Evidence for vertical phase separation in densely grafted, high-molecular-weight poly($N$-isopropylacrylamide) brushes in water

The detailed conformational change of poly($N$-isopropylacrylamide) (PNIPAM) brushes at high grafting density in ${\mathrm{D}}_2\mathrm{O}$ was investigated as a function of temperature using neutron reflection. PNIPAM chains were grafted at high surface density from gold and silicon oxide surfaces by atom transfer radical polymerization. Whereas single layer profiles were observed for temperatures below and above the transition region, bilayer profiles were observed for a narrow range of temperatures near the transition. This nonmonotonic change in the concentration profile with temperature is discussed in the context of theoretical models of vertical phase separation within a brush.

Figure 1

(a) Neutron reflectivity data from 71G for a series of temperatures in ${\mathrm{D}}_2\mathrm{O}$. The curves through the data correspond to best-fits segment concentration profiles shown in Fig. 2. (b) Expanded view of the lower q region that shows the nonmnotonic change in the data with temperature.

Figure 2

Best-fit segment concentration profiles for 71G for 41 °C (—), 32 °C (— —), 30 °C (—), 28 °C (– –), 20 °C (∙–∙–), in ${\mathrm{D}}_2\mathrm{O}$. The profile at 30 °C was obtained using a bilayer in the fitting analysis, whereas all other profiles were obtained using a single layer. The inset shows the best-fit concentration profiles using a bilayer function for all temperatures. The change in profile shape through the transition is still apparent.

Figure 3

Neutron reflectivity data from 71G at 30 °C in ${\mathrm{D}}_2\mathrm{O}$. The data are compared with the best-fit curves using a single layer profile (⋯) and a bilayer profile (—). The inset shows the ${\chi }^2$ values for all temperatures using both single layer (●) and bilayer (◯) functions in the fitting. For a single layer function, a large increase in ${\chi }^2$ occurs for the narrow range of temperatures from 31 to 29 °C. The improvement with a bilayer is only significant for that same temperature range.

Figure 4

(a) Neutron reflectivity data from 209S for a series of temperatures in ${\mathrm{D}}_2\mathrm{O}$. The curves through the data correspond to best-fits using a single layer to describe the PNIPAM brush. (b) Expanded view of the lower $q$ region for temperatures in the transition region. Solid lines correspond to fits using two layers to describe the PNIPAM brush, and dashed lines indicate fits using one layer to describe the PNIPAM brush.

Figure 5

(a) Best-fit segment concentration profiles for 209S at 27, 32, and 34 °C showing the dramatic change in profile shape in the transition region. (b) Best-fit segment concentration profiles for 209S at 31, 32, and 33 °C showing the gradual change in the shape of the bilayer profile through the transition.

Figure 6

(a) First moment of the segment volume fraction profiles for 71G as a function of temperature upon cooling from 41 °C. The symbols indicate results using bilayer (◯) and single layer (●) functions in the fits. (b) Comparison of the change in first moment for heating (◻) versus cooling (∎) 71G through the transition. The maximum occurs at a slightly higher temperature when heating through the transition. (c) First moment of the segment volume fraction profiles for 209S as a function of temperature upon heating from 20 °C. The symbols indicate results using bilayer (◯) and single layer (●) functions in the fits.