Development of Ruthenium Oxide Modified Polyethersulfone Membranes for Improvement of Antifouling Performance Including Decomposition Kinetic of Polymer

In this study, RuO2-embedded PES membrane was prepared and it was used for protein separation. The antifouling properties of the fabricated composite membranes were also investigated using bovine serum albumin (BSA) as protein solution. The mean roughness increased proportionally by introducing RuO2 particles. The porosity of the composite membranes was higher than that of the pristine PES membrane. On the other hand, composite membranes has smaller average pore size after addition of RuO2 particles. The blending of RuO2 particles to the PES membrane caused to increase the hydrophilicity. The contact angle was measured 76.67 ± 1.21°, 73.23 ± 0.84°, 70.28 ± 0.77°, and 67.13 ± 0.80° for pristine PES, PES/RuO2 0.50 wt%., PES/RuO2 0.75 wt%., and PES/RuO2 1.00 wt%, respectively. The pure water flux of the membranes decreased from 439.7 to 379.3 L/m2/h for the pristine PES and PES/RuO2 1.00 wt%. The pore size was calculated as 16.47 nm for the pristine PES and pore size decreased up to 6.05 nm when RuO2 particles increased up to 1.00 wt%. BSA fluxes were 84.1 ± 2.1, 86.3 ± 2.5, and 93.9 ± 3.2 L/m2/h for pristine, PES/RuO2 0.50 wt%, and PES/RuO2 0.75 wt% membranes, respectively. PES/RuO2 1.00 wt%. membrane supplied the lowest BSA flux (73.6 ± 3.1 L/m2/h). BSA rejection efficiencies increased from 45.5 ± 1.8% to 92.6 ± 1.5% when blended RuO2 particles increased from 0 to 1.00 wt%. The results depicted that Rir values decreased while Rr values increased after the blending of RuO2. The thermal studies of the PES/RuO2 membranes were also performed by DTA/TG. The Activation Energy (Ea) values of the PES/RuO2 membranes were found to be 57.67-641.34 kJ/mol for Flynn-Wall-Ozawa (FWO) and 55.13–659.10 kJ/mol for Kissenger-Akahira-Sunose (KAS).


Introduction
Industrial wastewater often contains hazardous macromolecules and environmentally unfriendly contaminants. For this reason, permanent damages arising from the toxicity and carcinogenicity of many contaminants in wastewater are significantly taken into consideration. Generally, three approaches are widely used to treat wastewater including some treatments such as physical, chemical, and biological [1]. Membranes are widely used in different industries such as medical, pharmaceutical, chemical, food, and textile. In these days, membrane processes are used extensively to obtain water with high quality for domestic or industrial demands, purification and reuse of wastewater, and removal or recovery of toxic or valuable components from different industrial wastewaters [2]. Pressure-driven membrane process has gained great interest in several industrial activities such as textile, chemical, water softening, pharmaceutical, etc. [3][4].
Recent studies have been targeted on choosing a suitable polymer and adding an productive contribution to the structure of the membrane to evolve the properties of that membranes [4][5][6]. Polyethersulfone (PES) is well-known as one of the most practicable polymers in laboratory scale and field operations for the producing of polymeric membranes. The high thermal steadiness, well mechanical properties, and perfect heat-aging resistance of the PES polymer have motivated researchers to use it in several membrane processes [7]. However, the natural hydrophobicity of PES limits water flux permeability and the antifouling properties [8,9]. To handle these problems, numerous metal oxide nano materials such as silver [5], cobalt [6], iron oxide [7], alumina [10], titanium oxide [11] have attracted great attention as effective additives [10]. Ruthenium (IV) oxide (RuO 2 ) is an inorganic compound that is applied well done in different areas in chemistry such as in electrocatalysis and heterogeneous catalysis. Like most other platinum group metals, ruthenium is quite resistant to chemical reaction [12]. RuO 2 has some superior properties such as indissoluble structure in nature, adsorption capacity in pH range (3)(4)(5)(6)(7)(8)(9)(10)(11), and no explorable dissolution in the presence of 0.1 M HCl [13].
According to the best of our knowledge, there is no research paper has been published with RuO 2 blended PES membrane. Accordingly, we focused to produce a new mixed matrix membrane blended with RuO 2 , to investigate the separation ability of the fabricated membranes against protein, and to measure the anti-fouling properties of membranes against BSA. Moreover, thermal analyses and decomposition kinetics of pristine PES and PES/RuO 2 blended composite membranes were also investigated.

