Catalyzed M–C coupling reactions in the synthesis of σ-(pyridylethynyl)dicarbonylcyclopentadienyliron complexes

The reactions between terminal ethynylpyridines, (trimethylsilyl)ethynylpyridines and cyclopentadienyliron dicarbonyl iodide were studied under Pd/Cu-catalyzed conditions to develop a synthetic approach to the σ-alkynyl iron complexes Cp(CO)2Fe–C 
<svg xmlns="http://www.w3.org/2000/svg" version="1.0" width="23.636364pt" height="16.000000pt" viewBox="0 0 23.636364 16.000000" preserveAspectRatio="xMidYMid meet"><metadata>
Created by potrace 1.16, written by Peter Selinger 2001-2019
</metadata><g transform="translate(1.000000,15.000000) scale(0.015909,-0.015909)" fill="currentColor" stroke="none"><path d="M80 600 l0 -40 600 0 600 0 0 40 0 40 -600 0 -600 0 0 -40z M80 440 l0 -40 600 0 600 0 0 40 0 40 -600 0 -600 0 0 -40z M80 280 l0 -40 600 0 600 0 0 40 0 40 -600 0 -600 0 0 -40z"/></g></svg>
 C–R (R = ortho-, meta-, para-pyridyl). Depending on the catalyst and reagents used, the yields of the desired σ-pyridylethynyl complexes varied from 40 to 95%. In some cases the reactions with ortho-ethynylpyridine gave as byproduct the unexpected binuclear FePd μ-pyridylvinylidene complex [Cp(CO)Fe{μ2-η1(Cα):η1(Cα)-κ1(N)-Cα 
<svg xmlns="http://www.w3.org/2000/svg" version="1.0" width="13.200000pt" height="16.000000pt" viewBox="0 0 13.200000 16.000000" preserveAspectRatio="xMidYMid meet"><metadata>
Created by potrace 1.16, written by Peter Selinger 2001-2019
</metadata><g transform="translate(1.000000,15.000000) scale(0.017500,-0.017500)" fill="currentColor" stroke="none"><path d="M0 440 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z M0 280 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z"/></g></svg>
 Cβ(H)(o-C5H4N)}(μ-CO)PdI]. The conditions, catalysts, and reagents that provide the highest yields of the desired σ-pyridylethynyl iron compounds were determined. The methods developed allowed the synthesis of the corresponding σ-4-benzothiadiazolylethynyl complex Cp(CO)2Fe–CC–(4-C6H3N2S) as well. Eventually, synthetic approaches to σ-alkynyl iron complexes of the type Cp(CO)2Fe–CC–R (R = ortho-, meta-, para-pyridyl, 4-benzothiadiazol-2,1,3-yl) based on the Pd/Cu-catalyzed cross-coupling reactions were elaborated.


Introduction
Metal s-alkynyl complexes displaying such peculiar characteristics as linear geometry, high stability, and p-unsaturated character have been demonstrated to constitute promising building blocks for the design of materials, which can possess such properties as optical nonlinearity, 1-5 light-emission, [6][7][8][9][10][11][12] and electrical conductivity. [13][14][15][16][17][18] Moreover, they are an important class of coordination compounds because of their relevance in synthetic chemistry [19][20][21][22][23][24] and proton reduction catalysis. [25][26][27] A variety of methods for the synthesis of transition metal acetylides have been developed. 24 The most common synthetic route to them is transmetallation, where a generated [M-C^C-R] species [M ¼ Cu(I), Ag(I), Au(I), alkali-metal (Li, Na), or an alkaline-earth-metal (MgX, etc.)] acts as an alkynyl transfer reagent to transition-metal halide complexes L n MX (X ¼ I, Br, Cl). [28][29][30][31][32][33][34] However, in some cases, such reactions may give low yields of desired products and a range of by-products, for example, complexes in which the copper or the silver fragments are p-coordinated to the transferred alkynyl ligand. [35][36][37][38] Other general strategy to the preparation of transition metal acetylides takes advantage of the facility of some transition metal complexes to catalyze M-C coupling reactions. The most general route to group 10 metal s-alkynyl derivatives is based on copper(I)-catalyzed dehydrogalogenation reactions between an appropriate metal halide complex and a terminal alkyne in an amine solvent. [39][40][41][42] This method is applicable to the synthesis of alkynyl, polyynyls, and polyyndiyls of tungsten, molybdenum, iron, ruthenium, rhodium, and iridium. 5,[43][44][45][46][47] In some cases, the CuI-catalyzed reactions of transition metal halides with stannyl acetylenes may be performed in the absence of amines. [48][49][50][51] Palladium catalysts, currently being an indispensable tool of organic synthesis, can also be used in the M-C^C-bond formation. The team of Claudio Lo Sterzo demonstrated that, similarly to organic electrophiles, transition metal iodides undergo coupling with trialkyltin acetylides, in the presence of palladium to form alkynyl complexes of ruthenium, iron, tungsten, and molybdenum. [52][53][54][55] The role of palladium catalysts in promoting these transformations was also investigated. 52,[56][57][58][59] Although, the Lo Sterzo approach was shown to be a valuable route to alkynyl complexes, the need for preparation of tin reagents and removal of tin impurities limits the appeal of this method. To overcome these disadvantages, one can use an organometallic analogue of the Sonogashira protocol, 60,61 where transition metal halides react with terminal alkynes to form metal alkynyls via the Pd/Cu-or Pd-catalyzed dehydrohalogenation route, which provides milder reaction conditions and facilitates purication of products. Despite the seeming availability of this approach for the synthesis of metal acetylides, it was exploited only once by Oshima 62 in the synthesis of salkynyl iron complexes Cp(CO) 2 Fe-C^C-Ar.
