Palladium-Catalyzed Heck Cross-Coupling Reactions in Water: A Comprehensive Review
- 271 Downloads
Palladium-catalyzed cross-coupling reactions have emerged as one of the most versatile tools in organic chemistry. Extensive efforts were made to adapt these reactions to aqueous media, not only for the purpose of environmental conservation but also to expand the scope, increase the efficiency and implement bio-compatible protocols. Among different palladium cross-coupling reactions, the Heck reaction turned out to be the most challenging in an aqueous environment. This led to various original developments in catalyst design. This review summarizes the different approaches pursued to perform Heck reactions in neat water as well as aqueous mixtures. Both, homogeneous and immobilized catalysts, including nanoparticles are presented herein.
KeywordsCross-coupling Palladium Aqueous catalysis Green chemistry Heck–Mizoroki reaction
The Heck cross-coupling reaction (HCR hereafter) was discovered independently by T. Mizoroki in 1971 and R. F. Heck in 1972 and is generally referred to as the palladium-catalyzed arylation of olefins. In the presence of catalytic amounts of palladium, an aryl (pseudo)halide reacts with an alkene in the presence of a base to afford the corresponding arylated alkene (Scheme 1) [1, 2]. What first started as a method for the synthesis of stilbenes from styrenes quickly developed into one of the most versatile and efficient reactions to synthesize substituted olefins both in academia and in industry. Together with Ei-ichi Negishi and Akira Suzuki, Richard F. Heck was awarded the Nobel Prize in Chemistry in 2010 “for palladium-catalyzed cross-couplings in organic synthesis” . Among the reviews covering Pd-catalyzed cross-coupling reactions [4, 5, 6, 7], one should mention the one by Beletskaya and co-workers  as it covers some of the most unique properties of the HCR.
Beside reducing organic waste and avoiding hazardous reaction conditions, running Pd-catalyzed cross-coupling reactions in water presents many benefits. Indeed, the use of water as a solvent offers the following attractive features: (i) it may simplify the purification of the product; (ii) it allows for an easy recovery and recycling of the catalyst  (iii) as water is one of the most polar solvents, it can accelerate catalysis by promoting the migratory insertion and protecting the catalyst by displacing labile ligands from its coordination sphere; (iv) the use of water may invert the selectivity of intramolecular HCR and provide an environment for protective-group free synthesis of complex building blocks [10, 11]; (v) finally, efficient and robust aqueous Pd-catalyzed cross-coupling reactions have found use in bioorthogonal chemistry in vivo [12, 13, 14, 15, 16, 17].
In 1988, Beletskaya and co-workers discovered a ligand-free Pd(OAc)2-catalyzed reaction of water soluble arylhalides with acrylic acid in water in the presence of K2CO3 as base. High yields (> 95%) were obtained using moderate reaction conditions (50–100 °C) which, at that time, were comparable to reactions performed under anhydrous conditions in polar aprotic solvents including MeCN, DMF, DMA, NMP, etc. [18, 19]. Calabrese and co-workers reported in 1990 the use of a sulfonated water-soluble phosphine ligand for aqueous Pd-catalyzed HCR (Scheme 2) . Since 2000, many sophisticated Pd-based catalysts were developed for the aqueous Suzuki–Miyaura cross-coupling (SMC). However, their use in aqueous HCR remains limited [21, 22, 23, 24].
To the best of our knowledge, no comprehensive review on aqueous HCR has been published to date. More specialized reviews on HCR usually cover specific topics such as nucleoside synthesis or the development of water-soluble phosphine ligands etc. [9, 11, 17, 25]. This review strives to provide a comprehensive overview of Pd-catalyzed aqueous HCR. This review is divided in three sections: (i) HCR in water; (ii) HCR in aqueous mixtures; (iii) HCR using immobilized catalytic systems.
2 HCR in Neat Water
2.1 Ligand-Free HCR
In solution, metal ions are always bound to one or several ligands forming a coordination complex. The ligands can either be the solvent, a base or a large organic molecule specifically designed for the purpose of binding and supporting a metal ion. The latter is commonly referred to a Lewis-basic “ligand” used for coupling reactions. “Ligand-free” HCR is generally referred to a reaction whereby no such organic ligands are added to the standard palladium salts. Performing efficient ligand-free HCR in water is desirable both from an ecological and an economic point of view. However, examples are quite sparse. Most protocols use Pd(OAc)2 or PdCl2 in the presence of either an inorganic base (e.g. K2CO3, NaHCO3 or NaOAc) or an amine (i.e. NEt3). Reaction temperatures vary between 50 and 400 °C (e.g. in supercritical water). Typical yields for these reactions are comparable to those reported in the presence of additional ligands. The reaction scope however is often quite limited. Unlike other cross-coupling reactions, the HCR does not require the presence of tightly bound ancillary ligands. The role of the ligand consists mostly of ensuring a defined coordination sphere to stabilize the reactive Pd-species. A strong ligand also suppresses palladium leaching into the solution in the form of either Pd-black or Pd-nanoparticles.
As mentioned above, the first ligand-free HCR in water was reported by Beletskaya and co-workers [18, 19]. The scope was initially limited to acrylic acids, but later extended to styrenes . Efficient coupling for the latter was only achieved for electron-deficient aryl iodides. For other less reactive substrates including aryl bromides, the presence of phosphine ligands was required.
In 1995, Parsons and co-workers investigated the use of superheated and supercritical water (SW) for the ligand-free homogeneous HCR (Scheme 3) [27, 28]. As the temperature rises, the dielectric constant of water drops significantly from 80.1 (20 °C) to 6 (374 °C) due to the decrease of hydrogen-bond strength. At elevated temperatures, water behaves much like diethylether. The corresponding yields were often lower than those reported in organic solvents, possibly due to the deactivation of the catalyst. Side reactions, including hydrogenolysis and hydrogenation were often reported as well. Advantages of performing the HCR in SW include: (i) increased sensitivity towards electronic effects and steric hindrance; (ii) facile reduction of the precatalyst to Pd(0) and (iii) short reaction times. The scope of the HRC in SW was subsequently extended by Gron and co-workers to include cyclic alkenes [29, 30]. Yields were generally moderate and side reactions constituted a severe issue.
Using the unique features of SW, the group of Ikushima reported the coupling of iodobenzene to styrene in the absence of a catalyst [31, 32]. The reaction conditions included triply distilled water, KOAc and careful tuning of temperature and pressure around the critical point. A mechanism was suggested whereby the bifunctionality of SW plays a critical role. The hydroxide ions removes the β-hydrogen of styrene forming a carbanion, which then attacks the aryl halide at the electrophilic carbon. The base neutralizes HI and facilitates the removal of iodide via an ion–dipole intermediate. Interestingly, the use of strong inorganic bases promoted the formation of phenol, ultimately leading to diphenylether instead of the HCR product. Transition metals traces in the ppb range contained in the base used for the reaction may however be the true catalyst . In the past, several “transition-metal-free” cross-coupling protocols were shown to be, in fact, catalyzed by such trace-palladium contaminants [34, 35].
The diastereoselective synthesis of (E)-cinnamonitriles via HCR between acrylonitrile and aryl iodides in water was reported by Cai and co-workers (Scheme 4) . Adding Bu4NBr to Pd(OAc)2 and NaHCO3 afforded good yields (≥ 72%) at moderate temperatures (80–90 °C).
Amini et al. investigated the effect of the solvent on the HCR of aryl iodides and butyl acrylates . Upon screening different copper-, palladium- and nickel salts, Pd(OAc)2 proved most effective, affording quantitative yields. However, a low yield (32%) was obtained in the presence of styrene as olefinic partner. This challenge could be overcome by using a DMSO:water (1:1) mixture to afford a quantitative yield.
