Application of POCOP Pincer Nickel Complexes to the Catalytic Hydroboration of Carbon Dioxide

: The reduction of CO 2 is of great importance. In this paper, different types of bis(phosphinite) (POCOP) pincer nickel complexes, [2,6-(R 2 PO) 2 C 6 H 3 ]NiX (R = t Bu, i Pr, Ph; X = SH, N 3 , NCS), were applied to the catalytic hydroboration of CO 2 with catecholborane (HBcat). It was found that pincer complexes with t Bu 2 P or i Pr 2 P phosphine arms are active catalysts for this reaction in which CO 2 was successfully reduced to a methanol derivative (CH 3 OBcat) with a maximum turnover frequency of 1908 h − 1 at room temperature under an atmospheric pressure of CO 2 . However, complexes with phenyl-substituted phosphine arms failed to catalyze this reaction—the catalysts decomposed under the catalytic conditions. Complexes with i Pr 2 P phosphine arms are more active catalysts compared with the corresponding complexes with t Bu 2 P phosphine arms. For complexes with the same phosphine arms, the catalytic activity follows the series of mercapto complex (X = SH) ≈ azido complex (X = N 3 ) >> isothiocyanato complex (X = NCS). It is believed that all of these catalytic active complexes are catalyst precursors which generate the nickel hydride complex [2,6-(R 2 PO) 2 C 6 H 3 ]NiH in situ, and the nickel hydride complex is the active species to catalyze this reaction.

As an earth-abundant cost-effective late transition metal, nickel has been extensively used to catalyze a variety of chemical transformations since the last century. Nowadays, homogeneous nickel catalysis is becoming more and more important in modern synthetic chemistry because of the diverse novel catalytic reactivities of nickel complexes [30][31][32].
We have been interested in the chemistry of transition metal complexes bearing bis(phosphinite) (POCOP) pincer ligands [14,15,26,27,[33][34][35][36][37][38]. Recently, we applied some POCOP pincer nickel thiolato complexes, [2,6-(R 2 PO) 2 C 6 H 3 ]NiY (R = t Bu, i Pr; Y = SC 6 H 4 -p-OCH 3 , SC 6 H 4 -p-CH 3 , SPh, SC 6 H 4 -p-CF 3 , SCH 2 Ph), to the catalytic reduction of CO 2 with catecholborane (HBcat) [27]. Carbon dioxide was effectively reduced to CH 3 OBcat under very mild conditions, and turnover frequencies (TOFs) up to 2400 h −1 were observed. Although POCOP pincer nickel hydrides are also good catalysts for this reaction under the same conditions [14][15][16], the catalytic activity of the nickel thiolato complexes are better than that of the corresponding nickel hydride complexes. It was concluded that the thiolato complexes may work as pre-catalysts with the in situ generated nickel hydride complexes being the real active species. The superior catalytic activity of the thiolato complexes may be due to some in situ generated thiolato species that act as co-catalysts. To confirm this speculation, we initiated a more extensive study for this hydroboration reaction. In this paper, we used different types of POCOP pincer ligated nickel complexes with different phosphine arms and different auxiliary ligands, [2,6-(R 2 PO) 2 C 6 H 3 ]NiX (Scheme 1), to catalyze this reaction under the same conditions. It was found that pincer complexes with t Bu 2 P or i Pr 2 P arms are active catalysts and CO 2 was reduced to CH 3 OBcat at room temperature under an atmospheric pressure of CO 2 . Turnover frequencies up to 1908 h −1 were observed. Complexes with i Pr 2 P arms are more active catalysts compared with the corresponding complexes with t Bu 2 P arms. For complexes with the same phosphine arms, the catalytic activity follows the series of mercapto complex (X = SH) ≈ azido complex (X = N 3 ) >> isothiocyanato complex (X = NCS). The results are consistent with the speculation that these different types of pincer nickel complexes are catalyst precursors and the in situ generated pincer nickel hydrides are the actual active species to catalyze the hydroboration of carbon dioxide with catecholborane. Scheme 1. The nickel complexes used for the catalytic hydroboration of CO 2 with HBcat.