As a part of our studies on proton reduction catalysis, 25 we searched for a facile synthetic route to a series of s-pyridylethynyl iron complexes Cp(CO) 2 Fe-C^C-(n-C 5 H 4 N) (n ¼ ortho (1), meta (2), para (3)) containing two Lewis base centers (C b of ethynyl and N atom of pyridine). The efficient preparation of the iron arylethynyls complexes Cp(CO) 2 Fe-C^C-Ar by Pd/ Cu-catalyzed cross-coupling of Cp(CO) 2 FeI with terminal arylacetylenes 62 inspired us to apply the Oshima protocol for the preparation of 1-3. Unfortunately, our rst attempt to obtain the ortho-pyridylethynyl iron complex Cp(CO) 2 Fe-C^C-(2-C 5 H 4 N) (1) using the Oshima conditions did not give the target substance. Upon an increase in the reaction temperature to 60 C an unexpected binuclear FePd m-pyridylvinylidene (4) was obtained in yield of 2%, but still without traces of 1 (Scheme 1). The 12% yield of complex 4 was achieved by using one equivalent of PdCl 2 without copper iodide and pure diisopropylamine as a solvent 63 (Scheme 1). Attempts to synthesize the iron derivatives of meta-and parapyridylethynyles under described conditions 62 were also unsuccessful; the iron acetylide compounds were obtained in very low yields (less than 5%) although no formation of the side binuclear FePd products was observed in these cases.
For that reasons, we became interested in developing a simple and reliable synthetic approach to s-acetylide iron complexes 1-3 based on the Pd/Cu-catalyzed cross-coupling reactions. We proposed that the inability to obtain the pyridylethynyl iron complexes 1-3 from Cp(CO) 2 FeI and ethynylpyridines under the Oshima conditions was primarily caused by a suppression of the formation and transfer of pyridylethynyl moieties to the palladium atom (Pd/Cu-catalyzed reactions) or of the formation of this moiety at the Fe-Pd center from the pcoordinated alkyne (Cu-free reaction). Here we report on the cross-coupling of terminal ethynylpyridines and (trimetylsilyl) ethynylpyridines with cyclopentadienyliron dicarbonyl iodide under different Pd/Cu-catalyzed conditions to obtain the target iron pyridylethynyl complexes.

Results and discussion
Development of Pd/Cu-catalyzed approaches to the synthesis of Cp(CO) 2 Fe-C^C-(n-C 5 H 4 N) (n ¼ ortho (1), meta (2), para (3)) and Cp(CO) 2 Fe-C^C-(4-C 6 H 3 N 2 S) (5) Following on from the hypothesized reasons of the inability to obtain the ethynylpyridyl iron complexes Cp(CO) 2 Fe-C^C-(n- (3)), we realized that facilitation of the transmetallation step (the formation and the transfer of acetylide species) in Pd/Cu-catalyzed reaction between Cp(CO) 2 FeI and ethynylpyridines should allow the preparation of the desired products 1-3. Therefore, two approaches were proposed to solve this problem: (i) the use of (trimethylsilyl)ethynylpyridines that in the presence of uoride ion generate anionic pentacoordinate silicate species [Me 3 Si(F)-C^C-Pyr] À , in which the alkynyl group is more nucleophilic and should be smoothly transferred to a palladium catalyst (Hiyama coupling); 64 (ii) the application of stronger base than secondary and tertiary amines such as 1,8-diazabicyclo [5.4.0] undec-7-ene (DBU) to facilitate the acetylide species formation in Pd/Cu-catalyzed reactions of Cp(CO) 2 FeI with terminal ethynylpyridines. Our preliminary experiments suggested that both approaches can be used.