By using microwave irradiation, Len and Hervé developed an efficient ligand-free HCR in water for the synthesis of a variety of nucleosides . Interestingly, the base of choice was NEt3 which smoothly reduces Pd(OAc)2 to form the active Pd(0)-species. To illustrate the versatility of aqueous-phase total synthesis, they reported the synthesis of (E)-5-(2-bromovinyl)-2′-deoxyuridine (BVDU) in water (Scheme 5).
2.2 Palladium Nanoparticles
Over two decades ago, Beller et al. reported that palladium nano-particles (PdNPs) derived from Pd(CF3COCHCOCF3)2 are effective catalysts for the coupling of activated aryl bromides in HCR . Reetz and Westermann reported mechanistic studies on PdNPs in phosphine-free Pd-catalyzed coupling reactions based on transmission electron microscopy . They showed that PdNPs (average particle size 1.6 nm) are formed upon thermolytic decomposition of Pd(OAc)2 in polar solvents. It was proposed that tetraalkylammonium salts stabilize these particles, which are thought to be the actual catalytic species for HCR. Since then, a heated debate is ongoing concerning the nature of the active species in ligand-free HCR and ligand-free SMCs. Whether the defect sites in PdNPs, leached Pd(0) or Pd(II) are the catalytically active species remains an open question [40, 42, 43, 44]. It may well be that both the PdNPs and leached species are involved in the catalytic cycle. This may depend on multiple reaction parameters, reflecting the complexity of these reactions (Scheme 6) . For this reason, this section only covers publications explicitly mentioning PdNPs as the catalyst, although it can be assumed that many “ligand-free” HCR may involve nano-sized palladium colloids.
Relying on UV–Vis spectroscopy and transmission electron microscopy, Bhattacharya et al. investigated the formation of PdNPs in water and their reactivity in the HCR . Upon screening various surfactants, cetyltrimethylammonium bromide (CTAB) proved to be the best additive as it provides a strong micellar effect in combination with suitable stabilization for PdNPs.
Heating a solution of PdCl2 with CTAB in water to 80 °C did not display any change in the UV–Vis spectrum. Upon addition of methyl acrylate, the color changed and characteristic absorption bands appeared, diagnostic of the formation of PdNPs. The corresponding PdNPs efficiently catalyzed the arylation of methyl acrylate, styrene, acrylonitrile and ethyl acrylate with various aryl iodides and aryl bromides. Catalyst loadings could be decreased to 0.1 mol% without a significant drop in yield. In contrast to the findings of Reetz and de Vries for catalysis in organic solvents, decreasing the Pd-loading did not lead to an increase in reactivity [39, 40].
Subsequently, Wang and co-workers reported that ultrasonic irradiation could be used to perform HCR in water below room temperature . They demonstrated that ultrasonic irradiation in combination with methyl acrylate was crucial for the formation of PdNPs from PdCl2. The absence of either led to no PdNPs or aggregation to inactive PdNPs. A protocol was reported on the regioselective functionalization (i.e. para) of electron-rich aromatics containing iodo substituents on both the para- and ortho-positions. Such selectivity was observed only at mild temperatures. At 90 °C, both positions were functionalized. The corresponding catalyst could be reused several times without significant erosion of the yield.
Recently, Tsai and co-workers reported the Pd-catalyzed mono- and double HCR of aryl halides and dialkyl vinylphosphonates in water under air (Scheme 7) . Optimization revealed the superior activity of PdCl2(NH3)2 in conjunction with diisopropyl amine (DIPA). Catalysis was performed in water at 80 °C for aryliodides and at 120 °C for arylbromides. According to the proposed mechanism, the palladium precatalyst is reduced to form PdNPs in situ.
Weissleder and co-workers published an extensive study on the use of PdNPs as cellular catalysts for the allycarbamate cleavage and intramolecular HCR in vivo . The complex PdCl2(TFP)2 [TFP = tris(2-furyl)phosphine] was identified as the best catalyst to activate the chemotherapeutic prodrug doxorubicin (DOX) in buffer. This catalyst was then encapsulated with PLGA and PLGA-PEG via nano-precipitation. The resulting PdNPs catalyzed the in vitro uncaging of prodrugs and fluorophores (Scheme 8). The PdNPs were also used for the localized chemical uncaging in mice tumors by injecting the PdNPs intravenously or intra-peritoneally. In vivo imaging revealed that PdNPs selectively uncage prodrugs in the vicinity of the tumors. High tumor-cell uptake and confinement in disparate cell populations are proposed to be the reason for the enhanced activity of PdNPs around tumors. This study highlights the potential of biocompatible Pd-catalyzed cross-coupling reactions and the potential of implementing these under near physiological conditions.
2.3 Phosphine Ligands
The development of efficient ligands is arguably the single most important parameter to improve the scope and the efficiency of palladium-catalyzed cross-coupling reactions. In this context, phosphine ligands have played a key role [3, 48]. Their widespread use in aqueous catalysis remained rather sparse however. To address the frequent water insolubility of phosphines, two complementary approaches have been pursued: (i) use of a surfactant such as different quaternary ammonium salts (QX), or (ii) the introduction of charged moieties including sulfonates or ammonium salts to the phosphine ligand. Both strategies lead to increased solubility and stability of the catalytic systems, which are often reflected in improved activity of the catalyst.
In 1994, Jeffrey investigated the effect of different solvents and QXs on the HCR (Scheme 9) . Interestingly, QXs only lead to an increase in efficiency in aqueous solvents. In addition, the use of QXs allowed the Pd-catalyzed arylation of methyl acrylate in water in high yields (98%) at room temperature. Despite its limited scope, this was the first report of the efficient use of phosphine ligands in water for a Heck reaction. It also highlighted the necessity of an alkali metal carbonate in combination with a QX. Compared to the ligand-free protocol of Beletskaya, this method allows to perform HCR at moderate temperatures.
Walker and co-workers expanded Jeffrey’s protocol to the coupling of methyl acrylate or thiophene to 5-halogenated-2,4-dimethoxypyrimidines (Scheme 10) . They reported that the selectivity of the thienylation was influenced by the reaction temperature and the pressure.
Sulfonated phosphine ligands are usually synthesized under harsh reaction conditions leading to intractable mixtures. To circumvent this challenge, Hiemstry and co-workers synthesized a small library of sulfonated dibenzofuran-based phosphine ligands (Scheme 11) . Combined with palladium, these soluble ligands catalyzed intra- and intermolecular HCR in water as well as in water:acetonitrile (1:1).
The sulfonated dibenzofuryl-phosphine ligands were also used in other aqueous coupling reactions including the Pd-catalyzed SMC and the Rh-catalyzed hydroformylation. However, better yields were obtained with the sulfonated triphenylphosphine (PTTS).
Using microwave-assisted Pd-catalyzed HCR, Wang et al. reported a simple protocol to selectively couple aryl iodides to various olefins [52, 53]. Inspired by Jeffrey’s protocol, the reaction procedure used TBAB and K2CO3. However, Pd(PPh3)2Cl2 was used as the catalyst instead of Pd(OAc)2. Microwave irradiation was compared to conventional heating for the coupling of seven different substrates. Full conversion in minutes compared to hours and higher yields were reported, highlighting the versatility of microwave heating for HCR reactions in water.
A Pd-catalyzed HCR in water was reported by Yokoyama et al. as a critical step in the total synthesis of clavicipitic acid . Strikingly, performing the HCR in organic solvents or under Jeffrey’s conditions gave unsatisfactory results. Optimized reaction conditions included the use of TPPTS, an inorganic base (either K2CO3 or NaOH) and high reaction temperatures (130 °C) in a sealed tube. This strategy allowed for the protective-group free synthesis of optically active clavicipitic acid (Scheme 12). Water was crucial for the success of the synthesis as it suppressed the racemization and decomposition of the free aminoacid under the strongly alkaline conditions. Moreover, tuning of the pH allowed changing the selectivity from an HCR product to an N-allylation. This elegant synthesis highlights the versatility of aqueous HCR for the synthesis of complex organic architectures.
