Introduction

All-carbon tetrasubstituted olefins extensively exist in natural products, leading drugs and functional organic materials1,2. They are also key precursors for various useful transformations such as hydrogenation, epoxidation, reductive Heck and other processes that generate contiguous highly substituted C(sp3) centers3. The importance of such tetrasubstituted olefins in both academia and industry has stimulated the enduring efforts towards exploiting the catalytic methods for their syntheses4,5,6,7,8,9,10,11,12. In this context, the transition metal-catalyzed difunctionalization of alkynes via carbometallation stands for the most efficient and modular route to synthesize fully substituted olefins3,13,14,15,16,17,18,19,20,21,22, but still suffers from the difficulty in achieving regio-/stereocontrol when applied to electronically or sterically unbiased substrates. In light of this, the alternative and general methods for the regio-/stereoselective construction of all-carbon tetrasubstituted olefins remain highly went after.

Since the pioneering work of Hallberg23,24, the chelation-assisted intermolecular Heck reaction has been prospering and stood out as a powerful strategy to get stereodefined, highly functionalized olefins25. Terminal alkenes bearing a removable heteroatom-tethered 2-pyridyl or 2-pyrimidyl have been developed by Itami and Yoshida26,27 to implement stereoselective sequential Heck arylation, followed with cross-coupling reactions to produce triaryl- and tetraaryl-alkenes with defined double bond geometry (Fig. 1a). Later on, Studer et al.28 presented a nitroxide-mediated sequential oxidative Heck arylation for the stereoselective synthesis of tetrasubstituted olefins from easily available acrylates. Evidence proved that the directing group (DG) is not necessary for this sequential arylation (Fig. 1b). In another scenario, an alkenyl Catellani reaction has been utilized for modular and regioselective synthesis of all-carbon tetrasubstituted olefins by Dong29 (Fig. 1c), the use of a modified norbornene was critical to suppressing the undesired cyclopropanation pathway. By now, great efforts have been made on chelation-assisted alkenyl C–H activation30,31, but mainly focused on exploring the reactivity of specific sites in various substrates, especially the C–H alkenylation, little work has been reported on sequential alkenyl C–H functionalization, not to mention the modular synthesis of fully substituted olefins32,33,34,35, which would be even more a hostage to the steric hindrance.

Fig. 1: Modular and regio-/stereoselective synthesis of all-carbon tetrasubstituted olefins.
figure 1

a Yoshida’s chelation-assisted regio-/stereoselective synthesis of tetra-arylated alkenes. b Studer’s sequential oxidative Heck arylation for the stereoselective synthesis of tetrasubstituted olefins. c Dong’s modular and regioselective synthesis of tetrasubstituted olefins enabled by an alkenyl Catellani reaction. d This work: native carbonyl group directed stereospecific sequential arylation of alkenes via a pivotal E/Z isomerization. Pd, palladium. Ar, aryl. DIBAL-H, Diisobutylalumium Hydride. E, electrophile. Nu, nucleophile. DG, directing group.

Full size image

As a long-term interest in challenging the C–H functionalization of simple raw materials assisted by a native DG, we envisioned whether we can take advantage of both the flexibility of chelation-assisted Heck arylation and the excellent regio-/stereocontrol of directed alkenyl C–H activation to implement a multistep sequential arylation process, and furthermore, its application in modular synthesis of fully substituted olefins would be a good touchstone. However, in practice, the chelation effect given by a native DG is generally too weak to support a thorough multi-fold Heck arylation, in another hand such a native DG directed distal alkenyl C–H functionalization is also nontrivial. To make the assumption come true, some specific research ideas are designed as follows: (i) olefinic substrates bearing weakly coordinating DGs would provide a chelation effect to bring the metal in close proximity to the alkenyl, (ii) a flexible-alkyl-chain-tethered DG could promote the formation and dissociation of less rigid bidentate complexes (Fig. 1d, I & II) possessing high conformational degree of freedom, (iii) the weak chelation effect may not direct C–H activation but rather stabilizes the alkenyl-metal complex and facilitates the subsequent transformation36,37,38, (iv) the chelation effect may facilitate a reversible Pd(II)-catalyzed internal alkene E/Z isomerization39 to provide one more potential reactive site (Fig. 1d, (E)-3 & (Z)-3), (v) this platform would contribute to improving chemo-/regio-/stereo-selectivity under mild conditions as well, and vi) using a native group as the weakly coordinating auxiliary would be benefit for simplifying syntheses and further derivatization.