Synthesis and Characterization of the Ni Catalysts
Complexes 1a [37], 1b [38], 2a-c [35] and 3a-c [35] were synthesized as described in the literatures. Complex 1c has never been reported before and was synthesized by the reaction between [2,6-(Ph 2 PO) 2 C 6 H 3 ]NiCl and NaSH in THF/methanol at room temperature (Scheme 2). As a new complex, 1c was fully characterized by nuclear magnetic resonance spectroscopy (NMR), Fourier Transform infrared spectroscopy (FTIR), X-ray crystallography and elemental analysis. Complex 1c is an air-sensitive orange solid and is soluble in toluene and dichloromethane. The 1 H, 13 C{ 1 H} and 31 P{ 1 H} NMR spectra of 1c were in good agreement with the structure depicted in Scheme 2. Complex 1c showed the expected spectra characteristics for a POCOP pincer ligated nickel complex [14][15][16]27,[33][34][35][36][37][38][39][40][41][42]. Similar to other reported transition metal mercapto complexes [37,38,[43][44][45], the 1 H NMR resonance of the -SH proton of 1c appeared in a relatively high-field (−0.76 ppm) as a triplet. An H-P coupling constant of 19.7 Hz was observed. The 31 P{ 1 H} NMR spectrum of 1c displayed one singlet, implying an identical chemical environment for the two phosphine arms in solutions. The FTIR spectrum of 1c displayed a weak absorption at 2546 cm −1 which was attributed to the S-H stretching vibration.
As shown in Figure 1, the square-planar geometry for the Ni center of 1c is distorted; this is quite common for POCOP pincer nickel complexes [14][15][16]27,[33][34][35][36][37][38][39][40][41][42]. Both the C ipso -Ni and Ni-S bond lengths in complex 1c are comparable to those of the related pincer nickel mercapto complexes with t Bu 2 P or i Pr 2 P arms [37,38]. This indicates that the C ipso -Ni and Ni-S bond strengths are not considerably affected by the substituents on the phosphine arms.

