In order to improve the permeate flux of photocatalytic membranes, we present an approach for coupling TiO2 with ceramic hollow fiber membranes. The ceramic hollow fiber membranes with high permeate flux were fabricated by a controlled wet-spinning process using polyethersulfone (PESf) and ceramic powder as precursors and 1-methyl-2-pyrrolidinone as solvent, and the subsequent TiO2 coating was performed by a dip-coating process using tetra-n-butyl titanate as precursor. It has been found that the PESf/ceramic powder ratio could influence the structure of the membranes. Here the as-prepared TiO2 hollow fiber membranes had a pure water flux of 4,450 L/(m2·h). The performance of the TiO2 hollow fiber membrane was evaluated using humic acid (HA) as a test substance. The results demonstrated that this membrane exhibited a higher permeate flux under UV irradiation than in the dark and the HA removal efficiency was enhanced. The approach described here provides an operable route to the development of high-permeable photocatalytic membranes for water treatment.

INTRODUCTION

Fresh water resources available on earth account for only about 0.26% of the total water resources. Unfortunately, part of the surface water has been polluted by the drainage of industrial wastewater, the emissions of an agricultural nonpoint source, or the discharge of domestic sewage effluent (Sujaul et al. 2013; Miao et al. 2015; Schreiber et al. 2015). Recently, TiO2 photocatalysis has attracted great interest because it is environment-friendly and provides a promising pathway for potential applications in breaking down many kinds of organic pollutants, such as persistent toxic substances, dyes, pesticides, and herbicides under UV light irradiation. (Hoffmann et al. 1995; Chen & Mao 2007; Yu et al. 2009a, b; Chong et al. 2010). However, there are issues with TiO2 photocatalysis involving the recycling and reuse of the TiO2; these are inherent drawbacks that restrict practical application of this technology (Yu et al. 2001; Zhang et al. 2006).

Membrane separation achieves the separation, purification, and concentration of different components by using membrane selectivity; this has the advantages of high efficiency, energy savings, and convenient operation (Molinari et al. 2002; Sholl & Johnson 2006). However, the traditional membrane has the drawback of rapid membrane fouling. Considering the advantages of combining TiO2 photocatalysis and membrane separation, a TiO2-based photocatalytic membrane has been proposed as a possible solution that would provide the dual functions of membrane separation and photocatalytic pollutant degradation concurrently (Choi et al. 2006; Zhang et al. 2006; Albu et al. 2007). In particular, the photocatalysis function could greatly reduce membrane fouling. (Ma et al. 2009; Zhang et al. 2014)

Until now, most of the reported TiO2-based photocatalytic membranes have been limited to single-channel tube membranes or flat sheet membranes (Wang et al. 2008; Ma et al. 2010), whose permeate flux is relatively low. Hollow fiber membranes with high specific surface area have shown more promise in increasing the permeate flux of membranes (Kingsbury & Li 2009; Yu et al. 2009a, b; Leo et al. 2011; Haworth et al. 2012; Wu et al. 2013; Fan et al. 2015). For instance, Razmjou et al. (2012) reported that the pure water flux of a TiO2 nanoparticle-modified polyethersulfone ultrafiltration hollow fiber membrane was about 60 L/(m2·h). Zhang et al. (2014) reported that a nanostructured TiO2 hollow fiber photocatalytic membrane exhibited the membrane flux of 87.7 L/(m2·h). Especially, a porous yttria-stabilized zirconia hollow fiber membrane, although without the photocatalytic property, has been reported to show a extremely high pure water permeate flux of 10.89 m3/(m2·h·bar) at a transmembrane pressure difference of 1.5 bar (Zhang et al. 2015). Motivated by this finding above, we expect that the permeate flux of TiO2 hollow fiber membranes could still be enhanced by optimizing the hollow fiber formation process. Therefore, in the present work, the TiO2 hollow fiber membrane was prepared by dry-jet wet-spinning and dip-coating, and the photocatalytic activity of this membrane was evaluated using humic acid (HA) as a test substance.

MATERIALS AND METHODS

Chemicals and materials

Ceramic powder and polyethersulfone (PESf) were provided by Shanghai Kaidu Industrial Development Co., Ltd. 1-Methyl-2-pyrrolidinone (NMP), tetra-n-butyl titanate ([CH3(CH2)3O]4Ti), ethanol (C2H5OH), and diethanolamine (NH(C2H4OH)2) were obtained from Tianjin Fuyu Fine Chemical Co., Ltd. HA was provided by MP Biomedicals, Inc. (Eschwege, Germany).