Here, we describe our research of carbonyl-directed stepwise double arylation of 1,1-disubstituted olefins, via a sequential Mizoroki-Heck, E/Z isomerization and C–H activation (3 steps in 2 pots), implementing the modular synthesis of stereodefined 1,1,2-triarylated all-carbon tetrasubstituted olefins (Fig. 1d). Some common carbonyl groups can direct the reaction, particularly the aldehyde function works well as a weakly coordinating DG and remains in final products. Research indicates that the first Heck reaction can furnish a terminal mono-arylated product (E)-3 in high stereoselectivity as we expected. Bearing a weakly coordinating DG such as aldehyde, ketone or ester, alkene (E)-3 is unable to take a second Heck arylation to give the target molecule 4 (Fig. 1d, Path b), or undergo C–H activation via a twisted endo-metallacyclic intermediate III to generate the β-trans-arylated product iso-4 (Fig. 1d, Path c). As an alternative, a reversible Pd(II)-catalyzed isomerization from (E)-3 to its (Z)-isomer and subsequent vicinal group-directed alkenyl C–H arylation works well to form the target molecule 4 in a stereoconvergent fashion (Fig. 1d, Path a). Pd(II)-catalyzed internal alkene isomerization has been recognized as a ubiquitous transformation for more than 60 years40, recently its combination with asymmetric catalysis has already been exploited in stereoconvergent synthesis41,42,43,44; however, its latent capacity in affecting stereoselective alkenyl C(sp2)–H functionalization seems to be neglected, particularly the thermodynamically unfavorable E to Z isomerization was believed to have potential practical value45.

Results and discussion

Model substrates selection

Initially, methyl 4-phenylpent-4-enoate 1a bearing a non-conjugated, weakly coordinating group was chosen as the model substrate. It has a pre-installed α-aryl group and is easily available (Supplementary Methods 1.2). Under palladium/mono-N-protected amino acid (MPAA)46 cooperative catalysis, 1,1-disubstituted olefin 1a could undergo β-diarylation with methyl 4-iodobenzoate 2a to form the tetrasubstituted olefin 5a in an isolated yield of 36% (Fig. 2a). During the screening of DGs under the same conditions, surprisingly, alkenyl aldehyde 1b emerged as the best candidate to give the corresponding diarylated product 5b in 40% yield (Fig. 2a). The aldehyde function has rarely been used as the actual DG in palladium catalysis, even in some latest reports on alkene functionalization, the aldehyde could only direct the reaction in case of the formation of the transient directing group47,48,49,50,51,52 (TDG). Thus the olefin 1b bearing a flexible-alkyl-chain-tethered aldehyde was chosen as the model substrate to test our hypothesis.

Fig. 2: Reaction condition optimization and substrate scope for β-mono-arylation.
figure 2

a Model substrates selection. Reactions were run on a 0.2 mmol scale, yields are isolated yields. b Selected optimization for β-mono-arylation. Reactions were run on a 0.2 mmol scale, yields were determined by 1H-NMR spectroscopic analysis using 1,1,2,2-tetrachloroethane (0.1 mmol) as the internal standard, isolated yields are given in parentheses, E/Z ratios were determined by 1H-NMR spectroscopic analysis of the reaction crude. a18% 5b. b20% 3a”. c Substrate generality for β-mono-arylation. All data represent the average of (more than) two independent experiments, reactions were run on a 0.2 mmol scale, yields are isolated yields, E/Z ratios were determined by 1H-NMR spectroscopic analysis of the reaction crude. cGram-scale reaction. d45 °C. eSingle-crystal structure of 3f, determined by X-ray diffraction. w/o, without.