Preparation of Ruthenium Oxide (RuO 2 )-Blended PES Membranes
Phase inversion method has been used for preparing of pristine and ruthenium oxide-blended PES membranes, where NPM was used as the solvent. The detailed analysis can be found in elsewhere [14]. Table 1 shows that the composition of casting solution.

Characterization of the Prepared Composite Membranes
The gravimetric method is used for calculate the overall porosity (ε) and the mean pore radius (r p ) is calculated by the Guerout-Elford-Ferry equation [14].
The surface morphology of the PES membranes was investigated by scanning electron microscopy (SEM, FEI, Quanta 650 Field Emission SEM). Atomic force microscopy (AFM) was used to characterize the surface roughness properties of the prepared membranes. The AFM analyses were performed on an AFM microscope (Park System XE-100 SPM). The square shape pieces of the membranes (approximately 1 cm 2 ) were cut and scanned in the tapping mode in the air. The standard deviation of all the height values within the given area is used for calculating of the average of the surface roughness (R a ) value. Three different points were tested and the average values of R a were given as result. The surface hydrophilicity of the fabricated membranes was carried out by water contact angle measurement (Attension Theta Lite). The thermogravimetry/differential thermal analysis (TG/DTA) curves were obtained via Seiko II TG/DTA 7200 Instrument. The system was set at 10 °C/min heating and 100 mL/min N 2 gas flow rate. TG curves were obtained at different heating rates (5, 10, and 15 o C/min). Kinetic analysis of the composite membranes were performed using these curves. Platinum crucible was used as a pan. Al 2 O 3 was used as reference and the samples were weighed in the range of 5 to 10 mg. Operation of some characterization items can be referred to the literature [15][16][17][18][19][20][21].

Protein Rejection Studies of the Membranes
A dead-end filtration system (Sterlitech, HP4750 Stirred Cell) was used to test the efficiency of the prepared membranes. The system was filled with the 100 mg/L of BSA solution which was prepared in Phosphate Buffer Solution (PBS, 50 mM, pH 7.4 ± 0.1) to prevent denaturation of protein. BSA was filtrated at 5 bar for 120 min. The protein concentration of the permeate (C p ) and the feed (C f ) were tested by the Lowry method [22]. Total fouling ratio (R t ), reversible fouling ratio (R r ), and irreversible fouling ratio (R ir ) were calculated by Eqs. (1)-(3), respectively [23]. Moreover, flux recovery ratio (FRR) was calculated by Eq. (4).

Physical Characterization of RuO 2 -Blended PES Membranes
The surface SEM micrographs of the fabricated membranes are shown in Fig. 1. A homogenous dispersion without significant agglomeration was obtained for the composite membranes containing lower RuO 2 particles. On the other hand, agglomeration of the RuO 2 particles was observed on the surface of PES/RuO 2 1.0 wt% composite membrane. The energy dispersive spectrometry (EDS) mapping depicted that the RuO 2 powders were well-spread inside of the structure of PES membranes (Fig. 2). AFM analyses of the fabricated composite membranes are shown in Fig. 3. Pristine membrane showed lower mean roughness compared to RuO 2 blended membranes ( Table 2). The mean roughness increased proportionally by introducing RuO 2 particles. This outcome can be the result of mixing different RuO 2 ratios in the casting solution. The strong interaction between RuO 2 and NMP was influenced by the addition of RuO 2 to the PES casting solution [24]. This result is consistent with the calculated membrane pore diameter shown in Table 3. The calculated pore size decreased from 16.47 nm to 6.05 nm when blended RuO 2 particles increased from 0 to 1.00 wt%. However, BSA fluxes were 84.1 ± 2.1, 86.3 ± 2.5, and 93.9 ± 3.2 L/m 2 /h for pristine, PES/RuO 2 0.50 wt%, PES/RuO 2 0.75 wt% membranes, respectively. PES/RuO 2 1.00 wt%. membrane supplied the lowest BSA flux (73.6 ± 3.1 L/m 2 /h) (Fig. 4B). The BSA rejection capasity of the membranes was also tested (Fig. 4 C). BSA rejection efficiencies increased from 45.5 ± 1.8% to 92.6 ± 1.5% when blended RuO 2 particles increased from 0 to 1.00 wt%. Both higher hydrophilicity and the smaller pore size of the composite membranes improved BSA rejection.

