Lukas J. Gooßen

Carboxylic Acids as Substrates in Homogeneous Catalysis

In organic molecules carboxylic acid groups are among the most common functionalities. Activated derivatives of carboxylic acids have long served as versatile connection points in derivatizations and in the construction of carbon frameworks. In more recent years numerous catalytic transformations have been discovered which have made it possible for carboxylic acids to be used as building blocks without the need for additional activation steps. A large number of different product classes have become accessible from this single functionality along multifaceted reaction pathways. The frontispiece illustrates an important reason for this: In the catalytic cycles carbon monoxide gas can be released from acyl metal complexes, and gaseous carbon dioxide from carboxylate complexes, with different organometallic species being formed in each case. Thus, carboxylic acids can be used as synthetic equivalents of acyl, aryl, or alkyl halides, as well as organometallic reagents. This review provides an overview of interesting catalytic transformations of carboxylic acids and a number of derivatives accessible from them in situ. It serves to provide an invitation to complement, refine, and use these new methods in organic synthesis.

Carboxylates as sources of carbon nucleophiles and electrophiles: comparison of decarboxylative and decarbonylative pathways

This tutorial review provides a comparison between the concepts of catalytic decarboxylative and decarbonylative couplings for the ipso-substitution of carboxylate groups, and illustrates their potential benefits over alternative C–C bond-forming reactions. Redox-neutral decarboxylative reactions allow generating organometallic species with nucleophilic reactivity via the extrusion of carbon dioxide from metal carboxylates. Such C–C bond activating processes provide a way of employing carboxylate salts as substitutes for the traditional sources of carbon nucleophiles, i.e. stoichiometric organometallic reagents. If the decarboxylation of carboxylic acids is performed under oxidative conditions, organometallic species with electrophilic reactivity are obtained instead. These can alternatively be accessed via the extrusion of carbon monoxide from acyl–metal species generated via the oxidative addition of activated carboxylic acid derivatives (e.g. acid chlorides, anhydrides or esters) to metal complexes. In the latter two reaction types, carboxylic acids thus become substitutes for organohalides. The complementary redox-neutral and oxidative decarboxylative and decarbonylative reaction modes allow the broad use of carboxylic acids as substrates in C–C bond-forming reactions. Their applicability, scope and limitations are discussed using the examples of Heck reactions, cross-couplings and direct arylations.

Decarboxylative Cross‐Coupling of Aryl Tosylates with Aromatic Carboxylate Salts

Metal-catalyzed coupling reactions are effective synthetic tools for the formation of C C bonds between nucleophilic and electrophilic substrates at positions predefined by leaving groups. Recently, decarboxylative coupling reactions have emerged as powerful alternatives for regioselective C C bond formation, thus providing new protocols for Heck-type reactions, oxidative arylations, redox-neutral cross-coupling reactions, and allylations. The redox-neutral decarboxylative coupling reactions that have been developed in our research group aim at replacing sensitive and costly organometallic reagents, which are traditionally used as nucleophilic coupling partners, with stable, inexpensive and widely available carboxylate salts. 7] In this type of reaction, a copper(I) or silver(I) catalyst mediates the extrusion of CO2 from the carboxylates while a palladium complex catalyzes the coupling of the resulting carbon nucleophiles with carbon electrophiles (Scheme 1). In view of the high performance level of traditional crosscouplings, a widespread practical application of decarboxylative couplings hinges on a broad substrate scope, the use of inexpensive and readily available carbon electrophiles, and mild reaction conditions. The early protocols allowed the coupling of diversely functionalized ortho-substituted benzoates, heterocyclic arenecarboxylates, and a-oxocarboxylates with a wide variety of aryl and heteroaryl halides. 8,9] However, the strongly coordinating halide ions that were formed in the process were found to impede the decarboxylation of other arenecarboxylates. A decisive extension of the substrate scope covering the full range of substitution patterns including metaand para-substituted arenecarboxylates was achieved by new catalysts that allowed the use of aryl triflates as carbon electrophiles. The coupling of these substrates releases only noncoordinating anions that do not hinder the decarboxylation at the copper center. Unfortunately, the practical utility of this protocol is limited by the expense and the sensitivity of aryl triflates. The use of the inexpensive and more robust aryl ptoluenesulfonates (tosylates) is of profound interest for all types of cross-coupling reactions, and substantial effort has been devoted to the development of catalyst systems capable of activating them. In earlier protocols, nickel complexes were mostly used as catalysts, until a new class of bulky, electron-rich phosphines was discovered that strongly facilitates the oxidative addition of aryl tosylates to palladium catalysts. In recent years, aryl tosylates have successfully been employed as substrates in, for example, Stille, Suzuki, and Kumada coupling reactions, aminations, and ortho-arylations. The use of aryl tosylates as substrates in decarboxylative coupling reactions should have an even higher synthetic impact, when considering that a low coordinating ability of the leaving group is an essential prerequisite for accessing the full range of carboxylic acid substrates. We started the development of the catalyst for the desired decarboxylative cross-coupling (see Scheme in Table 1) with a series of protodecarboxylation reactions in the presence of phosphine ligands to identify phosphines that would not interfere with the decarboxylation step. Fortunately, the conversion of 3-nitrobenzoic acid into nitrobenzene using Cu2O/1,10-phenanthroline (phen) catalysts was not affected by the electron-rich, sterically demanding phosphines that are known to activate unreactive leaving groups (Scheme 2). We next investigated the performance of palladium complexes with such ligands as catalysts in the decarboxylative coupling of potassium 2-nitrobenzoate (1a) with 4-tolyl tosylate (2a) in combination with a Cu2O/phen co-catalyst (Table 1). Scheme 1. Cu/Pd-catalyzed decarboxylative cross-coupling. M = Ag, Cu; R = (hetero)aryl, vinyl, acyl; R’= (hetero)aryl; X = I, Br, Cl, OTf. Tf= trifluoromethanesulfonyl.