Full size image

β-Mono-arylation in EtOH

To explore the proposed sequential stereospecific arylation (Fig. 1d), the chelation-assisted β-mono-arylation of alkenyl aldehyde 1b with methyl 4-iodobenzoate 2a was first studied (Fig. 2b). After careful evaluation of various reaction parameters (Supplementary Methods 1.3.1), ultimately, in the presence of 5 mol% Pd(OAc)2 and 10 mol% N-Cbz-aspartic acid (Cbz-Asp-OH) as a ligand, and one equivalent of Ag2CO3 as the halide scavenger, after 24 h reaction at room temperature (25 °C) in ethanol under air, the desired mono-arylation product 3a was isolated in 85% yield and approximately 8:1 cis/trans (E/Z) selectivity (Fig. 2b, entry 1). It was found that ethanol (or methanol) would give the highest yield and good E-selectivity when used as the solvent (entries 2 ~ 4). Control experiments indicate that palladium catalyst and silver salt are essential to this reaction, a moderate yield was observed in the absence of ligand, and increasing the temperature led to a diminished yield and selectivity (entries 5–8). When replacing the aldehyde function with a methyl group, without (w/o) the chelation effect, a diarylation 3a” was isolated as the main product, indicating the importance of a DG in controlling the direction of β-H elimination (entry 9). Generally, MPAAs and limited phosphine-based ligands benefit the reaction (Supplementary Table 1).

With the optimized conditions for the chelation-assisted E-selective β-mono-arylation, substrate scope was investigated (Fig. 2c). In the reaction with alkenyl aldehyde 1b, aryl iodides bearing para- or meta-electron withdrawing groups (e.g., ester, halogen, trifluoromethyl, nitro and cyano) afforded the desired products in good yields and stereoselectivities, electron donating groups (e.g., alkyl, alkyloxy) were also well tolerated (3a ~ 3n). Note that reactions involving a functional group with coordinating property (e.g., nitro, cyano) need higher temperature (3f & 3g). The derivative of (+)-Fenchol worked well to provide the corresponding product 3j. The steric factors appeared to be crucial for this reaction, for example, it is hard to get compound 3o bearing ortho-fluorophenyl at room temperature, yet it could be obtained in 64% yield at a higher temperature (45 °C). Some bulky polycyclic aryl iodides (e.g., naphthyl) worked fine as well (3p & 3q). Heteroaryl (e.g., benzofuryl, pyrrolyl, indolyl) coupling partners were not compatible, probably due to their relatively strong coordinating property. The absolute configuration of 3f was determined by single-crystal X-ray diffraction, while other products were assigned based on NMR similarities.

The alkenyl substrate was then explored (Fig. 2c). Modifying the pre-installed phenyl group with an electron-withdrawing group or a large substituent led to an obviously decreased yield (3r & 3s), illustrating the reduced electron density of the double bond is unfavorable for the reaction, and so does the increased steric hindrance close to the coordinating center. Adding one methylene unit to the alkyl chain furnished the β-mono-arylated product 3t in a slightly decreased yield and higher selectivity. In these reactions, selectivities (E/Z) are all close to 10:1.

Consecutive double arylation in HFIP

After introducing the first aryl by Heck reaction, we turned our attention to the second arylation to construct the target 1,1,2-triarylated olefins. As the consecutive diarylation in one pot was obtained in our preliminary study (Fig. 2a), we envisioned that the same condition might be suitable for the second arylation of compounds 3. As a result, the consecutive diarylation of alkenyl aldehyde 1b with methyl 4-iodobenzoate 2a was studied (Fig. 3a).

Fig. 3: Reaction condition optimization and substrate scope for consecutive and stepwise double arylation.
figure 3

a Selected optimization for β-diarylation. Reactions were run on a 0.1 mmol scale, yields were determined by 1H-NMR spectroscopic analysis using 1,1,2,2-tetrachloroethane (0.1 mmol) as the internal standard, isolated yields are given in parentheses. a43% 3a (E/Z = 5:1, determined by 1H-NMR spectroscopic analysis of the reaction crude). b < 10% 6a. c23% 3a (E/Z = 3:2). dSingle-crystal structure of 6a, determined by X-ray diffraction. b Substrate generality for β-diarylation. All data represent the average of (more than) two independent experiments, reactions were run on a 0.2 mmol scale, yields are isolated yields. eGram-scale reaction. fwithout ArB(OH)2. gIsolated yield after esterification with methyl iodide. c Substrate generality for stepwise double arylation. All data represent the average of (more than) two independent experiments, reactions were run on a 0.2 mmol scale, yields are isolated yields. hConfirmed by NOESY. iSingle-crystal structure of 4s, determined by X-ray diffraction. w/o, without. NOE, Nuclear Overhauser Effect.