In the same year, Sinou and Rabeyrin reported the asymmetric arylation of 2,3- dihydrofuran with aryl triflates in water under very mild conditions in the presence of surfactants . After evaluating various reactions parameters including: surfactants different atropoisomeric ligands, triflates and catalyst loadings, an enantioselective protocol was developed achieving good conversions (up to 73%), good enantiomeric excesses (up to 67% ee) and perfect regioselectivity (Scheme 13). All reactions were performed in water at 45 °C with Me2(C16H33)N+(CH2)3SO3 − (HDAPS) as surfactant.
Lipshutz and co-workers reported on a series of surfactants based on vitamin E for various cross-coupling reactions [56, 57, 58, 59]. These surfactants consist of a non-natural α-tocopherol unit, a succinic acid or sebacic acid linker and polyethyleneglycol (Scheme 14). In water, these surfactants assemble into “nanocompartments” that catalyze various cross-coupling reactions at room temperature. These catalytic systems offer a very attractive alternative to the use of organic solvents.
One major drawback of using surfactants, could be the potential contamination of the product by the surfactant resulting from the challenge of decanting of emulsions. For this reason, Zhao and co-workers developed cholate-functionalized phosphine ligands capable of forming a hydrophobic micro-environment around the catalytic center . Cholic acid is a natural surfactant present in the liver that is used to emulsify lipids and cholesterols. The corresponding phosphine ligand was synthesized from methyl cholate in five steps and combines the advantages of phosphine ligands and surfactants (Scheme 15). The catalytic system achieved high yields over a broad range of substrates. It allowed to couple styrenes with various aryl iodides at moderate temperature (40 °C). Reactions performed in water in the presence of air strongly favored the more hydrophobic substrates in competition reactions. No selectivity was observed using organic solvents. Although limited to aryl iodides, this system proved broad and efficient.
2.4 N-Donor Ligands, O-Donor Ligands and Palladacycles
Historically, phosphine ligands have dominated the field of Pd-catalyzed cross-couplings thanks to their versatility and high activity. However, their cost, limited stability and challenging recovery increased the incentive to develop more stable phosphine-free ligands including nitrogen- or oxygen-based ligands and mixtures thereof as well as N-heterocyclic carbene ligands (NHC), etc. Direct comparison between phosphine- and other types of ligands are unfortunately rarely presented in the literature for aqueous HCR.
In the late nineties, various bidentate pyridine-containing ligands were developed, all displaying excellent coordination properties towards palladium [61, 62, 63]. These complexes showed high activity in organic solvents even under aerobic conditions. Inspired by this, Nájera and co-workers synthesized a modular, water-soluble di-2-pyridylmethylamine Pd-complex (Scheme 16) . In combination with tetrabutylammoniumbromide (TBAB) and diisopropylamine (DIPA), high yields (≥ 95%) were achieved for the coupling of aryl iodides and aryl bromides to butyl acrylates or m-chlorostyrenes in water. The rates and the yields of the reactions performed in water were higher than those carried out in DMF.
High turnover numbers (TON) were achieved with the phosphine-free cyclopalladated ferrocenylimine (2a) (Scheme 17) . The complex was air stable and water insoluble but highly active in combination with NEt3 and TBAB. Various aryl iodides and aryl bromides could be coupled to different acrylates in very high yields: up to 278,000 TONs were reported. Reactions were performed at reflux and were limited to acrylates as coupling partners. By using a related palladacycle (2b), Nájera and Botella reported the controlled mono and double HCR of α,β-unsaturated carbonyl compounds in water with TON’s up to 59,000 (Scheme 18) . By varying the amount of aryl iodide, the reaction could be steered to produce either the mono- or diarylated product. The base of choice was Cy2NH, which significantly outperformed NEt3 or DIPA. Interestingly, when using simple Pd(OAc)2 instead of 2b, higher yields were obtained for the mono arylation. However, Pd(OAc)2 was only moderately active for the double Heck arylation affording yields of ≤ 13% compared to ≥ 58% when using ferrocenyl catalyst 2b for the same reaction. Generally, extended reaction times (up to 24 h) and high temperatures (up to 120 °C) were required.
With the aim of decreasing the reaction temperature and the reaction time using phosphine-free catalysts for HCR in water, Bumagin and Evdokimov developed a protocol using 4-dimethylaminopyridine (DMAP) in combination with K2CO3 . The corresponding complex, PdCl2(DMAP)2 catalyzed the coupling of aryl iodides and aryl bromides with acrylic acid at 100 °C giving full conversion within minutes.
Efficient HCRs in water at room temperature were achieved by Thore and co-workers . The sulfonated pyridine-based ligand 3a (Scheme 19) in combination with Pd(OAc)2 achieved higher yields in water than in DMF, DME, NMP or aqueous mixtures. This protocol was applied to a broad range of aryl iodides and aryl bromides. All yields exceeded 90% after 4 h. To the best of our knowledge, this is the only protocol using N-donor ligands that affords high yields in water at room temperature without the use of an additional base. So far, this protocol was not extended to olefinic coupling partners other than α,β-unsaturated carbonyl compounds. Compared to the numerous reactions using Pd(OAc)2 presented above, this protocol allows to significantly decrease the temperature to a point where it becomes interesting for applications in a biological context.
Similar studies using higher temperatures but much broader substrate scope were reported by Tsai and co-workers . By using cationic Pd(II)/2,2′-bipyridyl systems in water, TON’s of up to 920,000 were achieved. The cationic ligand 3b displayed good solubility in water and stability in air for both Pd- as well as Rh-complexes. The high reaction temperatures (140 °C) allowed to use very low catalyst loadings (0.0001 mol%). This method gave high yields for 50 different substrate compositions including some challenging styrene analogues.
With the aim of expanding the field of microwave-assisted organic synthesis, Singh and Allam reported a cheap and easy method using Pd(l-proline)2 in combination with TBAB for HCR in water . The precatalyst was prepared by mixing Pd(OAc)2 and l-proline in the presence of NEt3 in methanol at room temperature. Within minutes, challenging substrates including 1-octene or cyclohexene smoothly reacted with aryl iodides and aryl bromides in high yields. This is one of the few methods reporting an efficient coupling of benzyl chloride via HCR in water.
To improve the recyclability and to facilitate purification, Kumar, Khungar and co-workers synthesized an ionic liquid-tagged Schiff base Pd-complex (Scheme 20) . The inclusion of an ionic liquid appendage on the Pd-complex afforded a soluble easy to recover catalyst for Heck- and Suzuki reactions in water. Generally, neutral catalysts tend to leach into the organic phase preventing their recycling. This catalyst however could be recovered and reused six times without significant loss in activity. In addition, this is one of the few reports on HCR in water with an aryl chloride substrate. This is particularly interesting, as aryl chlorides are usually significantly cheaper than their iodide- or bromide counterparts, which allows for better atom economy.
Ahmed and Waheed synthesized various ionic coumarin-derived ligands for the SMC and the HCR in water . All ligands (5a–5e) were synthesized in four steps from ammonia, benzaldehydes and 2-naphthols (Scheme 21). The binding mode towards palladium was not specified by the authors. In combination with TBAB and K2CO3, various aryl iodides and aryl bromides were coupled with methyl acrylate or styrene at 80 °C in water. As demonstrated by Cai and co-workers , Pd(OAc)2 alone catalyzes various HCR under similar conditions. The addition of an ionic ligand however allowed to broaden the reaction scope and to recycle the catalyst. In this case, the catalyst remained reasonably active (60% yield) after the 4th recovery.
