Reaction mechanism of ruthenium-catalyzed azide

0 downloads 0 Views 1MB Size Report
Istanbul Technical University, Faculty of Science and Letters, Department of Chemistry, Maslak, Istanbul 34469, .... are relative free energies with respect to the lowest energy ..... [4] C.F.T. Joan, H. Elizabeth, M. Beatrice, P.B. Daniel, Antimicrob.
Journal of Organometallic Chemistry 724 (2013) 167e176

Contents lists available at SciVerse ScienceDirect

Journal of Organometallic Chemistry journal homepage: www.elsevier.com/locate/jorganchem

Reaction mechanism of ruthenium-catalyzed azideealkyne cycloaddition reaction: A DFT study Esra Boz, Nurcan S¸. Tüzün* Istanbul Technical University, Faculty of Science and Letters, Department of Chemistry, Maslak, Istanbul 34469, Turkey

a r t i c l e i n f o

a b s t r a c t

Article history: Received 22 August 2012 Received in revised form 5 November 2012 Accepted 8 November 2012

The ruthenium catalyzed reaction of alkynes with azides (RuAAC) have enabled a facile route to obtain substituted triazoles, in contrast to thermal 1,3-dipolar cycloaddition which requires high activation barrier and is slow with limited regioselectivity. In this study, ruthenium catalyzed azideealkyne cycloaddition reaction mechanism has been modeled by quantum mechanical methods. The reactions of benzyl azide and substituted alkynes have been modeled by following model mechanism with quantum mechanical calculations at the B3LYP/6-31G* level of theory with LANL2DZ on Ru. The reaction has been investigated for terminal and internal alkynes, with Cp (cyclopentadiene) and Cp* (pentamethylcyclopentadiene) ligand and compared with the experimental results. The calculations in this study have reproduced the experimental regioselectivity and allowed us to account on the electronic and steric effects. Both thermodynamic and kinetic parameters appeared to be important in these reactions. The thermodynamic stability of the Rueazideealkyne complexes and the relative ease of the complex to undergo reaction determined the product distribution. Ó 2012 Elsevier B.V. All rights reserved.

Keywords: Click chemistry Catalysis Ruthenium Cycloaddition Reaction mechanism DFT

1. Introduction 1,2,3-triazoles are the members of the azole family, which have been widely used especially in biomedical applications including treatment of tumors [1,2], HIV [3], allergy [4], fungal infection [5,6] and microbial diseases [7e11]. One of the synthetic methods to assemble 1,2,3-triazoles is the Huisgen 1,3-dipolar cycloaddition of azides and alkynes (AAC) [12,13]. This reaction has a high activation barrier, although it is exothermic and consequently the reaction rate is very low even at high temperatures. With unsymmetrical alkynes, a regioisomeric 1:1 ratio of 1,4- and 1,5-disubstituted-1,2,3-triazoles is formed (Fig. 1) via Huisgen cycloaddition [14,15]. In 2001, the coppercatalyzed 1,3-dipolar azideealkyne cycloaddition (CuAAC) process has emerged as the premier example of click chemistry by Sharpless to describe a set of ‘near-perfect’ bond-forming reactions, useful for rapid assembly of molecules with desired function [16]. The main advantages of this reaction are reported independently by Sharpless [17] and Meldal [18]. Click transformations give rise to products with very high reaction rates in very high yields with little or no byproducts and are easily performed. The process

* Corresponding author. Tel.: þ90 212 285 3163; fax: þ90 212 285 6386. E-mail address: [email protected] (N.S¸. Tüzün). 0022-328X/$ e see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jorganchem.2012.11.011

proceeds well in many protic and aprotic solvents, including water [19]. Additionally, internal alkynes are not active in CuAAC reactions, because stepwise catalytic cycle is thought to begin with deprotonation of the alkyne to form CuI acetylide [17,18,20]. The success of the copper catalyst has invoked the scientists to work on different metal analogs. Several metals which are effective in alkyne transformation reactions have been tested, including different transition metals, Ag(I), Pd(0/II), Pt(II), Au(I/III) and Hg(II), however, reasonable yield, rate increase or selectivity have not been observed [19]. The catalytic effect of ruthenium complexes on alkynes is known, so several ruthenium complexes have been searched for a catalyst in azideealkyne cycloaddition [21e23]. In 2005, ruthenium catalysis was found as a complementary reaction to CuAAC which has a different regioselectivity. By using Ru catalyst, 1,5disubstituted triazoles are synthesized from both terminal and internal alkynes. It has been shown by the experiments that the catalytic activity and regioselectivity of the Ru complexes differ with the ligands around the central metal atom [24]. For example, in cycloaddition of benzyl azide and phenylacetylene in THF, an acetate complex ligand like Ru(OAc)2(PPh3)2 gives completely 1,4-disubstituted triazole product with a yield of 46%. RuCl2(PPh3)3 and RuHCl(CO)(PPh3)3 complexes have also given 1,4-disubstituted triazoles with less than 5% yields. Similarly, they produce trace amount of

168

E. Boz, N.S¸. Tüzün / Journal of Organometallic Chemistry 724 (2013) 167e176

Ru-azideealkyne starting complex, calculated at 298 K, unless otherwise stated. The cartesian coordinates of all discussed structures are presented in Supporting information. For single point solvent calculations Tomassi’s Polarized Continuum Model (PCM) has been used in tetrahydrofuran (THF) [37,38]. For comparative purposes, the azideealkyne cycloaddition without Ru catalysis have also been modeled for all studied sets. The details are presented in the Supporting information. 3. Results and discussion

Fig. 1. 1,3-dipolar cycloaddition of azides and alkynes.

