Introduction

Alkenes represent one of the most important classes of organic compounds. Although numerous methods exist for alkene preparation, the synthesis of polysubstituted alkenes in a regio- and stereochemically defined manner still poses a significant challenge to synthetic organic chemists1,2,3,4. In principle, six isomers exist for fully substituted alkenes with four different R groups. The control of regio- or stereoselectivity, therefore, is critical when planning a synthetic method. As an extended π-system, tetraarylethenes have been valuable targets of synthetic endeavor due to their useful physical, structural, and electronic properties5,6,7. Nevertheless, given the typically small electronic and steric differences between the individual aryl groups (compared to the difference between aryl and alkyl groups), the selective synthesis of tetraarylethenes bearing four different aromatic substituents is inherently challenging. While the classical double bond-forming methods such as olefin metathesis8, Wittig9 and McMurry10 reactions are effective for tetraarylethene synthesis, the stereoselectivity of these protocols is typically a problem1,2,3,4. Platform synthesis is a powerful strategy to realize the synthesis of complex molecules in a programable and diversity-oriented format. In this regard, the advances of metal catalysis have inspired tremendous efforts towards modular tetraarylethene construction starting from alkenes11,12,13,14,15,16,17 or alkynes18,19,20,21,22,23,24,25,26. Nevertheless, lengthy synthetic operations, the poor functional group tolerance arising from the participation of organolithium or Grignard reagents, and the often-encountered poor regioselectivity strongly limits their practical application.

A notable contribution by the Larock group revealed a straightforward tetraarylethene synthesis via a palladium-catalysed three-component coupling of readily available aryl iodides and arylboronic acids with internal alkynes (Fig. 1a)27,28. The ability to simultaneously introduce two aryl groups in one single synthetic operation makes this protocol appealing. Mechanistically, the three-component reaction proceeds via an oxidative addition/migratory insertion/reductive elimination sequence. While the stereoselectivity, derived from cis-arylpalladation, is generally excellent, the regioselectivity is problematic when unsymmetrical diaryl alkynes are employed. In addition, terminal alkynes are ineffective coupling partners due to the competitive Sonogashira coupling side reaction.

Fig. 1
figure 1

Related work and reaction design. a Overview of Larock’s three-component coupling reaction. b Palladium-catalysed functionalization of alkenyl MIDA boronates. c This work: palladium-catalysed syn-diarylation of arylethynyl MIDA (N-methyliminodiacetyl) boronates and its follow-up transformations

Full size image

We considered a variation of Larock’s protocol as a promising solution to the truly flexible tetraarylethene synthesis. To maximize the flexibility of the reaction, a high-degree control of the direction of the migratory insertion step should be realized. To this end, we envisioned that the use of an alkyne substrate prefunctionalized by attaching a transformable masking group to the triple bond is a promising approach. Ideally, the masking group should confer sufficient electronic bias to the triple bond, thereby rendering the migratory insertion regioselective. Nevertheless, the drastic alternation of the electron density on the triple bond should be avoided, otherwise the formation of biarenes via the direct Suzuki–Miyaura coupling of aryl iodide with arylboronic acid would compete as a side reaction. In addition, this masking group should be applicable to a follow-up cross-coupling reaction to install the final aryl group, but remain intact under the three-component coupling conditions with the palladium catalyst present. As such, the identification of proper reaction conditions to enable orthogonal reactivities is necessary but represents a challenge. Guided by the above-mentioned concept, we reasoned the sp3-hybridized MIDA (N-methyliminodiacetyl) boron might be an optimal choice29,30,31,32. Early studies reveal the notable stability of MIDA boron towards diverse reaction conditions. Importantly, this moiety, as a protected form of boronic acid, is transformable to other functional groups under a one-pot deprotection/coupling reaction condition. Advances from us, and others have demonstrated that MIDA boron functionalized unsaturated systems (including alkenes33,34,35,36,37,38,39,40,41,42 and alkynes43,44,45,46,47,48,49,50) are intriguing synthons for boron-retentive transformations51,52,53. Specifically, Yudin and Dudding54, and Guérinot and Cossy55, have independently observed the unique electronic directing effect of MIDA boron in palladium-catalysed Heck reactions and Wacker oxidations of alkenyl MIDA boronates, respectively (Fig. 1b). Computational analysis reveals a drastic electron donation from the C−Pd σ bond to a boron centered p-orbital the transition state due to the hemilabile nature of the MIDA B−N dative bond56. A similar regioselectivity trend was also observed in our previous rhodium(III)-catalysed C–H annulation reaction with alkyne MIDA boronates44.

