60th Annual Report on Research 2015

Samarium diiodide is one of the most useful reagents employed by organic chemists.  The utility of this reagent is greatly enhanced by the addition of electron rich cosolvents.  HMPA (1, Scheme 1) is the most common cosolvent to be employed and the majority of modern SmI₂ chemistry is accomplished with this cosolvent (Scheme 1).  Unfortunately HMPA is a mutagen which has leaded to an extended search for suitable alternatives.  The goals of this work are to find SmI₂ activators which are both safer and more effective.  We have recently reported that TPPA (2) is approximately an order of magnitude better at activating SmI₂ at two of its key synthetic processes, reduction of carbon-halogen bonds and reduction of ketones.  Our most recent published work involves the use of deprotonatable phosphoramides such as DPMPA (3).  Complexes formed from SmI₂ and four equivalents of the anion derived from deprotonation of 3 (DPMPA^-) proved to be > 100X more reactive than the SmI₂/HMPA complex as determined by the reduction of 1-chlorodecane.

Scheme 1. Phosphoramides.

Recently we have focused on extending this approach to ureas.  Advantages of ureas over phosphoramides include lack of mutagenicity, lower molecular weight, and ease of synthesis.  Due to the lower valency of the central atom, the conjugate bases of ureas will potentially form less sterically encumbered complexes with SmI₂.  The most common urea used in synthesis, DMPU, is known to be a modest activator for SmI₂.  After a series of screening reductions of 1-chlorodecane, we have decided to focus on three deprotonatable ureas as shown in Scheme 2.

Scheme 2. Deprotonatable Ureas.

Triethylurea (TEU, 4) was selected for its demonstrated ability to activate SmI₂ for R-X reduction as well as its ease and low cost of synthesis.  N-Butylimidazolidinone (BI, 5) and N-propylpropylene urea (PPU, 6) are better activators than TEU but made from more costly starting materials.  Because of the centrality of TEU to our program, we engaged in an optimization study for the reduction of 1-chlordecane (Table 1).  It was determined that the order of addition mattered greatly.  Adding SmI₂ last (Entry 2, Method B) works much better than adding 1-chlorodecane last (Entry 1, Method A).  Although the reaction proceeds without a proton source, the addition of t-amyl alcohol does increase the yield.  Interestingly, even two equivalents of TEU^- per equivalent of SmI₂ in the presence of t-amyl alcohol is sufficient for near quantitative reduction of 1-chlorodecane in 1 hour (Table 1, Entry 7). Table1.  The Optimization of 1-Chlorodecane Reduction with SmI₂/TEU^-.

We used this newly optimized procedure (analogous to Entry 7 in Table 1) to reduce several alkyl chlorides and even fluorides.  Chlororiboside 7 and pyridyl fluoride 9 are illustrative (Scheme 3).

Scheme 3.  Selected R-X Reductions.

With less reactive alkyl and aryl halides the better activators BI^- and PPU^- must be employed (Scheme 4).  In this case the strongly electron donating amino substituents slows the initial reduction necessitating the use of stronger activators for synthetically useful reactions.

Scheme 4.  The reduction of 4-N,N-Dibutyl-1-chlorobenzene.

We are also interested in the ability of our SmI₂/TEU^- complexes to effect radical cyclization of appropriately substituted haloalkenes.  A series of naphthyl-based substrates were cyclized (Table 2).  Both the 2:1 and 4:1 complexes were employed.  The proton source t-amyl alcohol was also used with the 2:1 complex.  Side reactions include premature reduction to afford reduced uncyclized products 14a and b, reductive deallylation of the O-allyl species leading to b-naphthol (15), and overreduction to generate naphthalene (16).  In spite of this trifecta of side reactions, conditions could be crafted which produced > 60% yield of the cyclized product in each case.

Table 2.  Cyclization of Haloalkenylnaphthalenes.

We are also involved in radical cyclizations of aniline and pyridine-based systems.  The two substrates shown in Scheme 6 are typical.  Substrate 12 has a substantial tendency to afford the reduced uncyclized product 14 when R is H or Et.  With the larger Bu and Hex substituents the cyclization product 13 is favored.

Scheme 5.  Other radical cyclizations.

Current work in this laboratory focuses on the use of other reluctant radical precursors, as well as the development of new SmI₂ reactions.