Case Study of N‐ i Pr versus N‐Mes Substituted NHC Ligands in Nickel Chemistry: The Coordination and Cyclotrimerization of Alkynes at [Ni(NHC)2]

Abstract A case study on the effect of the employment of two different NHC ligands in complexes [Ni(NHC)2] (NHC= i Pr2ImMe 1Me , Mes2Im 2) and their behavior towards alkynes is reported. The reaction of a mixture of [Ni2( i Pr2ImMe)4(μ‐(η 2  : η2)‐COD)] B/ [Ni( i Pr2ImMe)2(η4‐COD)] B’ or [Ni(Mes2Im)2] 2, respectively, with alkynes afforded complexes [Ni(NHC)2(η2‐alkyne)] (NHC= i Pr2ImMe: alkyne=MeC≡CMe 3, H7C3C≡CC3H7 4, PhC≡CPh 5, MeOOCC≡CCOOMe 6, Me3SiC≡CSiMe3 7, PhC≡CMe 8, HC≡CC3H7 9, HC≡CPh 10, HC≡C(p‐Tol) 11, HC≡C(4‐ t Bu‐C6H4) 12, HC≡CCOOMe 13; NHC=Mes2Im: alkyne=MeC≡CMe 14, MeOOCC≡CCOOMe 15, PhC≡CMe 16, HC≡C(4‐ t Bu‐C6H4) 17, HC≡CCOOMe 18). Unusual rearrangement products 11 a and 12 a were identified for the complexes of the terminal alkynes HC≡C(p‐Tol) and HC≡C(4‐ t Bu‐C6H4), 11 and 12, which were formed by addition of a C−H bond of one of the NHC N‐ i Pr methyl groups to the C≡C triple bond of the coordinated alkyne. Complex 2 catalyzes the cyclotrimerization of 2‐butyne, 4‐octyne, diphenylacetylene, dimethyl acetylendicarboxylate, 1‐pentyne, phenylacetylene and methyl propiolate at ambient conditions, whereas 1Me is not a good catalyst. The reaction of 2 with 2‐butyne was monitored in some detail, which led to a mechanistic proposal for the cyclotrimerization at [Ni(NHC)2]. DFT calculations reveal that the differences between 1M e and 2 for alkyne cyclotrimerization lie in the energy profile of the initiation steps, which is very shallow for 2, and each step is associated with only a moderate energy change. The higher stability of 3 compared to 14 is attributed to a better electron transfer from the NHC to the metal to the alkyne ligand for the N‐alkyl substituted NHC, to enhanced Ni‐alkyne backbonding due to a smaller CNHC−Ni−CNHC bite angle, and to less steric repulsion of the smaller NHC i Pr2ImMe.


Introduction
Transition-metal-catalyzed [2 + 2 + 2] cycloaddition reactions are elegant, atom-efficient and group tolerant processes which involve the formation of several CÀ C bonds in a single step. [1] These reactions offer the convenient access to a wide variety of carbocycles and heterocycles, mostly aromatic, starting from simple and inexpensive substrates. [1] After Reppe et al. provided their pioneering report on the first cyclopolymerization of acetylene using a mixture of NiBr 2 and CaC 2 as the precatalyst, [2] many different unsaturated substrates such as alkynes, diynes, alkenes, imines, isocyanates, isothiocyanates and CO 2 were transformed in cycloaddition reactions to yield highly substituted derivatives of benzenes, pyridines, pyridones, pyrones, thiopyridones and cyclohexanes. Since then, catalytic systems such as NiBr 2 /dppe in the presence of Zn powder or [Ni(COD) 2 ]based systems have been applied to many substrates. [1b-l,3] Nickel complexes of N-heterocyclic carbenes (NHCs) were also explored in cycloaddition reactions in the last two decades, mainly by Louie [3a,b] and Montgomery [3c,4] and co-workers. The Louie group commonly employed an in situ prepared catalyst system using [Ni(COD) 2 ] as a nickel source and two equivalents of a sterically bulky and electron rich NHC ligand such as Dipp 2 Im (= 1, 3-bis{2,6-di-iso-propylphenyl}-imidazolin-2-ylidene) or Dipp 2 Im H2 (= 1,3-bis{2,6-di-iso-propylphenyl}-imidazolidin-2ylidene), that supposedly forms complexes of the type [Ni-(NHC) 2 ] or [Ni(NHC)] as the pre-catalyst. These catalyst systems are highly efficient in the cyclization of different carbohydrates such as diynes or alkynes with ketones, aldehydes, nitriles, isocyanates and other substrates. [3a,b,5] For example, the cycloaddition of alkynes or diynes with isocyanates to afford 2pyridones and pyrimidinediones is highly efficient and occurs with a high degree of chemoselectivity if a 1 : 1 mixture of [Ni(COD) 2 ]/Dipp 2 Im H2 was used as catalyst. [6] For this Ni/NHCcatalyst system, alkyne cyclotrimerization was largely inhibited. [6] However, differences in reactivity, yield, and selectivity have been observed in these Ni/NHC-catalyzed cycloaddition reactions depending on the NHC ligand applied. The influence of the electronic and steric properties of the NHC ligand employed, for example Dipp 2 Im vs. Dipp 2 Im H2 vs. Mes 2 Im (= 1,3-dimesitylimidazolin-2-ylidene), to different cyclization reactions seems currently not to be completely understood. [7] However, Montgomery et al. demonstrated that stereo-electronic properties of NHC ligands play a crucial role for the regioselectivity observed for related nickel catalyzed allene hydrosilylation and reductive coupling reactions of aldehydes and alkynes. [8,9] The regioselectivity of the latter is supposedly controlled by steric repulsion between the NHC ligand