Over the last three decades, late-transition metal complexes such as pincer-type metal complexes have attracted greatattention in catalytic chemistry^[1-7]. The pincer-type metal complexes catalyzed cross-coupling reactions between ary/alkyl halide and organometallic reagents have been widely developed in the field of organic chemistry^[8-12]. Among the most studied ligands are pincer DCD^[13-14] (D a donor atom such as P, N, or O, S) systems and NNN^[15-16], PNP^[17-19], SNS^[20] NNO^[21] and N-heterocyclic carbene (NHCs)^[22-23], which have found wide applications in bond activation because of their unprecedented stability at high temperatures and catalytic activity. In addition, their ligand framework can be easily modified in order to tune the electronic properties of the metal and the steric hindrance around it.

In fact, among catalysts with metal ions as active sites, the Pd complexes are the most commonly used catalysts for cross-coupling reactions such as the Suzuki reaction, Negishi reaction, Kumada reaction, and Heck reaction because of their advantages that relatively mild reaction conditions, applicable to substrates of wide scope and displaying high turnover^[24-26]. Despite their high utility, in the process of activation, precipitation of palladium black is harmful to the catalyst and leads to be inefficient, other drawbacks: palladium and the necessary ligands are expensive, palladium compounds are toxic and leaching of the palladium catalyst into final products can be problematic.

Compared with Pd, Ni catalysts are much cheaper and are the most promising alternative to palladium. Some nickel complexes^[27-29] have been shown to be catalytically active for region-and stereo-selectivity and be tolerate to a wide range of functional groups, so the development of new nickel complexes with catalytic efficiency would be of practical importance for the synthetic industry. And thus the search for appropriate ligands for tuning the generation of highly active Ni catalyst precursors to catch up with or replace expensive Pd catalysts is of current interest.

Recently, a series of nickel halide complexes containing tridentate N^N^N ligands have been employed to catalyze ethylene oligomerization^[30-31]. Upon activation with methylaluminoxane, the nickel complexes display good catalytic activities in ethylene oligomerization. And in the previous report the author suggested that the steric bulk of the aryl groups in the backbone created a reaction channel which made the ethylene coordination to the Ni atom more efficient and then makes the activity increase^[24].Such an activity increase with increased steric bulk has also been observed in olefin polymerization reactions using zirconocene/MAO as catalysts^[32]. Herein the activities of nickel complexes (Scheme 1) supported by tridentate [N^N^N] ligands with one phenanthroline linked to pyrazoly donors in C—C coupling reactions have been studied.

1 Experimental 1.1 General Procedures

All solvents and reagents were of analytical grade and were used as purchased without further purification. Unless otherwise mentioned, all non aqueous reactions and manipulations were carried out in a nitrogen atmosphere using standard Schlenk techniques, and chemicals were used as received without any further purification. Solvents were dried by sodium/benzophenone and distilled under nitrogen prior to use. GC analysis was performed with an SP-2100 gas chromatography spectrometer equipped with a flame ionization detector and an SE-30 silica capillary column. Retention times were compared to commercially obtained compounds. Decane was used as an internal standard.

1.2 Synthesis of the ligands L¹-L⁴

2-(3-R²-5-R³-pyrazol)-9-R¹-1, 10-phenanthroline ligands (L¹-L⁴) were prepared according to the previously reported procedures^[30]. In a typical experiment, a mixture of 2-hydrazine-1, 10-phenanthroline or 2-hydrazine-9-aryl-1, 10-phenanthroline (3.0 mmol) and the diketone (3.6 mmol), i.e. acetylacetone (for L¹), benzoylacetone (for L²), or dibenzoylmethane (for L³ and L⁴), was stirred in dry methanol (15 mL) containing a few drops of acetic acid for 6~12 h. The solvent was removed and the residue was recrystallized from chloroform/activated charcoal to give the products.

1.3 Synthesis of the nickel complexes

The nickel(Ⅱ) complexes were synthesized by the following general procedure: a solution of the ligand in THF was added to an equimolar amount of NiCl₂·6H₂O in THF. The reaction mixture was stirred at room temperature for 12 h. The resultant precipitate was collected, washed with THF and diethyl ether, and dried in vacuum. All of the complexes were prepared in high yield in this manner.

