1
Introduction: Carbon Monoxide as Synthon in Organic Synthesis

Bartolo Gabriele

University of Calabria, Laboratory of Industrial and Synthetic Organic Chemistry (LISOC), Department of Chemistry and Chemical Technologies, Via Pietro Bucci 12/C, 87036 Arcavacata di Rende, Italy

This book discusses the synthesis of carbonylated compounds by introducing the carbonyl function into an organic substrate (carbonylation) employing the simplest C-1 unit as carbonylating agent, carbon monoxide. Carbon monoxide is a largely available feedstock. It is produced industrially by partial oxidation of petroleum hydrocarbons and steam reforming of light hydrocarbons (including natural gas) or gasification of coal to give syngas (CO and H2) [1]. In the future, it is expected that a growing amount of carbon monoxide will be available from renewable feedstocks, such as biowastes and CO2 [2]. CO is also the simplest unit that, upon insertion into an organic substrate, can be directly transformed, without atom loss, into a carbonyl group. It is therefore a desirable and useful building block in synthesis to produce high value-added industrially relevant molecules and fine chemicals.

Carbonylation reactions were disclosed in the 1930s by the seminal works of Roelen [3] and Reppe [4] (who also coined the term "carbonylation") for industrial applications. Since then, the scientific progress in this field has been enormous, thanks, in particular, to the development of more and more selective and efficient catalysts. These catalysts are able to promote a plethora of carbonylations under mild conditions, which can be applied to a large variety of organic substrates. Accordingly, carbonylations with CO have become increasingly more and more important, at the industrial and academic level, as testified by the considerable number of books [5-12] and reviews [13-115] dedicated to this topic and by the increasing number of industrial patents and scientific publications.

As said above, the incorporation of carbon monoxide into an organic substrate to give a carbonyl compound is called carbonylation. Interestingly, during the last years, considerable effort has been made by the scientific community to use CO surrogates as indirect carbonylating agents or as in situ sources of CO (both in industry and in academia, to avoid the direct handling of gaseous and toxic CO). In this book, several representative examples of CO surrogates will also be presented and discussed. Although CO surrogates can be interesting from a practical point of view, it should still be considered that carbon monoxide is cheaper than its surrogates and that carbonylations with CO occur with a higher atom economy.

Carbon monoxide possesses the strongest bond currently known (257.3?kcal/mol) [116]. This bond is only weakly polarized in the direction of carbon (the experimental dipole moment is 0.122?D) [117]. These characteristics make carbon monoxide a relatively stable and inert molecule. Consequently, CO can be attacked by highly reactive species, such as free radicals, carbocations, and strong nucleophiles (like alkoxides, amide anions, and organolithium reagents) (Scheme 1.1). These reactions form acyl radicals, acyl carbocations, and [NuCO]- intermediates (alkoxycarbonyl anions, carbamoyl anions, and acyl anions). They evolve toward forming the final carbonylation product depending on the nature of reactants and reaction conditions (Scheme 1.1).

Scheme 1.1 Reactions of carbon monoxide with free radicals, carbocations, or strong nucleophiles (such as RO-, R2N-, RLi).

However, the most common way to activate CO in carbonylation reactions under relatively mild conditions is metal coordination. In fact, upon coordination to a metal center M, the carbon atom becomes more electrophilic. It accordingly becomes susceptible to attack even by a relatively weak nucleophile, either external or coordinated to the metal (Scheme 1.2; formal charges are omitted for clarity). When occurring within the coordination sphere of the metal, this process is called migratory insertion. In this case, the metal also favors the attack to coordinated CO for entropic reasons. In either case (external attack, Scheme 1.2a, or migratory insertion, Scheme 1.2b), the coordinated carbon monoxide is transformed into a species in which the carbonyl group is bonded to M, and whose particular structure depends, apart from the metal, on the nature of the nucleophile. Thus, if the nucleophilic species is a carbon group (alkyl, alkenyl, or aryl) s-bonded to the metal undergoing migratory insertion, an acyl- or aroyl-metal species is formed. On the other hand, oxygen and nitrogen nucleophiles will lead to hydroxycarbonyl-, alkoxycarbonyl-, or carbamoyl-metal complexes, respectively (Scheme 1.2).