N-heterocyclic carbene ligands (NHC, 5a, 5b) display similar electronic properties to phosphines: very good sigma-donor ligands combined with limited π-acceptor properties. Since their first characterization in 1991 and their use as ligands in a cross-coupling reactions 4 years later (5c and 5d), NHC’s were extensively used in catalysis as they emerged as a versatile alternative to the phosphines (Scheme 22) [73, 74]. In contrast to many phosphines, NHC’s tend to be easy to prepare, easy to functionalize, air-, temperature- and moisture stable and tightly bind to metals in all oxidation states [75, 76, 77].
Despite the obvious advantages of NHC-ligands in aqueous environment and their broad use for various cross-coupling reactions in organic solvents, it took 20 years until the first report of a HCR using NHC-ligands in water was published. In 2011, Tu and co-workers reported the use of a pyridine-bridged bisbenzimidazolydinene-Pd-pincer complex for an HCR in water (Scheme 23) . The complex (6) catalyzed the coupling of different aryl iodides with tert-butyl acrylate in high yields. The molecular nature of the NHC-Pd catalyst was confirmed through different catalyst-poisoning experiments using mercury.
Luo and Lo reported the synthesis of a bis-NHC-Pd catalyst (7) from caffeine. It catalyzes various cross-coupling reactions in water (Scheme 24) . High yields were obtained for the coupling of aryl iodides and aryl bromides to methyl acrylate. Again, the catalyst was used in combination with a surfactant (Brij), a base (NEt3, KOH or K2CO3) and near reflux temperatures (90 °C) in water. This protocol allows for a simple and cheap catalyst synthesis in combination with a green coupling procedure.
Originally designed for the SMC, the proline-derived catalyst 8 prepared by Shao and co-workers catalyzes the coupling of various challenging aryl bromides including bromo-thiophenes and bromo-pyridines with acrylates and styrenes in good yields (Scheme 25) [80, 81]. Compared to the SMC protocol, high temperatures were required (100–120 °C) to achieve good yields in the HCR. Although the catalyst is highly air- and moisture stable, all reactions were performed under a nitrogen atmosphere. Importantly, this is one of the few reports achieving high yields for an HCR in water without the use of surfactants or surfactant-like ligands.
In 2012, SanMartin, Dominguez and co-workers investigated the reactivity of CNC palladium pincer complexes (9a and 9b) for different cross-coupling reactions (Scheme 26) . High solubility in water was achieved by the incorporation of polar COOMe and COOH groups on the ligand. A more polar version of Tu’s pyridine-bridged bisbenzimidazolydinene pincer complex (6) was reported. Despite the increased solubility, addition of a surfactant (e.g. TBAF) was necessary to achieve reasonable yields for the HCR of styrenes with aryl bromides. The CNC pincer complexes performed generally better than the unsymmetrical PCN pincer complex (9c) both in water and in DMF. General reaction conditions included the use of K2CO3, high reaction temperatures (130 °C) and long reaction times (22 h). The complexes 9a and 9b afforded moderate to excellent yields in DMF but only moderate yield in water. Interestingly, the more polar complex 9b performed worse in water than its ester counterpart 9a, suggesting that the solubility of the catalyst may not be the sole critical factor.
Following a different approach to improve solubility of HCR catalysts, Huynh co-workers developed an efficient method to oxidize dinuclear thiolato-NHC Pd(II) complexes with Oxone to afford water-soluble NHC-SO3-derivatives (Scheme 27, 10a–c) . From the reported ligands, 10a and 10c proved to be excellent ligands for the Pd-catalyzed HCR of tert-butyl acrylate with 4-bromobenzaldehyde in water. In combination with NEt3 and TBAB, quantitative yields were obtained with a catalyst loading of only 0.05 mol%. However, other aryl bromides including 4-bromotoluene and 4-bromoanisole afforded low yields. No coupling was observed for aryl chlorides.
With the goal of preparing an NHC-ligand with surfactant properties, Taira et al. implemented an oligoethylene glycol monomethyl ether chain and a n-dodecyl chain on either nitrogen of the NHC ring (11a) (Scheme 28) . The corresponding Pd-complex forms emulsions, thus decreasing interfacial tension and solubilizing the organic substrates. The Pd center is localized between the hydrophobic interior and the hydrophilic exterior thus forming a reactive interface. The corresponding Pd-complex was highly active, allowing to produce styrene from iodobenzene at moderate temperatures (70 °C) in DMF as well as in water. Other aryl iodides were coupled to afford either styrene or methyl acrylate in good yields. However, HCR with aryl bromides gave low yields and no coupling was observed for aryl chlorides. The importance of the surfactant properties was highlighted by synthesizing a hydrophobic analogue (11b), which was only moderately active for HCR in water.
Crabtree and co-workers reported abnormal NHC complexes (aNHC) whereby the C-4 or C-5 position of the NHC binds to the metal (classic NHC complexes bind via the C-2 atom) . Those aNHCs opened a new field in ligand design as they displayed excellent activity at very low catalyst loadings. Mandal and co-workers synthesized an abnormal NHC palladium dimer for aqueous oxidative HCR at room temperature (Scheme 29) . The dimer 12 was readily prepared from asymmetric, sterically-crowded phenylimidazolium salts, thus only allowing for abnormal binding at the C-5 carbon. In combination with benzoquinone as the oxidant and TFA, good to very good yields were obtained for 25 different β-arylated products including electron deficient olefins such as acrylates and electron-rich olefins such as vinyl acetates. All reactions were performed in water at room temperature with boronic acids instead of the aryl halide coupling partner.
3 HCR in Aqueous Mixtures
The addition of water to a cross-coupling reaction typically taking place under exclusion of air and moisture may seem counter-intuitive, yet water can have significant benefits as an additive or co-solvent in terms of efficiency and selectivity. Water, thanks to its high polarity can have an accelerating effect on the oxidative addition as well as displace labile ligands, thus removing tightly bound iodide ions from the metal ion. Performing HCR in water has one major drawback however: the limited solubility of many substrates in water. Thus, several HCR protocols were developed relying on aqueous mixtures, such as water:DMF, water:EtOH or water:CH3CN.
As early as 1993, Daves and Zhang reported on the critical importance of water as a co-solvent on HCR that do not proceed in neat organic solvents . Pd-mediated cross-coupling of nitrogen-heterocyclic halo-derivatives with 2,3-dihydrofuran is challenging as most attempts in water or neat organic solvents yielded unsatisfying results. However, when performed in a 1:1 EtOH:H2O mixture, afforded up to 83% yield without the use of any ligand. Various nitrogen heterocyclic compounds including nucleosides (13a, 13b) were synthesized with this procedure at low reaction temperatures (25–50 °C) (Scheme 30).
In 2003, Hallberg and co-workers reported a methodology to selectively functionalize methylcyclopentanone at the α-position of the ketone in a stereoselective manner via ligand-free aqueous HCR (Scheme 31) . First, the methylcyclopentanone was subjected to an acid-catalyzed acetalization-elimination protocol. The corresponding cyclic enolether was linked to an enantiopure pyrrolidine moiety. The regioisomer was regio-and stereoselectively arylated in aqueous DMF via a ligand-free HCR. The reaction mechanism was suggested to proceed via a diastereopure six-membered palladacycle. Finally, an acid-mediated hydrolysis of the functionalized cyclic enol ether afforded the corresponding 2-aryl-2-methylcyclopentanones in very good yields. Reactions were performed at 70–100 °C in air.
3.2 Phosphine Ligands
With the aim of modifying the molecules of life in their natural environment, Casalnuovo and Calabrese investigated the use of TPPTS for various aqueous cross-coupling reactions including HCR . For the first time, the Pd(PPh2(m-C6H5SO3M))3 (M = Na+, K+) complex was structurally characterized. Mild reaction conditions with temperatures ranging from 25 to 80 °C were used to afford good to excellent for various cross-coupling reactions in water:EtOH, water:CH3CN, water:MeOH mixtures or in biphasic mixtures of water:MeOH:benzene.