1.4 General Procedures for the C—C coupling reaction

The C—C coupling reactions were performed in 20 mL Schlenk tube in a nitrogen. In a typical experiment, required amount of catalyst, toluene and substrate were charged into the reactor and form a homogeneous solution. To the stirred solution was added dropwise a solution of PhMgBr in THF at room temperature. Stirring was continued at the desire reaction temperature within the desire reaction time. The reaction was ceased by addition of water (5 mL). The mixture was extracted with Et₂O (3 × 5 mL), and the combined organic layers were dried over anhydrous Na₂SO₄. The Na₂SO₄ was removed by filtration and washed with Et₂O.

1.5 X-ray crystal structure determination

Diffraction data for the complexes were collected on a Bruker SMART APEX Ⅱ diffractometer at room temperature (293 K) with graphite-monochromated Mo Kα radiation (λ = 0.710 73 Å). An empirical absorption correction using SADABS was applied for all data. The structures were solved by direct methods using the SHELXS program. All non-hydrogen atoms were refined anisotropically by full-matrix least-squares on F₂ by the use of the program SHELXL. The hydrogen atoms bonded to carbon were included in idealized geometric positions with thermal parameters equivalent to 1.2 times those of the atom to which they were attached. Crystallographic data for the complexes 4 are listed in Table 1.

2 Results and discussion 2.1 X-ray Crystallographic Study

Blue single crystals of complex 4 suitable for single crystal X-ray diffraction were obtained through slow diffusion of diethyl ether into its methanol solution. An ORTEP view of the molecular structure of complex 4 is presented in Figure 1 along with the selected bond lengths and angles. The complex 4 reveals a 1:1 me- tal-to-ligand ratio with the metal center. And our catalysts are prepared in dry CH₂Cl₂, so the structure can represent the employed that of catalysts. The nickel center is coordinated a distorted trigonal bipyramidal geometry (The Addison τ= 0.16 for 4) by two nitrogen atoms of phenanthroline, one pyrozely nitrogen, and two chlorine atoms in which N2 of the ligand and the two chlorine atoms define the equatorial plane, while the N1 and N2 atoms occupy the two coordination site of the axial. The center metal Ni(Ⅱ) deviate from the equatorial 0.085 Å, the central Ni-N2 [1.973(2)Å] bond is shorter than the terminal Ni-N1 [2.183(2) Å] and Ni-N4 [2.201(2) Å] bonds. In addition, compared with its analogue bromine (Ni-N2, 1.963(4) Å and Ni-N4, 2.199(4) Å)^[23], clearly, the length of Ni-N2 and Ni-N4 in 4 are shorter. While the N1-Ni-N2 bite angle [79.50(8)°] is wider than the N2-Ni-N4 bite angle [75.72(8)°], the Cl-Ni-Cl angle in 4 [145.53(3)°] is similar to the analogue bromine [146.17(3)°]. And the mesityl group is nearly perpendicular to the phenanthroline plane in 4 (dihedral angle 89.5°), is also similar to the analogue bromine (89.2°), meanwhile the dihedral angle between the pyrazolyl ring and the phenanthroline plane is 15.1° in 4, slightly smaller than the analogue bromine (17.0°). In the previous report, we found the skeleton of the chloride complexes with the formation of a more directional environment, bulkier aryl substituents can suppress chain-transfer reactions (typically-H elimination) and therefore increase the activity ^[30]. It suggests that if the steric bulk of the aryl groups create a reaction channel which also may makes the organometallic reagents coordination to the Ni atom more efficient and further easily reacted with other aryl halides.

2.2 Catalytic Properties 2.2.1 Optimization of reaction conditions

With these considerations in mind, our initial work focused on the model reaction of p-MeC₆H₄Br with PhMgBr to optimize the reaction conditions. These results are summarized in Table 2. With toluene as the solvent and at room temperature, we first examined the effect of catalyst precursors. To our delight, the C—C coupling reaction of p-MeC₆H₄Br with PhMgBr could proceed smoothly in the presence of NiCl₂L¹ complex (Table 2, entry 1). To achieve better results, various commercially available salt additives^[33] were introduced to these systems, and the results proved that 1 equiv. PPh₃ was the best choice (Table 2, entry 3). Meanwhile, we conducted the experiment in situ by using ligand L¹ and (DME)NiCl₂, turning out to be the yield of p-MePh-Ph as well as NiCl₂L¹ complex (Table 2, entry 2). Subsequently, different solvents (non-polar or polar solvents)were examined. The results showed that toluene and dioxane were proved to be the best for the formation of p-MePh-Ph. After these optimized conditions were developed, the effects of dosage of NiCl₂L¹ complex and temperature were studied as well. Investigation of temperature revealed that getting the same yield needed to shorter time. Besides, with either an increase or decrease of the amount of the NiCl₂L¹ complex, inferior results were observed.

