Introduction

1.1.1 Aims of the Book

Since its introduction to the synthetic chemistry community in 1977 by Kagan, samarium diiodide (SmI2) has captured the imagination of organic chemists and has become one of the most important reducing agents available in the laboratory. The main chapters of this practical guide deal with the remarkable ability of SmI2 to transform functional groups selectively and to orchestrate carbon–carbon bond formation. Other chapters deal with our understanding of mechanism and additive effects in reactions mediated by SmI2 – an area that should still be considered as very much a work in progress. The final chapter of the book deals with selected emerging areas in the use of SmI2 in synthesis and reflects the authors' research interests.

The book aims to steer a difficult course by providing a sufficient level of detail and new developments without burying the basics. Representative procedures have been included to encourage the reader to take their first steps into the fascinating organic chemistry of SmI2.

1.1.2 Further Reading

The many excellent reviews on the use of SmI2 in organic chemistry are a rich source of additional information. The major reviews and their authors are categorised according to their coverage below in Figure 1.1.

1.2 Introducing the Reagent

1.2.1 Working with SmI2

Samarium(II) iodide (SmI2) is commercially available as a solution in THF or can be prepared readily using one of several straightforward methods that have been described (see Chapter 2, Section 2.1). SmI2 is air sensitive, but is tolerant of water and can be handled using standard syringe techniques. Reactions are typically carried out in THF although the use of other solvents has been investigated. Part of the reagent's popularity arises from its ability to mediate both radical and anionic processes and sequences involving both. As a result, it has been utilised in a wide range of synthetic transformations ranging from functional group interconversions to carbon–carbon bond-forming reactions. In addition, the reagent is often highly chemoselective and transformations instigated by SmI2 tend to proceed with high degrees of stereoselectivity. Further adding to its appeal, the reactivity, chemoselectivity and stereo-selectivity of SmI2 can be manipulated and fine-tuned by the addition of various salts and cosolvents to the reaction mixture (see Chapter 2, Section 2.2).

1.2.2 Electronic Configuration of Sm(II)

Samarium, like all lanthanide elements, preferentially exists in the + 3 oxidation state. The loss of the three outermost electrons, namely the 5d1, 6s2 electrons, results in enhanced thermodynamic stability in which a closed-shell Xe-like electronic configuration is adopted. The + 2 oxidation state is most relevant for samarium (f6, near half-filled), europium (f7, half-filled), thulium (f13, nearly filled) and ytterbium (f14, filled). In order to attain the more stable + 3 oxidation state, SmI2 readily gives up its final outer-shell electron, in a thermodynamically driven process, making it a very powerful and synthetically useful single-electron transfer reagent.

1.2.3 Reduction Potential

The redox potential of the SmI2-SmI2+ couple has been established through the use of linear sweep and cyclic voltammetry and was found to be approximately -1.41V, determined for a solution of SmI2 in THF. Although the preparation of SmI2 in a variety of different solvents is known, THF is by far the most common solvent associated with its use. It would be expected, however, that the observed reduction potential of the reagent will vary greatly depending on the choice of solvent in which the measurement is carried out owing to the differences in the strength and number of solvent interactions.

It is possible to manipulate the reduction potential of SmI2 through the use of various additives; most commonly these are found to be molecules containing neutral or Lewis basic oxygen functionalities. The most common example of this is the use of HMPA (hexamethylphosphoramide) as an additive, for which it was determined that, upon addition of 4 equiv, the reduction potential of SmI2 in THF is increased to approximately -1.79 V, thereby significantly increasing its potency as a single-electron transfer reagent. Similar effects are also observed in the presence of alcohols and even water, for which it has been reported that upon the inclusion of 500 equiv, the reduction potential can be increased as far as -1.9 V. The use and mechanistic role of such additives will be discussed in more detail in Chapter 2, Section 2.2.

1.2.4 Coordination Chemistry

Lanthanides typically adopt coordination numbers greater than six, depending largely on the size of the lanthanide ion and on the size of the ligands. SmI2 is an oxophilic reagent and, as such, much of the coordination chemistry observed involves the close association of oxygenated molecules, including solvents and substrates, to the metal centre.

It has been established that in a solution of THF, SmI2 exists in a hepta-coordinate geometry in which five THF molecules are equatorially bound to the central Sm(II) ion through their oxygen lone pairs, with iodide ligands axial. The coordination number and geometry of SmI2 is variable depending on the nature of the ligands involved. For example, in the [SmI2(HMPA)4] complex alluded to previously, the more sterically demanding HMPA ligands are positioned equatorially around the Sm(II) ion, but this time only four ligands are involved and an octahedral (hexacoordinate) geometry results. Complexes in which the coordination number is as high as eight ([FORMULA OMITTED]) and nine ([FORMULA OMITTED]) have also been identified.