Full size image

After extensive research, the optimal conditions (catalyst, ligand, base, additive, etc.) have been determined eventually (Supplementary Methods 1.3.2). The desired diarylated product 5b was isolated in 77% yield with 10 mol% Pd(OAc)2 as the catalyst, 30 mol% N-Fmoc-phenylglycine (Fmoc-Phg-OH) as a ligand, 0.5 equivalent of aryl boronic acid as an additive and two equivalents of Ag2CO3 as the halide scavenger, after 24 h reaction at 45 °C in HFIP under air (Fig. 3a, entry 1). Hexafluoroisopropanol (HFIP) displays a unique solvent effect in facilitating diarylation53 (entry 2). Control experiments show that palladium catalyst and silver salt are crucial to this reaction (entries 3 ~ 4). The absence of the aryl boronic acid caused a little decrease in the yield, meanwhile mono-arylated α,β-unsaturated aldehyde 6a was obtained as a byproduct in less than 10% yield (entry 5). The aryl boronic acid could play an effect in slightly promoting the reaction in the absence of ligand (entries 6 ~ 7). In contrast to the obvious mono-arylation in ethanol without the ligand (Fig. 2b, entry 7), herein, there’s almost no reaction in the absence of ligand and additive (entry 7). Replacing aryl iodide with aryl boronic acid led to the formation of mono-arylated product 3a in only 23% yield (entry 8). Other sterically hindered aryl boronic acids, such as 2,6-dimethylphenyl boronic acid and 2,6-diisopropylphenyl boronic acid, were tested, the results were comparable with that obtained without using aryl boronic acids (Supplementary Table 16). These details reveal some important features of reaction mechanism (vide infra). Moreover, a comparable result was obtained at room temperature (25 °C, entry 9). Very little amount of the goal product could be observed when using non-chelating substrate (entry 10).

Under the optimized conditions, a variety of para- and meta-substituted aryl iodides could react with alkenyl aldehyde 1b in one-pot to give the desired diarylation products in moderate to good yields (Fig. 3b), combined with a little mono-arylation. Both electron-poor and -rich aryl groups are tolerated (5b ~ 5s), generally, the presence of an electron withdrawing group can promote the oxidative addition of aryl iodide to palladium, leading to a higher yield than that derived from an electron-rich analogue. Due to the steric hindrance, ortho-fluoro-aryl iodide could only generate the mono-arylated product in approximate 40% yield (3o). The derivative of anti-inflammatory and analgesic drug Ketoprofen was applicable in this reaction (5m). Some multi-substituted aryl iodides were also tolerated, delivering the diarylation products in moderate to low yields (5t ~ 5v). Similar to the β-mono-arylation (Fig. 2c), heteroaryl iodides were not compatible in this reaction, and even the cyano group.

Further investigation of alkenes was performed under optimal conditions with aryl iodide 2a (Fig. 3b). In contrast to mono-arylated products 3r and 3s (Fig. 2c), it seems that steric factors appeared to be more dominant for diarylation (5w & 5x) than electronic effects. For example, the tert-butyl substituted alkenyl substrate can only be diarylated in a low yield (5x), likely due to its aggravated distortion in structure. Some common native groups can also work as coordinating auxiliary to direct the diarylation (e.g., ester 5a, ketone 5y, carboxylic acid 5z) except for the aldehyde. Replacing aldehyde with hydroxy led to a complicated result, such as arylation, isomerization, migration, oxidation, and etc, amide substrate failed to give any product and could be recovered almost completely, the substrate bearing a longer alkyl chain was not compatible either (Supplementary Fig. 1, 1c, f & g). The results suggest that weakly coordinating DGs and a proper length of alkyl chain are definitely important for the consecutive diarylation.

Stepwise double arylation

Having established the efficient assembling methods for β-position mono- and di-arylation, we next examined the stepwise double arylation using different aryl iodides via the combination of these two approaches (Fig. 3c). Experimental results showed the strategy is feasible with a slight modification of optimal conditions. 1,1-Disubstituted olefin 1 was first subjected to the chelation-assisted E-selective β-mono-arylation in ethanol, after simple purification by filtrating over silica gel, the second β-arylation was furnished in HFIP to generate various stereodefined 1,1,2-triarylated tetrasubstituted olefins.