and the alkyne substituents in the first, rate determining oxidative addition step. [9e] We reported earlier that complexes [Ni 2 (NHC) 4 (μ-(η 2 : η 2 )-COD)] of alkyl substituted NHCs such as i Pr 2 Im (= 1,3-di-isopropyl-imidazolin-2-ylidene) or n Pr 2 Im, which act as a source of [Ni(NHC) 2 ], are efficient catalysts for the insertion of diphenyl acetylene into the CÀ C bond of biphenylene leading to 9, 10di(phenyl)phenanthrene. [10] The reaction rate of the formation of 9,10-di(phenyl)phenanthrene depends on the steric demand of the NHC employed, giving the highest rates for the sterically most hindered NHC used. However, alkyne cyclooligomerization was suppressed at the reaction conditions employed (60-80°C) for diphenyl acetylene, but excess of other alkynes (3-hexyne or 2-butyne) afforded traces of the cyclooligomerization product. As we are currently interested to evaluate the differences in the reactivity of complexes [Ni(NHC) 2 ] of NHCs of different size, [11] we decided to (re-)evaluate the reactivity of complexes [Ni-(NHC) 2 ] with alkynes in some detail.
As all the work presented so far point to a decisive role of the sterics of the NHC ligand, we decided to reduce the steric demand of the N-aryl substituted NHC on going from Dipp to Mes substituted NHC and to increase the steric demand of the N-alkyl substituted NHC by backbone methylation. It has been demonstrated previously that backbone substitution at the C4 and C5 position of the imidazole framework, for example by methylation, greatly effects the stereo-electronics of the NHC ligands as repulsion between the C4/C5 methyl group and the N-organyl substituent leads to smaller C carbene -N-C substituent angles. [7,12] Thus, the NHCs we use for this study are Mes 2 Im and i Pr 2 Im Me (= 1,3-di-iso-propyl-4,5-dimethylimidazolin-2-ylidene).

Results and Discussion
The reaction pathways and the results of key-processes in transition metal chemistry and catalysis, such as oxidative addition, reductive elimination, migratory insertion, transmetalation, and β-hydride elimination, depend decisively on the sterics of the (NHC) co-ligands used and on the degree of electron transfer from the metal to the substrates and thus to the nature, sterics and number of co-ligands. [13] We recently investigated differences in the reactivity of the NHC-stabilized nickel(0) complexes [Ni 2 ( i Pr 2 Im) 4 (μ-(η 2 : η 2 )-COD)] A [10] as a source of [Ni( i Pr 2 Im) 2 ] 1 and [Ni(Mes 2 Im) 2 ] 2 in some detail. [11] In course of our work on CÀ F bond activation and catalytic defluoroborylation of polyfluoroarenes using the complexes A [14] and 2, [15] we provided evidence from experiment and theory that, depending on the NHC ligand used, the insertion of [Ni(NHC) 2 ] into the CÀ F bond of hexafluorobenzene proceeds via a concerted oxidative addition pathway for the small NHC i Pr 2 Im and via a radical pathway for the more bulky NHC Mes 2 Im. Additionally, we found for both mechanisms a competitive NHC-assisted reaction pathway which seems to be of general importance in transition metal NHC chemistry. [11a] Furthermore, we provided a detailed study on the steric influence of NHCs of different size on the stabilization of nickel π-complexes, since such complexes are very important intermediates in many different catalytic cycles. [16] Therefore we investigated the reaction of [Ni 2 ( i Pr 2 Im) 4 (μ-(η 2 : η 2 )-COD)] A, i.e., [Ni( i Pr 2 Im) 2 ] 1, and [Ni(Mes 2 Im) 2 ] 2 with different olefines, aldehydes and ketones, which led to the formation of complexes of the type [Ni-(NHC) 2 (η 2 -R 2 C=CR 2 )], [Ni(NHC) 2 (η 2 -O=CHR)] and [Ni(NHC) 2 (η 2 -O=CR 2 )]. Whereas A readily formed alkene complexes with olefins of different size, complex 2 reacted only with the smallest olefin ethylene or with activated acceptor olefins such as acrylates. Thus, the NHC nitrogen substituent influences the reactivity for steric reasons. However, these studies also pointed at the fact that substrate binding and electron transfer in bis-NHC nickel complexes can be fine-tuned very well beyond the accessibility of the metal center by steric protection and complex stability with respect to co-ligand or NHC dissociation. A subtle influence of sterics to the electronic behavior of [Ni(NHC) 2 ] lies in the C NHC À MÀ C NHC bite-angle the NHC ligands will adopt in the final product and in the propensity of the complexes [Ni(NHC) 2 ] to get involved into radical electron transfer processes. [17] Herein we want to expand this study on the reactivity of NHC-stabilized nickel complexes towards simple alkynes using [Ni(Mes 2 Im) 2 ] 2 and suitable sources of [Ni( i Pr 2 Im Me ) 2 ] 1 Me . As mentioned above, we reported some alkyne complexes [Ni( i Pr 2 Im) 2 (η 2 -RÀ C�CÀ R') starting from [Ni-( i Pr 2 Im) 2 ] 1, earlier, [10,18] which were included in this study if appropriate.
