Experiments were performed using a synthetic seawater as the feed solution. A high salinity feedwater was used since a possible use for the membranes would be as pretreatment to seawater desalination. To mimic seawater, a feed water composed of 2 mg/L of each suwannee River fulvic and humic acids, along with 0.1 mM CaCl2 as a representative of divalent cations, 0.1mM NaHCO3 as buffer system, 1M NaCl as background electrolyte, and 1mg/L of SiO2 was used. Several experiments were performed with both modified and unmodified membranes to evaluate the effectiveness of graft polymerization. Filtration experiments with different time intervals were performed on both unmodified and modified membranes at a constant pressure of 1.38 bar (20 psi). Experiments were run for 1, 5, 15 and 30 minutes, and for 1, 2, 4, 6 and 8 hours. These experiments with different time periods were performed to determine the extent of fouling on the membrane leading to flux decline. Figures 5.19 – 5.22 show variations in flux for the unmodified and modified membranes for the 1-minute, 15- minute, 6-hour and 8-hour filtration experiments.
Figure 5.19: Influence of PEG grafting on membrane flux: 1 min run.
40 60 80 100 120 140 160
0 50 100 150 200 250 300
Flux, L/(m2-hr)
Time, min
Unmodified Membrane Precompaction Modified Membrane Precompaction Unmodified Membrane Filtration Modified Membrane Filtration
77
Figure 5.20: Influence of PEG grafting on membrane flux: 15 min run.
Figure 5.21: Influence of PEG grafting on membrane flux: 6 hour run.
40 60 80 100 120 140 160
0 100 200 300
Flux, L/(m2 -hr)
Time, min
Unmodified Membrane Precompaction Modified Membrane Precompaction Unmodified Membrane Filtration Modified Membrane Filtration
40 60 80 100 120 140 160
0 200 400 600 800
Flux, L/(m2-hr)
Time, min
Unmodified Membrane Precompaction Modified Membrane Precompaction Unmodified Membrane Precompaction Modified Membrane Precompaction
78
Figure 5.22: Influence of PEG grafting on membrane flux: 8 hour run.
Instantaneous fouling by the synthetic seawater was determined by the 1-minute experiment. The permeability of unmodified and modified membranes at the beginning of precompaction averaged 117 and 140 L/m2h, respectively. Flux decline during precompaction was comparable for both modified and unmodified membranes as these values averaged 18 and 20 L/m2h, respectively. Steady-state fluxes after precompaction, or just before the beginning of feed water filtration, were nearly 105 L/m2h for unmodified membranes and 119–124 L/m2h for modified membranes. At this time, the synthetic seawater was filtered for one minute to determine the instantaneous fouling.
Flux decline due to instantaneous fouling of the unmodified membrane was nearly 100%
greater than that of the modified membrane, which also always stabilized at 15% higher fluxes than the unmodified membrane, as shown in Figure 5.19. Flux decline during the 15-minute experiment is shown in Figure 5.20. Initial water fluxes for unmodified and
0 20 40 60 80 100 120 140 160
0 200 400 600 800 1000
Flux, L/(m2-hr)
Time, min
Modified Membrane Precompaction Unmodified Membrane Precompaction Modified Membrane Filtration Unmodified Membrane Precompaction
79
modified membranes were 93 and 103 L/m2h, respectively. Flux values at the end of the 15-minute filtration run were 90 and 102 L/m2h for unmodified and modified membranes, respectively. Percentage flux drop measured for unmodified and modified membranes averaged 4.1 and 1.2, respectively.
Similar trends of higher fluxes and lower flux drops for modified membranes were observed for all filtration runs. Flux decline results for the two long-term experiments, 6-hour run and 8-hour run, are presented in Figures 5.21 and 5.22, respectively. Data for the remaining filtration runs, between 15 minutes and 6 hours, is presented in the Appendix A. Long-term filtration runs showed a trend in which flux declined at a steeper rate for unmodified membranes than for modified membranes. The flux decline rate was measured by plotting a normalized percentage flux decline [(final flux – initial flux)/(initial flux)] at the end of a given run against its total filtration time.
Normalized percentage flux declines were then plotted as a function of time for both unmodified and modified membranes, and regressed with the intercept forced through zero (Figure 5.23). This graph shows a higher rate of flux decline, i.e., the slope of the trend line, for unmodified membranes (5.3%/hr) as compared to modified membranes (4.3%/hour). Both unmodified and modified membranes displayed TOC rejections of nearly 48% as shown in Figure 5.24.
80
Figure 5.23: Measure of flux drop with time, i.e., rate of flux drop variation for unmodified and modified membranes.
Figure 5.24: Influence of PEG grafting on membrane TOC rejection: 6 hour run.
y = 5.3231x
R² = 0.9454 y = 4.3014x
R² = 0.9195
05 1015 2025 3035 4045 50
0 2 4 6 8 10
% Flux drop
Filtration period, hours
Unmodified Modified
0 10 20 30 40 50 60 70 80 90 100
0 50 100 150 200 250 300 350
% of Rejection
Time, min Unmodified Modified
81
To evaluate the change in flux due to the addition of hydrophilic layer of PEG for a given sample, several unmodified membrane samples were precompacted and modified only after precompaction (see Figure 5.25). Each sample was then tested after modification and compared to its unmodified membrane sample. Figure 5.25 shows that modification after precompaction led to a 25% increase in flux. This supported that the increase in flux was due to the PEG grafted layer and not to experimental biases.
Figure 5.25: 6-hour run showing the increase in the flux due to modification. Two unmodified membranes were precompacted, after which one of them was modified.
AFM analysis of fouled samples of both modified and unmodified membranes indicated uniform deposition of material on modified membranes. Unmodified membranes experienced severe instantaneous fouling compared to modified membrane, as shown in Figure 5.26. For a given duration of filtration, unmodified membranes had
40 60 80 100 120 140 160
0 100 200 300 400 500 600 700
Time (min)
Flux (L/sqm/hr)
Unmodified Precompaction Unmodified Filtration Precompacted&modified Modified Filtration
82
higher roughness values than those of modified membranes, as shown in Figure 5.27 and Table 5.3, indicating lower fouling susceptibility for the modified membrane. Higher flux data, shown in Figures 5.19 – 5.22, also corroborates this behavior in modified membranes showing the efficiency of the added PEG layer in promoting lower fouling during the filtration operations.
Figure 5.26 A: Isometric view fouled unmodified membrane – 1 min run: Instantaneous fouling.
Figure 5.26 B: Isometric view fouled modified membrane – 1 min run: Instantaneous fouling.
83
Figure 5.27 A: Isometric view fouled unmodified membrane – 8 hour run: Long term fouling.
Figure 5.27 B: Isometric view fouled modified membrane – 8 hour run: Long term fouling.
84
Table 5.3: Roughness and peak count values of fouled membranes.
Membrane Roughness, nm Peak count
Unmodified: Prefiltration 9.04 156
Modified: Prefiltration 9.09 150
Unmodified: 1 min run 31.34 155
Modified: 1 min run 20.54 157
Unmodified: 1 hour run 21.68 164
Modified: 1 hour run 16.59 164
Unmodified: 8 hour run 64.82 162
Modified: 8 hour run 44.82 162