Nickel-Catalyzed Csp2 -OMe Functionalization for Chemoselective Aromatic Homologation En Route to Nanographenes.

A Ni-catalyzed Csp2 -OMe ortho-functionalization methodology to form chemoselectively alkyne monoannulation or aromatic homologation products is reported as a novel protocol towards the valorisation of substrates containing Csp2 -OMe units. Double activation of Csp2 -OMe and Csp2 -F bonds is also demonstrated. Further use of aromatic homologation products towards the synthesis of nanographene-like compounds is described.

Introduction

Activation of Csp2−O bonds of aryl ethers by transition metals is much more difficult than those of aryl halides, thus their use as electrophilic counterparts in cross coupling protocols is much more limited. This is probably due to the reluctance of the C−OMe bond towards oxidative addition and the lower propensity of the methoxy residue to act as a leaving group. However, among all transition metals, Ni0‐catalyzed Csp2−O activation has become the methodology of choice to effect the cleavage of Csp2−OR (R=Me, Ph) bonds. Simple model Csp2−OMe substrates have been used as electrophiles in Kumada, Negishi and Suzuki cross coupling catalysis (Figure 1a). Also, the cleavage of Csp2−OMe to Csp2‐H has also been achieved via reductive protocols with silanes or in the presence of H2. On the other hand, Csp2−OPh bond activation have been recently engaged into Ni‐catalyzed C−O/N−H annulation of aromatic amides with alkynes for the production of isoquinolinones, by using Csp2−OPh substrates bearing N‐phenyl‐benzamide directing groups (Figure 1b). Concerning the mechanism of Csp2−OMe activation, it is generally believed that oxidative addition at nickel(0) species is the key step in the process, although current knowledge is still limited. In addition to the oxidative addition mechanism, non‐classical modes of activation should be taken into account depending on the nature of the nucleophile and the ligand used. For instance, the involvement of in situ formed nickel(I) species has also been proposed in Ni0(COD)2/PCy3‐catalyzed reductive cleavage of aryl ethers using hydrosilane.[ , ] A very relevant raw source of Csp2−OR moieties is lignin, which constitutes up to 30 wt % of wood‐based biomass and is considered the largest source of renewable aromatics. However, lignin decomposition is hampered by the difficulty in converting it into synthetically useful monomeric units by activation of the uniting Csp2−OR (R=Me, Alkyl, Ar) and C−C bonds. Nevertheless, the inert Csp2−O bonds of aryl ethers contained in lignin remain in the monomers and new methodologies are needed to unlock this bottleneck of the lignin valorisation process.

Figure 1

(a) Nickel‐catalyzed Csp2−OMe activation reaction using preactivated R−M nucleophiles to form Csp2−Csp2 coupling and reductive cleavage. (b) Directing group‐assisted nickel‐catalyzed Csp2−OR functionalization with alkynes.

Recently we reported a nickel‐catalyzed Csp2−F functionalization with internal alkynes to form either alkyne monoannulation or aromatic homologation products in a chemodivergent manner, demonstrating the ability of the 8‐aminoquinoline directing group to activate the strong Csp2−F bonds. Here we report an analogous Ni‐catalyzed methodology to form alkyne monoannulation or aromatic homologation products via the functionalization of strong Csp2−OMe bonds, as a novel protocol towards the valorization of lignin monomers containing Csp2−OMe units.

Results and Discussion

We first focused on whether Csp2−OMe could be functionalized with internal alkynes using a similar Ni0‐based methodology as for the Csp2−F substrates recently reported by us. We screened as directing group (DG) the 8‐aminoquinoline (8‐AQ)[ , ] (1 aa), 2‐pyridylisopropylamine (PIP) (1 ba) and 2‐pyridylmethylamine (PM) (1 ca) units in ortho‐position to the Csp2−OMe substrate and diphenylacetylene (Scheme 1). Only positive outcome of the reaction was found for 8‐AQ‐containing substrate 1 aa, affording the aromatic homologation product 2 aa in 47 % yield and the alkyne monoannulation product 3 aa in 27 % yield. These results point out to the requirement of a rigid and bidentate DG for the reaction to proceed.