The complex [Ni 2 ( i Pr 2 Im Me ) 4 (μ-(η 2 : η 2 )-COD)] B of the backbone methylated NHC i Pr 2 Im Me was synthesized -as reported for A -from the reaction of [Ni(COD) 2 ] with two equivalents of i Pr 2 Im Me (Scheme 1). As observed for A, the yellow solid obtained consists of two complexes, the dinuclear reaction product B and the mononuclear complex [Ni( i Pr 2 Im Me ) 2 (η 4 -COD)] B' as a by-product in various amounts (up to approximately 40 %). As B and B' typically show identical reactivity with respect to alkynes (the same was observed previously for A and its mononuclear counterpart [Ni( i Pr 2 Im) 2 (η 4 -COD)]), we did not further purify the mixture for the following reactions.
Dinuclear B and mononuclear B' can be distinguished easily in their 1 H and 13 C{ 1 H} NMR spectra. The resonances of the NHC ligand of B were detected as a broad doublet at 1.42 ppm for the iso-propyl methyl protons, a singlet at 1.88 ppm for the backbone methyl protons and a septet at 6.03 ppm for the isopropyl methine protons, whereas sharp resonances were found for the NHC ligand of complex B' at 1.33 ppm (d), 1.86 ppm (s) and 5.90 ppm (sept.). In the 13 C{ 1 H} NMR spectra the resonances
The complexes 3-13 were isolated as yellow or orange-red, air and moisture sensitive powders and were characterized using 1 H NMR, 13 C{ 1 H} NMR and IR spectroscopy (see Supporting Information). The complexes were obtained as analytically pure material except for the complexes of the terminal alkynes 1pentyne and phenylacetylene, [Ni( i Pr 2 Im Me ) 2 (η 2 -HC�CC 3 H 7 )] 9 and [Ni( i Pr 2 Im Me ) 2 (η 2 -HC�CPh)] 10, which are only stable in solution and decompose upon removal of the solvent. The reactions of B/B' with alkynes proceeded in quantitative yield if performed on NMR scale; the yield of isolated 7, however, is rather low due to losses in the crystallization process to get analytically pure material. Important 1 H and 13 C{ 1 H} NMR data of the compounds 3-13 are summarized in Table 1. In the 1 H NMR and 13 C{ 1 H} NMR spectra the signals for the NHC ligands were observed in the typical regions expected, and for the complexes 8-13 of unsymmetrical or terminal alkynes the set of NHC resonances is doubled due to a lowering of the complexes' symmetry. Each alkyne proton of 9-13 is shifted upon
Many of the complexes 3-13 are unstable upon heating and the result of thermal exposure in solution depends on the alkyne ligand coordinated. While [Ni( i Pr 2 Im Me ) 2 (η 2 -PhC�CPh)] 5 and [Ni( i Pr 2 Im Me ) 2 (η 2 -MeOOCC�CCOOMe)] 6 are stable in solution at 100°C for days, complexes [Ni( i Pr 2 Im Me ) 2 (η 2 -MeC�CMe)] 3 and [Ni( i Pr 2 Im Me ) 2 (η 2 -HC�CPh)] 10 decompose already at room temperature, but much more rapidly upon heating with formation of so far unidentified products. Although we could not identify many of the decomposition products, for the thermal decomposition of the terminal alkyne complexes [Ni( i Pr 2 Im Me ) 2 (η 2 -HC�C(p-Tol))] 11 and [Ni( i Pr 2 Im Me ) 2 (η 2 -HC�C(4-t Bu-C 6 H 4 ))] 12 we characterized the rearrangement products 11 a and 12 a (Scheme 2 and Figure 3) after heating of benzene or toluene solutions of these complexes to 60°C for 72 h. In addition to 11 a or 12 a other, so far unidentified sideproducts were formed. However, the complexes 11 a and 12 a result from an interesting addition of a CÀ H bond of one of the NHC N-iso-propyl substituent methyl groups across the C�C triple bond of the coordinated alkyne (Scheme 2).
