Enhancement of anti-fouling properties and

filtration performance of poly(ethersulfone)

ultrafiltration membranes by incorporation of

nanoporous titania nanoparticles

Ze-Xian Low†, Zhouyou Wang†, Sookwan Leong†, Amir Razmjou†, Ludovic F. Dumée§,

Xiwang Zhang†, Huanting Wang†,*

†Department of Chemical Engineering, Monash University, Clayton VIC 3800, Australia

§Institute for Frontier Materials, Deakin University, Pigdons Road, 3216 Waurn Ponds,

Victoria, Australia

Nanoporous titania nanoparticles (NTNs) were synthesized and used as an additive at a low

concentration of 0.1 to 1 wt% in the fabrication of poly(ethersulfone) (PES) ultrafiltration

membranes via non-solvent induced phase separation. The structure and properties of

nanoparticles were characterized using nitrogen sorption, X-ray diffraction, scanning electron

microscopy and transmission electron microscopy. The NTNs have a size distribution with

particle size mainly less than 100 nm and have a BET surface area of about 100 m2 g-1. The

modified membranes were fabricated and investigated in terms of their pure water flux, solute

rejection, and fouling resistance. The water permeability and molecular weight cut-offs

(MWCO) of membranes were determined under constant pressure filtration in dead-end

mode at 100 kPa. Membrane fouling resistance was characterized under constant flux

operation using bovine serum albumin (BSA) as a model foulant. The membranes were

characterized in terms of morphology, porosity, pore size distribution, energy-dispersive X-

ray spectroscopy (EDX), contact angle goniometry, surface free energy and viscosity of the

dope solution. Overall, the modified membrane showed increased wettability and reduced

surface free energy and pore size. The modified UF membrane with 0.5 wt% NTN loading

exhibited improved fouling resistance (fouling rate of 0.58 kPa/ min compared to 0.70 kPa/

min of control membrane) with ~80% water flux recovery. The same membrane showed ~

20% increase in water flux and improvement in MWCO and narrower pore size distribution.

1. Introduction

Ultrafiltration (UF) has been widely used to remove suspended solids, proteins, bacteria

and macromolecules in various industrial processes including food, dairy, pharmaceutical,

biotechnological industries and water/ wastewater treatment.1, 2 As a consequence of the

growing demand, efforts have been focused on improving UF process performance, including

feed pre-treatment, membrane materials and module design, and process optimization.3 In

many cases the key to improved performance processes still relies on the structure and

properties of the membrane materials.3

The fast development of techniques and methods for producing nanostructured materials

such as nanoparticles or nanotubes has led to breakthroughs in membrane preparation for

enhanced solute rejection and solvent permeation.4 In particular, attempts to develop organic-

inorganic hybrid nanocomposite membranes in which nanoparticles are used as modifiers of

polymeric membranes have received much attention.5-31 The production of composite

membranes by incorporating nanomaterials has resulted in enhanced chemical, mechanical

and thermal stability, as well as increased fouling resistance and membrane filtration

performance compared to plain polymeric membrane. The improvements are attributed to the

changes to the microstructural features (porous structure, and pore size and pore size

distribution) of the membranes arising from localized interactions and interfaces between the

nanoparticles and the polymer chains during the membrane formation, and altered membrane

surface roughness and chemistry.4

The incorporation of nanoparticles (such as metal oxides) into the membrane structures is

typically performed by either surface modification route5-8 or casting solution blending

during the thin film formation.9-31 Surface modification of UF membranes can be achieved

via surface coating, adsorption or grafting, resulting in enhanced membrane surface

wettability, fouling resistance and other surface properties such as surface roughness.8

However, the substantial post-treatment processes are required in surface modification

techniques. By contrast, blending the main polymer constituent with additives is simple and

straightforward without involving expensive post-processing.32

Metal oxide nanoparticles used for the modification of polymeric membranes include

alumina (Al2O3),21, 28 silica (SiO2),

22, 23 zirconia (ZrO2),24, 25 zinc oxide (ZnO),26, 27 and titania

(TiO2).22, 29-31 More recently, various porous (0- and 1-dimensional) or layered (2-

dimensional) nanomaterials, from conventional mesoporous silica31, 33 and zeolite,34, 35 to

zeolitic imidazolate framework (ZIF),36, 37 carbon nanotube (CNT),38, 39 hollow carbon

sphere,40 and clay41, 42 have been studied for their potential as additives for polymeric

membranes. The use of porous nanomaterials could lead to interesting results different from

those of conventional dense inorganic materials. The incorporation of porous particles may

improve the filler-polymer contact.43 The polymer chains of diameter about ~1 nm44, 45 may

be able to penetrate into the pores of the particles, hence improving the interface of additive

and polymer matrix.43 In addition, the porous nature of the fillers may give them high affinity

to water, which could facilitate water transport through the membrane.31

TiO2 nanoparticles have been widely studied as membrane additive due to their natural

antibacterial and photocatalytic properties.46, 47 Membrane incorporated with TiO2

nanoparticles exhibited improved performance in terms of fouling resistance48, 49 and

photocatalytic properties.50 In most of these studies, TiO2 nanoparticle concentrations above

1 wt% (relative to polymer solution) were used in terms of harnessing their photocatalytic /

antibacterial activity.11-18 However, the combination with UV irradiation may introduce the

risk of polymeric membrane degradation, and UV irradiation is not practicable for hollow

fiber or spiral wound membrane modules. High loadings of nanoparticles may be

unfavourable considering the aggregation tendency of particles; and membrane pores could

be plugged by the particle aggregates during non-solvent induced phase separation (NIPS),

and this could adversely affect the membrane performance. Recently, Sotto et al. used low

TiO2 concentrations (below 0.4 wt%) in the membrane fabrication process and improved the

flux and fouling resistance of the membrane.49 At lower nanoparticle loadings, the tendency

of aggregation can be reduced.49 However, the membrane performance at a TiO2 nanoparticle

loading of below 1 wt% is still not entirely understood. The study on the low loading of

nanoparticles is scarce, and in the case of TiO2 nanoparticles only a few paper have been

published.49, 51, 52 Therefore, there is a pressing need for an in-depth systematic study into the

effects of titania at a concentration of below 1 wt% on UF membranes. The understanding of

such effects is fundamentally important for the development of high-performance of titania-

polymer composite membranes for practical applications. In addition, the concern for

commercial implementation of TiO2-polymer membranes arises from potential health and

environmental risks of TiO2 nanoparticles.53-59 The incorporation of TiO2 nanoparticles into

membrane may pose secondary contamination to the treated water caused by leaching of

TiO2 nanoparticles. As a result, more studies into membrane stability are required to ensure

the TiO2-based polymeric membrane can be safely used without posing environmental and

health risks.