Scheme 1

Screening of effective DG for the Ni‐catalyzed Csp2−OMe functionalization with diphenylacetylene (in parenthesis: isolated yields; nd=non detected).

The 2 aa/3 aa ratio of 1.7 obtained under these conditions was similar to the analogous Csp2−F cleavage to obtain the same products (2 aa/3 aa ratio=3.7), what suggested a similar mechanism of chemodivergent product formation. At this point, we pursued the optimization of the chemodivergence, and found that the 2 aa/3 aa ratio could be completely reversed; if, additionally to the standard conditions, LiOTf (2 equiv.) was added, the aromatic homologation product 2 aa was maximized to a 2 aa/3 aa ratio of 11.8 in a 59 % yield (3 h) (Scheme 2). However, when PPh3 (2 equiv.) was added in lieu of LiOTf, a switch of chemoselectivity occurred and 3 aa was obtained in 64 % yield within 3 h (2 aa/3 aa ratio=1 : 2.6). It is worth to note that by increasing the reaction time up to 24 h, the alkyne monoannulation product 3 aa was obtained in an exclusive manner due to the decomposition of 2 aa under these conditions.

Scheme 2

Chemodivergent behavior towards the aromatic homologation and the alkyne monoannulation product (2 aa/3 aa ratio in parenthesis).

Then, we turned our attention to substrates bearing a second −OMe group in para‐ (1 ab) and ortho‐ (1 ac) to the DG, in order to analyse the electronic and steric effects of the additional methoxy‐ group to the reaction outcome (Table 1). On the one hand, the yield for 2 ab (26 %) and 3 ab (22 %) decreased and the ratio 2 ab/3 ab was ∼1 when using substrate 1 ab. The same trend as in 1 aa was observed, and adding LiOTf, the formation of the aromatic homologation 2 ab was observed in a selective manner (32 % in 24 h). Instead, adding PPh3 the corresponding alkyne monoannulation product was formed in higher yields (70 % in 3 h). The higher reactivity observed towards the formation of the alkyne monoannulation product suggested that the aromatic homologation reaction was impeded by the steric effect of the second −OMe group in meta. On the other hand, although no reactivity was observed using 1 ac under the standard conditions (Table 1, entry 7), full chemoselectivity for the aromatic homologation product 2 ac was regained (48 % yield in 24 h, entry 8) when LiOTf was used as the additive.

Table 1

Csp2−OMe functionalization using different methoxyarene substrates.

Entry	Substrate 1 ax	Additives	Yield [%] of 2 ax [a]	Yield [%] of 3 ax [a]
1		without	47 %	27 %
2		LiOTf	38 % (59 %)	tr (tr)
3		PPh3	tr (25 %)	58 % (64 %)
4		without	26 % (tr)	22 % (tr)
5		LiOTf	32 % (tr)	tr (tr)
6		PPh3	tr (tr)	43 % (70 %)
7		without	tr	tr
8		LiOTf	48 % (26 %)	tr (tr)
9[b]		PPh3	–	–

[a] Yield calculated from 1H NMR of crude mixture using 1,3,5‐trimethoxybenzene as internal standard (in parenthesis, yield at 3 h); tr=traces. [b] Formation of product 4 in 25 % yield.

–

–

Strikingly, substrate 1 ac afforded the double activation of both −OMe groups in ortho to the DG when PPh3 was used, affording product 4 in 25 % yield (29 % using 50 mol% of Ni(COD)2), featuring the aromatic homologation and the alkyne monoannulation simultaneously (Figure 2a). Gratifyingly, the crystal structure of 4 was obtained by slow evaporation of a CHCl3 solution and consisted of a racemic mixture of the two enantiomeric helical species that arise from the isoquinolinone formation at the monoannulation step (Figure 2b).

Figure 2

(a) Double Csp2−OMe activation using substrate 1 ac to afford product 4 (yield based on alkyne). (b) Crystal structure of 4 (only shown one of the helical enantiomers of the racemic mixture).