The oxophilic nature of SmI2 is in many cases a highly beneficial quality: The coordination of two or more oxygenated reactive centres to samarium in both radical and ionic processes mediated by the reagent can often lead to high levels of diastereoselectivity in the formation of products. The coordination of Sm(II) and Sm(III) to oxygen donors on the substrate, solvent or cosolvent is a common theme that runs through each of the subsequent chapters.

The Reagent and the Effect of Additives

2.1 Preparing SmI2

Although SmI2 is commercially available as a 0.1 M solution in THF from several suppliers, it is also easy to prepare using one of several known procedures.

During Kagan's early work with SmI2, he developed a convenient method to prepare the reagent from samarium metal using 1,2-diiodoethane in THF (Scheme 2.1). Stirring this mixture under an inert atmosphere for several hours gave a 0.1–0.05 M solution of SmI2 as a characteristic dark blue solution with the formation of ethene as a by-product. This solution is stable for several days when stored under an inert atmosphere, particularly when a small amount of samarium metal is present in the solution. Since Kagan's early studies, diio-domethane has been used by many groups, including that of Molander, to oxidise samarium metal in place of 1,2-diiodoethane.

Representative procedure – preparation of SmI2 using Kagan's procedure. Samarium powder (3.00g, 0.02mol) was placed in a reaction flask under an inert atmosphere and a thoroughly degassed solution of 1,2-diiodoethane (2.82 g, 0.01 mol) in dry THF (250 ml) was slowly added. The mixture was stirred at room temperature until a dark blue SmI2 solution was obtained.

Several years after Kagan published his preliminary work on SmI2, Imamoto reported a more atom–efficient method for preparing SmI2 that used samarium metal and iodine in THF. Imamoto assumed that Kagan's route to the reagent proceeds by initial formation of samarium triiodide (SmI3), which is then further reduced by samarium metal to give SmI2 by a disproportionation process (Scheme 2.2).

Representative procedure – preparation of SmI2 using Imamoto's procedure. Samarium powder (1.00 g, 6.65 mmol) was placed in a reaction flask under an inert atmosphere and thoroughly degassed THF (55 ml) was added. Iodine (1.41 g, 5.56mmol) was added to the reaction mixture and the resulting suspension heated at 60 °C for 12 h to give a dark blue solution of SmI2.

In 1992, Ishii reported the generation of a SmI2 equivalent that displayed similar reactivity to SmI2 by treatment of samarium metal with TMSCl and NaI.

In recent years, exposing the preparation of the reagent from samarium metal and an oxidant to different stimuli has led to significant improvements in reaction time. Concellón utilized the sonication of samarium metal and iodo-form at room temperature to give a solution of SmI2 in THF in approximately 5min (Scheme 2.3). This approach was also used by Flowers to synthesise other Sm(II) species. It was also reported that using different oxidants, such as 1,2-diiodoethane, diiodomethane and iodine, works just as well with this technique.

Hilmersson achieved similar results by heating the reaction to 180 °C under microwave conditions for 5min (Scheme 2.3). Although this method is compatible with all of the oxidants described above, formation of SmI2 from samarium metal and iodine is preferred as no gas evolution accompanies the reaction.

As THF can act as a hydrogen donor towards radical species and can in some cases undergo ring opening in the presence of Lewis acidic samarium species, the solvent is not ideal for all SmI2-mediated transformations. This led various groups to investigate the formation of SmI2 in other solvents. For example, Kagan and Namy reported the formation of the reagent in tetra-hydropyran (THP), whereas Ruder reported the use of acetonitrile, and Tani prepared a solution in benzene and HMPA. Kagan and Namy showed that the use of SmI2 in THP allowed several transformations to be carried out that were not possible using the reagent in THF. For example, they succeeded in efficiently coupling acid chlorides, a transformation that was not possible using SmI2 in THF, and were able to form allylic and benzylic organosamariums, organometallic species that underwent decomposition in THF. More recently, Flowers utilised sonication to prepare SmI2 in low concentration (0.02–0.05 M) in acetonitrile, DME, 2-propanol, 2-methyl-2-propanol and 2-heptanol. Attempts to prepare samarium diiodide in ethereal solutions other than THF, including diethyl ether, tert-butyl methyl ether and dioxane, were unsuccessful.