In this paper, we report on a simple and effective method for preparing nanoporous titania

nanoparticles (NTN) as a modifier for PES UF membrane. NTN was produced via a modified

hydrothermal reaction, followed by calcination to improve its compatibility with PES. A

systematic study of the concentration effect of NTN in concentration range below 1 wt%

relative to dope solution on PES UF membrane was carried out to determine the effect of

NTN on the membrane performance. The amount of TiO2 in the membrane-treated water was

determined using inductively coupled plasma mass spectrometry (ICP-MS) to investigate the

risk of secondary contamination.

2. Materials and Methods

2.1 Materials

Polyvinylpyrrolidone (PVP; 40 kDa), polyethylene glycol (PEG; with MW of 35, 100 and

200 kDa), N-methyl-2-pyrrolidone (NMP), and bovine serum albumin (BSA; 66 kDa) were

purchased from Sigma-Aldrich, Australia. TiO2, (Degussa P25, about 20 nm) was purchased

from Evonik Degussa Chemicals (Germany). Sodium hydroxide (NaOH) pellets were

purchased from Merck Millipore, Australia. Hydrochloric acid was purchased from Ajax

Finechem Pty Ltd, Australia. Poly(ethersulfone) (PES; Ultrason E6020P, 51 kDa) was

purchased from BASF Co. Ltd., Germany. The water used for the experiments was purified

with a water purification system (Milli-Q integral water purification system, Merck Millipore

Australia) with a resistivity of 18.2 MΩ/cm. Distilled water was obtained from a laboratory

water distillation still (Labglass Aqua III).

2.2 Synthesis of Nanoporous Titania Nanoparticle (NTN)

The synthesis of NTN followed a modified process reported elsewhere.60 1.5 g of P25 was

added into 100 ml of 10 M NaOH solution. The suspension was mixed by mechanical stirring

and ultrasonication; and added into a 150 ml autoclave. The autoclave was sealed and heated

at 120 °C for 24 h, and then cooled to room temperature naturally. The suspension was then

washed with excessive 0.1M of hydrochloric acid and then Milli-Q water, the solid product

was recovered by repeated centrifugation in water. The recovered product was calcined at

450 °C at a ramp rate of 5 °C per min and holding time of 3 h, and white-coloured NTN were

obtained.

2.3 Preparation of the membranes

The PES and PES-NTN nanocomposite membranes were prepared via NIPS at room

temperature. Three different concentrations of NTN (0.1 wt%, 0.5 wt% and 1.0 wt%) were

used in this study. For instance, to prepare the PES casting solution with 1 wt% NTN, 0.125 g

of PVP powder and 0.250 g of NTN were first added into 20.5 g of NMP and mixed via

mechanical stirring and ultrasonication for 1 h to ensure good dispersion. 4.125 g of PES was

then added, and the solution was stirred overnight or until the PES was completely dissolved.

The solution was left to degas for 8 h before use. Table 1 shows the compositions of the

casting solutions.

Table 1. The compositions of the casting solutions for PES and PES-NTN nanocomposite

membranes.

Membrane PES (wt%) PVP (wt%) NMP (wt%) TiO2 (wt%)

Control 16.50 0.50 83.00 0.00

PES-NTN0.1 16.50 0.50 82.90 0.10

PES-NTN0.5 16.50 0.50 82.50 0.50

PES-NTN1.0 16.50 0.50 82.00 1.00

The membranes were cast on a glass plate using an adjustable micrometer film applicator

(stainless steel blade at a gap of 150 µm, Elcometer, USA) at room temperature. The phase

inversion step was carried out by immersing the membrane in a coagulation bath of distilled

water for 24 h. The membranes were then removed from the bath, rinsed thoroughly with

double-deionized water (DDI) water and stored in fresh DDI water for later use. 0.1 wt%, 0.5

wt%, and 1.0 wt% PES-NTN nanocomposite membranes were denoted as PES-NTN0.1,

PES-NTN0.5, and PES-NTN1.0, respectively.

2.4 Characterization of the Nanoporous Titania Nanoparticle (NTN)

The crystal structure of NTN was examined using powder X-ray diffraction (15 mA and 40

kV) at a scan rate of 2°/min and a step size of 0.02° (PXRD; Rigaku MiniFlex 600, Cu Kα

radiation, Japan). High resolution transmission electron microscopy (HRTEM) images and

selected-area electron diffraction (SAED) were obtained using a transmission electron

microscope (TEM; FEI Tecnai G2 T20 Twin TEM, USA). The dried NTN was observed by

scanning electron microscopy (SEM; FEI NOVA NanoSEM 450, USA). To measure

Brunauer, Emmett and Teller (BET) surface area of NTN, nitrogen sorption analysis was

carried out on a physisorption analyzer (Micromeritics ASAP 2020, USA) at liquid nitrogen

temperature, and the sample was degassed at 100°C for 12 h before the measurement.

Mesopore size distribution was calculated using the Barrett-Joyner-Halenda (BJH) method.

Particle size distribution was obtained by analyzing the FESEM images of the membrane top

surface using the Image J (http://imagej.nih.gov/ij/). The image was first filtered and

thresholded at the predetermined level T, which corresponds to the grey value between 0 and

255 to account for all the membrane pores. The image was then converted to binary image

and analyzed to determine the pore size distribution.

2.5 Membrane characterization

2.5.1 Viscosity of casting solution

The viscosity of the solution was measured at 25 °C using a Haake MARS III rotational

rheometer coupled with Haake C35P refrigerated bath with a Phoenix II controller (Thermo

Scientific, USA). Shear rate in the range of 10 to 50 s-1 was used. An average of multiple

measurements obtained from the program was reported for each sample.

2.5.2 FESEM and EDX analysis

Membrane surface and cross-sectional images were obtained using a field emission

scanning electron microscope (FESEM; FEI Magellan 400 and FEI Nova NanoSEM 450,

USA). Membrane samples were prepared by drying at room temperature and sputter-coating

with a 0.5 nm thickness of Pt (208 HR sputter coater, Cressington, UK). For imaging the

cross section, the membrane was fractured in liquid nitrogen to retain the membrane

structure. Elemental analysis of the membrane samples was conducted by energy-dispersive

X-ray (EDX) spectroscopy in a Nova NanoSEM 450 (Quantax 400 X-ray analysis system,

Bruker, USA).

2.5.3 Contact angle and surface free energy measurement

The water contact angles of the membranes surface were measured using a contact angle

goniometer by the sessile drop technique (OCA15, Dataphysics, Germany). Images were

taken at 1 s intervals for 10 s. An average of minimum 5 measurements was reported. To

calculate the surface free energy of membranes, Owens, Wendt, Rabel and Kaelble (OWRK)

method was applied.61 In this method, contact angles against at least two liquids with known

parameters are measured. The surface free energy of the membrane was calculated using the

OCA15 software package (Dataphysics, Germany). Table 2 shows the parameters of the three

liquids used in this work.