Catalytic Hydroboration of CO 2 with HBcat
In order to check the catalytic activities, a NMR tube reaction was carried out for each of the nickel complexes. Typically, the nickel complex (0.01 mmol) was mixed with HBcat (0.3 mmol) in C 6 D 6 (0.5 mL) in a NMR tube under a N 2 atmosphere. Carbon dioxide was then introduced and the reaction was monitored by NMR spectra at room temperature. For complexes 1a, 1b, 2a, 2b and 3b, the 11 B NMR spectra showed that HBcat was completely or partly consumed in 30 min. CH 3 OBcat and catBOBcat were formed as evidenced by the two singlets at 23.6 and 22.6 ppm [14,26,27,46,47] in the 11 B NMR spectra (see Figure 2 for example, the representative 11 B NMR spectra for the interactions of the nickel complexes with HBcat under CO 2 ). The formation of CH 3 OBcat was also confirmed by the singlet at 3.36 ppm [14,26,27,46] in the 1 H NMR spectra. Similar to the hydroboration of CO 2 with HBcat using the related nickel thiolato complexes as the catalysts [27], the 31 P{ 1 H} NMR spectra were too complex to be understood-many unidentified species were shown. For complex 3a, no reaction was observed after 60 min at temperatures up to 60 • C as evidenced by the 31 P{ 1 H} NMR spectra. For complexes 1c, 2c and 3c, black precipitate developed in 10 min and the 31 P{ 1 H} NMR spectra showed that the original nickel complexes decomposed completely; neither CH 3 OBcat nor catBOBcat were detected by the NMR spectra.  Flask reactions were carried out to further evaluate the catalytic activities of these complexes. Typically, the nickel complex (0.01 mmol), HBcat (5 mmol) and C 6 (CH 3 ) 6 (0.02 mmol, internal standard) were dissolved in benzene-d 6 (4 mL) under a nitrogen atmosphere. The solution was stirred at room temperature and CO 2 was bubbled through the solution at the same time. The reaction was stopped after a large amount of white precipitate developed. The reaction mixture was filtered and the clear liquid was analyzed by NMR spectroscopy. When complexes 3a and 3b were used as the catalysts, no white precipitate developed and the reaction was stopped after 2 h.
The isolated white precipitate was confirmed to be catBOBcat [14,26,27,46,47]. The 11 B NMR spectra of the clear reaction solutions indicated that HBcat was completely (for the reactions catalyzed 1a, 1b, 2a or 2b) or partly (for the reaction catalyzed by 3b) converted to CH 3 OBcat and catBOBcat. No product was detected for the reaction catalyzed by 3a. Based on the 1 H NMR integrations of the CH 3 OBcat methyl resonance (3.36 ppm) and the C 6 (CH 3 ) 6 methyl resonance (2.11 ppm), turnover number (TON) was calculated for each of the reactions. Table 1 summarizes the results of the above catalytic reactions. A representative 1 H NMR spectrum of the clear reaction solution that was used to determine the TON value is shown in Figure 3.  It can be seen from Table 1 that the mercapto and azido complexes have quite similar activities in catalyzing the hydroboration of CO 2 with HBcat and the catalytic activities are comparable to those of the corresponding palladium [26] and nickel [27] thiolato complexes reported previously. Complexes with isopropyl-substituted phosphine arms are more active than the corresponding complexes with tert-butyl substituted phosphine arms. This is in good agreement with the previous observation for the hydroboration of CO 2 with HBcat using the related nickel thiolato complexes as the catalysts [26,27]. Isothiocyanato complexes are far less active than the corresponding mercapto and azido complexes.
We previously reported that POCOP pincer ligated nickel mercapto complexes with t Bu 2 P or i Pr 2 P arms (1a and 1b) can interact with HBcat at room temperature to generate POCOP pincer nickel hydride species [38]. In order to obtain more information, we checked the interactions of the other complexes (1c, 2a-c and 3a-c) with HBcat. The nickel complexes were treated with an excess amount of HBcat in C 6 D 6 at room temperature in a sealed NMR tube and the reactions were monitored by NMR. For the interaction of 2a or 2b with HBcat, the hydride species [2,6-(R 2 PO) 2 C 6 H 3 ]NiH [48] was detected by the 31 P{ 1 H} NMR spectra in 20 min. For the interaction of 3b with HBcat, the hydride species was also observed in 2 h. However, no reaction was detected for the interaction of 3a with HBcat after 24 h. When 1c, 2c or 3c was treated with HBcat, black precipitate developed in 20 min and the original complex decomposed partly as evidenced by the 31 P{ 1 H} NMR spectra.
It should be noted that POCOP pincer nickel hydride species with entirely phenyl substituted phosphine arms cannot be synthesized because of the instability of this type of hydride complexes [48]. Therefore, the hydride species were not likely formed for the interaction of 1c, 2c or 3c with HBcat.
The above experimental results are consistent with the speculation that these different types of pincer nickel complexes are pre-catalysts for the hydroboration of CO 2 with HBcat and the in situ generated pincer nickel hydrides are the actual active species to catalyze this reaction. Any POCOP pincer nickel complexes that can generate POCOP pincer nickel hydrides under the catalytic conditions are active catalysts for the hydroboration of CO 2 with catecholborane. The easier the hydride species can be generated, the more active the catalyst. For the hydroboration of carbon dioxide with HBcat using pure nickel hydrides as the catalysts, less bulky substituents on the phosphine arms make the catalysts less active because of the formation of an inactive complex outside the catalytic cycle [14][15][16]. However, for the hydroboration of CO 2 with HBcat using a pre-catalyst, such as POCOP pincer ligated nickel thiolato, mercapto, azido or isothiocyanato complex, the catalytic activity is also dependent on the ease or the difficulty to generate the active hydride species. Smaller substituents on the phosphine arms make it easier to generate the active hydride species, although other issues, such as the electronic nature of the substituents on the phosphine arms [39,49,50] and the bond disassociation energy of the Ni-S or Ni-N bond [34,35], may also play important roles. On the other hand, the in situ generated other species, such as thiolato, azido and isothiosyanato species, may act as co-catalysts for this hydroboration reaction.