Two years later, Genet and co-workers significantly expanded the scope of aqueous Pd-catalyzed cross-coupling reactions using a similar procedure . Even milder reaction conditions (25–66 °C) in water:CH3CN or water:EtOH with 5% TPPTS and 2.5% Pd(OAc)2 were applied. Very high yields were achieved for the aqueous HCR of different aryl and allyl iodides with electron deficient as well as electron rich olefins. A slightly modified procedure was applied for Sonogashira cross-coupling, SMC and different Pd-calatylzed allylic substitutions mostly at room temperature. Compared to ligand-free HCR, the use of TPPTS allowed to increase the substrate scope while maintaining low reaction temperatures.
Encouraged by the findings of Jeffrey, Zhou and co-workers developed a methodology to arylate immobilized aryl iodides in aqueous mixtures via HCR . Various aryl iodides were coupled to TentaGel resins or Millipore PS-PEG-PAL resins and subsequently treated with different vinylation reagents. Good to very good yields were obtained in DMF:water. Moderate yields were obtained in water. Optimal reaction conditions included the use of Pd(OAc)2, PPh3, K2CO3, TBAC in a 9:1 mixture of DMF:water.
In 1997, Sinou reported a Pd-mediated intramolecular HCR on carbohydrate templates to synthesize enantiopure bicyclic and tricyclic compounds (Scheme 32) [91, 92]. Depending on the configuration at C-4, erythro structures afforded the bicyclic glycal while threo derivatives afforded a tetrahydrofuranic moiety. However, erythro structures bearing a vinylic or allylic alcohol at C-1 reacted further to yield the tricyclic structures. This protocol was applied to many carbohydrate templates giving access to important enantiopure synthons.
Further studies to functionalize sugars via HCR were reported by Hayashi et al. . The reaction of protected 2-bromo-d-glucal with methyl acrylate was performed in CH3CN, CH3CN:water and DMF:water with Pd(dba)2, P(o-tol)3 and a base. The reaction proceeded better in CH3CN (90% yield) compared to aqueous mixtures (≤ 84% yield) and even allowed for the coupling of unprotected 2-bromo-d-glucal (64% yield).
TPPTS, originally developed by Rhone-Poulenc for aqueous-phase hydroformylation, was widely applied for various aqueous cross-coupling reactions . However, in the case of aqueous HCR, harsh reaction conditions were required. Thus, Shaughnessy and Moore synthesized two analogues, tri(4,6-dimethyl-3-sulfonato-phenyl)phosphine trisodium (TXPTS) and tri(4-methoxy-6-methyl-3-sulfonatophenyl)phosphine trisodium (TMAPTS) with the aim of creating a more active system that catalyzes aqueous HCR under mild conditions (Scheme 33) . By comparing Pd-catalyzed HCR with TXPTS, TPPTS, TMAPTS and a ligand-free protocol for coupling styrene with aryl bromides, TXPTS proved by far the most active, affording up to three-times higher than TPPTS or TMAPTS. Essentially, no product was observed in the absence of ligands. All reactions were carried out with Pd(OAc)2 and Na2CO3 at 80 °C in a 1:1 mixture of CH3CN:water. TXPTS proved to be also an efficient ligand for the aqueous SMC giving high yields under even milder conditions (50 °C).
One year later, the same group reported the use of 2-(di-tert-butylphosphino)-ethyltrimethylammonium chloride (t-Bu-Amphos) and 4-(di-tert-butylphosphino)-N,N-dimethylpiperidinium chloride (t-Bu-Pip-phos) as effective ligands for various aqueous cross-coupling reactions (Scheme 34) . SMCs were possible at room temperature giving high yields for the coupling of a variety of aryl bromides with aryl boronic acids. The reaction temperature had to be increased for the Sonogashira coupling (50 °C) and the HCR (80 °C) to achieve good yields. Overall, t-Bu-Amphos outperformed the commercially used TPPTS in terms of reactivity. In addition, this system could be recycled with only a minor loss in activity.
New water-soluble phosphatriazene ligands were developed by the group of Kapdi . Both water soluble PTAPS and PTABS ligands (where PS is propane sulfonate and BS is butane sulfonate) were synthesized in one step from 1,3,5-triaza-7-phosphaadamantane (PTA) in high yields. PTABS displayed great activity for the SMC as well as the Sonogashira coupling under mild conditions in water or a mixture of water:CH3CN (1:1). Higher temperatures were required to obtain good yields for the HCR. The corresponding catalyst was used in the total synthesis of BVDU: the protective-group free coupling of 5-iodo-2′-deoxyuridine and methyl acrylate was implemented as a key step in this synthesis. Subsequent hydrolysis and NBS-mediated bromination gave BVDU in 72% overall yield (Scheme 35). Compared to the ligand-free microwave-assisted HCR reported Len and Hervé, this corresponds to a slightly lower yield (87 vs. 90%) .
3.3 N-Donor Ligands, O-Donor Ligands and Palladacycles
Only isolated examples of aqueous HCR between aryl-chlorides and olefins under mild conditions have been reported [70, 71]. A protocol reported by Jin and co-workers describes a 3 step-synthesis of β-diketiminatophosphane 14 and its use as a highly efficient ligand for the aqueous HCR and the Buchwald-Hartwig amination (Scheme 36) . Due to their inert C–Cl-bond, aryl chlorides display low reactivity in oxidative additions. As a consequence, elevated temperature and high catalyst loadings are often required. Strikingly, Jin’s protocol allowed to couple a large variety of aryl chlorides –including 2-chloropyridines, 2-chlorothiophenes, 2-chlorofurans and 2-chloro-N-methylpyrrole– with acrylates or styrenes under mild conditions with good yields. General reaction conditions include the use of TBAB, K3PO4, a mixture of DMF and water (1:4) and reaction temperatures varying between 60 and 120 °C. An unsymmetrically divinylated-arene was prepared via one-pot double HCR of o-dichlorobenzene using the same protocol. This Pd-complex could also be used for Buchwald–Hartwig amination at room temperature, thus highlighting the versatility of this catalyst for a wide range of cross-coupling reactions.
The aqueous oxidative HCR was implemented as an attractive alternative to the well-established “click reaction” for bio-orthogonal ligations of proteins by Minnaard, Dekker and co-workers (Scheme 37) [99, 100]. For this purpose, the ligand bis(mesitylimino) acenaphthene (BIAN) was selected for the aqueous oxidative HCR. It clearly outperformed the commonly used dimethyl-1,10-phenanthroline (dmphen) ligand. The addition of water as a co-solvent with MeOH suppressed the formation of boroxides and allowed to decrease the amount of phenylboronic acid from 3.0 equivalents to 1.5 equivalents. A mixture of water:MeOH (1:9) in the presence of dioxygen was identified after extensive screening of different solvent compositions. Optimized reaction conditions allowed for the efficient coupling of challenging substrates including as cinnamaldehyde at room temperature under base-free conditions.
The oxidative HCR under mild aqueous mild conditions was further adapted for bio-orthogonal ligation . Although alkynes are widely used in bio-orthogonal ligation, they suffer from several drawbacks including their lack of stability, their high tendency for homo-coupling and potential inactivation by oxidative enzymes [103, 104, 105]. Alkenes on the other hand, allow for selective, efficient and stable ligation with proteins. Using Pd(OAc)2 in combination with BIAN, various aryl boronic acids were coupled with a mutant of 4-oxalocrotonate tautomerase (4-OT R61C) in a mixture of aqueous buffer and DMF at room temperature (Scheme 38). Furthermore, the fluorophore 3-(dansylamino) phenylboronic acid was selectively coupled with 4-OT R61C in a soluble fraction of lysates from RAW 264.7 macrophages. Performing the reaction in water under the same conditions resulted in very low yield as Pd(II) is believed to bind to proteins if not complexed tightly by ligands. This issue was addressed by substituting BIAN with EDTA . This protocol highlights the potential of aqueous HCR for biological applications.