Table 2. The parameters of OWRK method.61

Liquid SFT (mN/ m) σdisperse (mN/ m) Σpolar (mN/ m)

Milli-Q water 72.10 19.90 52.20

Glycerol 63.40 37.40 26.00

Poly(ethylene) glycol (200 Da) 43.50 29.90 13.60

2.5.4 Molecular weight cut-off (MWCO)

MWCO was characterized by measuring the rejection of poly(ethylene glycol) (PEG, 35,

100, and 200 kDa). A total organic carbon analyzer (TOC-LCSH/CSN with an auto-sampler

ASI-L, Shimadzu, Japan) was used to measure the amount of organic carbon in permeate to

determine the amount of PEG.

2.5.5 Pure water flux measurement

Water flux and flux recovery were measured in a dead-end cell (HP4750 Stirred Cell,

Sterlitech, USA). A schematic diagram of the bench-scale flux test setup is shown in Figure

1. The DDI water flux test was performed at 100 kPa and room temperature with an effective

membrane area of 1.4 × 10-3 m2. The membrane was pre-compacted at 150 kPa for at least 30

min until constant flux was reached. The initial water flux was determined under a constant

pressure of 100 kPa, as was PEG rejection when determining the rejection characteristics of

the membrane.

2.5.6 Membrane fouling resistance

Figure 1. Schematic diagram of a bench-scale flux test system. Constant pressure operation

was carried out without a pump (route A). In constant flux operation (route B), a pump was

used to control the permeate flux. Any pressure changes due to membrane fouling were

measured by the two pressure transducers.

Constant flux fouling test was carried out to determine the fouling resistance of the

membranes. The membranes were fouled by filtration of BSA (0.5 wt%, pH 7) at a constant

flux of 60 l.m-2.h-1 (LMH) for 50 min using a peristaltic pump (L/S Digital Drive, L/S Easy-

Load 3 pump head, peroxide-cured silicone tubing, L/S 13, Masterflex, USA) and fixed

nitrogen gas feed pressure. The transmembrane pressure (TMP) was measured using two

pressure transducers. The fouled membranes then underwent an in-place physical and

chemical cleaning cycle. For physical cleaning, the membranes in the cell were rinsed with

DDI water by half-filling the cell (150 ml), followed by pouring for 2 times. 300 mL of DDI

water was then added to the cell and stirred for 15 min. After that, the membrane was rinsed

for another 2 times. For chemical cleaning, 100 ml of NaOH solution (0.2 wt%) was added to

the cell and stirred for 25 min. After that, the membrane was rinsed twice to wash off the

solution. After every cleaning step, the water flux was measured at constant pressure of 100

kPa. Each membrane went through 3 fouling cycles at a constant flux and 3 cleaning cycles.

To evaluate the fouling performance of membranes, flux recovery (FR) of membranes was

calculated as follows:

𝐹𝑅(%) =𝐽𝐴𝐹𝐽𝐵𝐹

× 100 (1)

where JBF and JAF are the pure water flux of the membrane before and after the fouling and

cleaning, respectively. Fouling behaviour can be investigated by estimation of resistance of

membranes as shown below:

1) Total resistance (Rt)

𝑅𝑡 = 𝑅𝑚 + 𝑅𝑟 + 𝑅𝑖𝑟 (2)

where Rm is the intrinsic membrane resistance, Rir is the irreversible resistance, and Rr is the

reversible resistance.

2) Intrinsic membrane resistance (Rm)

𝑅𝑚 =𝑇𝑀𝑃

𝜇 × 𝐽𝐵𝐹

(3)

where TMP is the transmembrane pressure, and µ is the permeate viscosity.

3) Irreversible resistance (Rir)

𝑅𝑖𝑟 =𝑇𝑀𝑃

𝜇 × 𝐽𝐴𝐹− 𝑅𝑚

(4)

where JAF is the water flux at 100 kPa after physical and chemical cleaning.

4) Reversible resistance (Rr)

𝑅𝑟 =𝑇𝑀𝑃

𝜇 × 𝐽𝐹− 𝑅𝑚 − 𝑅𝑖𝑟

(5)

where JF is the BSA filtration flux which was set at 60 LMH in our experiment. TMP was

taken after 50 min BSA filtration.

2.5.7 Permeate analysis

The amount of TiO2 in the permeate was analyzed using a standard method (CM050A and

CM050C)62 used by Australian Laboratory Services (ALS; http://www.alsglobal.com/). 50 ml

of the permeate was collected for each membrane in the first 10 min. The sample was first

acid-digested and made up to volume with deionized water, and then analyzed via inductively

coupled plasma mass spectrometry (ICP-MS).

3. Results and Discussion

3.1 Structure and morphology of the nanoporous titania nanoparticles (NTN)

NTN was formed from calcination of TiO2 which form a nanoporous structure with pore

size ~20 nm (Figure 2a). Figure 2b shows the HRTEM image of NTN. The lattice spacing of

0.369 nm corresponds to the (101) plane of TiO2 anatase. The SAED ring patterns (Figure 2c)

and XRD (Figure 2d) indicate that the NTN consisted of mostly anatase phase and small

portion of rutile phase, which was similar to commercial P25. Figure 2e shows the SEM

image of NTN. The sizes of NTNs were mostly less than 100 nm (Figure 2f). NTN had a type

IV isotherm with a type H1 hysteresis loop (Figure 2g) and a BET surface area of 105.18 ±

0.34 m2 g-1 compared to 51.23 ± 0.18 m2 g-1 of P25. The higher surface area of NTN may

promote better compatibility with PES due to enhanced interaction with PES chains,

potentially between hydroxyl groups on the surface of NTN and aryl ether oxygen atom43, 63

or sulfone oxygen atom in PES.64, 65 Based on BJH model (Figure 2h), NTN had a pore size

distribution around 16- 31 nm. The pores, which may be the packing pores or the nanopores

in NTNs, may be penetrated by PES chains when mixed with the PES polymer which

improve the membrane stability.

Figure 2. (a) TEM image, (b) HRTEM image, (c) SAED, (d) XRD pattern, (e) SEM image,

(f) Particle size distribution based on SEM (inset), (g) nitrogen sorption isotherms, and (h)

pore size distribution of nanoporous titania nanoparticles (NTN).

3.2 Membrane morphology and surface property

Figure 3 shows the FESEM images of the top surface, bottom surface and cross-sections of

the membranes. The top and bottom surfaces of the membranes showed the typical

morphology of PES membranes synthesized via NIPS. Using image processing tool (Image

J) on the SEM images of the membrane top surface, a membrane surface pore size

distribution was obtained (Figure 4). All modified membranes (PES-NTN0.1, PES-NTN0.5

and PES-NTN1.0) show narrow pore size distribution between 2 – 6 nm, compared to 4 – 12

nm of the pristine membrane. That is, the modified membranes have smaller pore radius.