Scheme 1.2 Carbon monoxide coordinated to a metal center becomes more susceptible to nucleophilic attack, either intermolecularly (a) or intramolecularly (migratory insertion) (b).

These intermediates' fate will depend on the nature of the metal and of the reactants taking part in the carbonylation process and on reaction conditions. In most cases, the final carbonylated organic product is formed with the release of the metal, either in its original or in a different oxidation state. In the first case, a catalytic cycle is directly attained. In contrast, in the second case, the metal species must be reported in its original oxidation state (using a suitable redox agent) to achieve a catalytic process. For example, an acyl- or aroyl-metal intermediate R(CO)-M[+n]-X (X-?=?halide or other ligands, with M in the oxidation state [+n]) may undergo a nucleophilic attack by a nucleophile NuH (like water, alcohol, or an amine), with the formation of the carbonylated product R(CO)Nu (such as a carboxylic acid, an ester, or an amide), HX, and the reduced metal M[+(n-2)] (reductive displacement or nucleophilic displacement; Scheme 1.3).

Scheme 1.3 An acyl- or aroyl-metal intermediate (R?=?carbon group) undergoing reductive displacement (also called nucleophilic displacement).

If the metal initiated the process in its [+(n-2)] oxidation state (for example, by oxidative addition of R-X to the metal center to give R-M[+n]-X followed by CO migratory insertion), a catalytic cycle is directly achieved (Scheme 1.4). On the other hand, if the metal initiated the process in its [+n] oxidation state (for example, by metalation of R-H by M[+n]X2 with the formation of R-M[+n]-X?+?HX, followed by CO migratory insertion), the use of a suitable external oxidant is necessary to reconvert the reduced metal M[+(n-2)] to M[+n] and realize a catalytic process (Scheme 1.5). Clearly, from a practical and economical point of view, the occurrence of a carbonylative catalytic cycle is highly desirable. In the last decades, there has been considerable attention to developing more and more robust and efficient metal catalysts, also heterogeneous and/or with the possibility of being effectively recycled.

Scheme 1.4 An example of catalytic carbonylation process in which the metal is eliminated at the end of the process in its original oxidation state.

Scheme 1.5 An example of catalytic carbonylation process in which the metal is reduced at the end of the process and is reoxidized to its original oxidation state by the action of an external oxidant.

The nucleophilic attack of NuH to an acyl- or aroyl-metal intermediate (either inter- or intramolecular) is a common and important process by which the final carbonylated compound is delivered in a carbonylation reaction. This process is called reductive displacement or nucleophilic displacement. The exact mechanism this step may take place depends on reaction conditions, and, in particular, if the carbonylation process is done under acidic, neutral, or basic conditions. Under acidic and neutral conditions, the nucleophile tends to attack the carbonyl (possibly protonated) of the R(CO)-M[+n]-X complex, with the formation of a tetrahedral intermediate. This intermediate undergoes ß-H elimination from the H-O-C-MX moiety to give R(CO)Nu and a metal hydride species H-M[+n]-X, in equilibrium with M[+(n-2)]?+?HX (addition-elimination mechanism, Scheme 1.6a). On the other hand, under basic conditions, NuH (possibly in its anionic Nu- form) preferably attacks the metal center, with formal elimination of X- and formation of the R(CO)-M[+n]-Nu complex. Reductive elimination then leads to R(CO)Nu and M[+(n-2)] (ligand exchange mechanism; Scheme 1.6b). This latter case also occurs when the R(CO)-M[+n]-X species is attacked by an organometallic reagent R´M´ with the formation of M´X and R(CO)-M[+n]-R´ that undergoes reductive elimination to give R(CO)R´ (as occurs in the...