The palladium(II) complexes bis(triphenylphosphine)palladium(II) dichloride PdCl 2 (PPh 3 ) 2 and bis(acetonitrile)palladium dichloride PdCl 2 (NCMe) 2 , as well as tris(dibenzylideneacetone) dipalladium(0) Pd 2 (dba) 3 were tested as catalysts in this study. However, in our rsts experiments an application of PdCl 2 (-PPh 3 ) 2 as catalyst was found to results in an unexpected substitution of CO ligands by PPh 3 in [Cp(CO) 2 Fe] fragments to give the side-products Cp(CO)(PPh 3 )FeI and Cp(CO)(PPh 3 )Fe-C^C-(2-C 5 H 4 N), thereby making the separation of the reaction mixture more difficult. Therefore, in our following experiments we applied only such palladium catalysts that do not contain ligands capable to substitute carbonyl groups in the initial and the target compounds, namely Pd 2 (dba) 3 and PdCl 2 (NCMe) 2 . The inuence of the presence of CuI cocatalyst on the reaction outcome was also examined.
The "ethynylpyridine" approach was found to be much more effective, the complexes 1-3 were obtained in yields from 67 to 96%. The best catalytic system in this case was PdCl 2 (NCMe) 2 / CuI, with which iron-ethynylpyridine coupling proceeds efficiently at room temperature in 20 minutes (2 mol% of Pd catalysts and 20 mol% of CuI) or at 60 C in 30 minutes with reduced catalysts loadings (1 and 5 mol%, respectively) ( Table 2).
The presence of CuI in these conditions is important for high-yield synthesis of 1-3. The pyridylethynylation smoothly proceeds only for the Pd(II)/CuI-catalyzed reactions, whereas in the absence of CuI cocatalyst (Table 2, entries 5, 9 and 13) the conversion and the yield of corresponding s-pyridylethynyl complex decreased, but the yield of the dimerization product [Cp(CO) 2 Fe] 2 increased. Moreover, in these conditions the reaction with 2-ethynylpyridine gave m-pyridylvinylidene complex 4 as a side product in 4% yield.
Similar results were obtained for the reactions between Cp(CO) 2 FeI and ethynylpyridines catalyzed by the zero-valent palladium complex Pd 2 (dba) 3 (Table 3). When the Pd 2 (dba) 3 / CuI pair was applied, the target pyridylethynyl complexes 1-3 were produced in high yields (84, 86 and 87% for 1, 2 and 3, respectively) and the yields of the dimerization product [Cp(CO) 2 Fe] 2 were in a range of 6-8%. However, under copper-free conditions here, the yield of the target complexes decreased by about 20%, while the yield of [Cp(CO) 2 Fe] 2 increased to 20% and, also, m-pyridylvinylidene complex 4 was obtained in the case of the reaction of Cp(CO) 2 FeI and H-C^C-(2-C 5 H 4 N) ( Table 3, entries 2 and 4). Thus, the binuclear complex 4 is also produced under the conditions of catalytic formation of complex 1. It should be noted that for the reactions between Cp(CO) 2 FeI and ethynylpyridines we decided to apply the same temperature and Pd 2 (dba) 3 loadings that gave the highest yields of 1-3 in the ethynylation of Cp(CO) 2 FeI with [(trimethylsilyl) ethynyl]pyridines (entries 3, 8, 11 in Table 1).
Therefore, all reactions given in Table 3 were conducted at 36 C with 10 mol% of Pd 2 (dba) 3 .

The proposed catalytic cycles
Previously, the Lo Sterzo group clearly demonstrated a close analogy between the Pd-catalyzed carbon-carbon and metal-  Table 3 entries 2, 4) clearly shows that the reaction pathways leading to 1 and 4 may have common intermediates. Moreover, one can assume that in the reaction performed with the stoichiometric ratios of Pd 2 (dba) 3 , Cp(CO) 2 FeI and H-C^C-(2-C 5 H 4 N), the ratio of 1 to 4 should change in favor of the latter. Indeed, the reaction of Cp(CO) 2 FeI with 2-ethynylpyridine in the presence of 1 equiv. of Pd 2 (dba) 3 in triethylamine as a solvent   To reveal the mechanism of this process, an additional study is needed, since there are several possible reaction pathways for its proceeding. It is worth noting that the examples of the acetylene-to-vinylidene tautomerism involving two adjacent metal centers are known and described in several works, 66-70 but they are much less studied in contrast to the well-understood acetylene-tovinylidene rearrangement mediated by a single metal center. 71,72 It is also worth to note that the ortho-position of the nitrogen atom in 2-ethynylpyridine plays an important role in the process of formation of the binuclear FePd m-pyridylvinylidene complex 4, as the Pd-catalyzed reactions of Cp(CO) 2 FeI with 3-and 4-ethynylpyridines simply resulted in a decrease of yields of 2 and 3, and didn't give any binuclear products (Table 2 entries 9 and 13).