3.4 Palladium Nanoparticles
Replacing expensive and toxic ligands with abundant carbohydrates lead to an environmentally friendly procedure developed by Camp et al. (Scheme 39) . A catalytic system consisting of Pd(OAc)2, glucose and NEt3 in a mixture of CH3CN and water allowed to couple various aryl iodides via aqueous HCR, aqueous SMC and aqueous Sonogashira cross-coupling reaction to the corresponding cinnamates or biaryl products. The addition of glucose, increased the yield significantly from 18 to 97% in the HCR of iodobenzene with methyl acrylate. It was proposed that glucose promotes the reduction of Pd(II) to form PdNPs and stabilizes these to afford stable and defined colloids. The yields were similar to those relying on tetraalkylammonium salts. Further, Bhattacharya, Sengupta and co-workers could show that methyl acrylate alone reduces Pd(II) to PdNPs. Further additives are however necessary to avoid Pd-black formation . The use of carbohydrates as reducing agent and ligand for PdNPs offer a cheap alternative that allows the recycling of the catalyst.
4 HCR Using Immobilized Catalytic Systems
The use of immobilized catalytic systems (ICS) where the catalyst is anchored to a solid support such as zeolites, molecular sieves, metal oxides, clay etc. offers an attractive means to recycle precious catalyst and thus reduce some of the costs associated with homogeneous catalysis. In particular, ICS are attracting significant attention in the context of continuous-flow systems . The identification of the nature of active catalyst in immobilized catalytic systems is significantly more challenging than for their homogenous counterparts [108, 109]. This review focuses on catalytic applications rather than mechanistic details, including the nature of the precatalyst for aqueous HCR. Excellent reviews detailing mechanistic considerations of immobilized catalytic systems have been published [110, 111].
4.1 Silica-Based Support
In 1989 Arhancet et al. introduced a new class of heterogeneous catalysts by impregnating narrow-pore-sized silica with Rh-phosphine salts . Various alkenes were hydroformylated in water without any detectable leaching. Glass-bead technology combines high activity, high selectivity with facile recyclability in aqueous media . Despite the attractiveness of this support, only limited examples of aqueous HCR have been published to the best of our knowledge.
Pursuing a similar strategy, Williams and co-workers impregnated “hydrophobic-silica” with different phosphine-palladium species to create a recyclable heterogeneous catalyst for aqueous HCR and allylic substitution (Scheme 40) . The corresponding catalyst was used to couple various nitrogen-containing aryl bromides with methyl acrylate. Only limited leaching of palladium was detected. Higher yields were achieved with tri-o-tolylphosphine compared to triphenylphosphine. Chiral phosphines were also tested for the enantioselective allylic substitution.
A novel bifunctional silica-based palladium catalyst (SBA-R/Im-NH2-Pd) was reported by Nasab and Kiasat for aqueous HCR (Scheme 41) . Different electron-rich and electron-deficient aryl iodides and aryl bromides were coupled in high yields with various alkenes in water at 90 °C. The catalyst was recycled up to five times with only a slight loss in yield (98–88%). Analysis of the crude reaction mixture via ICP-MS indicated negligible Pd-leaching. Furthermore, a hot filtration test highlighted that, after removal of the heterogeneous catalyst, the reaction did not proceed. Both tests support the hypothesis that the immobilized PdNP are the catalytically active species. The immobilized catalyst compared favorably to a similar system operating in organic solvents: for the HCR of phenyl iodide with styrene, the SBA-R/Im-NH2-Pd catalyst outperformed other systems that operate in organic solvents and require higher temperatures and longer reaction times. Short reaction times, moderate reaction temperatures, straightforward catalyst recycling combined with good yields highlight the economic and ecological benefits of such heterogeneous catalysts for aqueous HCR.
4.2 Organic Polymer-Based Supports
Organic polymers have found wide applications for the immobilization of precious metal catalysts. The use of copolymers is of especial interest for aqueous catalysis as they can be designed to display amphiphilic properties, including a hydrophobic block where the homogeneous catalyst is anchored. In the presence of hydrophobic substrates, this allows to increase the effective molarity of the substrates in the proximity of the metal center. In the late 90 s, Uozumi and co-workers designed various PEG-PS copolymers linked to palladium-triarylphosphines as supported catalysts for different reactions including aqueous HCR [116, 117, 118, 119]. Pd-PEP (Scheme 42) was used to prepare cinnamic acid from iodobenzene and acrylic acid at room temperature in high yield (92%). Reactions were carried out in water with reaction times of 20 h with a 3 mol% catalyst loading. The product was purified by filtration and extraction of the reaction mixture. In a subsequent study, related systems were applied and optimized for the aqueous HCR of various aryl halides and alkenes affording high yields using 10 mol% catalyst .
Although Uozumi’s system gave high yields and good recyclability, it was lacking activity and thus required high catalyst loadings. To overcome this challenge, the group of Weberskirch designed new amphiphilic, water-soluble diblock copolymers with a covalently linked NHC-Pd catalyst (Scheme 43) [121, 122]. These supported catalysts showed very high TOF numbers of up to 2700 h−1 for the coupling of iodobenzene with styrene in water. The supported catalysts were prepared using a cationic ring-opening polymerization of functionalized 2-oxazolines and 2-methyl-2-oxazoline as reported by Aoi and Okada . The corresponding monomers consist of the NHC-Pd catalyst, a flexible alkyl spacer and 2-oxazoline. A thorough study of the effect of reaction conditions for the aqueous HCR as well as the aqueous SMC was performed. Optimized reaction conditions for the HCR included low catalyst loading (0.67 mol%), reaction times (30 min) at 110 °C. The base of choice was K2CO3 as it significantly outperformed other bases including: NEt3, NBu3 and KOAc. Although this protocol was efficient for coupling aryl iodides, mainly dehalogenation was observed for aryl bromides suggesting that the catalyst may not be active enough for less reactive haloarenes. It was again shown, that such Pd-based catalysts were significantly more active for the aqueous SMC than for aqueous HCR affording TOF’s of up to 5200 h−1 at 90 °C for the coupling of aryl bromides with phenylboronic acid.
With the goal of combining the advantages of ionic liquids and heterogeneous catalysis, Qiao et al. prepared different imidazolium-styrene copolymers linked to PdNPs for the aqueous HCR (Scheme 44) . The copolymers were synthesized from 1-vinyl-3-butylimidazolium chloride, a common ionic liquid, and styrene by the use of AIBN. Unlike Weberkirch’s approach, Qiao didn’t include the catalyst in the monomer, but introduced it to the assembled copolymer by stirring the copolymer and Pd(OAc)2 in ethanol at room temperature for 24 h. TEM and electron diffraction analysis suggested that palladium was bound as a PdNP rather than as an NHC-complex. The coupling of ethyl acrylate with iodobenzene was selected as model reaction. Interestingly, a black precipitate was observed in reaction mixtures consisting of water and ethanol as well as for reaction temperatures above 100 °C. Yields of up to 86% were obtained under optimized reaction conditions (water, NEt3, 100 °C). Recycling proved possible with 83% yield for the fourth run. Various electron-rich and electron-deficient aryl iodides were coupled with different alkenes in moderate to very good yield. Although this protocol did not outperform Weberskirchs catalysts in terms of TOF [121, 122], it displayed very good yields for a broad range of substrates, combined with good recyclability.
