Fractals, being "exactly the same at every scale or nearly the same at different scales" as defined by Benoit B. Mandelbrot, are complicated yet fascinating patterns that are important in aesthetics, mathematics, science and engineering. Extended molecular fractals formed by the self-assembly of small-molecule components have long been pursued but, to the best of our knowledge, not achieved. To tackle this challenge we designed and made two aromatic bromo compounds (4,4″-dibromo-1,1':3',1″-terphenyl and 4,4‴-dibromo-1,1':3',1″:4″,1‴-quaterphenyl) to serve as building blocks. The formation of synergistic halogen and hydrogen bonds between these molecules is the driving force to assemble successfully a whole series of defect-free molecular fractals, specifically Sierpiński triangles, on a Ag(111) surface below 80 K. Several critical points that govern the preparation of the molecular Sierpiński triangles were scrutinized experimentally and revealed explicitly. This new strategy may be applied to prepare and explore various planar molecular fractals at surfaces.

Self-assembly of trimesic acid (TMA) displayed remarkable abundance over its full coverage range on gold under ultrahigh vacuum conditions. Experiments showed that previously well-reported “chicken wire” and “flower” structures were actually two special cases within its full coverage. All observed assembling structures formed hexagonal porous networks that could be well-described by a unified model in which the TMA molecules inside the half unit cells (equilateral triangles) were bound via trimeric hydrogen bonds and all half unit cells were connected to each other via dimeric hydrogen bonds. These porous networks possessed pores of 1.1 ± 0.1 nm in diameter, and the interpore distance was tunable from 1.6 nm on at a step size of ∼0.93 nm. Energetics analysis unveiled that the assembling structures less than one molecular layer was optimally driven by maximization of the dimeric hydrogen bonds.

Molecule-based functional devices on surfaces may take advantage of bistable molecular switches. The conformational dynamics and efficiency of switches are radically different on surfaces compared to the liquid phase. We present a design of molecular layers which enables bistable switching on a surface and, for the first time, demonstrate control of a single switch in a dense and ordered array at the spatial limit. Up and down motion of a central Sn ion through the frame of a phthalocyanine molecule is achieved via resonant electron or hole injection into molecular orbitals.