Noteworthy features are that (i) double arylation step-by-step with the same aryl iodide gives a comparable result with that derived from a one-pot diarylation in HFIP (e.g., 5h), (ii) both electron-donating and -withdrawing group substituted aryl iodides are tolerated, however, an electron-donating group generally leads to low yield (4m & 4n), (iii) the steric hindrance is the main factor constraining reactions (4q), the method is not applicable for bulky coupling partners (e.g., naphthyl, tert-butyl-aryl and multi-substituted aryl), (iv) changing the feeding sequence of aryl iodides could furnish the syntheses of Z- or E-type isomers separately (4i ~ 4l), (v) the pre-installed α-aryl group can be varied as well (4r ~ 4t), (vi) some native groups can also direct the stepwise double arylation to give the corresponding 1,1,2-triarylated olefins (e.g., ester 4u, ketone 4v), (vii) the absolute configuration of 4s was determined by single-crystal X-ray diffraction, the configuration of 4r was confirmed by analysis of NOE (Nuclear Overhauser Effect) interactions, and above all, (viii) the E-dominated mixed isomers 3 obtained in the first mono-arylation undergo a stereoconvergent transformation, an E to Z geometric isomerization combining with a subsequent arylation generates a single isomer 4, any other possible isomers were not observed. Mechanistic studies proved the second arylation occurred in HFIP is not a Heck reaction (vide infra).

Mechanistic studies

It is noteworthy that a poor E/Z ratio was obtained in isolated β-mono-arylation product in HFIP (Fig. 2b, entry 2; Fig. 3a, entry 8; Fig. 3b, 3o). A series of control experiments were performed for mechanistic investigation (Fig. 4a~d, Supplementary Discussion 2.2). Parallel experiments of β-mono-arylation and β-diarylation of model substrates employing a smaller amount of aryl iodide 2a were conducted separately (Fig. 4a). In contrast to the high E/Z selectivity observed in the β-mono-arylation (E/Z = 13:1), the mono-arylated product 3a was also obtained in the β-diarylation experiment, albeit with a poor stereoselectivity (E/Z = 5:7), the latter was obviously distinguished from a normal Heck reaction which prefers to form the thermodynamically more stable E geometry. To gain insights into the reaction mechanism, the mono-arylated product 3b (with an E/Z ratio of 10:1) obtained under the standard β-mono-arylation conditions was subjected to different control experiments without using aryl iodides, obvious E to Z isomerization was observed under palladium catalysis assisted by HFIP, aryl boronic acid, ligand and silver carbonate (Fig. 4b). And further, the mixed isomers 3b (with an E/Z ratio of 5:6) could be transformed completely to furnish a single isomer 4w under the modified β-diarylation conditions (Fig. 4c). The results look likely to be consistent with alkene isomerization-then-C–H arylation. Parallel reactions of substrates 1a and deuterated-1a with aryl iodide 2a in HFIP were conducted (Fig. 4d), the KIE (kinetic isotope effect, kH/kD = 2.8) data obtained from their initial rate constants indicates that the C–H cleavage step is possible the rate-determined step for the arylation occurred in HFIP; in particular, double bond migration of 1a was also observed during the reaction, suggesting that an allyl-Pd(II) intermediate is probably involved, nevertheless, terminal alkenyl C(sp2)–H arylation seems to remain dominant. By comparison, the intermolecular competitive experiment in ethanol gave a value of kH/kD = 1 (Fig. 4d).

Fig. 4: Mechanistic studies.
figure 4

a Parallel experiments of β-mono-arylation and β-diarylation of model substrates employing a smaller amount of aryl iodide 2a, a poor stereoselectivity of mono-arylation was obtained in the latter case, suggesting that an E/Z isomerization is probably involved in the catalytic reaction in HFIP. b E/Z isomerization of the internal alkene 3b. Obvious E to Z isomerization was observed under palladium acetate catalysis assisted by ligand or silver carbonate in HFIP. c The second arylation of the mixed isomers 3b (E/Z = 5:6). The formation of the single isomer 4w demonstrates that the second arylation in HFIP is a stereoconvergent procedure, probably consisting of a reversible Pd(II)-catalyzed E to Z isomerization of 3b and subsequent vicinal group-directed alkenyl C–H arylation of (Z)-3b. d Deuterium KIE (kinetic isotope effect) experiments. A KIE data of kH/kD = 2.8 was given in parallel experiments of 1a and deuterated-1a with 2a in HFIP, indicating that the C–H cleavage step might be the rate-determined step for the arylation occurred in HFIP; a value of kH/kD = 1 was obtained in intermolecular competitive KIE experiments of 1a and deuterated-1a in ethanol, more in line with the characteristics of Heck reaction. e Outline of a possible pathway for the stepwise double arylation, some ligands and counter ions were omitted or simplified for clarity. A nucleopalladation via intermediate III is proposed as a key step to provide a stereoconvergent pathway, and a subsequent vicinal group-directed alkenyl C–H arylation via the CMD (concerted metalation deprotonation) mechanism collaborates well to provide the final product.