1,5-disubstituted triazoles or none. When cyclopentadiene (Cp) ligand is used with Ru, a mixture of 1,4 and 1,5-disubstituted triazoles can be obtained if the reaction takes place [25]. Switching Cp component to Cp* (the pentamethyl analog of Cp) resulted 100% conversion to 1,5-disubstituted triazole [25]. Adding several ligands to [Cp*Ru] complex gives similar results to obtain 1,5-disubstituted triazoles [24]. The aim of this study is to elucidate the mechanism of Ru catalyzed azideealkyne reaction and to explain the effect of the catalyst on the mechanism. With this aim, a set of Ru-catalyzed azideealkyne reaction has been studied by quantum mechanical methods (Fig. 2). In the literature, a reaction mechanism has been proposed (Fig. 3) and tested with calculations on a model reaction, where propyne and methyl azide have reacted in the presence of CpRuCl(PPh3)2 [24]. Later, another computational and experimental study has concentrated on the reaction of terminal alkynes with azides in the presence of 16-electron ruthenium complexes [26]. In this study, the mechanism will be studied with both Cp and Cp* ligands with terminal and internal alkynes, considering the possible reaction pathways. These efforts will enable to shed light on the reaction mechanism in detail, which is important in the sense that finding the reaction mechanism will enable one to play with the effects and lead to obtain tailor-made products. 2. Methodology In the context of this study, all the possible reaction mechanisms, including intermediates and transition states, have been modeled and discussed in terms of relative energies obtained from quantum-mechanical calculations. The DFT method employing the B3LYP functional with the 6-31G* basis set has been used to carry out the full optimization of the compounds of interest in the gas phase with the G09 package [27e30]. For Ru, LANL2DZ effective core potential has been used. In the literature it is stated that this methodology gives successful results for Ru metal and the DFT methodology with the B3LYP functional has been shown to give reliable results in transition metals, including Ru-catalyzed reactions [31e34]. The stationary points were analyzed by vibrational frequency calculations. All transition states were verified to be saddle points by one imaginary frequency belonging to the reaction coordinate. For all transition state structures the intrinsic reaction coordinate (IRC) was followed to validate the expected reactants and products [35,36]. The energies discussed in the manuscript are relative free energies with respect to the lowest energy

The cycloaddition reaction of azide and alkyne with and without Ru catalyst has been modeled by the DFT calculations to elucidate the RuAAC reaction mechanism, the effect of ligands and the substrate on the regioselectivity. In the first part of this study, a model reaction has been studied in order to eliminate the steric effects and to solely concentrate on the electronic properties of the reaction mechanism. For this purpose, ruthenium catalyzed reaction of methyl-azide and propyne (Fig. 2, entry 1) has been modeled. In the second part of the study, the proposed mechanism from the model study has been applied to another set of experimental work; the reaction of benzyl azide with phenylacetylene (Fig. 2, entry 2). In the third part, Cp and Cp* ligated ruthenium catalysts have been used to elucidate the effect of methyl groups on the cyclopentadiene ring (Fig. 2, entry 3). In the last part, the reaction has been modeled for internal-alkynes with different functional groups studied experimentally by Majireck et al. [39] (Fig. 2, entries 4e6). Both internal and terminal alkynes can reveal triazoles via Ru catalysis unlike the case with copper, which invoked the idea of a pcomplex between the Ru metal and the olefinic bond. It has been reported that the RuAAC prevents oligomerization of alkynes which has been shown to take place via ruthenacyclopentadiene intermediate [24]. The reaction of (2-azidoethyl)benzene with Cp*RuCl(COD) has given inactive tetraazadiene complex. By adding only azide reactant will result giving hydrobenzamide, benzyl-benzylideneamine, benzonitrile products by using different Ru catalysts [40]. In a recent computational and experimental study, the order of adding reagents has also been investigated in detail [26]. These ideas led to the finding that “the azide should not be added to the catalyst before the alkyne” [24]. Thus, the mechanism is assumed to start with the catalystazideealkyne complex. At the first step of mechanism shown in Fig. 3, spectator ligands are displaced and a Rueazideealkyne complex is produced. This step has been found to be exergonic by 15.1 kcal/mol by considering tri-phenylphosphines as ligands in our calculations. In the Rueazideealkyne complex, there are two sites for Ru to bind to the nitrogen atom: either from the secondary (A and B, Fig. 4) or from the primary nitrogen (C and D, Fig. 4). It is stated in the literature that bonding can take from both sides, although bonding proximal to carbon is a more commonly observed bonding site [25]. The alkyne group can coordinate to the metal center in two orientations; the functional group directed toward the azide group or away from the azide group. In this study, these four configurations named AeD have been modeled for their all possible conformers (Fig. 4) and the reaction mechanisms starting with the global minima of each type have been studied. Three dimensional structures have been demonstrated for only one of the studied sets (for Fig. 2, entry 2) but the cartesian coordinates of all structures are presented in the Supporting information. In the proposed catalytic cycle (Fig. 3), at step A, spectator ligands are displaced and the azideealkyne groups are coordinated to the Ru catalyst. After this step, terminal nitrogen atom on the azide group attacks to the alkyne and a ruthenacycle intermediate is formed. At step C, alkyne group is separated from the Ru metal