We reported recently that NHC ligands are no good spectator ligands in cobalt NHC half sandwich alkyne chemistry and that they react in the coordination sphere of cobalt with terminal alkynes under coupling of the NHC and the alkyne ligand. [23a] Related decomposition pathways involving coordinated alkynes and NHC ligands are also known. [23] For the alkyne complexes of [Ni(NHC) 2 ] we did not observe this kind of NHC alkyne coupling so far, but the complexes 11 a and 12 a were formed via an intramolecular CÀ C coupling reaction of the NHC N-substituent. Formally, a hydrogen atom is transferred from the nearest N-iso-propyl methyl group of the NHC ligand to the coordinated alkyne carbon atom. The terminal alkyne carbon thus couples with the iso-propyl methyl carbon with formation of a 6-membered metallacycle and reduction of the C�C triple bond to an η 2 -(C,C)-coordinated alkene.

Chemistry-A European Journal
Full Paper doi.org/10.1002/chem.202103093 Red crystals of compound 11 a were isolated for a complete characterization of this complex including X-ray analysis, while 12 a was only characterized in situ via the characteristic 1 H NMR resonances in the NMR spectrum (see Figure 3, for the full NMR spectra see Supporting Information Figures S34-S40). In each case, the resonances of the olefinic protons of 11 a and 12 a were detected as a doublet at 3.85 ppm (C=CHR) for the proton at C 1 (see Scheme 2 and Figure 3) and a doublet of doublets of doublets at 2.91 ppm for the proton at C 2 . The two diastereotopic protons of the CH 2 group at C 3 give rise to two separate resonances at 2.64 ppm (ddd) and 2.78 ppm (ddd), while the former iPr methine proton was detected as a broad multiplet at 3.99 ppm. The three remaining iso-propyl methine protons of the NHC ligands give rise to three partially overlapping and broadened septets in the range between 5.30 ppm and 5.90 ppm. In the 13 C{ 1 H} NMR spectrum of complex 11 a the resonances of the olefinic carbon atoms are shifted towards higher fields compared to complex 11 and were detected at 34.1 ppm (C 2 ) and 51.9 ppm (C 1 ). The signals for the C 3 carbon atom and the former iso-propyl methine carbon C 4 were observed at 40.2 and 54.1 ppm, respectively. The carbene carbon atom resonance of the NHC ligand involved in the metallacycle is also shifted to higher fields at 191.7 ppm, whereas the resonance of the unaffected NHC carbon atom was found at 204.5 ppm.

Chemistry-A European Journal
Full Paper doi.org /10.1002/chem.202103093 line with Ni-C olefin distances observed for A and B and related compounds. The C1À C2 separation of 1.439(2) Å is consistent with C=C bond lengths observed for other [Ni(NHC) 2 (η 2 -olefin)] complexes. [11b] The nickel atom, the olefin carbon atoms C1, C2 and the NHC carbon atom C7 are perfectly aligned in a plane and the intact NHC ligand is nearly perfectly perpendicular oriented to this plane (88.58(9)°). The NHC ligand of the metallacycle (i.e., plane N3À C6À N4) is twisted towards the plane C1À Ni1À C2 with an angle of 32.51(11)°. The olefin adopts transconfiguration with angles of 121.19(13)°(C1À C2À C3) and 123.28(13)°(C2À C1À C8) between the C=C-bond vector and the substituents. The C2À C3 distance of the new bond between the olefin and the iso-propyl carbon atom is 1.516(2) Å and thus clearly a single bond. The 6-membered metallacycle adopts a distorted chair-conformation.
