Previous Research
Undergraduate Research: University of Virginia
My first research experience was at UVA under MG Finn. The group discovered two new reactions, shown below:
Using the same reagents, allenes can be formed from an aromatic non-enolizable aldehyde, and dienes can be formed from an enolizable aldehyde. I was able to successfully react p-tolualdehyde (Ar = C6H5CH3) and p-anisaldehyde (Ar = C6H5OCH3) to get allenes in good yields. I was also able to react hexanal (R = C4H9) and cyclohexane carboxaldehyde (R = cyc) to get dienes in good yields. However, phenylacetaldehyde, diphenylethanal, and 2-propionaldehyde failed to give any significant products.
A large part of my project was to optimize yields. I tried varying the solvent, nitrogen base, and phosphorus ylide, but the yields got worse. Cooling the temperature did not give any consistent effect. After I left, a way was found to make allenes using two different aldehydes, making this reaction one of the best for allene synthesis. I got my first publication out of this work, shown below.
A new condensation synthesis of allenes and dienesGraduate Research: Stanford University
In graduate school I worked for Barry Trost working on the ruthenium-catalyzed ene reaction shown below.
Notice that the reaction gives two different products, depending on which side of the alkyne bonds to the alkene. This reaction had been studied before, and my work expanded on this knowledge. I spent a lot of time studying the macrocyclization (forming large rings) reaction. If R1 and R3 are part of the same chain, then doing the ene reaction will form a ring.
Using the substrates below, I tested for the possible ring sizes which the reaction can produce.
The best ring size for the reaction is 25 atoms, which is quite difficult for most other reactions, making this a very good way to make large rings. The isomer ratio is consistently about 4:1, and doesn't depend on ring size. I made and tested the substances below, and found that they gave about the same yields and isomer ratios as the compounds above.
I then tried putting different groups on the alkyne to test the effect on the isomer ratio.
The term A-value refers to steric size. Notice that the size of R does have a significant effect on the isomer ratio. In general, smaller substituents tend to favor isomer A, with the interesting exception of R = H. I used the CAChe computer modeling system to see if I could find an electronic effect on the isomer ratio, but the computer found no significant effect.
Of course, most of the actual work in lab was making these substrates to test. Most were not too difficult, but some were tricky.
Once we got a good handle on what the first ruthenium catalyst could do, it was time to see if the catalyst system could be improved. First, I tried adding a bunch of different ligands, shown below.
Unfortunately, none of these ligands had any impact on the isomer ratio. Their only effect was to lower the yields. I tried a variety of other additives, none of which had any positive impact. It is worth noting that the reaction can tolerate acids, but not bases.
I assumed that the ligands had little effect because they dissociate in solution, so my next idea was to attach groups directly to the cyclopentadienyl ring, so that they could not come off during the reaction. It tried all of the complexes listed below for the reaction. I synthesized all of them myself except for 2. The complexes 1, 3, 4, and 8 were known compounds made before, but the others were all made for the first time by me.
Complex 1 is the one used in the macrocyclization studies. It pains me to admit that the majority of the complexes shown did not improve the yield or isomer ratio of the reaction. The Cp* complexes 3 and 8 did give a significantly different ratio than 1, but I still got mixtures of both isomers. My pride and joy was complex 11 which gave a 96:4 isomer ratio in the case of a terminal alkyne, the only significant improvement over 1. Here is how I made it:
Many of these complexes were rather tricky to make, and I got to learn some organometallic chemistry. Trost, for reasons not particularly clear to me, never published any of this work. However, it can be found in my thesis:
Sundermann, M. J. The Ruthenium Catalyzed Alder-ene Reaction: Macrocyclizations and Studies of Alternate Catalyts, 2000 at Stanford University
Post-Doctoral Research: Colorado State University
After graduate school, I worked as a post-doc under Louis Hegedus. Before I arrived, a way was found to make compounds called cyclams in six steps. Compound 13 shown below is a cyclam. The cyclam had been "capped" with 2,6-(bromomethyl)pyridine to give 14. When 14 was treated with copper (II), the metal complex 15 was formed.
My project was to make new capping reagents, find different metals other than copper to complex to the cyclams, and to start making oligomers of many complexed cyclams linked together.
Compound 16 (with pyridine) had already been used. I successfully capped a cyclam with 17 (with pyrazine). Despite many attempts, I was unable to synthesize 18. The purpose of the next two compounds 19, 20 was to cap two different cyclams together to make an oligomer. I made 19, but was not able to attach it to a cyclam. I was unable to synthesize 20. Both compounds 18 and 20 have simple structures, and one would think they would be easy to make. However, it is difficult to do reactions on electron-poor heterocycles. As of the time I did my research, no literature synthesis of either compound was availible, and it would have been nice to contribute one. Oh, well.
I attempted using metals other than copper to put inside the cyclams. I tried making complexes of 21 with nickel and ruthenium, but neither were successful. I was able to make the palladium complex of the uncapped cyclam 13, but not 21.
I was able to make dimers using a metal to attach two molecules together. For example, complexing 21 with copper, and treating it with Rh2(acac)4 in methanol or acetonitrile gives the compound shown below. The crystal structure is on the main page of this website.
I was also able to make a similar compound using a single ruthenium as the linker metal. I tried making dimers with palladium, chromium, and cobalt, but I could not isolate any desired compounds.
I did get one last interesting result. Usually, when inserting copper into a cyclam, such as making 15 from 14, the reaction is done thermally in hot methanol. When 21 is treated with excess copper and heated in the microwave, beautiful blue crystals of the compound below are formed. Notice that an extra copper forms the dimer. This time, it is not the pyrazine rings that are being attached, but the oxygens on the cyclam. Isn't chemistry neat?
This work was published
Synthesis, Complexation, and Coordination Oligomerization of 1,8-Pyrazine-Capped 5,12-Dioxocyclams