E. Boz, N.S¸. Tüzün / Journal of Organometallic Chemistry 724 (2013) 167e176

Entry 1 2 3 4 5 6

169

Ligand Cp*

R1 H

R2 CH3

R3 CH3

I:II

Cp* Cp Cp* Cp* Cp*

H H Ph Ph Ph

Ph Ph CH2OH COMe CH3

CH2Ph CH2Ph CH2Ph CH2Ph CH2Ph

0:100a 15:85a 0:100 b 100:0 b 38:62b

#

Fig. 2. Modeled reactions and their experimental products (#experimental data is not present, aRef. [24], bRef. [39]).

and bonded to the nitrogen atom to form a triazole ring. At the last step, triazole is departed from the catalyst and the catalyst is regenerated. This mechanism has been studied by DFT calculations for model compounds with unmethylated cyclopentadiene groups [24,41] and with methylated cyclopentadiene for the reaction of a terminal alkyne with benzyl azide [26]. In this study the mechanism depicted in Fig. 3 has been followed for all the sets in Fig. 2. In our calculations 5 types of transition state structures have been located for the catalytic cycle. The first one is (TS1, Fig. 5) for forming the ruthenacycle intermediate from Rueazideealkyne complex (Scheme 1). Once the intermediate int1 is formed, it can go through TS2 and form a 6-membered ruthenacycle intermediate (int2) which further transforms to int4 via TS4 or it can directly form a 5-membered triazole (int3) via TS3. Alternatively, the 6membered ruthenacycle (int2) can form directly from the starting Rueazideealkyne complex via TS5 (Scheme 2). Throughout the discussion that follows, these transition state structures and intermediates will be referred to whenever needed.

Fig. 3. Catalytic cycle of the RuAAC reaction [24].

In entry 2 case, the minimum starting complex was found to be structure A (Fig. 4), where Ru bonds from the 2 nitrogen and the phenyl group is directed away from the azide group so that the repulsive steric interactions are avoided (Fig. 6). Reaction path shown in Scheme 1 has been followed by taking A as the starting complex for entry 2. The 3-dimensional structures belonging to the reaction mechanisms of entry 2 are presented in Fig. 6 as a representative of all the studied sets. The terminal nitrogen of azide group attacks the closest alkyne carbon via A_TS1. In this transition state structure, Ru bonds more strongly to the unsaturated carbons; alkyne carbons open up toward a double bond length (1.26  A in A and 1.30  A in A_TS1, Fig. 6) and the terminal carbon and the terminal nitrogen come closer (3.30  A in A and 2.26  A in A_TS1). After A_int1 (Fig. 6) forms, Ru metal will like to exempt from the cyclic structure. It can proceed via A_TS3 to form metaletriazole complex, A_int3. In A_int3, Ru stands out of the triazole cycle and bonds to triazole from one of the nitrogen sides. When nitrogen bonds to phenyl substituted carbon in the transition state A_TS3, it loses its bond with Ru, which forces Ru to coordinate to the other terminal nitrogen. The formation of Ru-triazole (A_int3) is highly exergonic. The triazole product will be formed by the regeneration of Ru catalyst. An alternative path that can take place after A_int1 is isomerization of this complex to a slightly more stable six-membered ruthenacycle intermediate A_int2 via A_TS2. This transformation takes place via a small free energy barrier of 1.8 kcal/mol. A transition state following A_int2 (depicted as A_TS4 in Scheme 1) could not be located to exempt Ru from the ring. The relative energy of A_int1 is very close to that of A_int2 and the barriers are very small in both directions, thus these two structures tend to stay in equilibrium. Product formation which is highly exergonic is expected to take place via A_TS3. The product obtained from this path reveals the experimentally observed 1,5-product. As an alternative path, attack of nitrogen to distal carbon in A geometry, leading to 1,4-product was also considered. The first transition state for this addition requires 17.9 kcal/mol of energy barrier, eliminating it to be a possible path, which is in accordance with the experimental product distribution. In the B configuration of entry 2 (Fig. 4), Ru is bonded to the 2 nitrogen of the azide as in A but the phenyl group is now at the reaction center. The orientation of the phenyl substituent on alkyne causes steric interaction and lowers B’s stability as compared to A (1.1 kcal/mol higher than A). The reaction mechanism starting with B for entry 2, is demonstrated in Scheme 2. The reaction exhibits a different sequence (B/B_TS5/B_int2/B_TS4/B_int4) than that of A. With B, the B_int1 intermediate could not be formed. In this unwanted structure, the phenyl ring of benzyl azide would cause steric interactions with the cyclopentadienyl ring.