2.2.2 Effect of the ligand environment

Changes in the ligand environments led to different catalytic beha- viors^[30-31]. For example, Wang^[18] group has recently reported that ligands containing electron-rich phosphorus atoms were suitable for the coupling reactions of aryl chlorides, the extent of which is strongly dependent upon the electronic nature of the phosphine ligand in the catalytic action. Based on above mentioned idea, We investigated the effect on the catalytic activities of these ligands coordinated environment in detail. The results are summarized in Table 2. The R² and R³ substituents on the pyrazolinyl ring performed slight influence on their activities. It was observed that their catalytic activities increased with increasing the steric hindrance of their ligands, the regulation was easily confirmed in the catalytic systems of the complexes 1 (Me), 2 (Me, Ph) and 3 (Ph, Ph). Their activities increased in the order 1 >2 >3 (entries 1, 18 and 19 in Table 2). This case could be caused by electronic influences of ligands. Meanwhile, the results showed that the activity with the R¹ substituent as Mes on the phenanthroline backbone (4) was optimum (entry 20 in Table 2), This could be attributed to the more bulky groups at the 9-site of the phenanthroline backbone which may facilitate the course of oxidation addition and then reductive elimination.

2.2.3 Studies about the scope of substrates

With the optimized conditions in hand, the scope of aryl ha- lides was investigated (Table 3). For example, complex 4 acted as an efficient catalyst for the substrate of p-MeC₆H₄I, reaching 90% yield within 24 h at room temperature (Table 3, entry 6). Under same conditions, when p-CF₃C₆H₄I as a substrate gave excellent conversations (yield up to 98%, Table 3, entry 7). On the same experimental conditions, complex 4 also applies to deactivated substrate such as 4-ioideanisol, but the yield of the corresponding product is a bit lower, reaching 63% yields for deactived aryl iodides over a period of 24 h (Table 3, entry 5). Probably because of the nature of the substituents (electron-withdrawing or electron-donating substituents) in the aromatic substrates, which play a key role on the activation of C—I bonds. Unfortunately, when deactivated substrates 4-bromotoluene and 4-bromoanisole were involved, only moderate or trace amounts of products were observed. In expectation, due to low reactivity of aryl chlorides, nickel complexes are almost inefficient even for the coupling reaction of activated 4-chlorotoluene and PhMgBr. It is easily explain that C—Cl bond (Ph—X bond dissociation energies (BDEs)^[34]: Cl, 402 kJ/mol; Br, 339 kJ/mol; I, 272 kJ/mol) much stronger. In addition, the steric hindrance of the substituent has little influence on the product yield in the C—C coupling reaction.

2.2.4 Studies about the kind of products and catalytic action

We also studied the products in the C—C cross coupling reactions. For example, Table 3, entry 5, apart from the desired product 4-methoxybiphenyl, homo-couplings and dehalogenation are observed in the reaction mixtures by GC. It is the by-products of the organic groups that occurred in the C—C coupling reactions, leading to difficulties in purification and yields of the main cross-coupled product. From the phenomenon of the experiments PhI with PhMgBr, we surmised that reaction firstly occurred between catalysts and PhMgBr (the color of reaction mixture immediately changed brown) but no reaction occurred between PhI by GC (no disappear). This is rationally understood since the generation of catalytically active species from the PhI would be more difficult than from the PhMgBr. It was in agreement with the mechanism for the alkyl-alkyl Kumada coupling catalyzed by a Ni(Ⅱ) pincer complex recently proposed by Hu's group^[15].

3 Conclusions

In summary, four Ni(Ⅱ) dichloride complexes of 2, 9-subtituted-1, 10-phenanthrolin derivatives confirmed that Ni(Ⅱ) complexes in apenta-coordinated fashion hold promise as catalysts for C—C coupling reactions. The steric bulk of the ligands significantly affected the catalytic behavior of the complexes. And complex 4 exhibited highly catalytic activities in the C—C coupling reaction of aryl iodide with aryl Grignard reagents at room temperature. As result, it is possible to enhance the catalytic activities by tuning the size of substituent on pyrazolyl of these complexes.