Based on these results, lower molecular weight cut-offs (MWCO) of modified membranes

are expected. The membranes were dried in air and sputter-coated with 0.5 nm of Pt, thus the

pore size obtained via SEM image analysis may appear smaller than the actual pore size of

the membrane. However, since all membranes undergone the same drying and coating step,

the pore size distribution analysis provides useful information about the membrane top

surface, which is not attained via molecular probe rejection experiment. The membrane cross-

sections show typical asymmetrical structure, of finger-like structure on the top of the matrix,

extending to a sponge-like structure in the sub-layer and macrovoidic bottom layer. All NTN-

modified PES membrane show elongated and broader tiny finger-like structures of size 0.4 to

0.7 µm on the top section of the membrane, which extends to larger channels of 3.0 – 7.5 µm.

The sponge-like bottom layer on the other hand, becomes thicker with NTN loading. As the

loading of NTN were increased, more macrovoids were observed in the sub-layer of the

membrane (more pronounced in PES-NTN1.0), and the membrane thickness increased with

NTN loading, from ~70 µm to ~100 µm (Figure 5). PES-NTN1.0 exhibited increased large

macrovoids and membrane thickness. The bottom surfaces of the modified membranes,

especially PES-NTN1.0, showed increased large pores compared to pristine membrane. This

might be due to the increased viscosity of the casting solution which slows down the phase

inversion process, creating large macrovoids and thus, bigger pores in the bottom layer of the

membrane. For PES-NTN0.1 and PES-NTN0.5, the increase in viscosity is less significant

compared to the increase in water diffusion during NIPS due to the presence of TiO2 with

higher hydrophilic nature. The rapid water diffusion promotes instantaneous demixing and

creates more elongated finger-like pores. The effect near the bottom layer of the membrane is

less pronounced which might be due to the migration of TiO2 to the membrane surface. The

porosity and the water uptake of the membrane on the other hand, remained relatively

unchanged after incorporating NTN (Figure 5). This may be due to the nanopores in NTN

being filled with PES chains, thus not contributing to the membrane porosity. These

interactions not only improve the stability of NTN in the membrane, but also restrict the

migration of NTN in the polymer matrix to the filtrate when used, posing secondary

contamination.

Figure 3. SEM images of top, cross-section and bottom layer of PES and NTN-modified

membranes.

Figure 4. Surface pore size distribution of control and PES-NTN membrane based on SEM

top surface analysis.

Figure 5. Porosity, thickness, water uptake and viscosity of control and PES-NTN

membrane.

The NTN in the membrane matrix was confirmed using EDX spectrometry (Figure 6). At

high magnification, the SEM field of view becomes much smaller, and spotting of

nanoparticles from PES matrix can be challenging. Thus, PES-NTN1.0 which has the highest

loading of NTN among all samples was analysed. Particles found on PES-NTN1.0 cross-

section (Figure 6) were confirmed using EDX mapping on the whole region, where Ti

element was marked in red, indicating the successful incorporation of NTN.

Figure 6. EDX elemental mapping of PES- NTN1.0 membrane cross section. The red region

represents the Ti element.

Figure 7 shows the contact angle of various liquid and surface free energy of control and

modified membrane. The wettability of NTN-modified PES membrane surfaces was

characterized by measuring their water contact angles. The water contact angle of the

membranes decreased by around ~15% when NTN were added up to 1 wt%, indicating that

the surface of the modified membranes was slightly more hydrophilic compared to the native

PES membranes. The increase in wettability might be attributed to the migration of NTN

during NIPS. The polyethylene glycol and glycerol contact angles, on the other hand

remained relatively unaffected. Based on the contact angles of the 3 selected liquids, the

surface free energy was obtained via OWRK method.61 The surface free energy decreased

with increasing loading of NTN from 42.9 mN m-1 for pristine membrane to 40.2 mN m-1 for

PES-NTN1.0. The decrease in contact angle and surface free energy may improve the fouling

resistance of the membrane by reducing the propensity of membrane to interact with foulant.

Figure 7. Contact angle of various liquid and surface free energy of control and NTN-

modified membrane.

3.3 Membrane Performance

3.3.1 Pure water flux

Figure 8 shows the average pure water fluxes at 100 kPa for control and NTN-modified

membrane. In general, as the loading increases, the water flux increases. PES-NTN0.5

showed the highest pure water flux of 446 LMH which may be due to increased wettability,

reduced tortuosity and increase surface pore density. When NTN was increased to 1 wt%, the

improvement over the flux was reduced and flux similar to control membrane was attained.

This may be due to the undesirable effect caused by increased NTN agglomeration or pore

plugging. Loading beyond 1 wt% as of many reported studies may produce higher flux

compared to pristine membrane. However, the flux improvement is more likely due to the

increased surface pore size which reduces solute rejection, i.e. flux increase with the loss of

solution rejection. In our case, the modified membranes exhibited smaller pore size (Figure 9)

and a narrower pore size distribution in smaller pore range (Figure 4) compared to pristine

membrane. The abundance of pore in lower range of pore size distribution could be due to the

escape of small NTN aggregates during NIPS, while large NTN aggregates remained trapped

within the polymer matrix due to restricted mobility and entanglement by PES chains.

Figure 8. Pure water flux and rejection of PEG (100 kDa) of control and NTN-modified

membrane.

3.3.2 Molecular weight cut-off (MWCO)

The rejection of PEG with molecular weight of 100 kDa is shown in Figure 8. The NTN-

modified membranes showed higher rejection of 100 kDa, with PES-NTN0.1 having the

highest rejection (92.8%). This could be due to two reasons — (1) increased phase exchange

rate (demixing rate) near the top surface as NTN have higher affinity to water than PES; (2)

migration of small NTN (polymer-poor phase), leaving behind more pores of their own size

on membrane surface. The final pore sizes created by such phenomenon may be smaller after

completion of NIPS. Molecular weight cut-off curves for the control and NTN-blended PES

membranes are shown in Figure 9. The results show that the MWCO (90% PEG rejection) of

the membranes shifted below 100 kDa (corresponds to pore size of ~16.5 nm66) as NTN

loading was increased from 0.1 wt% to 1.0 wt%. Also, the rejections of PEG with molecular

weight of 35 kDa were increased for all modified membrane, further confirming the reduced

surface pore size. Compared to membrane pore size based on SEM analysis, the pores are

larger due to — (1) shrinkage of pores as the membrane was dried; (2) Existence of larger

pores or defects that decrease the solute rejection.

Figure 9. Molecular weights cut-off (MWCO) for control and NTN-modified membrane.

3.3.3 Fouling Resistance

The anti-fouling property of the membranes was investigated under constant flux operation.