170

E. Boz, N.S¸. Tüzün / Journal of Organometallic Chemistry 724 (2013) 167e176

Ru Cl R2

R1

N

Ru

R3 Cl R1

N

N

N

A

Cl R1

R2

R3

Ru N

Ligand

1

Cp*

2

R1 N N

Cl R2

N N

R3

C Entry

N

B

Ru

R3

R2 N

N

D

Functional groups

EA

EB

EC

ED

R1=H, R2=CH3, R3= CH3

0.0

2.5

1.6

1.4

Cp*

R1=H, R2= Ph, R3=CH2Ph

0.0

1.1

0.8

1.9

3

Cp

R1=H, R2= Ph, R3=CH2Ph

0.9

0.2

0.0

2.3

4

Cp*

R1=Ph, R2= CH2OH, R3= CH2Ph

1.4

6.8

5.6

0.0

5

Cp*

R1=Ph, R2= COCH3, R3= CH2Ph

1.0

0.2

0.6

0.0

6

Cp*

R1=Ph, R2= CH3, R3= CH2Ph

0.6

3.1

1.8

0.0

Fig. 4. Modeled Rueazideealkyne complexes and their relative ZPE corrected electronic energies (EA, EB, EC, ED) in kcal/mol.

Additionally, the phenyl substituted carbon atom of alkyne is reluctant to form a strained three-membered ring and undergo a dramatic change in its sp2 hybrid nature via pyramidilization and bonding to Ru, as it was present in A_int1 structure. Thus, the transition state goes directly from B_TS5 to the ruthenacycle intermediate B_int2 in a concerted fashion without formation of B_int1 (Fig. 6). The B_int2 could be transformed to 1,4disubstituted product through B_TS4, in an exergonic way. Following the same geometrical parameters of A for concerted addition of B did not reveal a product. Alternatively, addition of

TS1

TS5

TS2

nitrogen to distal carbon in B geometry, which would lead to 1,5product, have required 3.5 kcal/mol higher barrier. C complex would result in the formation of 1,5-regioisomer and D would give 1,4-regioisomer, if they react at all. In the pathways following C and D, it is found out that the formation of the first intermediates-which were found as the highest points on the free energy diagrams, have very high activation barriers (21.2 kcal/mol and 16.5 kcal/mol for C and D, respectively, Table 1) as compared to that of A and B, thus the paths have not been carried from these configurations.

TS3

Fig. 5. 2D structures of different types of transition states.

TS4

E. Boz, N.S¸. Tüzün / Journal of Organometallic Chemistry 724 (2013) 167e176

171

Cp* R3

Cl

Ru N

R2

Relative Free Energy (Kcal/mol)

R2

Cl

N R1

Cp* Cl

Cp* N

Ru N N

A_TS_1

R3

Cl

6.7 4.1 8.1

R1 N

R2

Ru

N

R3

Ru

N

A_TS_3

N

R1

3.9 -2.3 4.9

N R1 N

R3

N

R2

Cp*

A_TS_4

Cp*

A 0.0 0.0 0.9

R3

Cl Cp* Ru

N N

R2 R1

N

N N

R2 R3

Cl

Ru

R1

N

Cp*

Cp*

A_TS_2 -7.4 -15.1 -9.6

Cl

Ru

Cp*

R3

N

Ru N

Ru

Cl

N

N

R2

R1

A_int_2 -10.5 -17.5 -15.2

-9.2 -17.7 -11.4

Cl

N

R3

N

R1

A_int_1

N

N

R3

N

R2

R2

A_int_3

R1

A_int_4

-63.1 -72.4 -62.1

Reaction Coordinate Scheme 1. Energy profile for the Ru catalyzed azide alkyne cycloaddition of A configuration for entries 2 and 3. (plain text: entry 2: R1 ¼ H, R2 ¼ Ph, R3 ¼ CH2Ph, italic: entry 2 in THF, bold: entry 3: R1 ¼ H, R2 ¼ Ph, R3 ¼ CH2Ph with Cp ligand).

For entry 2, product formation is not expected via BeD configurations, since A is the lowest energy configuration, thus more abundant and requires the lowest activation barrier (13.1 kcal/mol) as compared to that of B (16.1 kcal/mol), C (21.2 kcal/mol) and D (16.5 kcal/mol). This is also consistent with the experimental observation that only 1,5-product is formed in the reaction. The reaction paths for entry 2 have also been considered in THF as solvent (Schemes 1 and 2) The relative energies are

systematically lower than the gas phase results for both starting configurations, intermediates and transition states. However, comparison of energies have shown that THF has not changed the preference for A configuration and the path following it. Compared to the uncatalyzed reaction of this system, there is a dramatic change in the activation energy (30.9 kcal/mol vs 13.1 kcal/mol) as compared to the metal catalyzed 1,5-triazole product formation, in accord with the experimental increase in rate.

