Pd(0)-Catalyzed Carbon-Carbon Bond Formation

Pd(0)-CATALYZED CARBON–CARBON BOND FORMATION

The most important catalytic reactions for the formation of carbon–carbon bonds involve the chemistry of Pd(0). Complexes of zero-valent palladium such as Pd(PPh₃)₄ are available commercially or can be prepared in situ by the reaction of Pd(II) salts [e.g., Pd(OAc)₂, PdCl₂, etc.] with phosphines or other reductants. The most stable complexes are those in which the sum of d electrons from the metal and electrons donated by ligands totals 18. Palladium(0) has a d¹⁰ electron configuration and thus normally coordinates with four ligands which each donate a pair of electrons to the metal. Such complexes are said to be coordinatively saturated and tend to be stable and relatively unreactive. However, dissociation of one or two ligands in solution produces a 16- or 14-electron complex which is reactive and seeks to regain the 18-electron configuration.

Palladium(0)-catalyzed transformations generally involve three steps: oxida-tive addition, insertion or transmetallation (really a special type of insertion), and reductive elimination. Together they comprise a pathway for the formation of new carbon–carbon bonds. Oxidative addition takes place when a coordinatively unsaturated Pd(0) species cleaves a covalent bond to give a new complex in which the palladium is oxidized to Pd(II). Typically dissociation of two phos-phine ligands to a 14-electron complex is the first step followed by oxidative addition to give a 16-electron Pd(II) complex.

Oxidative addition

This is quite analogous to the formation of a Grignard reagent by the oxidative addition of Mg(0) to an alkyl halide. What is remarkable is the generality and functional tolerance of the palladium process. A variety of bonds undergo oxida-tive addition with Pd(0). Bonds from carbon to halogen and other good leaving groups such as sulfonates, esters, and phosphonates are used most often (and often referred to as C–X or R–X bonds), but many other bond types are known to react. Even though the product of oxidative insertion has a carbon–palladium bond, this bond is unaffected by most functional groups. Thus alcohols, amines, amides, esters, ketones, aldehydes, and even carboxylic acids can be present in the substrate without interfering with the addition reaction or subsequent reactions. This is a truly phenomenal tolerance for functionality!

Another interesting facet of Pd(0) oxidative insertion is the chemoselectivity of the process. The most reactive bonds are vinyl and aryl C–X bonds, whereas with most other metals these are the least reactive bond types. Palladium(0) also inserts into allylic halides and esters, acid halides, and several other bonds but reacts only sluggishly with C–X bonds to saturated carbon. Taken together these characteristics make Pd(0) chemistry nearly unique.

The second step is insertion or transmetallation. An insertion reaction occurs when the palladium–carbon bond adds across a π bond to give a new organopal-ladium species. The types of π bonds normally reactive include alkenes, dienes, alkynes, carbon monoxide, and sometimes carbonyl π bonds. By far the most common reactions use alkenes and alkynes for the insertion reaction. This step results in a new carbon–carbon bond.

Insertion

R-Pd-X(PPh₃)₂ + A=B  → R-A-B-Pd-X(PPh₃)₂

for example,

R-Pd-X(PPh₃)₂ + CH₂ = CH₂ → R-CH₂-CH₂-Pd-X(PPH₃)₂

The regiochemistry of the insertion results from a combination of factors which are still being sorted out. It is possible to think of the carbon attached to palladium as electron rich, and it tends to attack the π system at the least electron rich position. Thus alkenes with electron-withdrawing groups react faster than alkenes with electron-donating groups. It is quite paradoxical, however, that alkenes, dienes, and alkynes react much more readily than carbonyl compounds, even though the latter are much more electron deficient.

Moreover there appears to be a steric bias which causes the R group to attack the least hindered end of the π system. In cases where the two ends of the π bond are similar or where electron-donating groups are attached, the regiochemistry can be very sensitive to the reaction conditions and the ligands that coordinate to palladium. At this point controlling the regiochemistry in such systems is more art than science! Nevertheless, in most cases it is possible to predict the regiochemistry with good success. Finally the stereochemistry of the insertion is syn; thus the insertion appears to be a concerted 1,2 addition across one face of the π system.