During the constant flux fouling test, transmembrane pressure (TMP) increases with

increasing fouling. A membrane with higher fouling resistance will show slower increase in

TMP. The membranes were pre-compacted before the fouling cycle to rule out an increase in

TMP due to morphological change of the membrane. Figure 10 shows the change in TMP

during 50 min BSA filtration at constant flux of 60 LMH. The actual TMP after 3 fouling

cycles are shown in Table 3. From Figure 10a, all modified membranes show lower TMP

increases, with PES-T0.5 showing the lowest fouling rate of 0.58 kPa/ min (0.70 kPa/ min for

control membrane), followed by PES-NTN0.1 and PES-NTN1.0. This is contributed by the

combined effect of reduced contact angle, free surface energy and surface pore size of the

modified membrane. After the first fouling and cleaning cycle, the change in TMP remained

almost the same for both the control and modified membranes, which indicate that the

modified membrane loses its anti-fouling properties after the cleaning process. We speculate

that NTN underwent the following reaction in NaOH solution during chemical cleaning

process:67

𝑇𝑖𝑂2(𝑠) + (4 − 𝑛)𝐻+ ⇄ 𝑇𝑖(𝑂𝐻)𝑛(4−𝑛)+ + (2 − 𝑛)𝐻2𝑂 (6)

where n refers to the state of Ti(IV) ion hydrolysis. The cleaned membrane therefore loses its

anti-fouling properties as TiO2 was converted to titanium hydroxide. For NTN-modified

membrane, other chemical cleaning agents besides NaOH solution or low concentration of

NaOH solution will need to be used to retain the effectiveness of the modified membrane in

fouling resistance.

Figure 10. Transmembrane pressure (TMP) change with time (0.5 wt% BSA, pH 7) for the

control PES and NTN-modified membranes at constant flux of 60 LMH for 50 min. (a) Cycle

1, (b) Cycle 2, and (c) Cycle 3.

Table 3 shows the flux recovery of the control and modified membrane after 3 cycles of

fouling and cleaning. All membranes showed BSA rejection of 95% and above during the

fouling tests and flux recovery of ~28 - 35% and ~67 - 80% after physical and chemical

cleaning, respectively. The flux recovery remained relatively unchanged after modification.

However, the pure water flux of PES-NTN0.5 remained higher than of that of the control

membrane after 3 fouling cycles.

To quantitatively examine the membrane fouling performance, flux recovery (FR), intrinsic

membrane resistance (Rm), reversible resistance (Rr), and irreversible resistance (Rir) were

calculated using Equations (1) ‒ (5). Table 4 shows the filtration resistances of the control

and modified membranes. The control PES membrane had highest intrinsic resistance among

all. The reversible resistance over total resistance ratios (Rr/RT) did not vary much after

modification, which were consistent with the overall flux recovery. All modified membranes

showed reduced irreversible resistance of the membrane.

Table 4. Filtration resistances of control and NTN-modified membranes.

Membrane Rm Rir Rr RT Rir/RT Rr/RT

PES 1.08 ± 0.04 0.64 ± 0.10 5.70 ± 0.17 7.42 ± 0.10 0.09 0.77

PESNTN-0.1 0.99 ± 0.01 0.45 ± 0.17 5.80 ± 0.40 7.24 ± 0.48 0.06 0.80

PESNTN-0.5 0.91 ± 0.02 0.37 ± 0.08 4.43 ± 0.28 5.70 ± 0.28 0.07 0.78

PESNTN-1.0 1.07 ± 0.01 0.42 ± 0.05 5.79 ± 0.16 7.29 ± 0.13 0.06 0.79

3.3.4 TiO2 amount in the permeate

TiO2 nanoparticles can produce free radical and exert a strong oxidizing ability.59 To

successfully implement the use of TiO2 nanoparticles as additives, the additives must not be

released to the permeate during filtration. Table 5 shows the amount of TiO2 of permeates for

both control and modified membranes. The result shows that the amount of TiO2 in water is

less than 0.001 ppm (or 1 ppb) for all membranes tested. The result confirms that most of the

TiO2 used in modifying the membrane is retained in the membrane during filtration. Thus,

secondary contamination of the treated water is negligible.

Table 5. Amount of TiO2 in the permeate (ppm) for various membranes tested.

Sample Amount of NTN in

membranea (wt%)

Amount of TiO2 in permeate

(ppm)

PES 0.0 <0.001

PES-NTN0.1 0.6 <0.001

PES-NTN0.5 2.9 <0.001

PES-NTN1.0 5.7 <0.001

aBased on amount of TiO2 added into the casting solution.

A systematic study of the filtration performance of NTN-modified membranes provides a

more complete understanding of membranes with low concentration of TiO2 below 1%.

Previous reports adopted techniques such as water flow rate method52 or rely on single

molecular probe (BSA)51 to determine the pore size or the rejection property of modified

membrane under low concentration of TiO2. By using poly(ethylene) glycol of different

molecular weight for rejection test, a more accurate pore size of the membranes can be

obtained. NTN-modified membranes show a decrease in the pore size and despite the

decrease, the modified membrane showed improved flux, which could be due to increased

wettability and increased surface pore density. Compared to other systems with similar TiO2

loading,49, 51, 52 NTN-modified membranes exhibited improved both flux and rejection. In the

work of Wu et al.,52 the improved flux was coupled with increased pore size (reduced

rejection). Although rejection test was not carried out, significant increase in the membrane

surface pore size was observed in their SEM images.52 In the case of TiO2-modified

nanofiltration membrane system (with PES concentration between 25 to 32 %), despite the

improvement over the flux and fouling resistance, the modified membranes with TiO2

concentration below 1% exhibited decreased rejection of methylene blue. The poorer

rejection could be due to increased pore size and porosity.49

Membrane fouling resistance is another important parameter to be considered when

determining modified membrane performance. Previous findings for membranes modified

with low loading of TiO2 revealed that the fouling resistance was improved, despite the low

loading.49, 51, 52 However, the long term fouling resistance of the modified membranes was not

known. The early studies relied on single fouling test and no chemical cleaning method was

used to remove the foulants on the membrane surface. We addressed these by adopting a

more complete fouling test. Our results are in agreement with the aforementioned49, 51, 52 and

further confirmed the fouling resistance improvement. The improvement was not observed

after the first cycle which could be due to the chemical reaction with the cleaning agent

(NaOH). Taking all the findings available into consideration (of the other groups and this

work), the low concentration approach is promising. By further using porous material in low

concentration, we are able to improve not only the flux and fouling resistance of the modified

membranes, but also their rejection properties.

One possible drawback of such inorganic-blended polymeric membrane is the potential

leaching of additives in the membranes when used. The leaching potential needs to be

investigated before the implementation of these membranes in separation process as the

leaching of inorganics may cause secondary contamination to the permeate. However, such

study is not available in the case of TiO2-blended polymeric membrane. Even though the

concentration of NTN used in this work is comparably low, the potential of releasing the

NTNs into the treated water cannot be ruled out. The trapped NTNs in the membrane matrix

may be released when the filtration system is pressurized, altering the structure of the

membrane due to compaction. Our results indicated that the amount of TiO2 in permeates was

similar to the feed water and thus the membranes are safe to be used for filtration.

