Figure 1. Design of linear metal-organic ligands L1-L5 and corresponding supramolecules C1-C5.
The story of this article started four years ago when I was still at Texas State University. Inspired by the complex structure of protein, I wanted to introduce sequence into the building blocks of metallo-supramolecules. At that time, most of the supramolecules were constructed with simple building blocks through coordination-driven self-assembly. Even right now, using simple building blocks to construct complex structures with similar complexity as Nature does is still the mainstream in self-assembly field.
The first structure I designed was C4 in Figure 1. To construct C4, we need two terpyridine ligands, A and B, in 1:1 ratio to assemble with metal ions. However, if we directly mixed these two ligands with metal ions, we would not able to obtain the target structure due to the self-sorting of individual ligand. For example, B would assemble to a small trimer structure (C1); whereas, the left A would form random supramolecular polymers. To avoid self-sorting, we needed to bridge A and B first using metal ions with strong coordination, i.e., Ru(II). Then the obtained metal-organic building block A-Ru-B can assemble with Zn(II) ions to form C4. Note that Zn(II) has very weak and highly reversible coordination with terpyridine to facilitate the self-assembly.
After finishing the design, I waited for more than one year to have the first graduate student, Yuanfang Ying in Fall 2015 to work on this project. The idea was simple but the synthesis and separation were not easy. She spent almost one year on the synthesis and self-assembly of C4. Meanwhile, I designed the second structure C5 with the combination of A and B. After the structure was drawn, I started to bridge some of them with the goal of building C5 with only one type of build block instead of using multi-components. It turned out that we needed a building block with longer sequence, A-Ru-A-Ru-B-Ru-B. Then in July 2016, my lab relocated to University of South Florida. Yuanfang chose to stay at Texas rather than moving to Florida. I had to give this long term project to another graduate student Bo Song.
Figure 2. Structure of supramolecule C5.
In the synthesis of building block for C5, it could generate two isomers, cis and trans forms. I almost gave up the separation, but Bo eventually isolated the major one with trans structure after one year’s struggling. Then mass spectrometry data told us we got the structure of C5 (Figure 2), among the largest metallo-supramolecules with molecular weight 38,066 Da. We then designed and synthesized C2 and C3 to make a full story. After finishing 5 structures, we found C2 was the most challenging one in synthesis although it has a relative small size. Typically, we would assume the large supramolecules are hard to make. I guess this is the most important thing we would like to share with the community.
The characterization of these large 2D supramolecules always bothered us. NMR became a nightmare because of the multiple types of terpyridine. Single crystal growth failed after numerous attempts. So we had to find other characterization. After another year of efforts, we were eventually able to get more structure information using ambient STM on graphite surface. Very recently, with low- temperature ultra-high-vacuum STM, we are able to characterize these supramolecules with atomic resolution. We will share the results in our further report.