Constructing Multiply Substituted Arenes Using Sequential Pd(II)-Catalyzed C–H Olefination

Abstract

Complementary catalytic systems have been developed in which the reactivity/selectivity balance in Pd(II)-catalyzed ortho-C–H olefination can be modulated through ligand control. This allows for sequential C–H functionalization for the rapid preparation of 1,2,3-trisubstituted arenes. Additionally, a rare example of iterative C–H activation, in which a newly installed functional group directs subsequent C–H activation has been demonstrated.

Divinylbenzene derivatives represent an important class of molecular building blocks in organic chemistry and materials science.^[1,2] Thus, their synthesis has been an active area of research.^[2] Indeed, shortly after the initial reports describing Pd(0)-catalyzed haloarene/olefin coupling (the Mizoroki–Heck reaction),^[3] Heck disclosed an example of divinylation of ortho-dibromobenzene.^[4] Since that time, Heck and others have studied differential olefination of arenes containing two different halides (or pseudohalides) to synthesize unsymmetrical divinylbenzenes.^[5,6]

In light of the recent advances in Pd(II)-catalyzed aryl C–H functionalization reactions,^[7] including those that directly couple aryl C–H bonds and olefins,^[8–17] an exciting method for divinylbenzene synthesis would be to use an arene substrate containing a synthetically versatile directing group (DG)^[7c] and perform sequential olefination of both ortho-C–H bonds (Scheme 1). In this way, two different alkenes could be expediently introduced in a position-selective fashion to form valuable 1,2,3-trisubstituted products,^[18] thereby obviating the complications associated with preparing dihaloarene starting materials.

Scheme 1.

Sequential directed C–H functionalization.

However, generally speaking, methods for sequential, directed ortho-C–H activation with Pd(II) remain underdeveloped^[19,20] owing to problems rooted in both selectivity and reactivity (Scheme 2). In terms of selectivity, a major challenge lies in forming the mono-functionalized product in high yield while avoiding functionalizing the second ortho-C–H bond to give the difunctionalized byproduct. Tactics to avoid this problem include substituting the ortho- or meta-positions to block the second C–H insertion, lowering the reaction temperature or time to suboptimal levels, or using reduced equivalents of one of the reactants. However, these maneuvers lead to limited substrate scope and/or compromised yields. Similarly, reactivity can also be a problem if the first new group is sterically bulky because it can attenuate the substrate's binding affinity for Pd(II). Alternatively, if the group is electron-withdrawing (e.g., halogen, olefin, or carboxylate), it can deactivate the aromatic ring for the second C–H activation step. Lastly, the original directing group and the newly installed functional group can coordinate to Pd(II) in a bidentate fashion, preventing the catalyst from accessing the second C–H bond.

Scheme 2.

The challenges of sequential directed C–H functionalization.

Reflecting on these two interrelated challenges, we became aware of a need for complementary catalytic systems in which the selectivity/reactivity balance of Pd(II) could be modulated through coordination of an external ligand. We report herein the realization of this goal in the case of sequential Pd(II)-catalyzed aryl C–H olefination. Furthermore, as part of our research program to develop expedient and versatile aryl C–H functionalization techniques, we demonstrate an example of iterative C–H functionalization, wherein one C–H activation reaction installs a directing group for a subsequent C–H functionalization reaction (Scheme 3).

Scheme 3.

Outline of this work.

To this end, we began by revisiting our recently disclosed ortho-C–H olefination reaction for phenylacetic acids.^[14] The original conditions gave good yields and high levels of mono-selectivity with electron-rich and electron-neutral substrates. However, when we took a product from this reaction and resubmitted it to the reaction conditions in the presence of a different olefin, we observed <10% conversion of the desired unsymmetrical diolefinated product. Furthermore, we took note of the fact that during our original screening studies to develop the mono-selective reaction, we never observed more than 30% of the diolefinated byproduct, even as we extensively surveyed different reaction conditions. In light of these findings, we concluded that we faced a problem of low reactivity and needed a more reactive catalyst for the installation of a second olefin.

Encouraged by our recent success in using amino acid ligands^[21] to enhance the reactivity of Pd(II) with electron-deficient substrates,^[14] we hypothesized that we could develop a generally applicable diolefination protocol using an optimized amino acid ligand and apply it as the second step in a two-step sequential olefination procedure. Gratifyingly, we found in our initial screening studies that an array of mono-N-protected amino acid ligands (10 mol%) were highly efficient in promoting diolefination with both electron-poor and electron-rich substrates, such as 4-(trifluoromethyl)phenylacetic acid (1a) and 4-methoxyphenylacetic acid (1b), fashioning the desired products in quantitative conversion after 48 h in the presence of ethyl acrylate (2a), Pd(OAc)₂ (5 mol%), KHCO₃ (2 equiv.), and tAmylOH under an O₂ (1 atm) at 90 °C.