4. Conclusions

PES membranes were modified with low concentrations of nanoporous titania

nanoparticles (NTNs) (between 0.1 wt% and 1 wt% relative to dope solution) in the

membrane fabrication process. NTN-modified membranes showed improved flux, reduced

pore size distribution, and improved solute rejection. The anti-fouling properties of the

modified membranes were also improved due to decreased contact angle, surface free energy

and pore size. Among all membranes, PES-NTN0.5 showed the highest improvement, with

water flux reaching 446 LMH and MWCO less than 100 kDa. The same membrane also

showed the highest fouling resistance (fouling rate of 0.58 kPa/ min compared to 0.70 kPa/

min of control membrane) with overall flux recovery of ~80 % after 3 fouling cycles under

constant flux operation. The amount of TiO2 in the permeate from the modified membranes

did not increase with NTN loading, and thus the incorporation of NTN did not cause

secondary contamination to treated solution. The strategy for incorporating a low

concentration of NTN is promising for the fabrication of polymer ultrafiltration membranes

with improved membrane performance.

Table 3. Water flux, TMP and flux recovery after 3 cycles of fouling-cleaning experiments for control and NTN-

modified membranes. NTN: nanoporous titania nanoparticles.

Membrane Water Fluxa

(LMH)

TMPb

(kPa)

Flux recovery after

physical cleaning (%)

Flux recovery after

chemical cleaning (%) Overall flux

recovery after

physical

cleaning (%)

Overall flux

recovery after

chemical

cleaning (%) I II III I II III

PES 375.3 ± 13.3 100.9 ± 1.4 40.4 35.7 29.9 84.5 78.0 75.3 35.3 ± 5.3 79.3 ± 4.7

PES-NTN0.1 407.7 ± 2.9 98.5 ± 6.5 28.6 26.4 27.2 75.8 64.7 60.3 27.4 ± 1.1 66.9 ± 8.0

PES-NTN0.5 446.0 ± 12.1 77.6 ± 3.9 37.0 30.0 30.4 80.9 83.2 73.8 32.5 ± 3.9 79.3 ± 4.9

PES-NTN1.0 376.3 ± 2.08 99.1 ± 1.7 31.4 30.3 28.8 78.1 73.1 73.9 30.1 ± 1.3 75.0 ± 2.7

a constant pressure of 100 kPa.

b final TMP after 50 min BSA filtration.

27

AUTHOR INFORMATION

Corresponding Author

*Email: huanting.wang@monash.edu

Tel.: +61 3 9905 3449

Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval

to the final version of the manuscript.

Funding Sources

This work was in part supported by The Australian Research Council (Project No.:

DP140101591).

ACKNOWLEDGMENT

The authors acknowledge use of the scanning electron microscopes at the Monash Centre for

Electron Microscopy. Z. L. is grateful for a Ph.D. top-up scholarship from the National Centre of

Excellence for Desalination Australia which is funded by the Australian Government through the

Water for the Future initiative.

REFERENCES

1. Baker, R., Membrane technology and applications. Wiley: 2012.

2. Shannon, M. A.; Bohn, P. W.; Elimelech, M.; Georgiadis, J. G.; Marinas, B. J.; Mayes,

A. M., Science and technology for water purification in the coming decades. Nature 2008, 452,

(7185), 301-310.

3. Susanto, H.; Ulbricht, M., Characteristics, performance and stability of polyethersulfone

ultrafiltration membranes prepared by phase separation method using different macromolecular

additives. Journal of Membrane Science 2009, 327, (1–2), 125-135.

4. Pal, A.; Dey, T. K.; Singhal, A.; Bindal, R. C.; Tewari, P. K., Nano-ZnO impregnated

inorganic-polymer hybrid thinfilm nanocomposite nanofiltration membranes: an investigation of

28

variation in structure, morphology and transport properties. RSC Advances 2015, 5, (43), 34134-

34151.

5. Rahimpour, A.; Madaeni, S. S.; Taheri, A. H.; Mansourpanah, Y., Coupling TiO2

nanoparticles with UV irradiation for modification of polyethersulfone ultrafiltration membranes.

Journal of Membrane Science 2008, 313, (1–2), 158-169.

6. Rahimpour, A.; Jahanshahi, M.; Mollahosseini, A.; Rajaeian, B., Structural and

performance properties of UV-assisted TiO2 deposited nano-composite PVDF/SPES

membranes. Desalination 2012, 285, (0), 31-38.

7. Razmjou, A.; Mansouri, J.; Chen, V.; Lim, M.; Amal, R., Titania nanocomposite

polyethersulfone ultrafiltration membranes fabricated using a low temperature hydrothermal

coating process. Journal of Membrane Science 2011, 380, (1–2), 98-113.

8. Rana, D.; Matsuura, T., Surface modifications for antifouling membranes. Chemical

Reviews 2010, 110, (4), 2448-2471.

9. Cao, X.; Ma, J.; Shi, X.; Ren, Z., Effect of TiO2 nanoparticle size on the performance of

PVDF membrane. Applied Surface Science 2006, 253, (4), 2003-2010.

10. Yi, X. S.; Yu, S. L.; Shi, W. X.; Sun, N.; Jin, L. M.; Wang, S.; Zhang, B.; Ma, C.; Sun, L.

P., The influence of important factors on ultrafiltration of oil/water emulsion using PVDF

membrane modified by nano-sized TiO2/Al2O3. Desalination 2011, 281, (0), 179-184.

11. Ngang, H. P.; Ahmad, A. L.; Low, S. C.; Ooi, B. S., Preparation of mixed-matrix

membranes for micellar enhanced ultrafiltration based on response surface methodology.

Desalination 2012, 293, (0), 7-20.

12. De Sitter, K.; Dotremont, C.; Genné, I.; Stoops, L., The use of nanoparticles as

alternative pore former for the production of more sustainable polyethersulfone ultrafiltration

membranes. Journal of Membrane Science 2014, 471, 168-178.

13. Razmjou, A.; Mansouri, J.; Chen, V., The effects of mechanical and chemical

modification of TiO2 nanoparticles on the surface chemistry, structure and fouling performance

of PES ultrafiltration membranes. Journal of Membrane Science 2011, 378, (1–2), 73-84.

14. Yang, Y.; Zhang, H.; Wang, P.; Zheng, Q.; Li, J., The influence of nano-sized TiO2

fillers on the morphologies and properties of PSF UF membrane. Journal of Membrane Science

2007, 288, (1–2), 231-238.

15. Razmjou, A.; Resosudarmo, A.; Holmes, R. L.; Li, H.; Mansouri, J.; Chen, V., The effect

of modified TiO2 nanoparticles on the polyethersulfone ultrafiltration hollow fiber membranes.

Desalination 2012, 287, (0), 271-280.