Herein, we report a palladium-catalysed regio- and stereoselective diarylation of arylethynyl MIDA boronates with aryl halides and arylboronic acids (Fig. 1c). The regioselectivity, guided by MIDA boron, is distinct from the previous observations when aryl alkyl acetylenes were used as substrates27,28. The starting arylethynyl MIDA boronates are readily accessible via a Sonogashira coupling of the corresponding aryl halides with commercially available ethynyl MIDA boronate. Synthetic manipulations of the MIDA boron moiety in the thus formed diarylated products via a stereospecific Suzuki–Miyaura coupling or base-promoted protodeborylation lead to a truly flexible and stereodefined synthesis of tetra- and triarylethenes.

Results

Screening of reaction conditions

We first investigated the feasibility of the palladium-catalysed coupling reaction of phenylethynyl MIDA boronate 1a with iodide 2a (2.0 equiv.) and boronic acid 3a (2.0 equiv.). Complexity may arise due to the presence of two different organoborons (the chemoselectivity issue), the possible competing direct Suzuki–Miyaura coupling, and the unclear regio- as well as stereoselectivity issues. Gratifyingly, with 5 mol % PdCl2(PhCN)2 as catalyst and KHCO3 as base in DMF at 40 °C27, the diarylated alkene product 4a was formed as a single regio- and stereoisomer in 42% yield (Table 1, entry 1). Interestingly, structural elucidation revealed that the phenyl and the 3,5-dichlorophenyl group resides trans to each other across the resulting carbon–carbon double bond, which is in contrast to the previous observations that the aryl group derived from the boronic acid preferentially locates at the position proximal to the aryl substituent of the alkyne, indicating the unique directing effect of MIDA boron. Biaryl derived from direct Suzuki–Miyaura coupling was detected as the major byproduct. The screening of different palladium-based catalysts proved that PdCl2(MeCN)2 to be ideal, giving a higher yield of 46% (entries 2–6). Switching from KHCO3 to Ag2CO3 led to a higher conversion of 1a (entries 7–10).The attempt to employ a variety of metal ligands in this reaction was proven to be unfruitful (see Supplementary Table 1). Other solvents were also screened (entries 12–14), with acetonitrile being a better choice (75% yield). It is worth to mention that a significant amount regioisomer 4a’ was formed in acetone (entry 13). Consistent with the previous observations, the use of H2O (15 equiv) as additive further improved the yield to 94% (entry 15). Interestingly, decreasing the catalyst loading to 1 mol % also deliverd the product in 63% yield (entry 16). Unlike the previous three-component coupling reaction wherein 100 °C was used, a temperature of 40 °C was sufficient for the reaction to proceed efficiently, demonstrating the mildness of our reaction. As a comparison, the reaction of three congeners of phenylethynyl MIDA boronate, including the corresponding pinacol boronate I, trifluoroboronate II, and B(dan) III all resulted in a rapid decomposition, with no boron-retentive diarylated products being found (Fig. 2). Interestingly, the silicon protected alkyne IV showed trivial reactivity in this reaction. IV was largely recovered after the reaction. Moreover, phenyl ethyl alkyne is applicable to our reaction conditions, but a low yield of 11% was obtained. The regioselectivity of this reaction is identical to Larock’s observation 27.

Table 1 Optimization studiesa
Full size table
Fig. 2
figure 2

Unsuccessful substrates in palladium-catalysed three-component reaction. For I–III no boron-containing diarylated products were obtained