Silver-catalysed protodecarboxylation of carboxylic acids.

A silver-based catalyst system has been discovered that effectively promotes the protodecarboxylation of various carboxylic acids at temperatures of 80-120 degrees C--more than 50 degrees C below those of the best known copper catalysts.

Sandmeyer Trifluoromethylation of Arenediazonium Tetrafluoroborates

The development of methods for the introduction of trifluoromethyl groups into functionalized molecules is of great importance due to their presence in many top-selling pharmaceuticals, agrochemicals, and functional materials. Trifluoromethyl groups are known to impart desirable properties, such as higher metabolic stability, increased lipophilicity, and stronger dipole moments to druglike molecules. Celecoxib, dutasteride, fluoxetine, and sitagliptin are some examples of top-selling pharmaceuticals featuring trifluoromethyl groups, and beflubutamid, diflufenican, and norfluazon examples of agrochemicals. However, traditional methods to access benzotrifluorides, for example, the Swarts reaction, typically require harsh conditions and have a low substrate scope, so that they are confined to the beginning of a synthetic sequence (Scheme 1a). Building on pioneering work on Cu– and Pd–perfluoroalkyl complexes by McLoughlin, Yagupolskii, Burton, Chambers, Grushin, and others, substantial progress has recently been made in the development of trifluoromethylation reactions that allow the selective introduction of CF3 groups into functionalized, late-stage synthetic intermediates. A wealth of new reactions has been disclosed, which can be roughly divided into five categories (Scheme 1b–f). The first are couplings of aryl halides with nucleophilic CF3 reagents (reaction type b), usually copper–CF3 complexes in stoichiometric amounts. These complexes may also be generated in situ from copper salts and Ruppert s reagent (CF3SiMe3), [7] fluoroform, potassium (trifluoromethyl)trimethoxyborate, trifluoroacetate salts, methyl trifluoroacetate, or fluorosulfonyldifluoroacetic acid. Grushin, Sanford, and Buchwald also disclosed trifluoromethylations of aryl halides based on palladium complexes. Palladium complexes also promote C H functionalizations of arenes with trifluoromethylating reagents (reaction type c). Examples are the ortho-trifluoromethylation of donor-substituted arenes with Umemoto s reagent described by Yu et al. and the Pd-catalyzed coupling of arenes with perfluoroalkyl iodides reported by Sanford et al. C H trifluoromethylations of heteroarenes have recently been reported also with nucleophilic trifluoromethylation reagents under oxidative conditions. Examples of couplings of aryl nucleophiles with electrophilic CF3 sources (reaction type d) include the coupling of arylboronic acids with Togni s and Umemoto s reagent disclosed by Shen and Liu, respectively. Sanford et al. employed a copper/ruthenium photocatalyst system to promote a radical trifluoromethylation of boronic acids. The copper-catalyzed syntheses of benzotrifluorides from boronic acids and CF3SiMe3 or K [CF3B(OMe)3] developed by Qing et al. and ourselves exemplify oxidative couplings of aryl nucleophiles with nucleophilic CF3 reagents (reaction type e). The radical trifluoromethylation of arenes (reaction type f) was pioneered by Langlois. Baran and MacMillan recently reported modern variants of this reaction concept based on peroxide or ruthenium initiators. From a practical standpoint, nucleophilic reagents are appealing for the introduction of trifluoromethyl groups for the following reasons. CF3SiMe3 and K [CF3B(OMe)3] are available in large quantities for a reasonable price, and are easy to store and handle. They are accessible not only from halofluorocarbons, but also from fluoroform, a by-product in the production of Teflon. One of the most widely used methods for the introduction of halides and related nucleophiles is the Sandmeyer reaction. Aromatic amines, which are available in great structural diversity, are diazotized using, for example, NaNO2 or organic nitrites. Upon treatment with the appropriate copper(I) halides, nitrogen gas is released, and a halide group is installed regiospecifically in the position Scheme 1. Strategies for the introduction of trifluoromethyl groups.