Scheme 3 sketches two reasonable reaction pathways for the rearrangement of [Ni( i Pr 2 Im Me ) 2 (η 2 -HC�C(p-Tol))] 11 to product 11 a. The first pathway (i) involves the rearrangement of the terminal alkyne ligand to a nickel vinylidene complex along the typical hydrido alkinyl route, which occurs with insertion of nickel into the CÀ H bond of the coordinated terminal alkyne ligand and subsequent hydride rearrangement to the β-C atom. [24] Insertion of the vinylidene into the NHC methyl CÀ H bond would lead then to complex 11 a. Another likely pathway (ii) involves a concerted or nickel mediated addition of the NHC methyl CÀ H bond across the C�C triple bond of the coordinated alkyne. DFT calculations (BP86//def2-TZVP(Ni)/def2-SVP(C,N,H)) reveal first of all that the rearrangement of 11 to yield 11 a is a surprisingly strong exothermic process (~E = À 102.6 kJ/mol), and that the corresponding nickel hydrido alkinyl (+ 65.5 kJ/mol) and nickel vinylidene (+ 49.2 kJ/ mol) complexes are significantly above the alkyne complex in energy, so that the barrier of process (i) is at least + 65.5 kJ/mol. For the pathway (ii), we investigated either a concerted or a nickel mediated CÀ H addition to the coordinated alkyne. However, we were not able to locate any transition state here and every attempt to model likely nickel hydrido intermediates resulted in the ground state geometry of 11 a. As DFT calculations gave no conclusive answer, we prepared complex 11 using deuterated p-tolylacetylene and repeated the rearrangement with the resulting complex 11-D. As shown in Scheme 3, the deuterium label of 11-D should appear in the final product at different positions, depending on the pathway involved. The vinylidene pathway should lead to deuterium at the former β-position of the coordinated alkyne (H atom marked in red in Scheme 3), the concerted/nickel-mediated addition should lead to deuterium at the former α-position of the coordinated alkyne (H atom marked in blue in Scheme 3). The result of the deuteration experiment revealed that the deuterium atom stays at the α-carbon atom C 2 (see Scheme 2 and Figure S40 of the Supporting Information) and therefore it is likely that the complexes 11 a and 12 a are formed according to a concerted or nickel-mediated CÀ H bond activation pathway with addition of the NHC methyl CÀ H bond to the triple bond, in accordance with pathway (ii) of Scheme 3.
As it is known that [Ni(NHC) 2 ] catalysts for cyclooligomerization reactions were prepared in situ from [Ni(COD) 2 ] and a bulky and electron rich NHC ligand such as Dipp 2 Im, Dipp 2 Im H2 or Mes 2 Im, [3b] we reacted isolated [Ni(Mes 2 Im) 2 ] 2 with alkynes. Initial NMR experiments revealed that complex 2 cyclotrimerizes 2-butyne quantitatively and therefore we investigated the catalytic activity and stereoselectivity of complex 2 in cyclotrimerization reactions using different internal and terminal alkynes (see Table 2). NMR spectra of the reactions of 2-butyne, 4-octyne, diphenylacetylene, dimethyl acetylendicarboxylate, 1pentyne, phenylacetylene and methyl propiolate with 5 mol% of [Ni(Mes 2 Im) 2 ] 2 in C 6 D 6 at 60°C were recorded and the consumption of the alkynes was monitored. The catalyst was then removed by filtration over a pad of silica gel and the products were analyzed using 1 H and 13 C{ 1 H} NMR spectroscopy as well as GC/MS. In all cases the cyclotrimerization of internal alkynes proceeded in quantitative yield on NMR scale (isolated yields were only determined for the preparation of hexaphenylbenzene, in this case the TON is 30) and no formation of side-products was detected, with exception of the cyclotrimerization of 1-pentyne, where traces of tetramerization products were observed (see Supporting Information). The reactions with terminal alkynes did not show any specific stereoselectivity and afforded mixtures of the 1,2,4-and 1,3,5-stereoisomers. The exact determination of the product ratio via integration of the 1 H NMR spectrum was only possible for the reaction of methyl propiolate due to overlapping NMR resonances for the products of the other alkynes. The use of internal alkynes yielded hexasubstituted benzene derivatives, and the cyclotrimerization of diphenylacetylene to give hexaphenylbenzene proceeded much faster compared to the cyclotrimerization of other alkynes (entry 3, Table 2). This reaction was finished after five minutes at room temperature using a small catalyst load of just 1 mol%. As the product is almost insoluble in C 6 D 6 it was isolated directly from the NMR tube as a colorless solid in 88 % yield.