Full size image

Synthesis of triarylalkenyl MIDA boronates

With the optimized conditions in hand, the scope on the aryl group of alkynyl MIDA boronates was first examined (Fig. 3a). Both electron-rich and electron-poor aryl substituents are compatible with the reaction conditions, providing the desired products in generally moderate to good yields. In addition, broad functional group tolerance was found, as evident by the survival of valuable methoxy (4n), bromo (4d), acetyl (4g, 4o), ester (4k, 4l, 4q), nitrile (4i), nitro (4h), trifluoromethyl (4j, 4u), and trifluoromethoxy (4m) groups in the reaction. The sterically hindered ortho-substituent (4p-4r) does not hamper the reactivity. Heterocycles, such as isoxazole (4w), thiophene (4x), indole (4y), thiozole (4z), benzothiophene (4aa), and carbozole (4ab) were also tolerated, although the electron-rich ones gave relatively lower yields. In all the cases, clean cis-addition to the alkyne was observed. However, in some cases, the formation of a minor amount of the regioisomer was found. The scope on the aryl iodides is also noteworthy. A variety of electron-rich or -neutral (hetero)aryl iodides, including the one bearing a free hydroxyl substituent (4ae), the one which is sterically shielded (4aj), were successfully employed. More interestingly, alkenyl bromides (4ap, 4aq) were also effective coupling partners to provide the desired 1,3-diene products in reasonable yields. The easy availability of arylboronic acids offered tremendous chances for the structural diversity of the product. Interestingly, in addition to the electron-neutral and electron-rich arylboronic acids, the electron-deficient ones (4as, 4at, 4av, 4az, 4ba, 4be, 4bg), which were poor coupling partners in Larock’s protocol, also worked very well in this chemistry. An alkenyl MIDA boronate (4bl) substituted with three heteroarenes was also constructed in reasonable yield. Unfortunately, the employment of arylboronic acids bearing an ortho-substituent (4bm) resulted in low yield using our present reaction conditions, revealing a limitation of the protocol. The structure of 4ba was unambiguously assigned by X-ray crystallographic analysis. In accordance to the observations mentioned above, an additive-based robustness screen demonstrated the broad functional group tolerance of the reaction conditions(Fig. 3b)57,58. A variety of synthetically important moieties,such as a ketone, aldehyde, free alcohol, amide, aryl bromide, acetal, or even challenging additives like aniline or a nitrile, were well tolerated. However, the screen indicated that terminal alkynes and alkenes along with sterically unshielded N-heterocycles are not compatible with the reaction conditions (see Supplementary Methods).

Fig. 3
figure 3

Synthesis of triarylalkenyl MIDA boronates. a Substrate scope. b Robustness screen. Reaction conditions: 1 (0.2 mmol), 2 (0.4 mmol), 3 (0.4 mmol), PdCl2(MeCN)2 (5 mol %), Ag2CO3 (0.2 mmol), H2O (3 mmol) in 2 mL MeCN at 40 °C for 3–5 h; ratio of regioisomers is determined by crude 1H-NMR and shown in the parentheses. i at 30 °C. ii at 25 °C

Full size image

Synthesis of tetraarylethenes

Having established the regio- and stereoselective synthesis of triarylalkenyl MIDA boronates, we set out to explore their capacity to be engaged in metal-catalysed cross-coupling reactions. As shown in Fig. 4a, we were pleased to find that under Burke’s slow-release reaction conditions, a series of triarylalkenyl MIDA boronates underwent Suzuki–Miyaura coupling smoothly with diverse (hetero)aryl iodides to deliver the tetraarylethenes in good to excellent yields. As expected, this reaction is stereospecific, thereby leading to a truly simple and flexible construction of differently substituted tetraarylethenes. A representative while interesting example is the effective synthesis of 5t bearing four heterocycles. To maximize the reaction efficiency, a telescoped synthesis was conducted. Thus, the arylethynyl MIDA boronate was first reacted with an aryl iodide and an arylboronic acid. After the consumption of alkyne, the reaction mixture was subjected directly to the follow-up Suzuki–Miyaura coupling with the second aryl iodide. To highlight the power of this protocol, three tetraarylethenes were obtained in reasonable yields over two steps upon one single purification (Fig. 4b) (See Supplementary Methods).