Low-temperature ag/pd-catalyzed decarboxylative cross-coupling of aryl triflates with aromatic carboxylate salts.

Metal-catalyzed decarboxylative couplings are evolving into powerful synthetic tools for the regioselective formation of C C bonds. New protocols for Heck-type reactions, oxidative arylations, redox-neutral couplings, and allylations have provided innovative atom-economic and wasteminimized pathways among others to biaryls, vinyl arenes, and aryl ketones starting from readily available carboxylic acids. These transformations have reached impressive performance levels in terms of selectivity, functional group tolerance, and yield. However, their practical applicability is still somewhat limited by the high reaction temperatures currently required in the decarboxylation step. The redox-neutral decarboxylative cross-couplings developed in our group allow a regioselective C C bond formation between aryl, heteroaryl, or acyl carboxylates and aryl halides to give biaryls or aryl ketones without resorting to stoichiometric amounts of organometallic reagents. Instead, the carbon nucleophiles are generated in situ via extrusion of CO2 at a copperor silver-based decarboxylation catalyst. They are then transmetalated to the palladium catalyst, where their coupling with the aryl electrophile takes place (Scheme 1). The decarboxylation cocatalyst is vital for the conversion of most carboxylic acids. Only a few particularly reactive derivatives such as certain heteroarene2-carboxylic acids or monoalkyl oxalates can be coupled by palladium alone, presumably by a different mechanism. So far, silver-based systems have not presented advantages over copper ones in terms of reaction temperature or scope of decarboxylative cross-couplings. In contrast to copper(I), silver(I) had to be employed in overstoichiometric amounts, because a salt metathesis between the potassium carboxylates 1 and silver halides a formed within the catalytic cycle is impossible, precluding further turnover of the silver catalyst. 10] However, when reevaluating the potential of silver catalysts for protodecarboxylation reactions, we discovered conditions under which silver salts mediate the extrusion of CO2 from certain arenecarboxylates with higher efficiency than copper complexes. The new silver-based protodecarboxylation proceeds at only 120 8C—a temperature more than 50 8C below that of the best known copper catalysts. Larrosa et al. independently discovered a similar protocol. In addition, we were able to show that decarboxylative cross-couplings can be performed using aryl electrophiles with non-coordinating leaving groups such as triflates or tosylates. We reasoned that the low affinity of these ions to silver(I) might enable the crucial salt metathesis between silver sulfonate salts a and potassium carboxylates 1 (Scheme 1). This prompted us to embark on the search for a new, low-temperature protocol for the decarboxylative cross-coupling of aryl sulfonates with potassium carboxylates using a Ag/Pd catalyst system. We based the catalyst development on the model reaction of potassium 2-nitrobenzoate (1 a) with 4-chlorophenyl triflate (2 a) (see Table 1). Using a catalyst analogous to the copper-based version but employing Ag2CO3 (5 mol %) instead of Cu2O/1,10-phenanthroline, modest turnover was ob[a] Prof. Dr. L. J. Gooßen, P. P. Lange, Dr. N. Rodr guez, C. Linder FB Chemie Organische Chemie, TU Kaiserslautern Erwin-Schrçdinger-Strasse Geb. 54, 67663 Kaiserslautern (Germany) Fax: (+49) 631-205-3921 E-mail : goossen@chemie.uni-kl.de Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.200903319. Scheme 1. Ag/Pd-catalyzed decarboxylative cross-coupling. R = (hetero)aryl; R’= (hetero)aryl; X =OTf, OTs.

New catalytic transformations of carboxylic acids

A series of metal-catalyzed processes are presented, in which carboxylic acids act as sources of either carbon nucleophiles or electrophiles, depending on the catalyst employed, the mode of activation, and the reaction conditions. A first reaction mode is the addition of carboxylic acids or amides over C-C multiple bonds, giving rise to enol esters or enamides, respectively. The challenge here is to control both the regio- and stereoselectivity of these reactions by the choice of the catalyst system. Alternatively, carboxylic acids can efficiently be decarboxylated using new Cu catalysts to give aryl-metal intermediates. Under protic conditions, these carbon nucleophiles give the corresponding arenes. If carboxylate salts are employed instead of the free acids, the aryl-metal species resulting from the catalytic decarboxylation can be utilized for the synthesis of biaryls in a novel cross-coupling reaction with aryl halides, thus replacing stoichiometric organometallic reagents. An activation with coupling reagents or simple conversion to esters allows the oxidative addition of carboxylic acids to transition-metal catalysts under formation of acyl-metal species, which can either be reduced to aldehydes, or coupled with nucleophiles. At elevated temperatures, such acyl-metal species decarbonylate, so that carboxylic acids become synthetic equivalents for aryl or alkyl halides, e.g., in Heck reactions.