Using ultrasonication, Rezaei developed a very efficient PEDOT nanofiber/PdNP composite capable of performing aqueous HCR at room temperature at very low catalyst loading (Scheme 45) . PEDOT [poly(3,4-ethylenedioxythiophene)] is a conductive polymer that recently attracted attention for various applications. Its reducing potential in the monomer state allowed its use for the in-situ reduction of Pd(II) salts to PdNPs before initiation of the polymerization. The corresponding PdNPs are incorporated between the PEDOT chains within a high-surface structure. Quantitative yields were obtained for the coupling of 4-methyl-iodobenzene with n-butyl acrylate with 0.05 mol% catalyst loading at room temperature in water after 25 min. Screening various organic solvents revealed that water was the best solvent for this system. The same protocol could be applied for aryl iodides and bromides with n-butyl acrylate. Strikingly, various aryl chlorides were successfully coupled with alkenes in moderate to very good yield. This is one of the rare protocols reported for the coupling of aryl chlorides with alkenes via aqueous HCR (Scheme 46). Furthermore, this is, to the best of our knowledge, the only report of performing this transformation at room temperature in water, albeit using ultrasonic irradiation.
4.3 Graphene or Graphene-Like Based Supports
Graphene-like supports are interesting alternatives to classical polymers as they possess a stacked sheet morphology, high thermal conductivity, high chemical stability and excellent mechanical strength. Their high-surface area and insolubility in most solvents make graphene-like supports a highly versatile support for the immobilization of homogenous catalysts.
The group of Wu prepared a Pd(II)-Schiff base complex immobilized covalently on hexagonal boron nitride (h-BN) as a highly efficient catalyst for aqueous HCR and aqueous SMC (Scheme 47) . The immobilization of PdNP’s on h-BN yielded a broad NP-size distribution and lead to significant leaching of palladium. Palladium was thus immobilized as Pd(II)-Schiff-base complex on furfurylimine located on the edge of the BN sheets. The immobilized system catalyzed the coupling a wide range of aryl iodides and aryl bromides with different alkenes in high to very high yields combined with excellent recyclability: > 90% conversion was obtained after the ninth recycling. The reaction conditions were similar to other heterogeneous protocols and included the use of K2CO3 as the base of choice, 5 mol% catalyst loadings, at 90 °C in water. Various pharmacologically active compounds such as Felbinac were prepared using h-BN/Fur/Pd(OAc) 2 as the catalyst for the critical HCR step. The good recyclability of this immobilized catalyst, combined with excellent yields makes it an attractive catalyst for industrial use.
Another approach to immobilize Pd-catalysts on graphene-like compounds for aqueous HCR was developed by Hezarkhani and Shaabani . Natural keratin, isolated from wool, was used as a linker to covalently attach Pd(II) tetrasulfophthalocyanine to modified graphene oxide nanosheets (Scheme 48). Palladium phthalocyanines are known to reversibly release the palladium precatalyst in a controlled manner. To suppress aggregation, phthalocyanines were covalently anchored to the graphene oxide. A hot filtration test in combination with atomic absorption spectrometry revealed leaching of palladium upon increasing the reaction temperature. The released palladium species could be recovered upon cooling the reaction mixture. Neglectable Pd-amounts (< 0.1 ppm) were observed in the reaction mixture after catalysis however. The immobilized catalyst was a good catalyst for HCR and Sonogashira coupling: different aryl bromides or iodides were coupled to either styrene or phenylacetylene in water. The coupling of chlorobenzene was more challenging and afforded low yields for both reactions. Low catalyst loadings (0.8 mol%), neglectable Pd-leaching and good recyclability (up to 5 runs) highlight the versatility of this reversible in-situ Pd catch and release catalytic system.
4.4 Coordination Polymer-Based Support
Coordination polymers consist of repeating units of coordination complexes combining the properties of metal complexes and polymers to form new hybrid materials with unique properties for a wide range of applications . The group of Sun expanded the application of coordination complexes to aqueous C–C coupling by synthesizing a series of new highly active heterobimetallic Pd/Ln coordination polymers based on 2,2′-bipyridine-4,4′-dicarboxylic acid . The combination of different Ln(III) (Ln = Nd, Sm, Eu and Dy) and Pd(II)-ions in a polymeric framework lead to an extremely stable (up to 400 °C) and highly active catalyst for aqueous HCR as well as for aqueous SMC. All coordination complexes performed similarly. Thus, the Pd/Sm was selected for further studies. The corresponding coordination complex outperformed the homogeneous catalyst in the aqueous HCR of 4-bromoacetophenone with styrene, highlighting the synergistic effect of samarium on the catalytic performance of the Pd-species. The electropositivity of lanthanides is hypothesized to affect the reductive elimination step. Unlike most protocols in this section, t-BuOK was used instead of K2CO3 as the base of choice. The catalyst performed poorly in pure solvents including toluene, water or DMF. Very good yields were obtained in a 1:1 mixture of DMF:water. Catalyst loadings could be reduced to 0.4 mol% affording to good to very good yield for the coupling of various aryl iodides and bromides to styrene. Again, low yields were observed when using aryl chlorides as substrate. Optimal reaction temperatures were found to be around 90 °C, as either, higher or lower temperatures caused a significant erosion in yield. Interestingly, similar observations concerning the optimal reaction temperature were reported for several immobilized catalytic systems for aqueous HCR presented herein.
4.5 Supramolecular Support
Cyclodextrins (CDs) are cyclic glucose oligomers first discovered by Villiers . The conical-shaped body forms a hydrophobic cavity that allows hosting of various drugs, flavors, perfumes as well as catalysts (Scheme 49). Importantly, CDs provide an enantiopure environment for catalysis. Through functionalization of the CDs with ligand moieties, various transition metal ions including Pd(II) can be anchored to the host, thus yielding a “metalloprotein mimic” .
Deng and coworkers developed an eco-friendly protocol using amino acid- modified cyclodextrins as ligands for aqueous HCR . The solution was saturated with CDs to improve substrate solubility. Different aryl iodides were coupled with styrenes using 0.1 mol% Pd(OAc)2 with moderate to very good yield. Out of the four different amino acids used (i.e. glycine, alanine, proline and phenylalanine) phenylalanine gave the best results (up to 95% yield). Some drawbacks associated with this novel protocol include: high temperatures (100 °C) prolonged reaction times (> 10 h), large excess of Li2CO3 (forty fold) as well as CD-saturated solutions.
In this context, the group of Jung reported an efficient protocol relying on a 2-aminopyridine-modified β-CD to anchor the Pd-complex . This method allowed for efficient HCR of various aryl iodides and bromides with different olefins in water with good to very good yields. Optimized reaction conditions include the use of CD/Pd-cat (3 mol%) with slight excess of K2CO3, refluxing reaction conditions and varying reaction times (4–12 h). The corresponding catalyst could be recovered and recycled for several runs by precipitation upon addition of acetone.
Combining the advantages of supramolecular hosts and polymer support, Filice, Palomo and co-workers prepared a new immobilized artificial metalloenzyme (ArM hereafter) for HCR (Scheme 50) . A phosphonate-substituted Pd-pincer complex was covalently linked to the catalytically competent serine residue of commercially available enzymes including lipase from Candida antarctica B (CAL-B). The metalloproteins were then immobilized on different supports including: Novozyme 435, Sepharose and Sepabeads (SP). All but one immobilized ArM were inactive: the SP-CHO-CAL-B-Pd hybrid proved active. The free Pd pincer complex catalyzed the HCR of iodobenzene with ethyl acrylate at 70 °C in a water:DMF mixture (1:3). It required higher temperatures in pure DMF (120 °C). The SP-CHO-CAL-B-Pd hybrid was extremely active at 70 °C even in pure DMF affording quantitative yields and TOF of up to 116 h−1. Modification of the Sephabead surface with long alkyl chains lead to an increase in activity (up to TOFs of 153 h−1). Introduction of an ethylenediamine moiety nearly completely suppressed catalytic activity (< 5% yield), thus highlighting the critical role of the support on catalysis. Thanks to the presence of an enantiopure host protein, asymmetric HCR between iodobenzene (or bromobenzene) with 2,3-dihydrofuran using optimized SP-CAL-B-C8-Pd afforded highly enantioenriched product (ee = 96.6%) event at elevated temperature (95% yield at 120 °C) in a water:DMF mixture (1:3) (Scheme 51). These findings highlight the potential of this yet underexplored field of immobilized ArMs.