Antifouling Performance
The most critical flux decrease (80.9%) caused by BSA fouling was observed due to the highest R t value of the pristine PES membrane (Fig. 5 A). However, R t values of the composite membranes decreased 79.2%, 76.1%, and 72.7% for PES/RuO 2 0.50, PES/RuO 2 0.75, and PES/RuO 2 1.00 wt%, respectively. Additionaly, Rir values decreased while Rr values increased after blending of RuO 2 (Fig. 5 A). Furthermore, FRR and R r /R ir ratios of the membranes are shown in Fig. 5B. The irreversible membrane fouling is displaced into reversible membrane fouling in the filtration process due to increase of R r /R ir ratio, which caused from incorporation of The mean pore size and overall porosity of the membranes are presented in Table 3. The results showed that the porosity of the composite membranes was higher than that of the pristine PES membrane. It could be elucidated with some parameters such as thermodynamic and kinetic during the phase inversion process were affected by ruthenium and oxygen groups and it impressed the structure of the composite membranes. Pristine membranes have higher average pore size than composite membranes. When the dope solution becomes much viscous, it limits the expansion of pores. For this reason, average pore sizes of RuO 2 blended membranes were reduced. The blending of RuO 2 particles to the PES membrane increased the hydrophilicity of the pristine membrane (from 76.67° to 67.13°). This can be explain with increasing of water molecules affinity by affected from the oxygen groups which are polar functional.

Pure Water Flux
The pure water flux of the fabricated membranes decreased from the pristine PES to PES/RuO 2 1.00 wt% (Fig. 4 A). A decrease in the flux from 439.7 ± 20.5 L/m 2 /h (for pristine membrane) to 379.3 ± 15.1 L/m 2 /h (for PES/RuO 2 1.00 wt% membrane) could be explained as the agglomeration of RuO 2 particles in the structure of composite membranes and balking the pores which caused smaller pore-sized membrane.

Thermal Analyses of PES/RuO 2 Membranes
The temperature ranges of degradation processes, DTA peaks, weight losses (%), and the leaving moieties in degradation processes for each step are presented in Table 4.
The TG curve of the PES membrane shows that the degradation occurs in three steps. Dehydration was monitored in the temperature range of 25 to 100 o C, and 2.78% mass loss is attributed to the water molecule leaving the PES membrane. The first degradation step was observed between 62 o C and 208 o C, with a weight loss of 7.11% (5.59%, theoretically). 54.35 moles of -HSO 3 groups left the structure at the end of the first step. The second degradation step is observed between 208 o C and 573 o C, and at this step, the 259 moles -C 6 H 4 groups and the 95.83 moles -HSO 3 groups leave the structure corresponding to the weight loss 47.68% (47.32%, theoretically). The third step was observed between 573 o C and 775 o C. In this step, leaving RuO 2 particles [25]. RuO 2 particles prevented irreversible BSA adsorption into composite membranes in order to form a hydration layer close the membrane surface [26]. FRR values increased from 39.6% for pristine PES membrane to 85.9% for RuO 2 1.00 wt% membrane.

Kinetic Analysis
The E a values of degradation steps for each α (conversion degree) was calculated via FWO and KAS methods. It was recommended by the ICTAC kinetic committee to set E a values between 0.05 and 0.95 in increments of 0.05. The equations are presented in Eqs. 5  The DTG graph shows the five-step degradation. It is seen that the second and third degradation steps overlap in the DTG curve. Since it is difficult to evaluate, the fourth and fifth degradation steps were considered as one step in accordance with the TG curve. The first degradation is observed between 62 o C and 208 o C and the second degradation step is observed between 286 o C and 579 o C. The third degradation takes place in the region between 578 o C and 778 o C. The maximum point of the strongest peak in the DTG graph gives the temperature at which degradation is fastest. The maximum degradation temperature is observed at 518 °C.