To determine the optimal ligand, we selected substrate 1a for screening. It is worth noting that in the absence of amino acid ligands, 1a was found to give only 13% yield of the mono-olefinated product 3a after 48 hours. We selected an abridged reaction time of 2 h in order to see the comparative kinetic behavior of the ligands. We found that Boc-Val-OH, Ac-Ile-OH, and Ac-Val-OH (entries 5, 11, and 12, Table 1) gave the highest conversion of 4a. Ac-Ile-OH was found to be the highly active; however, with this ligand the results varied substantially from trial to trial. Ac-Val-OH was found to be the best in terms of activity and reproducibility. Importantly, 1,4-benzoquinone (BQ), which we previously used in our mono-selective olefination procedure,^[14] was found to decrease the reaction rate. With Ac-Val-OH, increasing the reaction time from 2 h to 6 h improved the conversion from 63% to >99% (86% isolated yield) (entry 18). Importantly, with the reaction time extended to 24 h, 2 mol% catalyst could also be used, giving 82% isolated yield of 4a.

Table 1.

Ligand optimization.^[a]

Entry	Ligand	%			Entry	Ligand	%
		1a	3a	4a			1a	3a	4a
1	---	93	7	0	10	Boc-Ser-OH	77	23	0
2[b]	BQ	98	2	0	11 [c]	Ac-Val-OH	4	23	63
3[b]	BQ + Boc-Ile-OH	72	27	1	12 [c]	Ac-Ile-OH	3	43	54
4	Boc-Ile-OH	24	62	14	13	Ac-Leu-OH	54	43	3
5[c]	Boc-Val-OH	3	62	35	14	For-Ile-OH	77	23	0
6	Boc-Leu-OH	11	70	19	15	For-Leu-OH	68	31	1
7	Boc-tLeu-OH	81	19	0	16	Men-Leu-OH[d]	57	41	2
8	Boc-Ala-OH	25	65	10	17	Men-Val-OH[d]	64	35
9	Boc-Phe-OH	47	49	4	18 [e]	Ac-Val-OH	0	0	>99 (86)

^[a]The conversion was determined by ¹H NMR analysis of the crude reaction mixture.

^[b]5 mol% BQ.

^[c]Average of three trials.

^[d]Men = (−)-Menthyl(O₂C).

^[e]6 h.

Using these optimized conditions, a wide range of phenylacetic acids (1) were converted into the corresponding diolefinated products (4) (Table 2). The reaction worked well with electron-rich substrates containing ether and alkyl groups (4b, 4d, 4j–4m, and 4p) and with electron-poor substrates containing halogen atoms (4a, 4e–4h, 4n, and 4q). Sterically encumbered substrates that contained bulky substitutents at the α-position (4o–4q) or at the position meta to the carboxylic acid directing group (4j–4n) required extended reaction times (48 h), but still generally gave good to excellent yields. 1-Naphthylacetic acid was also found to be a reactive substrate for diolefination (4r). Though the yield was somewhat low (35%), the second olefination proceeds via an unusual seven-membered cyclopalladated intermediate and represents a unique example of remote C–H activation at the 8-position of a naphthalene ring with a directing group at the 1-position. The reaction was also optimized for hydrocinnamic acids (see Supporting Information). In these cases, Boc-Val-OH was found to be the best ligand. Higher catalyst loadings (10–15 mol%) and extended reaction times (48–96 h) were needed to improve the yield of the diolefinated products 4s and 4r. In nearly all of the cases in Table 2 where the isolated yield is below 80%, the remaining starting material had been completely converted to the mono-olefinated product (as evidenced by ¹H NMR of the crude reaction mixture).

Table 2.

Substrate scope for Pd(II)-catalyzed diolefination.^[a]

^[a]Isolated yield.

^[b]48 h.

^[c]15 mol% Pd(OAc)₂, 30 mol% Boc-Val-OH, 96 h.

^[d]10 mol% Pd(OAc)₂, 20 mol% Boc-Val-OH, 96 h.

Subsequently, different olefin coupling partners were examined (Table 3). Several acrylates (2a–2c) were found to be highly reactive. Styrene (2d) was also found to be a competent coupling partner; however the reaction required extended time (48 h) and higher Pd loading (10 mol%) (1d). Ethyl vinyl ketone (2e) was found to be effective, giving 50% yield of 4x after 6 h. Attempts to extend the reaction time to improve the yield of 4x led to substantial product decomposition. Other olefins, such as vinyl sulfones, vinyl phosphonates, and internal alkenes were unreactive.

Table 3.

Olefin scope for Pd(II)-catalyzed diolefination.^[a,b]

^[a]Isolated yield.

^[b]Reaction conditions: 2 equiv. olefin, 5 mol% Pd(OAc)₂, 10 mol% Ac-Val-OH, 2 equiv. KHCO₃, tAmylOH, 90 °C, 1 atm O₂, 6 h.

^[c]10 mol% Pd(OAc)₂, 20 mol% Ac-Val-OH, 48 h.