16. Wei, Y.; Chu, H.-Q.; Dong, B.-Z.; Li, X.; Xia, S.-J.; Qiang, Z.-M., Effect of TiO2

nanowire addition on PVDF ultrafiltration membrane performance. Desalination 2011, 272, (1–

3), 90-97.

17. Zhang, G.; Lu, S.; Zhang, L.; Meng, Q.; Shen, C.; Zhang, J., Novel polysulfone hybrid

ultrafiltration membrane prepared with TiO< sub> 2</sub>-g-HEMA and its antifouling

characteristics. Journal of Membrane Science 2013, 436, 163-173.

18. Zhao, S.; Wang, P.; Wang, C.; Sun, X.; Zhang, L., Thermostable PPESK/TiO< sub>

2</sub> nanocomposite ultrafiltration membrane for high temperature condensed water

treatment. Desalination 2012, 299, 35-43.

19. Bian, X.; Shi, L.; Yang, X.; Lu, X., Effect of Nano-TiO2 Particles on the Performance of

PVDF, PVDF-g-(Maleic anhydride), and PVDF-g-Poly (acryl amide) Membranes. Industrial &

Engineering Chemistry Research 2011, 50, (21), 12113-12123.

29

20. Rahimpour, A.; Jahanshahi, M.; Rajaeian, B.; Rahimnejad, M., TiO< sub> 2</sub>

entrapped nano-composite PVDF/SPES membranes: Preparation, characterization, antifouling

and antibacterial properties. Desalination 2011, 278, (1), 343-353.

21. Yan, L.; Li, Y. S.; Xiang, C. B., Preparation of poly (vinylidene fluoride)(pvdf)

ultrafiltration membrane modified by nano-sized alumina (Al 2 O 3) and its antifouling research.

Polymer 2005, 46, (18), 7701-7706.

22. Arthanareeswaran, G.; Devi, T. S.; Raajenthiren, M., Effect of silica particles on cellulose

acetate blend ultrafiltration membranes: Part I. Separation and Purification Technology 2008,

64, (1), 38-47.

23. Sun, M.; Su, Y.; Mu, C.; Jiang, Z., Improved antifouling property of PES ultrafiltration

membranes using additive of silica− PVP nanocomposite. Industrial & Engineering Chemistry

Research 2009, 49, (2), 790-796.

24. Arthanareeswaran, G.; Thanikaivelan, P., Fabrication of cellulose acetate–zirconia hybrid

membranes for ultrafiltration applications: Performance, structure and fouling analysis.

Separation and Purification Technology 2010, 74, (2), 230-235.

25. Bottino, A.; Capannelli, G.; Comite, A., Preparation and characterization of novel porous

PVDF-ZrO 2 composite membranes. Desalination 2002, 146, (1), 35-40.

26. Liang, S.; Xiao, K.; Mo, Y.; Huang, X., A novel ZnO nanoparticle blended

polyvinylidene fluoride membrane for anti-irreversible fouling. Journal of Membrane Science

2012, 394–395, (0), 184-192.

27. Hong, J.; He, Y., Effects of nano sized zinc oxide on the performance of PVDF

microfiltration membranes. Desalination 2012, 302, (0), 71-79.

28. Yan, L.; Li, Y. S.; Xiang, C. B.; Xianda, S., Effect of nano-sized Al< sub> 2</sub> O<

sub> 3</sub>-particle addition on PVDF ultrafiltration membrane performance. Journal of

Membrane Science 2006, 276, (1), 162-167.

29. Rahimpour, A.; Jahanshahi, M.; Khalili, S.; Mollahosseini, A.; Zirepour, A.; Rajaeian, B.,

Novel functionalized carbon nanotubes for improving the surface properties and performance of

polyethersulfone (PES) membrane. Desalination 2012, 286, 99-107.

30. Fan, Z.; Wang, Z.; Sun, N.; Wang, J.; Wang, S., Performance improvement of

polysulfone ultrafiltration membrane by blending with polyaniline nanofibers. Journal of

Membrane Science 2008, 320, (1), 363-371.

31. Huang, J.; Zhang, K.; Wang, K.; Xie, Z.; Ladewig, B.; Wang, H., Fabrication of

polyethersulfone-mesoporous silica nanocomposite ultrafiltration membranes with antifouling

properties. Journal of Membrane Science 2012, 423, 362-370.

32. Wang, D.-M.; Lai, J.-Y., Recent advances in preparation and morphology control of

polymeric membranes formed by nonsolvent induced phase separation. Current Opinion in

Chemical Engineering 2013, 2, (2), 229-237.

33. Martín, A.; Arsuaga, J. M.; Roldán, N.; de Abajo, J.; Martínez, A.; Sotto, A., Enhanced

ultrafiltration PES membranes doped with mesostructured functionalized silica particles.

Desalination 2015, 357, (0), 16-25.

34. Dong, L.-x.; Yang, H.-w.; Liu, S.-t.; Wang, X.-m.; Xie, Y. F., Fabrication and anti-

biofouling properties of alumina and zeolite nanoparticle embedded ultrafiltration membranes.

Desalination 2015, 365, (0), 70-78.

35. Han, R.; Zhang, S.; Liu, C.; Wang, Y.; Jian, X., Effect of NaA zeolite particle addition on

poly(phthalazinone ether sulfone ketone) composite ultrafiltration (UF) membrane performance.

Journal of Membrane Science 2009, 345, (1–2), 5-12.

30

36. Li, Y.; Wee, L. H.; Martens, J. A.; Vankelecom, I. F. J., ZIF-71 as a potential filler to

prepare pervaporation membranes for bio-alcohol recovery. Journal of Materials Chemistry A

2014, 2, (26), 10034-10040.

37. Low, Z.-X.; Razmjou, A.; Wang, K.; Gray, S.; Duke, M.; Wang, H., Effect of addition of

two-dimensional ZIF-L nanoflakes on the properties of polyethersulfone ultrafiltration

membrane. Journal of Membrane Science 2014, 460, 9-17.

38. Qiu, S.; Wu, L.; Pan, X.; Zhang, L.; Chen, H.; Gao, C., Preparation and properties of

functionalized carbon nanotube/PSF blend ultrafiltration membranes. Journal of Membrane

Science 2009, 342, (1–2), 165-172.

39. Celik, E.; Park, H.; Choi, H.; Choi, H., Carbon nanotube blended polyethersulfone

membranes for fouling control in water treatment. Water Research 2011, 45, (1), 274-282.

40. Gao, Y.; Qiao, Y.; Yang, S., Fabrication of PAN/PHCS adsorptive UF membranes with

enhanced performance for dichlorophenol removal from water. Journal of Applied Polymer

Science 2014, 131, (19), 40837.

41. Ma, Y.; Shi, F.; Wang, Z.; Wu, M.; Ma, J.; Gao, C., Preparation and characterization of

PSf/clay nanocomposite membranes with PEG 400 as a pore forming additive. Desalination

2012, 286, (0), 131-137.