General Information
All manipulations were performed under an inert gas atmosphere. Solvents were degassed and dried before use. C 6 D 6 was distilled from sodium and benzophenone before use. A Bruker Advance 400 MHz spectrometer (Swiss Bruker Corporation, Faellanden, Switzerland) was used for the NMR studies. For 1 H and 13 C NMR spectra, the residual solvent resonances were used to calibrate the chemical shift values internally. For 31 P and 11 B NMR spectra, H 3 PO 4 (85%) and BF 3 ·Et 2 O were used, respectively, to reference the chemical shift (δ 0) externally. Complexes [2,6-(Ph 2 PO) 2 C 6 H 3 ]NiCl] [51], 1a [37], 1b [38], 2a-c [35] and 3a-c [35] were prepared as described in the literatures.

Synthesis of [2,6-(Ph 2 PO) 2 C 6 H 3 ]NiSH (1c)
THF (10 mL) and MeOH (10 mL) were added to a flask containing [2,6-(Ph 2 PO) 2 C 6 H 3 ]NiCl (572 mg, 1 mmol) and NaSH (280 mg, 5 mmol). The flask was then sealed and the resulting mixture in the flask was stirred at room temperature. After 48 h, the reaction was stopped and the volatiles were removed under vacuum. The residue was extracted with toluene and filtered through a pad of Celite. Complex 1c was obtained as an orange solid (341 mg, 60% yield) after the solvent of the combined extractions were removed under vacuum. 1  Single crystals of 1c used for X-ray diffraction analysis were obtained from n-hexane/CH 2 Cl 2 solution at −10 • C. The identities of the resulting crystals were confirmed by 31 P{ 1 H} NMR spectroscopy before the X-ray diffraction analysis. The intensity data were collected at 103 K on a Bruker SMART6000 CCD diffractometer. A graphite-monochromated Mo Kα radiation (λ= 0.71073 Å) was used for the data collection. Details for the data processing and structure refinement were similar to our previously reported procedures [37]. A summary of crystallographic data and structure refinement for complex 1c is provided in Table 2.
CCDC 1869263 contains the supplementary crystallographic data for complex 1c. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre (The Cambridge Structural Database, Cambridge, UK).

Procedure for the Catalytic Hydroboration of CO 2
Flask experiments for the catalytic hydroboration of carbon dioxide were carried out at room temperature under an atmospheric pressure of carbon dioxide in benzene-d 6 with a catalyst-to-substrate ratio of 1:500. The nickel catalyst (0.01 mmol), HBcat (5.00 mmol), hexamethylbenzene (0.02 mmol) and benzene-d 6 (4 mL) were mixed in a flame-dried 50 mL Schlenk flask under N 2 . Then CO 2 was bubbled through the solution until a large amount of white precipitate developed. The mixture was stirred for an additional 2 min. The resulting white precipitate was then allowed to settle, and 0.6 mL of the clear liquid was transferred under nitrogen into a NMR tube. 11 B and 1 H NMR spectra were then recorded. The TON numbers were calculated based on the 1 H NMR integrations of the CH 3 OBcat methyl resonance at 3.36 ppm and the C 6 (CH 3 ) 6 (internal standard) methyl resonance at 2.11 ppm. The conversions were calculated from the 11 B NMR spectra.

Conclusions
In summary, we have applied three different types of POCOP pincer nickel complexes (i.e., mercapto complex, azido complex and isothiosyanato complex) to the catalytic hydroboration of carbon dioxide with catecholborane. CO 2 was successfully reduced to a methanol derivative (CH 3 OBcat) under an atmospheric pressure of CO 2 at room temperature with TOFs up to 1908 h −1 . It was found that pincer complexes with isopropyl-substituted phosphine arms are more active than the corresponding complexes with tert-butyl substituted phosphine arms. Complexes with phenyl-substituted phosphine arms decomposed under the catalytic conditions and failed to catalyze the reactions. Of the three types of complexes, the isothiosyanato complex is far less active than the corresponding mercapto and azido complexes. It is concluded that all of these catalytically active complexes are catalyst precursors which generate the nickel hydride complex [2,6-(R 2 PO) 2 C 6 H 3 ]NiH in situ under the catalytic conditions, and the nickel hydride complex is the active species to catalyze this reaction.

Conflicts of Interest:
The authors declare no conflict of interest. The funder had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.