To gain further insight into the mechanistic details of the catalysis we analyzed the reaction of 2 with 2-butyne. Therefore, we initially performed the reaction of 2 with a slight excess of

Chemistry-A European Journal
Full Paper doi.org/10.1002/chem.202103093 2-butyne (4 equiv.) in a Young NMR tube (see Figure 5a). Addition of the alkyne led to an immediate color change from deep violet, which is the color of 2, to bright yellow, which rapidly darkened after a few seconds. The analysis of the reaction mixture by NMR spectroscopy after five minutes at room temperature revealed the formation of hexamethylbenzene, free NHC Mes 2 Im and a new, well defined complex E. After 4 d at room temperature, some re-formation of complex 2 was observed, resonances of free Mes 2 Im were still detectable and the signals assigned to complex E started to decrease. Finally, heating of the sample for 4 h at 60°C led to a complete disappearance of the resonances for the NHC and for complex E and a full recovery of complex 2 plus the final cyclotrimerization product hexamethylbenzene was observed. The presence of uncoordinated carbene in the solution indicates that complex E might be a mono-NHC complex [(Mes 2 Im)Ni(η 6 -C 6 Me 6 )] E, stabilized by hexamethylbenzene. A similar arene-stabilized complex has been reported previously by Ogoshi et al. [25] for a larger NHC, i.e., [Ni(Dipp 2 Im)(η 6 -C 6 H 5 Me)]. Despite of several attempts, we were not able to isolate this complex. Furthermore, the absence of 2-butyne after five minutes at room temperature indicates that oligomerization proceeds very fast and quantitatively. To learn more details about this process, especially at which temperature the catalysis sets in, we additionally performed a variable temperature NMR experiment of the reaction from À 40°C to + 60°C in steps of 10°C (see Figure 5b). At À 40°C, the reaction mixture had a bright yellow color and the NMR spectrum showed the formation of the alkyne complex [Ni(Mes 2 Im) 2 (η 2 -MeC�CMe)] 14 (see below), similar as observed for complex B/B' with the smaller NHC ligand. Resonances of the trimerization product, free Mes 2 Im as well as the signals of complex E were already detected at temperatures of 0-10°C. Integration of the resonances was consistent with the formation of a mono-NHC arene complex [(Mes 2 Im)Ni(η 6 -C 6 Me 6 )]. After raising the temperature to 40°C, the alkyne was completely consumed, the resonance of hexamethylbenzene increased and both, the NHC Mes 2 Im as well as the complex [(Mes 2 Im)Ni(η 6 -C 6 Me 6 )] E, were detected. Finally, at 60°C, the recovery of complex 2 and the decrease of the resonances of the uncoordinated NHC and the mono-NHC complex E occurred. We also performed the reaction of [Ni( i Pr 2 Im Me ) 2 (η 2 -MeC�CMe)] 3 with an excess of 2-butyne, to see if 1 Me is also suitable for the catalytic trimerization. In contrast to complex 2 no cyclization was observed after 20 h at room temperature, but heating the reaction mixture to higher temperatures of 80°C and above led to slow transformation of 2-butyne to hexamethylbenzene.

Chemistry-A European Journal
Full Paper doi.org /10.1002/chem.202103093 observed previously for olefin complexes, [11b] the steric bulk of the NHC ligand Mes 2 Im of complex 2 limits the coordination of a third ligand to the nickel atom, which is in stark contrast to the behavior of complexes 1/1 Me . Alkynes with electron withdrawing substituents increase π-backbonding from the nickel atom to the alkyne and increase the stability of the alkyne complex in solution at room temperature. As noticed above, alkyl and/or aryl substituted alkynes lead to decomposition of the alkyne complexes with extrusion of one NHC ligand at temperatures slightly above 0°C. Unlike the complexes 3-13, the NMR spectra of the compounds 14-18 reveal remarkably broadened resonances for the bulkier NHC ligand Mes 2 Im due to hindered rotation, as it was previously reported by us for similar π-complexes with ketone or aldehyde ligands. [11b] Even the low temperature NMR spectra of 14, 16 and 17 revealed some signal broadening. Nevertheless, all characteristic resonances were assigned and the integration of the resonances is consistent with the presence of one alkyne ligand per two NHC ligands in complexes of the type [Ni(Mes 2 Im) 2 (η 2 -alkyne)]. Important 1 H and 13 C{ 1 H} NMR data of 14-18 are summarized in Table 3. In the 1 H NMR spectra of 14-18 the ortho and para mesityl methyl protons give rise to up to four broadened resonances in the range between 1.74 and 2.37 ppm. The alkyne protons of the compounds 17 and 18 each can be observed as a singlet at 6.11 ppm (17) and 6.94 ppm (18). In the 13 C{ 1 H} NMR spectra the signals of the carbene carbon atoms were detected in the range between 198.2 and 207.0 ppm. The resonances of the alkyne carbon atoms are shifted to lower fields upon coordination and were observed in the range between 118.6 and 136.7 ppm. The ν C�C stretching vibrations of the complexes 14-18 are shifted to lower wavenumbers in the range between 1701 cm À 1 and 1808 cm À 1 . Crystals suitable for X-ray diffraction of 14, 15, 16 and 17 were obtained by either storing a saturated solution of the complex in hexane or pentane at À 30°C or by layering a saturated benzene solution of the complex with hexane at room temperature (15). The molecular structures of 14, 15, 16 and 17 are provided in Figure 6 (for selected bond lengths and angles see Supporting Information Figures S6-S9). Important crystallographic data of these complexes and a comparison to the complexes [Ni(Me i PrIm) 2 (η 2 -PhC�CPh)] C, [10] [Ni( i Pr 2 Im) 2 (η 2 -MeC�CMe)] D, [10] [Ni( i Pr 2 Im Me ) 2 (η 2 -MeC�CMe)] 3, [Ni( i Pr 2 Im Me ) 2 (η 2 -

Chemistry-A European Journal
Full Paper doi.org /10.1002/chem.202103093 So what is the difference between [Ni( i Pr 2 Im) 2 ] 1 or [Ni-( i Pr 2 Im Me ) 2 ] 1 Me and [Ni(Mes 2 Im) 2 ] 2 in the behavior towards alkynes? All three compounds form alkyne complexes, but only the complexes of the sterically more encumbered Mes 2 Im ligand enter the catalytic cycle at ambient temperatures. To answer this question DFT calculations (BP86//def2-TZVP(Ni)/ def2-SVP(C,N,H)) have been performed on the initiation steps of the cyclotrimerization of 2-butyne with [Ni(NHC) 2 ] (NHC = i Pr 2 Im Me , Mes 2 Im; see Scheme 5). The results of these computations are given in Figure 7.