Degradation Kinetics of PES/RuO 2 0.75 wt% Membrane
The E a -α graphs for the PES/RuO 2 0.75 wt% membrane are given in Fig. 9(A-C).
In the first stage, the values of the E a decrease rapidly between a = 0.05 and 0.25 and increase slowly between α = 0.25 and 0.95, then rapidly decrease again. The E a values were found to be 57.67 kJ/mol for FWO and 55.15 kJ/mol for KAS. In the second stage, the E a values are increasing rapidly between α = 0.05 and 0.15. Then the E a values tend to increase irregularly with the increase in the value of α . The average E a values were calculated as 163.21 kJ/mol and 159.72 kJ/mol. In the third stage, E a values increase regularly between a = 0.05 and 0.65 and then they decrease irregularly with increasing a value. The average E a values were found to be 377.28 kJ/mol and 381.53 kJ/mol. The irregular increase and decrease observed in Ea values indicate that there is multistep degradation in all steps. The Ea values of the PES/RuO 2 0.75 wt% membrane are given in Table 5.
where, A is the pre-exponential factor, β is heating rate, T is absolute temperature, g(α) is an unknown function of conversion, and R is gas constant.
According to these equations, ln β vs. 1/T (for FWO) and ln β/T 2 vs. 1/T (for KAS) graphs are straight lines. The E a values are calculated from the slopes of these graphs. The E a values of all decomposition steps are given in Figs. 5, 6 and 7.

Dedradation Kinetics of Pristine PES Membrane
The E a -α graphs for the degradation stages of the PES membrane are presented in Fig. 7 A-C. It was detected that in the first stage the values of E a with α = 0.05 to 1.00 decreased very sharply in the range from 0.05 to 1.00. The E a values calculated via the FWO and KAS equations for this stage are 66.15 kJ/mol and 59.60 kJ/mol, respectively. In the second stage for α between 0.05 and 4.00 the E a values are increase sharply, and then rapidly decrease. The average E a 's were calculated to be 796.11 kJ/mol and 829.23 kJ/ mol, respectively. In the third step for α between 0.05 and 0.60 the E a values slowly decrease and beyond this point, it was observed to increase sharply. The average E a values were found to be 158.53 kJ/mol and 150.70 kJ/mol, respectively. Irregular increases and decreases in E a values due to the increase in alpha values show that all degradation steps have multi-step degradation kinetics. The E a values of the PES membrane are given in Table 5.

Degradation Kinetics of PES/RuO 2 0.50 wt% Membrane
The E a -α graphs for the PES/RuO 2 0.50 wt% membrane are given in Fig. 8(A-C). In the first stage, E a values tend to decrease and increase irregularly in the range a = 0.05 and 0.4. Then these values gradually decrease. According Scheme 1 The proposed degradation mechanism for PES membrane

Conclusions
In this study, RuO 2 -embedded PES membranes were fabricated by phase-inverion method and used for separation of protein. BSA was used to perform of the antifouling characteristing of the prepared membranes. All membranes were characterized by AFM, SEM-EDS, mean pore size, contact angle and porosity. The porosity of the composite membranes increased while the pore size decreased. Moreover, hydrophilicity of the composite membranes also increased after RuO 2 was spreaded into the PES membrane. The thermal analysis studies of the PES/RuO 2 membranes were also carried out via DTA/TG combined system. The E a values of the PES/RuO 2 membranes were calculated to be 57.67-641.34 kJ/mol for FWO and 55.13-659.10 kJ/mol for KAS. Furthermore, the composite membranes indicated good antifouling performance during BSA filtration. RuO 2 -blended PES membrane can be used to decrease membrane fouling.

Degradation Kinetic of PES/RuO 2 1.00 wt% Membrane
The E a -α graphs for the PES/RuO 2 1.00 wt% membrane are given in Fig. 10(A-C).
In the first step, the E a values decreased between a = 0.05 and 0.15 and increased up to a = 0.4. Later, the E a values kept mostly constant with the increase in a values. According to both methods (FWO and KAS), the E a values were calculated 39.43 kJ/mol and 35.43 kJ/mol. In the second stage, The E a values increased gradually between a = 0.05 and 0.90 and then tended to decrease. The E a values were found to be 150.78 kJ/mol and 144.03 kJ/mol. In the third stage, The E a values are almost constant between a = 0.05 and 0.60. it tended to increase after this point. The E a values were found to be 147.35 kJ/mol and 139.35 kJ/mol. The irregular distribution of E a values shows the tendency of multistep degradation at all steps. The E a values of the PES/RuO 2 1.00 wt% membrane are showed in Table 5. and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript.

Conflict of Interest
The authors whose names are listed immediately below certify that they have NO affiliations with or involvement in any organization or entity with any financial interest (such as honoraria;