We then focused our efforts in comparing the reactivity of the Csp2−F and Csp2−OMe groups. To this end, substrate 1 ad was synthesized bearing fluoride and methoxy groups in ortho to the amide motif (Scheme 3). Using the standard conditions and LiOTf, only the Csp2−F moiety was activated to afford 44 % of product 2 ad, with no traces of Csp2−OMe activation. The intramolecular monoannulation product was not detected. Since we could achieve the simultaneous activation of two Csp2−OMe bonds in ortho in 1 ac, and Csp2−F was more reactive than Csp2−OMe, we hypothesized that 1 ad could also undergo the simultaneous aromatic homologation and alkyne monoannulation to form 4. Indeed, after optimization studies, we found that product 4 was formed, when employing both LiOTf and PPh3.

Scheme 3

Competition experiment between Csp2−OMe vs. Csp2−F ortho‐functionalization in 1 ad towards the formation of aromatic homologation and the simultaneous double functionalization. [a] Isolated yield.

At the current stage of this investigation, we envision an analogous mechanism for the Csp2−OMe activation as the one described for Csp2−F in our previous report. Indeed, when substrate 1 ab was reacted with diphenylacetylene under the aromatic homologation conditions for 3 h, the square‐planar nine‐membered nickelacyclic intermediate complex 1 ab‐INT4‐E‐H was selectively formed in a 91 % NMR yield (Scheme 4). 2D NMR structural characterization clearly showed that 1 ab‐INT4‐E‐H featured two alkynes inserted. The species 1 ab‐INT4‐E‐H undergoes formation of the corresponding aromatic homologation product 2 ab in 32 % yield after 24 h (Table 1, entry 5). This low yield compared to the almost quantitative accumulation of 1 ab‐INT4‐E‐H might be related to the steric hindrance imposed by the −OMe group adjacent to the C−H activated in the aromatic homologation process, thus hampering the formation of 2 ab. As in the case of the Csp2−F functionalization, the presence of lithium ions is crucial to stabilize the LiOMe as leaving group.

Scheme 4

Isolation of the 1 ab‐INT4‐E‐H intermediate species. Yield calculated using 1,3,5‐trimethoxybenzene as internal standard and based on the total Ni content.

Remarkably, the 1,2,3,4‐tetraphenylnaphthalene unit in the homologation product 2 aa was envisioned as suitable to undergo Scholl or DDQ oxidative coupling to synthesize nanographene‐like compounds (7‐fused aromatic rings, pyrenoid type). We first attempted the FeCl3‐mediated Scholl reaction, but inconclusive results were obtained. On the contrary, DDQ‐based protocol afforded the oxidative fusion of the two phenyl moieties, along with in situ amide hydrolysis and esterification product 5 in 27 % yield (Scheme 5). It is worth mentioning that a 9‐fused ring compound was detected in the crude mixture (Figure S9), suggesting that it might also be accessible under modified experimental conditions. This type of transformation opens new opportunities in the use of these multi‐ring products as starting scaffolds for bottom‐up synthesis of unprecedented nanographene derivatives. We are currently working on widening the scope of diphenylacetylene derivatives in order to expand the size of nanographenes.

Scheme 5

Adjacent arene fusion reaction of 2 aa via DDQ‐mediated oxidative coupling to form nanographene monofunctionalized derivative 5.

Conclusion

In summary, we have developed a novel Ni0‐catalyzed methodology to achieve the Csp2−OMe functionalization using internal alkynes, forming chemoselectively either aromatic homologation or intramolecular monoannulation products upon fine optimization of the reaction. This methodology stands as a new tool to activate Csp2−OMe bonds showing the potential use of aryl‐alkyl ethers as alternative electrophiles to aryl halides. Furthermore, aromatic homologation products are proven as valid precursors towards bottom‐up nanographene‐like synthesis, as a further diversification of the possible uses of these compounds.

Experimental Section

See Supporting Information for materials, instrumentation, experimental procedures and spectroscopic characterization of all compounds. Deposition Number 2144669 (compound contains the supplementary crystallographic data for this paper. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe http://www.ccdc.cam.ac.uk/structures Access Structures service.

Conflict of interest

The authors declare no conflict of interest.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Click here for additional data file.