A comparison of the energy profile of the cyclotrimerization initiation steps of 2-butyne with [Ni( i Pr 2 Im Me ) 2 ] 1 Me (red) and [Ni(Mes 2 Im) 2 ] 2 (green) reveals that the profile is very shallow for 2 and each step is associated with a moderate energy change. The formation of the alkyne complexes [Ni( i Pr 2 Im Me ) 2 (η 2 -MeC�CMe)] 3 and [Ni(Mes 2 Im) 2 (η 2 -MeC�CMe)] 14 is connected with a very different energy gain, À 77.3 kJ/mol for 3 and only À 10.5 kJ/mol for 14. Assuming a dissociative process, the dissociation of the NHC ligand from [Ni(NHC) 2 (η 2 -MeC�CMe)] requires + 126.6 kJ/mol for the i Pr 2 Im Me complex, whereas for the Mes 2 Im complex only + 49 kJ/mol are needed. The attachment of another alkyne to [Ni(NHC)(η 2 -MeC�CMe)] is exothermic in both cases, À 41.3 kJ/mol for the formation of [Ni-(Mes 2 Im)(η 2 -MeC�CMe) 2 ] and À 83.1 kJ/mol for [Ni( i Pr 2 Im Me )(η 2 -MeC�CMe) 2 ]. Thus, the potential surface of the nickel complex [Ni(Mes 2 Im) 2 ] with both, low energy gain for alkyne addition and low energy loss for NHC dissociation, is nicely suited for catalysis, whereas for [Ni( i Pr 2 Im Me ) 2 ] 1 Me the alkyne complex seems to be too stable for further ligand loss (either alkyne or NHC) to enter a catalytic cycle at ambient temperatures.
As there is a distinct difference in the coordination of alkyne, specifically 2-butyne, to [Ni( i Pr 2 Im Me ) 2 ] 1 Me (red) and [Ni(Mes 2 Im) 2 ] 2 (green) it is interesting to track down the differences. Both ligands are different in their stereo-electronic features. For this purpose the steric demand of the NHCs i Pr 2 Im, i Pr 2 Im Me and Mes 2 Im expressed by their %V bur ("percent buried volume") [12,27] was re-evaluated on the basis of DFT geometry optimized structures (BP86//def2-TZVP(all)) of [(NHC)Ni(CO) 3 ]. With the aid of the Web application SambVca [28] we obtained % V bur ; values of i Pr 2 Im (26.5 %) < i Pr 2 Im Me (27.7 %) < Mes 2 Im (33.2 %), [29] for fixed NiÀ C carbene distances of 2.00 Å, which are perfectly in line with our experimental findings. The σ-donor and π-acceptor properties of the NHC ligands can be described via the TEP ("Tolman electronic parameter") [27,30] and 31 P or 77 Se NMR shifts of NHC phosphinidenes and selenourea compounds, [31] respectively. While our BP86//def2-TZVP(all)calculated TEP values reveal no significant differences for i Pr 2 Im (2054)~Mes 2 Im (2055)~iPr 2 Im Me (2056) in [(NHC)Ni(CO) 3 ], the πacceptor abilities of the NHCs increase in the order i Pr 2 Im (δ 31 P = À 61.2 ppm, δ 77 Se = À 3 ppm) < i Pr 2 Im Me (δ 77 Se = À 18 ppm) < Mes 2 Im (δ 31 P = À 23 ppm, δ 77 Se = + 27 ppm). [31] As a consequence of these different donor and acceptor properties of the NHC ligands used in [Ni(NHC) 2 ], different charges (see Figure 7; given are NBO charges) were calculated at nickel for the complexes [Ni( i Pr 2 Im Me ) 2 ] 1 Me (À 0.14) and [Ni(Mes 2 Im) 2 ] 2 (À 0.04). Thus, the nickel atom of 1 Me is more electron-rich compared to the metal atom of 2 and it should be expected that more electron density is transferred to the alkyne ligand of 1 Me . This is in line with the concept recently provided by Love and Kennepohl et al. for the stabilization of square planar d 10 nickel π-complexes bearing phosphine co-ligands. [32] These authors provided evidence that the stability of πcomplexes depends on the strength of the metal-to-ligand (alkyne or alkene) backbonding, which is critically influenced by charge transfer from the co-ligands (here the NHCs) via the metal atom to the π-acceptor ligand.
