Random hcp and fcc structures in thermoresponsive microgel crystals

Monodisperse thermoresponsive poly(N-isopropylacrylamide) (PNIPAM) microgel particles having a diameter of 520 nm were synthesized by free-radical precipitation polymerization and centrifuged to obtain a concentrated suspension. The centrifuged mother suspension was made to self-order into a crystalline state by repeated annealing beyond the volume phase transition (VPT) of the particles. We report here the three-dimensional (3D) real space structure, determined using a confocal laser scanning microscope, of PNIPAM microgel crystal samples prepared by two different recrystallized routes: (1) solidifying a shear melted colloidal liquid (referred as as-prepared sample) and (2) slow cooling of a colloidal liquid (referred as recrystallized sample). We have recorded images of several regions of the crystal with each region containing 15 horizontal crystal planes for determining the in-plane [two-dimensional (2D)] and 3D pair-correlation functions. The 2D pair-correlation function g(r) revealed hexagonal long-range order of particles in the layers with a lattice constant of 620 nm. The analysis of stacking sequence of layers recorded on as-prepared sample has revealed the existence of stacking disorder with an average stacking probability alpha similar to 0.42. This value of alpha together with the analysis of 3D pair-correlation function determined from particle positions revealed the structure of microgel crystals in the as-prepared sample to be random hexagonal close packing. We report the first observation of a split second peak in the 3D g(r) of the microgel crystals obtained from a shear melted liquid. Upon melting the sample above VPT and recrystallizing it the split second peak disappeared and the crystals are found to have a face centered cubic (fcc) structure with alpha similar to 0.95. From simulations, the split second peak is shown to arise from the displacement of some of the B-planes from the ideal hcp positions. The present results are discussed in light of those reported for charged and hard sphere colloidal crystals and plausible reasons for observing two different structures are also explained.