With this robust Pd(II)-catalyzed diolefination protocol in hand, we returned to our goal of effecting two sequential C–H olefination reactions to install two different alkenes (Table 4). We began by using our mono-selective C–H olefination reaction,^[14] which is compatible with electron-rich and electron-neutral substrates, to couple 1b and benzyl acrylate (2b), giving 3b in 70% isolated yield (over 2 g prepared). Using this intermediate, we attempted to perform a second olefination reaction with representative alkene coupling partners in the presence of our more active [Pd(II)–Ac-Val-OH] catalyst. Gratifyingly, we found that orthogonally protected acrylates could be smoothly coupled to give differentially protected aromatic triacids 5a and 5b and that styrene (2d) could also be installed in good yield (5c). Interestingly, in accordance with our earlier observation,^[14] when 1-hexene (2f) was used, the unique non-conjugated product 5d was obtained, representing a formal C–H allylation. The isolated yield was low due to decomposition of the product during the course of the reaction. In principle, all of the substrates used in Table 2 could be used for a similar two-step sequence. meta-Substituted acids offer another unique element of spatial control because the first olefin selectively reacts away from the substituent. The two-step preparation of these advanced unsymmetrical 1,2,3,5-tetrasubstituted arenes demonstrates the power of sequential aryl C–H functionalization reactions.

Table 4.

Sequential olefination.^[a,b]

^[a]Isolated yield.

^[b]Reaction conditions (1^st step): 2 equiv. 2b, 5 mol% Pd(OAc)₂, 5 mol% BQ, 2 equiv. KHCO₃, tAmylOH, 90 °C, 1 atm O₂, 48 h. Reaction conditions (2^nd step): 2 equiv. olefin, 5 mol% Pd(OAc)₂, 10 mol% Ac-Val-OH, 2 equiv. KHCO₃, tAmylOH, 90 °C, 1 atm O₂, 6 h.

^[c]1 equiv. 1-hexene (2f).

Finally, we sought to explore the possibility of performing iterative C–H functionalization, wherein a newly installed functional group would serve as the directing group for further C–H activation (Scheme 4). In particular, we envisioned using sequential olefination to install two different olefins and then, following functional group manipulations, to use one of the new moieties to direct an additional remote ortho-C–H functionalization reaction. To test this idea, we prepared 5e in good yield using our sequential olefination method (Scheme 4). Following methylation and hydrogenation, we obtained highly decorated hydrocinnamic acid 6. Gratifyingly, we found that treatment of 6 with benzyl acrylate (2b) (2 equiv.) in the presence of Pd(OAc)₂ (10 mol%), Boc-Val-OH (20 mol%), and tAmylOH for 12 h under O₂ (1 atm) effected C–H olefination to form 7 in 35% isolated yield. Attempts to extend the reaction time to improve the yield led to substantial product decomposition, such that only trace quantities of 7 remained in solution after 24 h. Though this yield is low, given the possibility of non-productive multidentate coordination of the substrate with Pd(II), a slow reaction rate can be expected. Moreover, this case represents the most complex setting in which C–H olefination of a hydrocinnamic acid has been attempted. In the absence of Boc-Val-OH, the reaction did not proceed.

Scheme 4.

Sequential and iterative C–H olefination to synthesize multiply substituted arenes. (See Supporting Information for experimental details.)

In summary, through the discovery and development of complementary catalytic systems that exhibit tunable reactivity and selectivity, we have demonstrated a sequential C–H olefination protocol for synthesizing complex divinylbenzene derivatives from simple starting materials. We first established robust reaction conditions to effect diolefination with a range of different phenylacetic and hydrocinnamic acids. We then applied these conditions as the second step in a two-step serquential C–H olefination sequence to prepare 1,2,3-trisubstitued arenes. Lastly, we used sequential C–H olefination to set the stage for a rare example of iterative C–H functionalization, wherein C–H activation installs a directing group for a subsequent C–H functionalization reaction.

Experimental Section

General procedure for Pd(II)-catalyzed diolefination of phenylacetic acids: A 50 mL Schlenk-type sealed tube (with a Teflon high pressure valve and side arm) equipped with a magnetic stir bar was charged with the phenylacetic acid substrate (0.5 mmol), Pd(OAc)₂ (5.6 mg, 0.025 mmol), KHCO₃ (100.1 mg, 1.0 mmol), Ac-Val-OH (8.0 mg, 0.05 mmol), the olefin coupling partner (1.0 mmol), and tAmylOH (2.5 mL). The reaction tube was capped, then evacuated briefly under high vacuum and charged with O₂ (1 atm, balloon) (×3). The reaction mixture was stirred at room temperature for 5 min, then at 90 °C for 6 h (or 48 h in the case of less reactive substrates, see Tables 2 and 3). Subsequently, the reaction vessel was cooled to 0 °C in an ice bath. A 2.0 N HCl solution (5 mL) was added, and the mixture was extracted with EtOAc (3 × 20 mL). The organic layers were combined, dried over Na₂SO₄, filtered, and concentrate in vacuo. The resulting residue was purified by silica gel flash column chromatography using hexanes/EtOAc (with 2–5% HOAc) as the eluent.