Cp* R3 Cl R1

Cp*

Relative Free Energy (Kcal/mol)

Cl

Ru N N N R2

R1

B

1.1 0.6 0.2

R3

Ru N N N R2

Cp* Cl

Ru

N N

B_ TS_5

R1

12.0 8.4 11.0

R3

N R2 B_ TS_ 4

Cp* R3 N Cl Ru N N

R1 R2

1.3 -3.5 3.8 Cp* R1

Cl Ru R2

B_ int_2

-14.8 -19.8 -13.6

R3 N N N

B _int_4

-51.3 -57.2 -48.2

Reaction Coordinate

Scheme 2. Energy profile for the Ru catalyzed azide alkyne cycloaddition of B configuration for entries 2 and 3. (plain text: entry 2: R1 ¼ H, R2 ¼ Ph, R3 ¼ CH2Ph, italic: entry 2 in THF, bold: entry 3: R1 ¼ H, R2 ¼ Ph, R3 ¼ CH2Ph with Cp ligand).

172

E. Boz, N.S¸. Tüzün / Journal of Organometallic Chemistry 724 (2013) 167e176

Fig. 6. 3D geometries of the selected structures of pathways A, B, C and D.

E. Boz, N.S¸. Tüzün / Journal of Organometallic Chemistry 724 (2013) 167e176

Fig. 6. (continued).

173

174

E. Boz, N.S¸. Tüzün / Journal of Organometallic Chemistry 724 (2013) 167e176

Table 1 The relative free energies of the Rueazideealkyne complexes C and D, and the transition states following C and D configurations, with respect to the lowest energy Rueazideealkyne complex. The numbers in parenthesis refer to solvent calculations in THF. Entry

EC

TS1

ED

TS1

1 2 3 4 5 6

1.6 0.8 (1.2) 0.0 5.6 0.6 1.8

23.2 22.0 (17.4) 23.2 26.0 26.8 22.8

1.4 1.9 (2.3) 2.3 0.0 0.0 0.0

17.9 18.4 (13.4) 21.4 26.6 25.1 20.3

the experimental 85% of 1,5-product formation. The B configuration that yields the experimental 1,4-product, has a significant abundancy and a slightly higher barrier (17.4 kcal/mol) which is still competitive to that of A. The product distribution of 1,5:1,4triazoles, 85:15 ratio can be attributed to the abundancy of the complexes (A and B) and the barriers, which lead to different regioselectivities. The low percent of A and B configurations that can lead to products and the int_2 intermediate behaving as a dead end for the reaction are consistent with the experimental low yield (50%) of the reaction. 3.2. The reactions of internal alkynes

3.1. Cp vs Cp* ligands In the literature, the reaction between propyne and methyl azide with [CpRuCl(PPh3)2] catalyst has been modeled [24]. In this study, the same reaction set has been modeled by using Cp* ring on the catalyst (Fig. 2, entry 1) since the Cp* ligand has been reported to be vitally important for the reaction to proceed and show regioselectivity if it does react at all [24]. To have a more realistic picture and comparison with the experimental results, the reaction coordinate of the benzyl azide and phenylacetylene reaction with Cp and Cp* ligands on Ru catalyst has been modeled and discussed (Fig. 2, entries 3 and 2, respectively). The reaction of benzyl azide and phenylacetylene with Cp ligand gives a mixture of 1,5 and 1,4triazoles in 85:15 ratio with 50% yield and with Cp* ligand 1,5 is the only product with a yield of 100%. The relative energies of the starting complexes with Cp and Cp* ligands show different relative energy orders (Fig. 4). The minimum energy structure with the Cp catalyst is found to be the C configuration which leads to 1,5triazole. However, the first activation barrier for transformation from this structure requires 23.2 kcal/mol (Table 1). The following structure is B (Fig. 4) with almost the same relative energy and results giving a 1,4-triazole product via 17.4 kcal/mol of barrier (Scheme 2). Product formation via A proceeds through A_int1, which can follow two paths to product formation. From A_int1, the reaction can proceed via A_TS3 and form Ru complexed 5membered triazole ring or it can go over A_TS2 transition geometry and form 6-membered ruthenacycle. However, a transition state following the 6-membered ruthenacycle (A_int2) could not be located. The A_int1 and A_int2 intermediates were reported to be important structures for the model study with Cp in earlier literature [24]. In our calculations with Cp*, these two intermediates have closer energies as compared to their non-methylated analogs, implying an equilibrium structure between the two intermediates. With Cp (entry 3), the energy difference between the intermediates A_int1 and A_int2 (3.8 kcal/mol) are higher than that of Cp* (Entry 2) (1.3 kcal/mol in vacuum and 0.2 kcal/mol in THF) (Scheme 1). It shows that methyl substitution on cyclopentadiene causes int1 and int2 intermediates to come closer in energy so that the 6membered ruthenacycle A_int2 does not behave as a dead end for the system and enables the reaction to proceed back from A_int1 via A_TS3 with a high yield. In the case of Cp, the equilibrium can be hardly shifted back to A_int1 and the reaction barrier is 3.2 kcal/mol higher than its Cp* analog. The minimum structure for entry 3 is C, followed by B, A and D. According to Boltzmann distribution of these four structures, the ratio of configurations in the reaction medium is A:B:C:D, 13:38:48:1. The activation barriers of first transition states for the oxidative coupling step for C and D structures are relatively high with 23.2 and 19.1 kcal/mol of energy, similar to their Cp* analogs. Thus, production of triazole is expected through the A and B structures. Considering the abundancy of A and its lower barrier (16.3 kcal/mol), this is the most favorable pathway, consistent with