Transmetallation occurs when compounds with bonds from carbon to several main-group elements (e.g., B, Al, Sn, Si, Hg) are present in the reaction mixture. The palladium intermediate from oxidative addition can undergo exchange of palladium and the main-group element. This essentially yields a second carbon ligand bonded to palladium. The most common compounds used for transmet-allation are tributyl tin compounds (R–SnBu₃) and boronic acids [R–B(OH)₂]. Again the most common and successful examples have the main-group element bonded to an aromatic ring or an alkene.

Transmetallation

for example,

The last step is reductive elimination in which the organic product is liberated and Pd(0) is regenerated to begin the catalytic cycle again. When there are two carbon ligands attached to palladium, as is the case when a transmetallation has occurred, the two carbon fragments couple with the expulsion of Pd(0). This occurs rapidly after a transmetallation and in these instances is the step in which carbon–carbon bond formation occurs.

Reductive elimination

R-Pd-R′(PPh₃)₂ → R-R′ + Pd(PPh₃)₂

for example,

A second common reductive elimination process termed β-hydride elimination occurs when there is a hydrogen atom β to the carbon–palladium bond, as in the case where an insertion reaction has taken place. The palladium atom inserts into the β carbon–hydrogen bond to give a palladium hydride species coordinated to the alkene. This is a reversible reaction and is akin to the process of alkene hydrogenation catalyzed by palladium. Dissociation of the alkene and elimination of HX gives back the Pd(0) catalyst. Since a strong acid is liberated in the β elimination, a base such as triethylamine is usually added to the reaction mixture to scavenge this acid. Although the formation of alkenes by β-hydride elimination is a facile process, it is not possible to form an alkyne or allene by β-hydride elimination from a vinyl palladium species.

 β-Hydride elimination

While the above reactions represent only a small fraction of the reactions known for palladium, they form the basis of a powerful methodology for building carbon structures. Several variations have been developed which utilize certain types of reactants and give particular types of products. All these variations, however, contain a common theme. In each case an electron-deficient reagent (e.g., a vinyl halide or aromatic triflate) reacts with an electron-rich reagent (e.g., an alkene, an organoborane, or an organotin) with the formation of a new carbon–carbon bond. In that sense these reactions are related to the reactions between carbon nucleophiles and carbon electrophiles discussed previously in this chapter. They are quite different, however, because they proceed only in the presence of Pd(0). In fact they proceed only in the coordination sphere of Pd(0). The ability of Pd(0) to catalyze these reactions is nearly unique! We will now examine some of the more common processes.

Heck Reaction

The Heck reaction involves the coupling of an organopalladium species formed by oxidative addition to an alkene followed by β-hydride elimination. The product is an alkene in which a vinyl hydrogen on the original alkene is replaced by the organic group on palladium. Thus aryl and alkenyl halides can be coupled to alkenes.

Because the by-product of the coupling is a strong acid, bases are usually added to the reaction mixture to scavenge it. For example, 4-iodobromobenzene can be coupled with methyl acrylate to give the 4-bromocinnamate ester in >68% yield. This reaction takes advantage of the faster oxidative addition to the carbon–iodine bond to give a single product.

The Heck reaction was discovered in the early 1970s and is extremely useful for rapidly assembling carbon skeletons. This reaction is unique to palladium! A great deal of information is known about the reaction. For example, the success of the reaction depends on each of the three steps involved. Electron-donating groups decrease the reactivity of alkenyl halides and triflates toward Pd(0), whereas electron-withdrawing group increase the rate of oxidative addition. In cases where Pd(II) salts are used, it is assumed that they are converted to Pd(0) by some redox process.

The insertion reaction is stereospecific and syn. Moreover the β-hydride elimi-nation is also syn. For acyclic alkenes there is free rotation in the organopalladium intermediate so that the more stable trans-alkene is formed. Electron-withdrawing groups in the alkene also increase the rate of the insertion reaction and give higher yields generally, but the reaction is limited to relatively sterically unhin-dered alkenes. In general, polar solvents such as DMF or acetonitrile are most commonly used. There are several common additives which aid in the reaction. These include lithium or tetraalkylammonium chlorides and bromide, silver salts, or cuprous iodide, but exactly how they function is unknown at present.