Full size image

On the basis of experimental results and literature precedents, a plausible mechanism of this chelation-assisted stepwise double arylation is proposed in Fig. 4e. Some ligands and counter ions were omitted for clarity. For the first β-mono-arylation in ethanol, a chelation-assisted intermolecular Heck reaction works smoothly to furnish the terminal arylated product 3 with a high E/Z selectivity. Theoretically, if the E-dominated mixture 3 still underwent the second arylation in HFIP via a Heck-type pathway, the final stereoselectivity after β-H elimination would not be obviously better than that observed in the first arylation. But in fact, the final 1,1,2-triarylated olefin 4 was obtained as a single isomer. Interestingly, a reversible Pd(II)-catalyzed E/Z isomerization of E-dominated-3 and subsequent vicinal group-directed alkenyl C–H arylation of (Z)-3 take place in one pot, in a manner of dynamic kinetic resolution to promote the isomerization and the final formation of the single isomer 4. Although the precise mechanism remains unclear at this stage, evidence suggests that both the directing group and the solvent HFIP display a critical role in alkene isomerization and subsequent C–H activation, thus an anti-nucleopalladation-then-bond rotation-then-β-OAc elimination pathway with the formation of intermediate III could well explain the results39.

For the alkenyl C(sp2)–H arylation that occurred in HFIP, the reactions preferably proceed via a Pd(II)/Pd(IV) catalysis under current conditions, particularly without using aryl boronic acids. What’s more, the key endo-metallacyclic intermediates V and VI are nontrivial in vicinal group directed alkenyl C–H activation, there is a possible interconversion between C-enolate/oxa-π-allyl/O-enolate palladium intermediates or carbonyl-π-palladium that makes the metallocycle V/VI no larger than a 7-membered ring. The presence of aryl boronic acid possibly promotes the formation of active aryl-Pd(II) species, thereby circumventing the issue of side reactions (6a in Fig. 3a) and accelerating the alkenyl C(sp2)–H arylation. However, due to its reductivity exhibited in catalytic circle, the aryl boronic acid cannot replace the role of aryl iodides, even if exogenous oxidants were added. The consecutive double arylation performed in HFIP might occur via a C–H arylation-then-isomerization-then-C–H arylation pathway.

Exploring optical properties

To further validate the potential use of synthesized tri-aryl all-carbon tetrasubstituted olefins in photoluminescence, we next investigated optical properties of compound 4s which is less soluble in common solvents. As depicted in Fig. 5a, b, the fluorescence microscopy images show that solid-state 4s emits strong yellow fluorescence with quantum yield about 11.3% under the excitation of light (λex = 460 nm). The solid-state 4s exhibits a broad absorption band in the visible region of 350–550 nm (with a maximum λab = 450 nm), and its fluorescence emission spectrum is centered at 560 nm (Fig. 5c). Compound 4s displays a large Stokes shift of 110 nm, which indicates the exciton self-quenching could be suppressed effectively.

Fig. 5: Optical properties of compound 4s.
figure 5

a Fluorescence image of the powder for compound 4s without light source. b Fluorescence image of the powder for compound 4s under excitation light (λex = 460 nm). c Solid-state UV/Vis absorption (UV/Vis) and fluorescence emission (FL) spectra of 4s.

Full size image

In summary, a remote-carbonyl-directed palladium-catalyzed sequential Heck/isomerization/C(sp2)–H arylation of 1,1-disubstituted olefins with aryl iodides has been established, various stereodefined 1,1,2-triarylated all-carbon tetrasubstituted olefins were prepared in a modular and stereoconvergent fashion. Several attractive features of this synthetic strategy should be noteworthy. First, all components assembled stem from readily available or easily prepared raw materials. Second, this protocol features mild reaction conditions, operationally simple procedures and diversified product structures. Third, various aryls can be introduced at the desired position simply by changing the coupling partners and adding in an appropriate order, a simple alteration of addition order results in producing possible isomers. Fourth, such a stereoconvergent, remote-chelation-assisted alkenyl C–H arylation open up a door for distal alkenyl C–H functionalization. Fifth, the use of weakly coordinating DGs, especially the direct use of aldehyde function, provides more possibility to achieve diversified transformations in an atom- and step-economic manner. Although some compatibility issues still exist (e.g., heteroaryl iodides), this strategy provides a highly flexible platform to access diverse tri-aryl all-carbon tetrasubstituted olefins which would find many uses in the development of functional organic materials and pharmaceuticals. Further efforts will focus on the investigation into other kinds of functionalization as well as expanding the applicable coupling reagents.