As a part of this study, RuAAC reaction with different internal alkynes, involving hydroxyl, acetyl and alkyl groups (entries 4e6, Fig. 2), giving 1,4,5-trisubstituted triazoles with high regioselectivity were studied. With internal alkynes, the energies of the configurations AeD become more complex due to the steric and electronic effect of substituents. In all the internal alkynes modeled in this study, D configuration is the lowest energy structure, unlike the case with terminal alkynes. In the D geometry, the azide is coordinated to Ru from its primary nitrogen and the functional groups on alkynes are directed away from the azide group. However, reaction taking place from this configuration (and also from C configuration) requires higher energies for even the first step in comparison to others since the sterically hindered addition requires bonding of the substituted olefinic carbon to substituted 2 nitrogen of azide. In the addition reaction of 3-phenylprop-2-yn-1-ol with benzyl azide (entry 4, Fig. 4), all the configurations’ relative energies, except that of A, are out of the error limits of methodology employed in this study. Although the reaction barrier from B configuration (12.3 kcal/mol, Scheme 4) has got the lowest energy of the whole set, product formation from B and C is not expected since these structures will not be present in the reaction medium due to their high relative energies. Reaction path following D requires the highest free energy of activation among the AeD configurations (Schemes 3 and 4, Table 1). Thus, product formation can be expected to take place from A structure, since it has a facile reaction path (DGz ¼ 13.9 kcal/mol, Scheme 3) and is present in the reaction medium, consistent with the experimental observation that 1,5-product forms exclusively. The reaction path for internal alkynes is different than the terminal alkynes modeled in this study. Starting from A, A_TS5 is found as the highest energy point on reaction coordinate and reveals the ruthenacycle intermediate A_int2 in a concerted fashion. The metal exempts itself from the cyclic structure via A_TS4 and forms the expected product. When the internal alkyne is present, the sterically hindered cyclic structure in A_int1 can hardly form due to the R1 substituent present at the bonding site. In entry 5, the reaction of eCOCH3 functional group substituted alkyne, namely 4-phenylbut-3-yn-2-one and benzyl azide has been considered and experimentally it only gives product I contrary to entry 4. The relative electronic energies of the possible configurations are all within the error limit of B3LYP and the lowest energy structure is D, followed by B which is only 0.2 kcal/mol higher in energy. C and D configurations are ruled out as in other sets since these configurations, require high activation barriers even for the first transition states (26.2 and 25.1 kcal/mol, respectively), again higher than that of the A and B, due to the attack of the secondary nitrogen to the substituted alkyne carbon. In the path following B (Scheme 4), the terminal nitrogen of azide group attacks to the closest alkyne carbon via B_TS1 and B_int1 intermediate is formed which enables Ru-triazole to form in a highly exergonic way (57.1 kcal/mol) via B_TS3. The low relative

E. Boz, N.S¸. Tüzün / Journal of Organometallic Chemistry 724 (2013) 167e176

175

Cp* R3 Cl R1

Cp*

Relative Free Energy (Kcal/mol)

Cl

Ru N N

R3

Ru N N N R2 Cl

Ru

A_TS_5

R2

N

R3

N R1

11.5

N

R1

Cp*

N R2

11.6 9.7

A_ TS_ 4

A

Cp*

1.4

1.0 0.6

Cl Ru

6.7 8.6 6.7

R3 N

R1

N N

Cp*

R2

R1

Cl Ru

R3

R2

N N N

A_ int_2 -7.2

A_ int_4

-7.8 -7.6

-51.4

-50.6 -51.1 Reaction Coordinate

Scheme 3. Energy profile for the Ru catalyzed azide and alkyne cycloaddition of A configuration for entries 4e6. (plain text: entry 4: R1 ¼ Ph, R2 ¼ CH2OH, R3 ¼ CH2Ph, italic: entry 5: R1 ¼ Ph, R2 ¼ COCH3, R3 ¼ CH2Ph, bold: entry 6: R1 ¼ Ph, R2 ¼ CH3, R3 ¼ CH2Ph).

Cp* R3

Cl

Ru N

R1

N Cp*

N

R2

Relative Free Energy (Kcal/mol)

Cp*

Cl R1

Ru N N R 2N

R3

Cl 12.0 10.0 11.1

R1

R3

N N

R2 N

B_TS_3

B 6.8 0.2 3.1

Ru

B_TS_1

Cp* R3

Cl

Ru

N

9.3 6.2 8.4 Cp*

N

R1 R2

N

Cl

Ru N

N N

B_int_1 -3.0 -7.5 -4.0

R2

R3

R1

B_int_3 -61.2 -56.9 -61.8

Reaction Coordinate Scheme 4. Energy profile for the Ru catalyzed azide and alkyne cycloaddition of B configuration for entries 4e6. (plain text: entry 4: R1 ¼ Ph, R2 ¼ CH2OH, R3 ¼ CH2Ph, italic: entry 5: R1 ¼ Ph, R2 ¼ COCH3, R3 ¼ CH2Ph, bold: entry 6: R1 ¼ Ph, R2 ¼ CH3, R3 ¼ CH2Ph).