The conversion of carbonyl compounds to their enol triflates provides a very simple way to couple the carbonyl carbon to an alkene. In general, however, aryl and vinyl iodides are the preferred substrates because of their ease of oxidative addition. Terminal alkynes are also good coupling partners.

Intramolecular versions of the Heck reaction are very useful for the construc-tion of ring systems. The entropic advantage of having both coupling partners present in the same molecule increases the efficiency of the insertion reaction and leads to efficient reactions. Moreover the intramolecular version can be carried out on hindered substituted alkenes, whereas the intermolecular Heck reaction is largely restricted to monosubstituted alkenes. These reactions illustrate the syn stereochemistry of both the insertion reaction and the elimination. A number of multicyclic natural products have been synthesized using intramolecular Heck reactions to assemble the skeletons, and this has become a powerful synthetic tool for such compounds.

Suzuki Coupling

The coupling of organoboron compounds with aryl or alkenyl halides is called the Suzuki reaction and was discovered in the early 1980s. This is a tremendously versatile method for joining two carbon fragments and is widely used in the com-mercial manufacture of pharmaceuticals, in the synthesis of compound libraries, and in drug discovery. After oxidative addition to the halide, the organopalladium intermediate undergoes transmetallation with the boronic acid or ester. The new carbon–carbon bond is formed in the reductive elimination which produces the product and regenerates the Pd(0) catalyst. A base must be present for the trans-metallation to proceed, and oxybases such as alkoxides, carbonates, or hydroxide are most commonly employed. The reaction is highly tolerant of a wide variety of functional groups and thus extremely versatile.

As noted for the Heck reaction, aryl, alkenyl, and alkynyl bromides, iodides, and triflates are best for the oxidative addition. However, aromatic, heteroaro-matic, alkenyl, and even alkyl boronic acids and esters can be coupled effectively. The reaction appears almost oblivious to other functional groups present!

Since the oxidative addition occurs with retention of configuration and the transmetallation is also stereospecific with retention, the method is extremely valuable for the stereoselective synthesis of conjugated dienes. The stereochemistry of the products is determined by the stereochemistry of the coupling precursors.

The required vinyl boranes and vinyl iodides can both be easily made by the hydroboration of alkynes with disiamyl borane (Sia). Thus the Suzuki reac-tion is an important methodology for the synthesis of conjugated polyene natu-ral products.

Stille Coupling

Stille coupling was also developed in the early 1980s and is similar to Suzuki coupling in its sequence. It is used to couple aryl or vinyl halides or triflates with organotin compounds via oxidative addition, transmetallation, and reduc-tive elimination. The oxidative addition reaction has the same requirements and preferences as discussed earlier for the Heck and Suzuki reactions. The reduc-tive elimination results in formation of the new carbon–carbon bond. The main difference is that the transmetallation reaction uses an organotin compound and occurs readily without the need for an oxygen base. Aryl, alkenyl, and alkyl stannanes are readily available. Usually only one of the groups on tin enters into the coupling reaction, and different groups transfer to palladium with different selectivities. Since simple alkyl groups have the lowest transfer rate, the most common tin reagents have three simple alkyl groups (usually methyl or n-butyl). The fourth group which is transferred is alkynyl, aryl, alkenyl, benzyl, or allyl.

Coupling of an aryl triflate with an arylstannane is a good method for the preparation of biaryls and other bis-aromatic species of all types. Coupling of vinyl groups takes place with retention of stereochemistry. Furthermore transfer of the allyl group occurs smoothly.

This is very robust chemistry that works very well with enol triflates. Intramolecular reactions have been used to close rings of many sizes, including large rings.

The use of (Me₃Sn)₂ provides a unique way to convert vinyl and aryl halides into the very tin reagents needed for subsequent Stille couplings!