42. Wang, P.; Ma, J.; Wang, Z.; Shi, F.; Liu, Q., Enhanced Separation Performance of

PVDF/PVP-g-MMT Nanocomposite Ultrafiltration Membrane Based on the NVP-Grafted

Polymerization Modification of Montmorillonite (MMT). Langmuir 2012, 28, (10), 4776-4786.

43. Zornoza, B.; Irusta, S.; Téllez, C.; Coronas, J., Mesoporous Silica Sphere−Polysulfone

Mixed Matrix Membranes for Gas Separation. Langmuir 2009, 25, (10), 5903-5909.

44. Ruben, G. C.; Stockmayer, W., Evidence for helical structures in poly (1-olefin sulfones)

by transmission electron microscopy. Proceedings of the National Academy of Sciences 1992,

89, (17), 7991-7995.

45. Boyer, R. F.; Miller, R. L., Polymer chain stiffness parameter, σ, and cross-sectional area

per chain. Macromolecules 1977, 10, (5), 1167-1169.

46. Syafei, A. D.; Lin, C.-F.; Wu, C.-H., Removal of natural organic matter by ultrafiltration

with TiO2-coated membrane under UV irradiation. Journal of Colloid and Interface Science

2008, 323, (1), 112-119.

47. Kim, S. H.; Kwak, S.-Y.; Sohn, B.-H.; Park, T. H., Design of TiO2 nanoparticle self-

assembled aromatic polyamide thin-film-composite (TFC) membrane as an approach to solve

biofouling problem. Journal of Membrane Science 2003, 211, (1), 157-165.

48. Kim, S. H.; Kwak, S.-Y.; Sohn, B.-H.; Park, T. H., Design of TiO 2 nanoparticle self-

assembled aromatic polyamide thin-film-composite (TFC) membrane as an approach to solve

biofouling problem. Journal of Membrane Science 2003, 211, (1), 157-165.

49. Sotto, A.; Boromand, A.; Balta, S.; Kim, J.; Van der Bruggen, B., Doping of

polyethersulfone nanofiltration membranes: antifouling effect observed at ultralow

concentrations of TiO2 nanoparticles. Journal of Materials Chemistry 2011, 21, (28), 10311-

10320.

50. Syafei, A. D.; Lin, C.-F.; Wu, C.-H., Removal of natural organic matter by ultrafiltration

with TiO 2-coated membrane under UV irradiation. Journal of colloid and interface science

2008, 323, (1), 112-119.

51. Arsuaga, J. M.; Sotto, A.; del Rosario, G.; Martínez, A.; Molina, S.; Teli, S. B.; de Abajo,

J., Influence of the type, size, and distribution of metal oxide particles on the properties of

nanocomposite ultrafiltration membranes. Journal of Membrane Science 2013, 428, 131-141.

31

52. Wu, G.; Gan, S.; Cui, L.; Xu, Y., Preparation and characterization of PES/TiO2

composite membranes. Applied Surface Science 2008, 254, (21), 7080-7086.

53. Ling, M.-P.; Chio, C.-P.; Chou, W.-C.; Chen, W.-Y.; Hsieh, N.-H.; Lin, Y.-J.; Liao, C.-

M., Assessing the potential exposure risk and control for airborne titanium dioxide and carbon

black nanoparticles in the workplace. Environmental Science and Pollution Research 2011, 18,

(6), 877-889.

54. Zhang, R.; Bai, Y.; Zhang, B.; Chen, L.; Yan, B., The potential health risk of titania

nanoparticles. Journal of Hazardous Materials 2012, 211–212, (0), 404-413.

55. Elsaesser, A.; Howard, C. V., Toxicology of nanoparticles. Advanced Drug Delivery

Reviews 2012, 64, (2), 129-137.

56. Menard, A.; Drobne, D.; Jemec, A., Ecotoxicity of nanosized TiO2. Review of in vivo

data. Environmental Pollution 2011, 159, (3), 677-684.

57. Kandlikar, M.; Ramachandran, G.; Maynard, A.; Murdock, B.; Toscano, W., Health risk

assessment for nanoparticles: A case for using expert judgment. In Nanotechnology and

Occupational Health, Maynard, A.; Pui, D. H., Eds. Springer Netherlands: 2007; pp 137-156.

58. Trouiller, B.; Reliene, R.; Westbrook, A.; Solaimani, P.; Schiestl, R. H., Titanium dioxide

nanoparticles induce DNA damage and genetic instability in vivo in mice. Cancer Research

2009, 69, (22), 8784-8789.

59. Wang, J.; Zhou, G.; Chen, C.; Yu, H.; Wang, T.; Ma, Y.; Jia, G.; Gao, Y.; Li, B.; Sun, J.,

Acute toxicity and biodistribution of different sized titanium dioxide particles in mice after oral

administration. Toxicology Letters 2007, 168, (2), 176-185.

60. Chen, X.; Mao, S. S., Titanium dioxide nanomaterials: synthesis, properties,

modifications, and applications. Chemical Reviews 2007, 107, (7), 2891-2959.

61. Owens, D. K.; Wendt, R., Estimation of the surface free energy of polymers. Journal of

Applied Polymer Science 1969, 13, (8), 1741-1747.

62. Rice, E. W.; Bridgewater, L.; Association, A. P. H., Standard methods for the

examination of water and wastewater. American Public Health Association Washington, DC:

2012.

63. Reid, B. D.; Ruiz-Trevino, F. A.; Musselman, I. H.; Balkus, K. J.; Ferraris, J. P., Gas

permeability properties of polysulfone membranes containing the mesoporous molecular sieve

MCM-41. Chemistry of Materials 2001, 13, (7), 2366-2373.

64. Luo, M.-L.; Zhao, J.-Q.; Tang, W.; Pu, C.-S., Hydrophilic modification of poly(ether

sulfone) ultrafiltration membrane surface by self-assembly of TiO2 nanoparticles. Applied

Surface Science 2005, 249, (1–4), 76-84.

65. Lee, S. J.; Han, S. W.; Yoon, M.; Kim, K., Adsorption characteristics of 4-

dimethylaminobenzoic acid on silver and titania: diffuse reflectance infrared Fourier transform

spectroscopy study. Vibrational Spectroscopy 2000, 24, (2), 265-275.

66. Tam, C. M.; Tremblay, A. Y., Membrane pore characterization—comparison between

single and multicomponent solute probe techniques. Journal of Membrane Science 1991, 57, (2–

3), 271-287.

67. Ziemniak, S. E.; Jones, M. E.; Combs, K. E. S., Solubility behavior of titanium(IV) oxide

in alkaline media at elevated temperatures. Journal of Solution Chemistry 1993, 22, (7), 601-623.

32

Insert Table of Contents Graphic and Synopsis Here