Fig. 4
figure 4

Synthesis of tetraarylethenes. a Pd-catalysed Suzuki-Miyaura coupling reaction to synthesize tetraarylethenes. b Synthesis of tetraarylethenes in one pot

Full size image

Synthesis of triarylethenes

Larock’s three-component coupling reaction is not applicable to triarylethenes synthesis since terminal alkynes are not viable substrates. We envisoned the lability of C–B bond towards hydrolysis may offer a chance to construct triarylethenes starting from triarylalkenyl MIDA boronates via protodeborylation. In other words, the MIDA boron moiety could be considered as a transient protecting group for terminal alkynes. As expected, our trials showed that by simple treatment of the alkenyl MIDA boronate products with K3PO4 (2.5 M, 7.5 equiv) in DMF at 40 °C, the corresponding triarylethenes could be obtained in high efficiency (Fig. 5a). This process is highly stereospecific with no isomerization of the double bond being found. The simplicity of the reaction conditions also rendered the telescoped synthesis of triarylethenes possible. Thus, by following a two-step, one-purification procedure, four triarylethenes were synthesized in moderate yields with good regio- and stereocontrol (Fig. 5b) (See Supplementary Methods).

Fig. 5
figure 5

Synthesis of triarylethenes. a Base-promoted protodeborylation to synthesize triarylethenes. b Synthesis of triarylethenes in one pot

Full size image

A gram-scale synthesis of 4a (1.5 g, 75%) was successful under the standard reaction conditions (Fig. 6a), demonstrating the potential utility of the reaction. Apart from the synthesis of tetraarylethenes and triarylethenes, other synthetic manipulation of the MIDA boron moiety in the cross-coupling products also allowed the stereodefined synthesis of diverse tetra-substituted alkenes (Fig. 6b). For instance, both alkenyl and alkynyl bromides were efficiently coupled with 4aa to give the corresponding extended π-systems (7 and 8)59. By treating with KHF2, the corresponding potassium trifluoroboronates 9 could be obtained60. In another vein, ligand exchange with pinacol under acidic conditions provided the pinacol boronate 10 in quantitative yield61. This sp2-B species 10 could then react with aliphatic halides under palladium catalysis to give a methylated (11) or benzylated (12) product62. In addition, a palladium-catalysed oxidative alkoxycarbonylation with carbon monoxide in MeOH allowed the construction of a fully substituted α,β-unsaturated ester 1363. Finally, a phenanthrene product (14) was obtained via a Suzuki–Miyaura coupling/oxidative cyclization cascade64.

Fig. 6
figure 6

Synthetic applications. a Gram-scale reaction b Other synthetic derivatizations of triarylalkenyl MIDA boronates. Reagents and conditions: i. beta-Bromostyrene (1.5 equiv), Pd(dba)2 (10 mol %), SPhos (10 mol %), K3PO4 (7.5 equiv), 1,4-dioxane (0.1 M), 40 °C; ii. ((3-bromoprop-2-yn-1-yl)oxy)benzene(1.5 equiv), Pd(dba)2 (10 mol %), SPhos (10 mol %), K3PO4 (7.5 equiv), 1,4-dioxane (0.1 M), 40 °C; iii. KHF2 (5.0 equiv), MeOH (0.025 M), 50 °C; iv. 2 M H2SO4 (3.0 equiv), pinacol (5.0 equiv), THF (0.1 M), RT; v. MeI (3.0 equiv), Pd(dba)2 (5 mol %), P(2-tol)3 (10 mol %), K2CO3 (2.0 equiv), DMF/H2O; vi. benzyl bromide (2.0 equiv), Pd(dba)2 (5 mol %), P(2-tol)3 (10 mol %), K2CO3 (2.0 equiv), DMF/H2O; vii. Pd(OAc)2 (10 mol %), PPh3 (20 mol %), PBQ (2 equiv), K3PO4 (5.0 equiv), CO (1 atm), MeOH (0.1 M), 60 °C; viii. p-iodotoluene (1.2 equiv), Pd(OAc)2 (5 mol %), SPhos (10 mol %), K3PO4 (7.5 equiv), THF (0.1 M), 60 °C; ix. DDQ (1 equiv), MeSO3H (1 mL), DCM (0.01 M), 0 °C

Full size image

Mechanistic proposal

The following reaction mechanism is proposed for this process (Fig. 6). Initially, Pd(II) is reduced to Pd(0), probably by arylboronic acid (Fig. 7a). The oxidative addition of aryl iodide to Pd(0) furnishes arylpalladium species A, which then undergoes cis-carbopalladation with arylethynyl MIDA boronate to deliver intermediate B. The regioselectivity, which is different from Larock’s observation, may arise from the stabilization due to the strong electron donation from the C−Pd σ bond to the p-orbital of boron in the transition state (Fig. 7b), although a simple steric reason that Pd favor the bulky B(MIDA) end is also possible. Thereafter, the transmetalation of B with Ar3B(OH)2 generates C, which upon reductive elimination produces triarylalkenyl MIDA boronate with simultaneous regeneration of the Pd(0) catalyst.