energy of B and the reaction path following B having the lowest barrier (13.7 kcal/mol in B vs 16.4 kcal/mol in A) enables formation of only product I which is in accord with the experimental observation. In entry 6 (Fig. 4), the alkyl group substituted alkyne and benzyl azide complex reveal D as the minimum energy configuration. However, the reaction barriers following this configuration and C as well, are high as in the previous sets. Although B having the lowest barrier (12.4 kcal/mol) of all configurations enables formation of product I, the relatively high energy of B, the higher abundancy and

the low reaction barrier (14.3 kcal/mol) of A configuration is consistent with the formation of both products I and II where the major product is coming from A. 4. Conclusion In this study Ru catalyzed azideealkyne reaction was modeled to elucidate the structural and electronic effects on the mechanism. Reaction was proposed to start with the Ruealkyneeazide complex which has 4 possible configurations. Among those configurations,

176

E. Boz, N.S¸. Tüzün / Journal of Organometallic Chemistry 724 (2013) 167e176

with the terminal alkynes, in the most preferred configuration, the azide binds to Ru from its secondary nitrogen and the alkyne substituent is directed away from the azide group (A type). With internal alkynes, D configuration is preferred where the azide is binded to Ru from the primary nitrogen and the functional group is on Cl side, allowing H-bonding interaction with the Cl ligand. The reactions following C and D configurations ended up with significantly higher reaction barriers as compared to that of A and B due to the steric effects stemming from the attack of the substituted alkyne carbon to substituted nitrogen. With terminal alkynes, reaction barriers following A configuration, enabling the least steric effects, are the lowest while in the case of internal alkynes modeled in this study, attack taking place from the functional group substituted site (B) configuration have lower barriers. With the internal alkynes, the reaction taking place from the phenyl substituted site (A configuration) requires higher energy since the attack enforces the delocalization from the alkyl carbons to be destroyed and the pyramidilization of the alkyne carbon. In internal alkynes, H-bonding interactions of the ligand Cl with H-bonding functional group on alkyne become important. The barrier of the reaction may become insignificant if the relative population of its starting configuration is low. For example, although the attack of unsubstituted nitrogen to eCH2OH substituted carbon required the lowest energies, the starting configuration has got too high relative energy. Thus, in the case of A and B configurations, there is a complex interplay of their relative energies and the barriers following them. The effect of methyl groups on cyclopentadiene ligand has also been investigated. The 85:15 product distribution of 1,5:1,4-triazoles with Cp on Ru, can be attributed to the similar abundancy of the complexes (A and B) which lead to different regioselectivities. The low percent of A and B configurations that can lead to products I and II, respectively and the int2 intermediate behaving as a dead end for the reaction with Cp are consistent with the experimental low yield (50%) of the reaction. The solvent calculations have not shown significant change in preferences although some of the structures have become relatively more or less stable in THF. It can be concluded that the regioselectivity is governed by the relative stability of the configurations that has formed and the relative ease of the reactions following these structures. Computed regioselectivity agreed well with the experimental product distribution, indicating the efficiency of the performance of the computational methodology employed. Acknowledgments The authors gratefully acknowledge National Center for High Performance Computing (Grant number: 10732009) and TUBITAK ULAKBIM, High Performance and Grid Computing Center (TRUBA Resources) for computer resources provided. Appendix A. Supplementary material Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jorganchem.2012.11.011. References [1] R. Alvarez, S. Velazquez, F. San, S. Aquaro, C. De, C.F. Perno, A. Karlsson, J. Balzarini, M.J. Camarasa, J. Med. Chem. 37 (1994) 4185e4194.