Taking into account the data obtained and the known mechanistic aspects of the cross-coupling reactions, a noncontradictory mechanistic scheme of the pyridylethynylation of cyclopentadienyliron dicarbonyl iodide was proposed (Scheme 4), which explains the reaction pattern and the possibility of alternative reaction pathway. For the Pd/Cu-catalyzed reactions (pathway A), the rst step of the catalytic cycle would be an oxidative addition of the Cp(CO) 2 FeI to the catalytically active Pd 0 L 2 species to give a binuclear FePd intermediate A. A subsequent transmetallation, where transfer of the pyridylacetylide moiety from the Cu-acetylide formed in the copper-cycle to the FePd complex A led to a complex B, which undergoes trans to cis isomerization to give a complex B 0 , a following reductive elimination of the latter results in the Fe-C cross-coupling products Cp(CO) 2 Fe-C^C-(n-C 5 H 4 N) (n ¼ ortho (1), meta (2), para (3)). The initial oxidative addition and the nal reductive elimination steps of a mechanism for the Cufree reaction (pathway B) are the same as for the Pd/Cucatalyzed one. However, an alternative ligand L substitution in the intermediate A by the alkyne can take place resulting in palkyne complex A 0 . Its subsequent deprotonation then occurs to give the complex B 0 . The generation of the complex 4 in this context can be considered as a side pathway C of the Pdcatalyzed reaction between Cp(CO) 2 FeI and 2-ethynylpyridine that is facilitated by the absence of transmetallating agents or appropriate base, and ortho-position of the nitrogen atom in the pyridine ring of the alkyne. The catalytic cycle for the complex 5 formation should be the same as for 1-3.
The results obtained in the case of the "(trimethylsilyl)ethynylpyridine" demonstrated that pyridylethynylation of Cp(CO) 2 FeI should proceed according to the reaction pathway commonly accepted for the Hiyama coupling, i.e. transfer of the pyridylacetylide moiety to the FePd complex A should goes through anionic pentacoordinate silicate species 64 (Scheme 5).

Characterization of the complexes
The IR and the 1 H and 13 C{ 1 H} NMR data for the complexes 1-3 and 5 were obtained. The NMR signals were assigned on the basis of 1 H-13 C correlations measured through HSQC and HMBC experiments, respectively ( Table 4). The structure of the complexes can be deduced from the combined NMR and IR data.
The IR spectra of 1-3 in CH 2 Cl 2 solution show two very strong absorptions at about 2043 and 1996 cm À1 that Scheme 4 Proposed mechanism of Pd/Cu-and Pd-catalyzed coupling reactions of cyclopentadienyliron dicarbonyl iodide and ethynylpyridines.

Scheme 5 Proposed mechanism of Pd-catalyzed coupling reactions of cyclopentadienyliron dicarbonyl iodide and [(trimethylsilyl)ethynyl] pyridines.
This journal is © The Royal Society of Chemistry 2020 RSC Adv., 2020, 10, 17014-17025 | 17019 correspond to the n(C^O) stretching modes of carbonyl groups at the iron atom. An additional absorption with strong intensity at about 2111 cm À1 is attributed to the n(C^C) stretching frequencies of the alkynyl ligand. The n(CO) and the n(C^C) frequencies of the complex 5 with s-4-ethynyl-2,1,3benzothiadiazole ligand are insignicantly shied to lowfrequency in comparison with 1-3.