These expectations can be confirmed by the experimental data obtained for the complexes [Ni( i Pr 2 Im) 2 (η 2 -MeC�CMe)] D, [10]
In total, we attribute the much higher stability of [Ni-( i Pr 2 Im Me ) 2 (η 2 -MeC�CMe)] 3 with respect to [Ni(Mes 2 Im) 2 (η 2 -MeC�CMe)] 14 to three main reasons: (i) electron transfer from the NHC to the metal to the alkyne ligand is higher for the Nalkyl compared to the N-aryl substituted NHC ligands in [Ni(NHC) 2 (η 2 -MeC�CMe)] due to different electron donor/acceptor properties of the NHC ligand. (ii) Electron transfer from the metal center to the alkyne ligand is enhanced for the N-alkyl compared to the N-aryl substituted NHC ligands due to their different steric size, as smaller NHC ligands (such as i Pr 2 Im Me or i Pr 2 Im) can adopt smaller C NHC À NiÀ C NHC bite angles, which leads to increased π-backdonation to the alkyne. (iii) Ligand dissociation is facilitated for the complex of the sterically more encumbered NHC ligand, i.e., [Ni(Mes 2 Im) 2 (η 2 -MeC�CMe)] 14 loses the NHC ligand more readily than [Ni( i Pr 2 Im) 2 (η 2 -MeC�CMe)] D [10] and [Ni( i Pr 2 Im Me ) 2 (η 2 -MeC�CMe)] 3. All these factors lead to a significantly enhanced stability of the alkyne complexes of the N-alkyl substituted NHCs and are thus the reason why these complexes are not catalytically active for alkyne oligomerization at ambient temperatures.
Complex 2 cyclotrimerizes alkynes at ambient conditions, which is in contrast to the behavior found for 1 Me or 1. NMR exploration of the reaction of 2 with 2-butyne gave evidence for the formation of the complexes [(Mes 2 Im)Ni(η 6 -C 6 Me 6 )] E and [Ni(Mes 2 Im) 2 (η 2 -MeC�CMe)] 14 as intermediates of the reaction. A mechanism for the NHC-nickel catalyzed cyclotrimerization of 2-butyne was proposed, which involves coordination of the alkyne to [Ni(Mes 2 Im) 2 ] 2 to yield [Ni(Mes 2 Im) 2 (η 2 -MeC�CMe)] 14 and [Ni(Mes 2 Im)(η 2 -MeC�Me) 2 ] F with loss of one NHC ligand as the initiation step of the catalysis. The efficient steps of the catalytic cycle involve addition of 2-butyne to [Ni(Mes 2 Im)(η 2 -MeC�Me) 2 ] F with cyclization to yield [(Mes 2 Im)Ni(η 6 -C 6 Me 6 )] E and re-formation of F with arene release. The re-coordination of the NHC ligand to the nickel atom of E with elimination of the aromatic trimerization product and recovery of complex 2 at the end of the catalysis is the termination of the catalytic cycle.
This study demonstrates for the example of bis-NHC nickel alkyne complexes and their reactivity how valuable NHCs are in the fine-tuning of substrate binding, electron transfer and reactivity. Although the differences in the TEP of both NHCs under investigation is small, the differences in the electron transfer of the complexes [Ni(NHC) 2 ] to a coordinated substrate are quite impressive. The increase of the steric demand of the NHC lead, of course, to a different accessibility of the metal center (steric protection) and to different complex stabilities as co-ligand/NHC dissociation is facilitated for the bulkier ligand. But we also demonstrate here that steric properties of the NHC significantly influence the donor properties of [M(NHC) 2 ]moieties by the C NHC À MÀ C NHC bite-angle NHC ligands of different size can adopt in the final product. Furthermore, we have shown previously [11a,34] that simple electron-transfer processes are possible if the substrate cannot bind to a (sterically encumbered) complex [M(NHC) 2 ] and that thus radical processes dominate its reactivity and catalysis. We anticipate that, as shown herein, further tuning of the NHC stereo-electronics, keeping [M(NHC) 2 ] units intact, will lead to the (further) design of catalysts which enter different reaction channels for similar (or even same) starting materials.