2.2 The Use of Additives and Cosolvents in SmI2 Reactions

One of the most fascinating features of SmI2 is the ability to modify its behaviour through the use of cosolvents or additives. For example, cosolvents or additives can be used to control the rate of reduction or the chemo- or stereoselectivity of reactions. Additives commonly utilised to fine-tune the reactivity of SmI2 can be classified into three major groups:

1. Lewis bases – HMPA and other electron-donor ligands, chelating ethers, etc.

2. Proton sources – predominantly alcohols and water.

3. Inorganic additives – NiI2, FeCl3, etc.

This section will focus on reactions that exemplify the use of particular additives and cosolvents and on the origin of the beneficial effects of the additives. Miscellaneous additives, not included in the three categories above, that promote the reactivity of SmI2 will be discussed at the end of the section. Further examples of the use of additives and cosolvents can be found in sub-sequent chapters.

2.2.1 Lewis Bases

Lewis bases, containing basic nitrogen or oxygen atoms, are important promoters of reactions mediated by SmI2. Among these additives HMPA has played an important role in the development of SmI2-mediated reactions since it accelerates a wide range of functional group conversions and bond-forming reactions. The use of HMPA in SmI2-mediated reductions not only increases the rate of reactions, but can also enhance the level of stereochemical control observed. Although numerous other Lewis basic additives have been used and alternative protocols have been developed, none yet approach the general utility of the SmI2–HMPA reagent system. Thus, in spite of the toxicity of HMPA, it remains the additive of choice for many reactions using SmI2. Illustrative examples of the use of HMPA as an additive are given below, and more detailed examples can be found in subsequent chapters.

The potential of the SmI2–HMPA pairing was first recognized by Inanaga, who discovered that the use of this additive significantly increased the rate of reduction of alkyl and aryl halides. The seminal work of Kagan showed that alkyl iodides are reduced by SmI2 at elevated temperatures but alkyl bromides are reduced at a very slow rate and alkyl chlorides are unreactive. Inanaga found that the addition of approximately 10% HMPA in THF dramatically enhanced the rate of reduction of alkyl halides: alkyl iodides and bromides were reduced at room temperature and even alkyl chlorides were reduced, albeit at elevated temperatures (Scheme 2.4). Synthetically relevant examples of alkyl halide reductions are described in Chapter 4, Section 4.2.

Inanaga showed that the presence of HMPA also accelerates the Barbier addition of alkyl halides to ketones and significantly improves the yield of the adducts. He also found that HMPA was a useful additive in the SmI2-mediated synthesis of lactones from bromo esters and ketones (Scheme 2.5). Since Inanaga's pioneering work, the Barbier reaction employing SmI2–HMPA has been employed widely in synthesis and is discussed in detail in Chapter 5, Section 5.4.

Although many early synthetic studies employed HMPA as a cosolvent, its mechanistic role remained unclear. Its role was later clarified by Molander, who studied the influence of HMPA concentration on the product distributions from the SmI2-mediated reductive cyclisations of unactivated olefinic ketones. The addition of HMPA was required to promote efficient ketyl-alkene cyclisation, and correlations between the concentration of HMPA, product ratios and diastereoselectivities were apparent (Scheme 2.6). In the absence of HMPA, attempted cyclisations led to the recovery of starting material 1, reduced side-product 3 and desired cyclisation product 2. Addition of 2 equiv of HMPA provided 2 and only a small fraction of 3. Further addition of HMPA (3–8 equiv) provided 2 exclusively (Scheme 2.6).

The first clear trend in the data was that HMPA effectively increased the reducing ability of SmI2. The absence of HMPA, or the presence of only small amounts (2 equiv), led to prolonged reaction times, while addition of a further 4–8 equiv of HMPA shortened reaction times. Molander suggested that HMPA may enhance the reducing ability in two ways: first, HMPA may dissociate SmI2 aggregates in THF, making the reductant more reactive, second, HMPA may perturb the electron-donating orbital of Sm(II) and raise its energy, thus increasing the Sm(II)/Sm(III) reduction potential. The second obvious trend in the data was the impact of HMPA concentration on product distribution. In the absence of HMPA, starting material 1 was recovered in addition to the desired product 2 and the side product 3. The addition of 2 equiv of HMPA led to a significant increase in the yield of 2. Molander also noted that 2 was obtained with improved diastereoselectivity on addition of HMPA. The addition of >4 equiv of HMPA provided 2 exclusively in high yield and with high diastereoselectivity. Based on these observations, Molander proposed that the addition of HMPA to SmI2 produces a sterically encumbered reductant that not only enhances the diastereoselectivity of reactions, but also stabilises reactive intermediates (ketyls, radicals) in close proximity to Sm–HMPA complexes, thereby preventing competing reaction processes such as hydrogen atom abstraction from THF that result in the formation of side product 3.