The 13 C nuclei of aand b-alkynyl carbons of 1-3 and 5 resonate in the regions d 112-117 and 95-103 ppm, respectively. It is noteworthy that these signals slightly vary with the nature of the substituents in s-alkynyl ligands. The downeld shi of C a nuclei by approximately 2 ppm is observed on moving from the meta-to para-and to ortho-pyridylethynyl iron complexes. However, this trend doesn't hold for the C b nuclei, here only the signal of 3 is downeld shied by 7 ppm compared with those of the 1 and 2. These correlations are apparently due to changes in p-electron distribution induced by the electronic effect of the pyridyl fragments. However, these effects can't be reduced to a single parameter (like only inductive or mesomeric effects), as was showed by Claude Lapinte team for cationic complexes of the type [Cp*(dppe)Fe-C^C-(n-C 5 H 4 N)][PF 6 ] (n ¼ 2, 3, 4). 73 At the same time, the 13 C chemical shis of the cyclopentadienyl and carbonyl groups coordinated to the iron atom are almost independent of their nature ( Table 4). The presence of the pyridyl substituents in 1-3 are indicated by signals between 7.02 and 8.50 ppm in 1 H NMR spectra, and between 119 and 152 ppm in 13 C NMR spectra of the complexes. The carbon and hydrogen atoms of the 2,1,3-benzothiadiazole group in 5 resonate in the regions d 118-156 and d 7.48-7.76 ppm in 13 C and 1 H NMR spectra, respectively. So, overall the IR and NMR spectra parameters of 1-3 and 5 are similar to those found for analogous cyclopentadienyliron dicarbonyl complexes with different s-alkynyl ligands. 34,54,62 The molecular structures of the complexes Cp(CO) 2 Fe-C^C-(2-C 5 H 4 N) (1), Cp(CO) 2 Fe-C^C-(3-C 5 H 4 N) (2), Cp(CO) 2 -Fe-C^C-(4-C 5 H 4 N) (3), and Cp(CO) 2 Fe-C^C-(4-C 6 H 3 N 2 S) (5) were solved on the basis of X-ray diffractometry data. Suitable crystals of 1-3 and 5 were grown from dichloromethane/hexane mixtures. The views of the structures are shown in Fig. 1, selected bond lengths and angles are given in Table 5. The crystal data and renement parameters are included in the ESI. † The proximity of crystals 1 and 2 cell parameters (Table S3 †) and small differences in the sequence of atoms in their molecules suggest that there are small differences in a packing of molecules. Indeed, the structures 1 and 2 are mirror-like each other, and the mirror plane is perpendicular to the cell axis b. The iron atom in 1-3 and 5 is coordinated by a cyclopentadienyl  ring in h 5 -fashion, two terminal CO ligands, and by s-alkynyl ligand to adopt a typical pseudo-octahedral geometry. The bond lengths and angles in 1-3 and 5 are close to each other and to those found in the known complexes of the type Cp(CO) 2 Fe-C^C-R. [74][75][76][77] The differences in geometry of the complexes' molecules are associated with the position of C^C-R moiety relative to the iron fragment. The pyridylethynyl ligands in 1 and 2, and the 4-ethynyl-2,1,3-benzothiadiazolyl ligand in 5 lean slightly toward one of the carbonyl group of the iron atom in such a way that the planes between ring of the substituent and [Fe(CO) 2 ] fragment exhibit angles about 81 . At the same time, in the complex 3 the para-pyridylethynyl ligand is almost perpendicular to the plane formed by two carbonyl ligands and the iron atom. Moreover, the angles between the planes of the substituent's ring in the s-alkynyl and the cyclopentadienyl ligands are about 70 in 3 and 5, whereas the analogous angles in 1 and 2 are ca. 85 . It is also of interest that the pyridyl substituents in 1 and 2 differ in their orientations: when the nitrogen atom in 1 is oriented to Cp ring, that in the complex 2 the nitrogen is placed opposite, i.e. oriented to carbonyl groups.

Conclusions
In this paper, two approaches were developed for the pyridylethynylation of cyclopentadienyliron dicarbonyl iodide based on Pd/Cu-and Pd-catalyzed Fe-C coupling of (trimethylsilyl)ethynylpyridines and ethynylpyridines with Cp(CO) 2 -FeI. Although these two approaches differ only in the nature of species participating in the transmetallation step (transfer of the pyridylacetylide moiety to catalytically active species), the second "ethynylpyridine" approach was found to be much more effective compared with the rst "(trimethylsilyl)ethynylpyridine" one. For example, the s-pyridylethynyl iron complexes Cp(CO) 2 Fe-C^C-(n-C 5 H 4 N) (n ¼ ortho (1), meta (2), para (3) The developed methodology can be extended to the synthesis of other s-alkynyl complexes of iron. So, the coupling of cyclopentadienyliron dicarbonyl iodide and 4-ethynyl-2,1,3benzothiadiazole under the same condition as above was found to proceed efficiently resulting in the complex Cp(CO) 2 -Fe-C^C-(4-C 6 H 3 N 2 S) (5). The spectroscopic and structural features of 1-3, 5 were described.