Segregation of mass at the periphery of $N$-isopropylacrylamide-co-acrylic-acid microgels at high temperatures

We investigate poly($N$-isopropylacrylamide) (pNIPAM) microgels randomly copolymerized with large mol % of protonated acrylic acid (AAc), finding that above the lower critical solution temperature the presence of the acid strongly disrupts pNIPAM's collapse, leading to unexpected new behavior at high temperatures. Specifically, we see a dramatic increase in the ratio between the radius of gyration and the hydrodynamic radius above the theoretical value for homogeneous spheres, and a corresponding increase of the network length scale, which we attribute to the presence of a heterogeneous polymer distribution that forms due to frustration of pNIPAM's coil-to-globule transition by the AAc. We analyze this phenomenon using a Debye-Bueche-like scattering contribution as opposed to the Lorentzian term often used, interpreting the results in terms of mass segregation at the particle periphery.

Figure 1

Hydrodynamic radius ($\blacksquare$) and radius of gyration ($\square$) as a function of temperature measured using 3DDLS and SLS, respectively. The inset shows the resulting values of $R_g/R_h$ across the same temperature range.

Figure 2

Experimental form factors from SLS ($\square$) and SANS ($\circ$), together with best fits to Eq. (1) (thick lines) and to just the first term in Eq. (1) (thin solid lines). (a) $T=49$ °C, (b) $T=14$ °C. Dotted lines show the Guinier fits to the low-$q$ SLS data. The inset in (a) shows a zoom of the SANS data together with the best fit to Eq. (1) using $\mu =1$.

Figure 3

Representative schematic of proposed microgel conformation changes as a function of temperature in °C. We emphasize that the heterogeneities in the schematic for 59 °C are not drawn to scale; the internal length scale $\xi$ is at least a factor of 10 smaller than the particle diameter. Hence, these heterogeneities are relevant at the polymer level and do not significantly affect the particle structure at larger length scales. This is why they are only apparent at the $q$ values probed with SANS.