Fig. 7
figure 7

Mechanistic hypothesis. a Possible reaction mechanism. b Rationale of regioselectivity

Full size image

Discussion

The remarkable stability combined with the unique electronic and steric properties of sp3-B MIDA boron renders arylethynyl MIDA boronates intriguingly competent substrates in palladium-catalysed three-component couplings with aryl iodides and arylboronic acids. The reaction offers a facile synthesis of triarylalkenyl borons in a regio- and stereoselective manner with a broad substrate scope under mild reaction conditions. Synthetic manipulation on the C−B bond in the product via the stereospecific Suzuki–Miyaura coupling and base-promoted protodeborylation reaction led to a step-economic and truly flexible synthesis of tetra- and triarylethenes, both of which are historically challenging synthetic targets. The value of the approach was further highlighted by the success of telescoped synthesis. Another notable feather is that all the starting materials, including arylethynyl MIDA boronates, are readily available. Other transformations of the formed triarylalkenyl MIDA boronates provided straightforward access to a number of tetra-substituted alkenes in good efficiency. Given the operational simplicity, the excellent control on regio- and stereoselectivity and the broad substrate scope of this method, we anticipate it will find numerous applications in organic synthesis.

Methods

General procedure for stereodefined synthesis of triarylalkenyl MIDA boronates

To a 15-mL Schlenk tube equipped with a stirring bar, were added 1 (0.2 mmol, 1.0 equiv), 2 (0.4 mmol, 2.0 equiv), 3 (0.4 mmol, 2.0 equiv), silver carbonate (55.2 mg, 0.2 mmol, 1.0 equiv), and PdCl2(MeCN)2 (2.6 mg, 5 mol %) in glove box. The reaction vessel was taken out, and 54 μL H2O and MeCN (2 mL) was added under N2 protection. The reaction mixture was stirred at 40 °C until 1 were consumed completely (about 3 h). The mixture was concentrated and directly subjected to flash column chromatography on silica gel (DCM/EtOAc 32:1 v/v) to afford the desired product.

General procedure for stereodefined synthesis of tetra-aryl substituted alkenes

To a 15-mL Schlenk tube, were added 4 (51.0 mg, 0.1 mmol, 1.0 equiv), aryl iodide 5 (0.11 mmol, 1.1 equiv), Pd(OAc)2 (1.1 mg, 0.005 mol), and SPhos (4.1 mg, 0.01 mmol), the reaction vial was sealed with a rubber septum. K3PO4 (3 M, 0.25 mL) and THF (0.75 mL) were added through syringe. The reaction mixture was stirred at 60 °C until 4 were completely consumed. After cooling to room temperature, the reaction mixture was concentrated in vacuo. Purification by flash chromatography (Petroleum ether/EtOAc 64:1 v/v) afforded the desired products 5.

General procedure for stereodefined synthesis of tri-aryl substistuted alkenes

To an oven-dried flask were added 4 (51.0 mg, 0.1 mmol, 1.0 equiv), K3PO4 (3 M, aq. 0.25 mL), and DMF (1.0 mL) through syringe. The solution was stirred at 40 °C until 4 was completely consumed. After cooling to room temperature, the reaction mixture was quenched with water (10 mL), extracted with EtOAc (3 × 10 mL). The combined organic layers were washed with brine (3 × 10 mL) and dried with Na2SO4 and concentrated in vacuo to give a crude residue. The crude product was subsequently loaded onto a silica gel column and purified by flash column chromatography (Petroleum ether/EtOAc 100:1 v/v) to afford the desired compound 6.

Synthetic transformations

Full procedures for synthetic transformations of compounds 7–14 are available in the Supplementary Methods and Supplementary Figs. 1-16.

Robustness screen

Please see Supplementary Table 2 and Supplementary Figures 17-18.

NMR spectra

1H, 13C, and 19F NMR spectra of purified compounds are available in Supplementary Figs. 19–130.

Crystallography

X-ray crystallographic CIF files for compounds 4ba and 6p are available in Supplementary Data 1 and 2.