[2] F. da Silva, C.B.V. de Souza, I.I.P. Frugulhetti, H.C. Castro, A.M.T. Souza, P.A. Abreu, F. Passamani, C.R. Rodrigues, V.F. Ferreira, Eur. J. Med. Chem. 44 (2009) 373e383. [3] D.R. Buckle, C.J.M. Rockell, H. Smith, B.A. Spicer, J. Med. Chem. 29 (1986) 2262e2267. [4] C.F.T. Joan, H. Elizabeth, M. Beatrice, P.B. Daniel, Antimicrob. Agents Chemother. 42 (1998) 313e318. [5] M.J. Genin, D.A. Allwine, D.J. Anderson, M.R. Barbachyn, D.E. Emmert, S.A. Garmon, D.R. Graber, K.C. Grega, J.B. Hester, D.K. Hutchinson, J. Morris, R.J. Reischer, C.W. Ford, G.E. Zurenko, J.C. Hamel, R.D. Schaadt, D. Stapert, B.H. Yagi, J. Med. Chem. 43 (2000) 953e970. [6] R. Périon, V. Ferrières, M.I. García- Moreno, C.O. Mellet, R. Duval, J.M.G. Fernández, D. Plusquellec, Tetrahedron 61 (2005) 9118e9128. [7] J. Shen, R. Woodward, J.P. Kedenburg, X.-W. Liu, M. Chen, L.-Y. Fang, D.-X. Sun, P.-G. Wang, J. Med. Chem. 51 (2009) 7417e7427. [8] D.-R. Hou, S. Alam, T.C. Kuan, M. Ramanathan, T.-P. Lin, M.-S. Huang, Bioorg. Med. Chem. Lett. 19 (2009) 1022e1025. [9] A.K. Jordã, V.F. Ferreira, E.S. Lima, M.C.B.V. de Souza, E.C.L. Carlos, H.C. Castro, R.B. Geraldo, C.R. Rodrigues, M.C.B. Almeida, A.C. Cunha, Bioorg. Med. Chem. 17 (2009) 3713e3719. [10] A.M. Thompson, A. Blaser, R.F. Anderson, S.S. Shinde, S.G. Franzblau, Z.-K. Ma, W.A. Denny, B.D. Palmer, J. Med. Chem. 52 (2009) 637e645. [11] M. Hoshino, Nature 186 (1960) 174e175. [12] K.V. Gothelf, K.A. Jorgensen, Chem. Rev. 98 (1998) 863e909. [13] J. Mulzer, Organic Synthesis Highlights (1991) 77e95. [14] J. Bastide, O. Henri-Rousseau, Bull. Soc. Chim. Fr. (1973) 2294e2296. [15] D. Clarke, R.W. Mares, H. McNab, J. Chem. Soc. Perkin. Trans. 1 (1997) 1799e 1804. [16] H.C. Kolb, M.G. Finn, K.B. Sharpless, Angew. Chem. Int. Ed. 40 (2001) 2004e 2021. [17] V.V. Rostovtsev, L.G. Green, V.V. Fokin, K.B. Sharpless, Angew. Chem. Int. Ed. 41 (2002) 2596e2599. [18] C.W. Tornøe, C. Christensen, M. Meldal, J. Org. Chem. 67 (2002) 3057e3064. [19] J.E. Hein, V.V. Fokin, Chem. Soc. Rev. 39 (2010) 1302e1315. [20] N.S¸. Tüzün, C. Özen, J. Mol. Graph. Model. 34 (2012) 101e107. [21] T. Naota, H. Takaya, S.I. Murahashi, Chem. Rev. 98 (1998) 2599e2660. [22] C. Bruneau, P.H. Dixneuf, Acc. Chem. Res. 32 (1999) 311e323. [23] B.M. Trost, F.D. Toste, A.B. Pinkerton, Chem. Rev. 101 (2001) 2067e2096. [24] B.C. Boren, S. Narayan, L.K. Rasmussen, L. Zhang, H. Zhao, Z. Lin, G. Jia, V. Fokin, J. Am. Chem. Soc. 130 (2008) 8923e8930. [25] L. Zhang, X. Chen, P. Xue, H.H.Y. Sun, I.D. Williams, K.B. Sharpless, V.V. Fokin, G. Jia, J. Am. Chem. Soc. 127 (2005) 15998e15999. [26] M. Lamberti, G.C. Fortman, A. Poater, J. Broggi, A.M.Z. Slawin, L. Cavallo, S.P. Nolan, Organometallics 31 (2012) 756e767. [27] C.T. Lee, W.T. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785e789. [28] A.D. Becke, J. Chem. Phys. 98 (1993) 5648e5652. [29] B. Miehlich, A. Savin, H. Stoll, H. Preuss, Chem. Phys. Lett. 157 (1989) 200e206. [30] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J.A. Montgomery Jr., J.E. Peralta, F. Ogliaro, M. Bearpark, J.J. Heyd, E. Brothers, K.N. Kudin, V.N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J.M. Millam, M. Klene, J.E. Knox, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, R.L. Martin, K. Morokuma, V.G. Zakrzewski, G.A. Voth, P. Salvador, J.J. Dannenberg, S. Dapprich, A.D. Daniels, O. Farkas, J.B. Foresman, J.V. Ortiz, J. Cioslowski, D.J. Fox, Gaussian Inc., Wallingford, CT, 2009. [31] S.Q. Niu, M.B. Hall, Chem. Rev. 100 (2000) 353e406. [32] M. Diedenhofen, T. Wagener, G. Frenking, The accuracy of quantum mechanical methods for the calculation of transition metal compounds, in: T. Cundari (Ed.), Computational Organometallic Chemistry, Marcel Dekker, New York, 2001, pp. 69e122. [33] N.E. Schultz, Y. Zhao, D.G. Truhlar, J. Phys. Chem. 109 (2005) 4388e4403. [34] F. Neese, J. Chem. Phys. 118 (2003) 3939e3948. [35] K. Fukui, J. Phys. Chem. 74 (1970) 4161e4163. [36] K. Fukui, Acc. Chem. Res. 14 (1981) 363e368. [37] J. Tomasi, B. Mennucci, E. Cancés, J. Mol. Struct. (THEOCHEM) 464 (1999) 211e226. [38] M. Cossi, G. Scalmani, N. Rega, V. Barone, J. Chem. Phys. 117 (2002) 43e54. [39] M.M. Majireck, S.M. Weinreb, J. Org. Chem. 71 (2006) 8680e8683. [40] J. Risse, R. Scopelliti, K. Severin, Organometallics 30 (2011) 3412e3418. [41] D.R. Hou, T.C. Kuan, Y.K. Li, R. Lee, K.W. Huang, Tetrahedron 66 (2010) 9415e 9420.