at the N1 and N3 azide-nitrogen atoms (see Table 6 and Figure 4), confirming that 5 participates as zwitterionic TAC in the 32CA reactions [57]. The electronic structure of Ac shows the presence of one V(C5) monosynaptic basin, with a high electron population of 3.37 e and two V(C4,C5) and V'(C4,C5) disynaptic basins, with a total population of 4.75 e. These valence basins can be attributed to the C4-C5 double bond and the C5 carbanionic center, reveals the ionic nature of Ac (see Table 6 and Figure 4).
The ELF analysis for the reactive complex (RC) shows non-appreciable topological changes with The topological analysis of the ELF of 5 is characterized by the presence of two V(N1,N2) and V'(N1,N2) disynaptic basins integrating a total population of 4.00 e, two V(N1) and V(N3) monosynaptic basins integrating 3.94 and 3.53 e and one V(N2,N3) disynaptic basin integrating 2.51 e. Accordingly, Lewis' bonding model for 5 is identified by one N1-N2 double bond, a N2-N3 single bond and two lone pairs at the N1 and N3 azide-nitrogen atoms (see Table 6 and Figure 4), confirming that 5 participates as zwitterionic TAC in the 32CA reactions [57]. The electronic structure of Ac shows the presence of one V(C5) monosynaptic basin, with a high electron population of 3.37 e and two V(C4,C5) and V'(C4,C5) disynaptic basins, with a total population of 4.75 e. These valence basins can be attributed to the C4-C5 double bond and the C5 carbanionic center, reveals the ionic nature of Ac (see Table 6 and Figure 4).
The ELF analysis for the reactive complex (RC) shows non-appreciable topological changes with respect to the separated reagents. For example, a similar behavior is located at the azide framework for the V(N2,N3) disynaptic basin and the total population of the V(N1,N2) and V (N1,N2) disynaptic basins increases by an equivalent amount. A similar behavior is found for the two V(N1) and V(N3) monosynaptic basins. The V(C4,C5) and V'(C4,C5) disynaptic basins in the alkyne framework are slightly depopulated by 0.03 and 0.05 e, respectively. Several changes occur in the azide framework at TS1, while a new V(N2) monosynaptic basin, associated with the N2 nitrogen non-bonding electron density, was created, integrating 2.60 e. The total population of the V(N2,N3) disynaptic basin, as well as that of the two V(N1,N2) and V (N1,N2) disynaptic basins, strongly decreased by 0.86 and 1.64 e, respectively. In addition, the V(N3) monosynaptic basin split into two V(N3) and V (N3) monosynaptic basins integrating 1.99 and 1.63 e, although their total population only increased by 0.04 e. Otherwise, the population of the V(N1) monosynaptic basin more markedly increased to 3.62 e, but the N1 nitrogen non-bonding electron density remained characterized by one single V(N1) monosynaptic basin. A strong depopulation of the two V(C4,C5) and V (C4,C5) disynaptic basins was achieved for the acetylide framework by a total of 0.81 e. The V(C5) monosynaptic basin reached 2.34 e and a new V(C4) monosynaptic basin was created integrating 0.39 e, which can be attributed to a C4 pseudoradical center. At CI, a new V(N1,C4) disynaptic basin is observed integrating 1.97 e and the disappearance of the V(C4) monosynaptic basin located at TS1, which is related to the new N1-C4 single bond by sharing part of the electron population of the V(N1) monosynaptic basin and that of the V(C4) one (Table 6). At TS2, the most relevant topological change is the merging of the two V(N3) and V (N3) monosynaptic basins into one single V(N3) monosynaptic basin 3.50 e, with respect to the ELF topological characterization of CI. In addition, a slight depopulation of 0.13 e towards the V(N2,N3) disynaptic basin, increasing to 2.01 e also occurred. It was also found that the population of the V(C5) monosynaptic basin, which was decreased at TS1, has decreased to 3.00 e. Finally, at AT, a new disynaptic basin V(N3,C5) was observed integrating 2.20 e. The creation of this disynaptic basin, which is associated to the new N3-C5 single bond, was accompanied by a notable depopulation of the V(N3) by 2.20 e and the increasing of the total population of V(C5) monosynaptic basins by 1.00 e. These results suggest that the creation of the V(N3,C5) disynaptic basin took place by displacement of part of the electron populations of both V(N3) to V(C5) monosynaptic basins. On the other hand, the two V(N1,N2) and V(N2,N3) disynaptic basins in the azide framework have decreased to less than 2 e and the V(N2) monosynaptic basins integrates 3.34 e. All these results show that the ionic nature of all chemical intermediates rules out any covalent interaction involving the copper(I)-complexes, as supported by the ELF topological analysis, reaffirming the zw-type mechanism of the Cu(I)-abt complex catalyzed AAC reactions.