4.6 Active Carbon-Based Support
Salunkhe and co-workers were the first to report on the use of palladium immobilized on activated carbon (Pd/C) in a hydrotropic solution containing 60% sodium xylene sulphonate (NaXS) for SMC and HCR. Various arene-diazonium salts were coupled with methyl acrylate, butyl acrylate or styrene with good to very good yields at room temperature. The immobilized catalyst was easily recovered and recycled with very minor loss in activity for up to three cycles. Although arene diazonium salts are not readily available and NaXS is costly, this method remains one of the most attractive protocols for aqueous Heck–Matsuda cross-coupling .
Developing reaction protocols capable of coupling aryl chlorides at room temperature and with a broad substrate scope remains a challenge. Compared to aryl iodides or bromides, the corresponding aryl chlorides most often require significantly harsher reaction conditions including: higher reaction temperatures, higher catalyst loading and prolonged reaction times. It remains challenging to reliably predict which catalyst and what conditions should be selected as a starting point in an optimization effort. Results are often hard to rationalize, and comparisons between different catalytic systems are scarce in the literature. While palladium catalyzed cross-coupling chemistry is well understood in organic solvents, the corresponding aqueous chemistry is significantly less investigated. The formation of PdNP is a serious issue, which further complicates the identification of the catalytically active species. As this review intends to not only summarize the steadily growing field of aqueous HCR but also to point out its untapped potential especially for biological applications, a table summarizing the most promising catalytic systems for HCR is presented below. Ten unique procedures are highlighted which show remarkable activity in combination with broad substrate scope or novel bio-orthogonal applications of HCR (Table 1).
Ten versatile and efficient or bio-orthogonal methods for aqueous HCR
18 products from the coupling of aryl iodides and aryl bromides with methyl acrylate or ethyl acrylate
0.5 eq of ligand, 3 mol% of Pd(OAc)2, neat water, RT, 3–4 h
This mild and simple method can be applied to a broad substrate scope giving consistently very good yields (≥ 90%) 
49 products from the coupling of aryl iodides with butyl acrylate, ethyl acrylate or styrene, 3 products from the coupling of aryl bromides with butyl acrylate
1-0.0001 mol% of catalyst with 2 eq of Bu3N, 0.5 eq of TBAB, neat water, 140 °C, 24 h
This highly active catalyst achieved TON’s of up to 920,000 for the coupling of various aryl iodides and is reusable for over five runs. Hower, it is only moderately active for the coupling of aryl bromides 
27 products from the coupling of aryl iodides, bromides and benzyl chlorides with various olefins
1 mol% of catalyst with 1 eq of TBAB, 0.1 eq of NaOAc, neat water at 80–140 °C for 10–50 min
Versatile microwave-assisted protocol presented giving good yields even challenging coupling partners such as benzyl chloride, cyclohexene or 1-octene within minutes 
17 products from the coupling of aryl iodides, bromides, chlorides with various olefins
1 mol% of catalyst with 2 eq K2CO3, neat water, 80 °C, 4–6 h
Imidazolium ionic liquid-tagged Pd Schiff base complex used as an efficient catalyst for HCR and Suzuki reaction. Good yields (≥ 76%) even for aryl chlorides were obtained on a moderate substrate scope 
33 products from the coupling of aryl chlorides and aryl dichlorides to various olefins
1 mol% of catalyst with 2 eq K3PO4, 0.5 eq TBAB in DMF:H2O (1:4), at 85–120 °C for 4–24 h
This is the broadest study regarding the coupling of aryl chlorides via HCR at aqueous conditions. Consistently good yields (≥ 73%) were obtained and selective one-pot double HCR was achieved 
16 products from the coupling of aryl iodides, bromides and chlorides with either butyl acrylate or styrene
0.05 mol% of catalyst with 1.2 eq K2CO3, neat water, RT, 170 W of ultrasonification
PdNP’s incorporated into PEDOT was shown to be a highly efficient and reusable catalyst for the HCR of even challenging substrates such as aryl chlorides at RT in neat water 
Performed in vivo in cancer cells or in vitro in PBS buffer at 37 °C
For the first time, HCR was used in vivo to target cancer cells via intracellular 6-(dimethylamino)-coumarin fluorescence. The corresponding catalyst was further used for selective drug-uncaging in mammals 
2 products from the coupling of iodo benzene to ethyl acrylate or 2,3-dihydrofuran
0.024 mol% catalyst with 1.7 eq of NEt3, in DMF:H2O (3:1) at 70–120 °C for 24–31 h
An achiral metal complex anchored to lipases immobilized on modified Sepabeads. This ArM achieved very high yield (95%) and ee’s (96.6) for the asymmetric HCR of iodo benzene with 2,3-dihydrofuran 
25 product from the coupling of aryl boronic acids with various electon-rich and electron-deficient olefins
1 mol% of catalyst with 1.2 eq of 1,4-benzoquinone, in H2O:TFA (2:1) at RT for 12 h
A selective and highly active aNHC-Pd-dimer used for the oxidative HCR under mild conditions and with a broad substrate scope 
19 products from the coupling of aryl boronic acids to various olefins
Furthermore, 7 products from the coupling of aryl boronic acids with modified mutants of 4-OTR61C
5 mol% of Pd(OAc)2 with 7 mol% of ligand, O2, in MeOH:H2O (9:1) at RT for 30 h
Modified procedure for protein labeling
At the turn of the millennium, Beletskaya and Cheprakov pointed out that Heck chemistry remained a multidimensional, unanswered, complex but fascinating puzzle . Since then, incremental improvements in Heck chemistry and particularly in aqueous HCR have been reported. Catalysts were developed capable of achieving high yields at room temperature, even in the presence of aryl chlorides as substrates. More recently, HCR in vivo was reported, which opens interesting avenues for chemical biology applications.
TRW thanks the Swiss National Science Foundation “Grant 200020 162348” for financial support.
Compliance with Ethical Standards
Conflict of interest
The authors declare no conflict of interest.
- 6.Heck RF (2005) Org React 27(2):345–390Google Scholar
- 7.Bonrath W, Létinois U, Netscher T, Schütz J (2009) Mizoroki-Heck reactions: modern solvent systems and reaction techniques. In: Oestreich M (ed) The Mizoroki-Heck reaction. Wiley, ChichesterGoogle Scholar
- 18.Bumagin NA, Andryuchova NP, Beletskaya IP (1988) Izv Akad Nauk SSSR 6:1449Google Scholar
- 40.Reetz MT, De Vries JG (2004) Chem Commun 1559Google Scholar
- 48.Kuntz EG (1987) CHEMTECH 17:570Google Scholar
- 57.Lipshutz BH, Ghorai S (2008) Aldrichimica Acta 41:59Google Scholar
- 70.Allam BK, Singh KN (2011) Synthesis 7:1125Google Scholar
- 78.Wang Z, Feng X, Fang W, Tu T (2011) Synlett 7:951Google Scholar
- 109.Davies IW, Matty L, Hughes DL, Reider PJ (2001) 123:10139Google Scholar
- 113.Anson MS, Leese MP, Tonks L, Williams JMJ (1998) J Chem Soc Dalton Trans 3529Google Scholar
- 117.Uozumi Y, Danjo H, Hayashi T (1999) J Org Chem 64:3884Google Scholar
- 130.Villiers A (1891) C R Hebd Seances Acad Sci 112:536Google Scholar