60th Annual Report on Research 2015

  Background   Since Casey’s report on the catalytic activity of 3 in 2007, there have been a growing number of publications reporting catalytic applications of 3 and other (cyclopentadienone)iron carbonyl derivatives.  We and others have discovered that cyclopentadienone substitution has a large impact on the catalyst’s reactivity. One major goal of our research has been to synthesize a series of (cyclopentadienone)iron tricarbonyl compounds with systematic variations in the cyclopentadienone substitution and test these compounds to determine their catalytic activities.  Our hope has been to discover a structure/activity relationship and to use it to develop more active catalysts.    During the first two years of the award we focused on synthesizing the compounds shown in figure 1 and exploring their reactivity in a set of oxidation and reduction reactions (shown in figure 2).  More catalysts than those shown in figure 1 were made, but initial catalytic studies revealed that not all of them were worth studying in greater detail. After developing GC-based analysis methods for each of the six screening reactions and finding appropriate catalyst loadings for each reaction, we explored the catalytic activity of our iron compounds.   

  Each screening reaction was chosen as a representative of a broader transformation: reactions 1 and 2 are benzylic and aliphatic alcohol oxidations, respectively; reactions 3 and 4 are reduction reactions with varying steric hindrance on the substrate; reaction 5 is a challenging substrate for these catalysts (presumably due to coordination of the nitrile to the iron center); and reaction 6 is an imine reduction.  

    Results from Year 3   Because the catalytic activity of iron compounds bearing cyclopentadienones without fused rings has not been extensively explored, we initially focused on testing compounds 7–11 (containing two phenyl groups in the 3- and 4-positions of the cyclopentadienone ring) in our six oxidation and reduction reactions. Knölker’s compound 1 had been the primary catalyst used in these types of reactions, so we were interested in comparing its reactivity to those of 7–11.  After running the reactions, we were surprised to find that compound 1 never afforded the highest GC yields or conversions for any of the six reactions.  For example, the average GC yield for the reduction of N-benzylideneaniline (reaction 6) with 1 is 27%, but it is 99% with catalysts 10 and 11.     There was no iron compound that was superior in all reactions, which suggests it may not be possible to develop a “magic bullet” catalyst.  The 3,4-diphenyl catalysts (7–11) generally worked better in the alcohol oxidation transformations (reactions 1 and 2) relative to those bearing fused rings in the 3,4-positions.  In acetophenone reduction (reaction 3), catalysts bearing less steric hindrance in the 2,5-positions afforded product in modestly higher GC yields, with catalyst 1 being the worst of those tested. Based on the proposed mechanism for these reactions, the substitution on the 2,5-positions of the cyclopentadienones would be expected to have a larger impact compared to the 3,4-positions, but in most reactions modifications to the 3,4-positions of the catalysts while holding the 2,5-positions constant caused changes in GC yields.  In the case of mesitaldehyde reduction (reaction 4), changes to the 3,4-positions of the cyclopentadienone impacted the catalysts much more than changes to the 2,5-positions.  Interestingly, the only reaction that showed a clear pattern based on the steric hindrance of the groups in the 2,5-positions was the imine reduction (reaction 6).  The smaller the group in the 2,5-positions, the better the GC yield.  For example, catalysts 5 and 11 (both bearing methyl groups in the 2,5-positions but different groups in the 3,4-positions) behaved almost identically.    After gaining some insights into how the structure impacted catalyst activity, we decided to focus on two areas: exploring in more detail how the different positions on the cyclopentadienone impact catalyst activity, and looking at the kinetics of these reactions.  To tackle the first objective, we synthesized the five catalysts shown in figure 3 and have begun screening those in our reactions. The data from these reactions, combined with our previous data, should address our questions about 2,5-substitution vs. 3,4-substitution.  We are examining the second objective by monitoring conversion/yield vs. time for a subset of our catalysts (1, 4, 6, 8, and 11) in reactions 2, 3, and 6.  While these studies are ongoing, they have already revealed information about the initial rates of catalysts versus their long-term stability that we were unable to obtain from the screening reactions.     This year we also revisited oxidative lactonizations of diols using these catalysts.  Initial studies have shown that we can lower the loadings to at least 2.5 mole % with certain catalysts and get excellent yields of valerolactone from 1,5-pentanediol.  Current studies are focused on determining the best catalyst and the optimal reaction conditions.     Impact on Undergraduate Research Program   Thirteen undergraduate students have worked on this project over the years, and seven of those have been directly impacted by the funding provided by the ACS PRF.  Of those seven, four are in chemistry graduate programs and the remaining three are still undergraduate students.  It has helped provide the resources to get this project to the point where we can collect a large amount of data in a relatively short amount of time, which is very beneficial to my research program moving forward.  Since receiving funding from the ACS PRF, we have presented our results at both regional and national conferences, including ACS meetings, and we are preparing our data for publication.