Methods

General procedure for β-mono-arylation

To an oven-dried 35 mL Schlenk tube with previously placed magnetic stir-bar were added aryl iodide 2 (0.3 mmol, 1.5 equiv.), Pd(OAc)2 (2.2 mg, 0.01 mmol, 5 mol%), Cbz-Asp-OH (5.3 mg, 0.02 mmol, 10 mol%), Ag2CO3 (55 mg, 0.2 mmol, 1 equiv.), followed by addition of EtOH (2 mL) and alkene substrate 1 (0.2 mmol). The tube was sealed with a screw cap and the reaction mixture was stirred vigorously at room temperature (25 °C). After stirring for 24 h, the resultant solution was filtered through a short pad of 1:1 mixture of Celite and silica gel, and the column was washed with ethyl acetate (15 mL). The combined organic solutions were concentrated under reduced pressure, and the residue was purified by flash column chromatography on silica gel to afford the desired product 3.

General procedure for consecutive double arylation

To an oven-dried 35 mL Schlenk tube with previously placed magnetic stir-bar were added aryl iodide 2 (0.5 mmol, 2.5 equiv.), aryl boronic acid (0.1 mmol, 0.5 equiv.), Pd(OAc)2 (4.5 mg, 0.02 mmol, 10 mol%), Fmoc-Phg-OH (22.4 mg, 0.06 mmol, 30 mol%), Ag2CO3 (110 mg, 0.4 mmol, 2 equiv.), followed by addition of HFIP (4 mL) and alkene substrate 1 (0.2 mmol). The tube was sealed with a screw cap and the reaction mixture was stirred vigorously on a hotplate at 45 °C for 24 h. After completion of the reaction, the resultant solution was filtered through a short pad of 1:1 mixture of Celite and silica gel, and the column was washed with ethyl acetate (15 mL). The combined organic solutions were concentrated under reduced pressure, and the residue was purified by flash column chromatography on silica gel to afford the desired product 5.

General procedure for stepwise double arylation

To an oven-dried 35 mL Schlenk tube with previously placed magnetic stir-bar were added aryl iodide (0.28 mmol, 1.4 equiv.), Pd(OAc)2 (2.2 mg, 0.01 mmol, 5 mol%), Cbz-Asp-OH (5.3 mg, 0.02 mmol, 10 mol%), Ag2CO3 (55 mg, 0.2 mmol, 1 equiv.), followed by addition of EtOH (2 mL) and alkene substrate 1 (0.2 mmol). The tube was sealed with a screw cap and the reaction mixture was stirred vigorously at room temperature (25 °C). After stirring for 24 h, the resultant solution was filtered through a short pad of 1:1 mixture of Celite and silica gel, and the column was washed with ethyl acetate (15 mL). The combined organic solutions were concentrated under reduced pressure to afford the mono-arylated product 3 which was used in the next step without further purification. To another oven-dried 35 mL Schlenk tube with previously placed magnetic stir-bar were added the second aryl iodide (0.3 mmol, 1.5 equiv.), aryl boronic acid (0.05 mmol, 0.25 equiv.), Pd(OAc)2 (2.2 mg, 0.01 mmol, 5 mol%), Fmoc-Phg-OH (11.2 mg, 0.03 mmol, 15 mol%), Ag2CO3 (55 mg, 0.2 mmol, 1 equiv.), followed by the addition of the crude mono-arylated product 3 dissolved in 2 ml HFIP. The tube was sealed with a screw cap and the reaction mixture was stirred vigorously on a hotplate at 45 °C for 24 h. After completion of the reaction, the resultant solution was filtered through a short pad of 1:1 mixture of Celite and silica gel, and the column was washed with ethyl acetate (15 mL). The combined organic solutions were concentrated under reduced pressure, and the residue was purified by flash column chromatography on silica gel to afford the desired double arylated product 4.