A research group led by Oleg Gang from Brookhaven National Laboratory has assembled a complete three-dimensional (3D) superlattice from nanosized cubes and spheres, using complementary DNA matching and shape-related directional interactions. Significantly, the researchers are able to change the type of the lattice by using octahedra instead of cubes and have shown that the final structure is a function of DNA flexibility and the relative cube-to-sphere sizes. Such controlled organization of nanoparticles is a highly sought-after goal in the field of emergent materials, where it promises tantalizing applications in areas such as metamaterials, photonics, and catalysis. The researchers reported their results in the April issue of Nature Communications (DOI: 10.1038/ncomms7912).
DNA has previously been used to assemble nanoparticles into 3D structures. These are usually chemically synthesized, single-stranded DNA, containing anywhere from ten to a few hundred bases. When complementary bases from two such strands meet, they form a duplex and therefore a bond. Assembly of DNA-functionalized spherical nanoparticles only yields one type of crystal lattice as the structure is determined by packing considerations alone. This is similar to how metals form structures—most metallic lattices are cubic. The rich diversity of crystal structures found in nature comes from the inherent directionality of covalent bonds. Gang and his team member Fang Lu reasoned that attaching DNA to a polyhedron such as a cube would cause the bonds to be perpendicular to the faces, such that they would form directional bonds.
A binary system of spherical (SNP) and cubic (CB) gold nanoparticles, each with 46 nm in size, were coated with complementary DNA and allowed to mix. Mixing in a 6:1 ratio generated clusters, while a 1:1 ratio formed lattices. The resulting 3D spatial arrangement of clusters was determined by tomography in a transmission electron microscope. Each CB was decorated at the center of each face by an SNP, which corresponds to a coordination number of six.
As expected, the use of octahedra instead of cubes resulted in eightfold coordination as the octahedron has eight faces. Annealing the 1:1 mixture at 39°C created an extended 3D lattice (see Figure) that was determined by small- angle x-ray scattering to be a NaCl type cubic lattice of unit cell size ≈100 nm. The superlattice based on the octahedron, in contrast, was found to be a CsCl type lattice. Thus, the shape symmetry of the polyhedra is found to translate predictably to the global superlattice structure.
“Directionality of interactions provided by the polyhedron is crucial for defining the lattice type. At the same time, DNA structure and SNP size control the morphology. Flexible DNA motifs are required for achieving crystalline organization,” Gang says; “However, excessively flexible DNA results only in short range order.”
Also interesting is the effect of size mismatch between the SNP and CB. Smaller SNPs of 27-nm edge length did not form extended global lattices. This can be understood by considering the local geometry of the interaction. The adhesive energy between the SNP and CB is correlated to the projected area of the SNP on the square face of the CB. If the SNPs are smaller than the cube, the projected area need not fall around the center of a face, making the position of the SNP and hence the bond direction ambiguous. An intermediate SNP of 38-nm diameter did form good crystals, but with a reduced lattice parameter, which implies that the DNA must be in a moderate state of compression.
“There has been a lot of work using DNA to crystallize spherical nanoparticles and to get directional interaction, but this research combines these two ideas for the first time,” says John Crocker of the University of Pennsylvania, who was not involved in the study.