Comparative Study of Copper‐ and Silver‐Catalyzed Protodecarboxylations of Carboxylic Acids

The protodecarboxylation of aromatic carboxylic acids by various copper and silver catalysts is investigated with the help of density functional calculations and experimental studies. The computational results reveal that the catalytic activity of copper(I)–1,10‐phenanthroline catalysts increases with the introduction of electron‐rich substituents at the phenanthroline ligand. They also predicted that for some substrates, silver complexes should possess a substantially higher decarboxylating activity than copper, which is confirmed by experimental studies, leading to the discovery of a silver(I) catalyst that effectively promotes the protodecarboxylation of various carboxylic acids at temperatures in the range of 80–120 °C—more than 50 °C below those of the best known copper(I) catalyst. The scope of the new system complements that of the copper(I)‐based method as it includes benzoates for example, with halogen or ether groups in the ortho positions.

Synthesis of Aryl Ethers from Benzoates through Carboxylate‐Directed CH‐Activating Alkoxylation with Concomitant Protodecarboxylation

The aryl ether functionality is a common motif in many classes of biologically active compounds and functional materials. Traditional methods to attach ether groups to aromatic rings require either activated substrates (SNAr) or harsh reaction conditions (Ullmann-type condensation), and are therefore not ideal for the late-stage functionalization of complex molecules. In the last two decades, the need for reliable and atomeconomic methods to access aryl ethers started a rapid development of new methodologies for C O bond formation. Important advances include copperand palladium-catalyzed couplings of aryl halides with alcohols or phenols, and copper-mediated oxidative coupling of aryl borates with phenols. These methods depend on the availability of starting materials with the desired substitution pattern, a limitation that may be overcome with C H activating alkoxylation processes. Several protocols for palladiumcatalyzed regioselective alkoxylations of arenes were recently reported with the use of pyridyl, N-methoxyimine, Nmethoxybenzamide, cyano, and anilide directing groups. Copper-based systems have been used for similar transformations, including oxygen atom transfer or catalytic acetoxylation reactions, but to the best of our knowledge, no copper-catalyzed C H activating alkoxylations of arenes have been reported to date. However, the possibility of directed copper-mediated aryl ether formation is supported by two experimental findings: Yu et al. observed the coupling of 2-phenylpyridine with 4-cyanophenol, and Ribas and Stahl found that methoxylation of a benzene ring within a macrocyclic ligand scaffold takes place in the presence of copper. In our studies of decarboxylative coupling reactions, we recently disclosed a decarboxylative etherification in which a C O bond is formed at the original position of the carboxylate group of benzoates. For a broad range of substrate combinations, the decarboxylative alkoxylation was found to proceed regiospecifically at the ipso position. Only for one substrate, we observed the formation of a by-product in which the C O bond had formed in the ortho position of the extruded carboxylic group. This by-product, although obtained in only modest yield and using stoichiometric amounts of silver and copper, pointed to the principal feasibility of a second, complementary pathway for alkoxylations with decarboxylation, namely the carboxylatedirected ortho alkoxylation of benzoates through C H activation, 19] followed by protodecarboxylation. If further developed, such a reaction concept could give access to aryl ethers with a complementary product range to decarboxylative ipso alkoxylations: meta-substituted aryl ethers would arise from paraor ortho-substituted benzoates, and para-substituted aryl ethers from meta-substituted benzoates (Scheme 1).

Palladium‐Catalyzed Synthesis of Aryl Ketones from Boronic Acids and Carboxylic Acids Activated in situ by Pivalic Anhydride

A new palladium-catalyzed cross-coupling reaction between arylboronic acids and mixed anhydrides, generated in situ from carboxylic acids and pivalic anhydride, is presented. Optimization of the new catalyst and the reaction conditions led to the development of a convenient one-pot ketone synthesis directly from carboxylic and boronic acids in the presence of different (phosphane)palladium complexes in wet THF at 60 °C. Systematic studies were performed to elucidate the reaction mechanism of this transformation. The scope and the limitations of the new process are demonstrated by the synthesis of 33 functionalized ketones. (© Wiley-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002)