Materials
All chemicals and reagents employed in this work were purchased from Sigma Aldrich and all organic solvents were used without additional purification. The reactions were monitored by thin-layer chromatography (TLC) carried out on commercial glass-backed TLC plates (aluminum plates coated with silica gel 60 F254, 0.25 mm thickness, Merck).
FT-IR spectra were performed on a FTIR spectrophotometer (FT-IR Nicolet 5700, ThermoFischer, Madrid, Spain). The NMR spectra were recorded on a BRUKER DRX-300 AVANCE spectrometer in CDCl 3 using the tetramethylsilane (TMS) as an internal standard.

Synthesis of [Cu(abt) 2 (OOCCH 3 ) 2 ] (1)
In a typical experiment, 0.10 g of abt was dissolved in 10 mL of methanol then 5 mL of Cu(OAc) 2 water solution (2 × 10 −2 M) was added slowly under continuous stirring at room temperature. The solution took a green color during the addition of the copper(II) acetate solution. A precipitate was formed and the mixture was stirred for 12 h at room temperature. The solid was filtered, washed with water and dried at 40 • C overnight to afford complex 1 as a green solid (Yield: 88%).

General Procedure for the CuAAC Reaction
Azides (0.6 mmol), alkynes (0.5 mmol) and complex 1 (2 mol%) were mixed with water (3 mL) and stirred at room temperature. After completion of the catalytic cycloaddition reaction, as confirmed by TLC analysis, the reaction mixture was diluted with diethyl ether and extracted three times. The combined organic layers were then concentrated under reduced pressure to give the corresponding triazole derivatives that were purified by simple recrystallization, if needed.

Computational Methods
All computational calculations were conducted using the Gaussian 09 program [58]. Optimized geometries were performed with B3LYP [59,60], while the LANL2DZ basis set was used for the Cu atom [61] and the 6-31G(d,p) basis set for the other atoms. Frequency calculations were performed to verify the stationary points to be real minima and each TS had only one single imaginary frequency at the same level. The solvent effects were performed with water as solvent using the polarizable continuum model (CPCM). All charge distributions were calculated by natural bond orbital (NBO) analysis [62]. All distances and energies are given in Å and kcal/mol, respectively.
The global electrophilicity (ω) and nucleophilicity (N) indexes were measured at the same level, and are given by the following simple expressions: ω = µ 2 /2η and N = E H − E H (TCE) [63,64]. The electronic chemical potential (µ) and chemical hardness (η) were approached in terms of the one-electron energies of the frontier molecular orbitals HOMO and LUMO, E H and E L , using the expressions µ = (E H + E L )/2 and η = (E L − E H ), respectively.
Both the electronic chemical potential (µ) and chemical hardness (η) may be further approached in terms of the one-electron energies of the frontier molecular orbitals HOMO and LUMO, E H and E L , using the expressions µ = (E H + E L )/2 and η = (E L − E H ), respectively.
Recently, Domingo et al. introduced a new local reactivity index for the Parr functions, namely, local electrophilic (P k + ), and nucleophilic (P k − ), obtained from the analysis of the Mulliken atomic spin density (ASD) of the corresponding reagents [65]. The ELF analysis was also carried out by means of Multiwfn software [66].

Conclusions
In summary, a copper(II) complex containing the heterocyclic 2-aminobenzothiazole (abt) as ligand was prepared and used as a precatalyst for the regioselective synthesis of the corresponding 1,4-disubstituted 1,2,3-triazole under CuAAC strict click chemistry regime. A variety of azides and alkynes were employed in the CuAAC, affording the desired products, which were easily separated in excellent yields. In addition, the 32CA reaction between 5 and 4 was chosen as reaction model and the CuAAC mechanism was studied at the B3LYP/6-31G(d,p) (LANL2DZ for Cu) computational level, choosing the abt as a ligand. The experimental findings are correctly explained by means of the local reactivity indexes obtained from the Parr functions, for regioselectivity synthesis of 1,4-triazoles. The reaction of Ac and 5 follows a two-step mechanism and an easy reaction in terms of energetics, which is in good agreement with the experimental observations. Importantly, the ionic nature of the starting Ac excludes any covalent interaction involving the copper(I) species throughout the reaction, as supported by the ELF topological analysis, indicating the zw-type pathway of the mechanism of CuAAC reactions, using abt as a ligand.