Preparation of ceramic hollow fiber membranes

The ceramic hollow fiber membranes were fabricated by a dry-jet wet-spinning process. To prepare the dry-jet wet-spinning solution, the ceramic powder was added to the mixed solution of PESf and NMP; the ratio of PESf to ceramic powder was set at 13%, 14%, 17% and 20%, respectively. This mixture was continuously stirred for about 2 h. As shown in Figure 1, the spinning solution was then extruded using a tube-in-orifice spinneret (outer diameter = 750 μm and inner diameter = 510 μm) to form a hollow fiber; the air gap between the spinneret and the coagulation bath used for the spinning process was 20 mm. Either spinning dope or pure water was used as the bore fluid at a rate of 25 mL/min or 12 mL/min, respectively. To eliminate the influence of NMP, the as-prepared ceramic hollow fiber membranes were dipped in pure water for 24 h, and the phase inversion formed completely on the inner side of the ceramic hollow fiber membranes. Finally, the ceramic hollow fiber membranes were calcined at 1,200 °C for 2 h with a heating/cooling rate of 5 °C/min.
Figure 1

Schematic diagram of experimental set-up for controlled wet-spinning.

Figure 1

Schematic diagram of experimental set-up for controlled wet-spinning.

Coating TiO2

The TiO2 hollow fiber membranes were obtained by coating TiO2 on the ceramic hollow fiber membranes using a dip-coating process. A TiO2 colloidal solution was prepared by adding 17 mL of tetra-n-butyl titanate and 4.5 mL of diethanolamine to 67 mL of absolute ethyl alcohol, and continuously stirring for 2 h. Subsequently, a mixture of 10 mL anhydrous ethanol and 1 mL pure water was added, and continuously stirred for 1 h (Yu et al. 2001). A dip coater (SYDC-100, Shanghai SAN-YAN Technology Co., Ltd, China) was used for coating TiO2 on the ceramic hollow fiber membranes. The dipping time was set to 15 min and the lifting speed was 1 mm/s. In this study, the impregnation of ceramic hollow fiber membranes with TiO2 solution was repeated 5, 7 or 9 times. The final products were calcined at 500 °C for 2 h with a heating/cooling rate of 2 °C/min.

Characterization

The morphology of the as-prepared products including ceramic hollow fiber membranes and TiO2 hollow fiber membranes was characterized using a scanning electron microscope (SEM; Quanta 200 FEG). The crystallinity of the prepared samples was determined by X-ray diffraction (XRD) using a diffractometer with Cu Kα radiation (Shimadzu LabX XRD-6000). A three-point bending test performed on an Instron 5543 universal material testing machine was used to measure the mechanical strength of the TiO2 hollow fiber membranes. In order to determine the flux of the membranes, one end of the membrane was sealed using silicone gel and the other end was connected to a permeate collector, which was connected to a vacuum pump. The absolute pressure at the mouth of the pump was 75 kPa as a driving force. The pure water flux of the TiO2 hollow fiber membranes was measured without and with UV irradiation. A 300 W high-pressure mercury lamp was used as the UV light source with a principal wavelength of 365 nm. The incident UV light intensity was 1.5 mW/cm2, which was measured by a radiometer (model UV-A). The flux of the hollow fiber membrane was calculated depending on the volume variation of the solution on the basis of the following equation (Zhang et al. 2014): 
formula
Here, J is the flux; V0 is the initial volume of the feed solution; Vt is the volume of the feed solution at time t; and A is the membrane surface area. The porosity of hollow fiber membranes with different ratios of PESf/ceramic powder is calculated based on the following equation: 
formula
Here, P is the porosity of the hollow fiber membranes; and are the weight of the hollow fiber membrane before and after dipping in pure water, respectively; ρ is the density of pure water; R and r are the outer and inner radii of the hollow fiber membrane, respectively; and L is the length of the hollow fiber membrane. The TiO2 mass-to-area ratio was calculated based on a previous report (Ma et al. 2009). The performance of the membrane was evaluated for treating HA (initial concentration of 10.5 mg/L) using a dead-end filtration unit (as shown in Figure 2). The removal rate of HA was calculated based on the variation of total organic carbon (TOC) in the solution. TOC was measured with a TOC analyzer (TOC-V CPH, Shimadzu). The rejection of HA was calculated using the formula (Jyothi et al. 2014): 
formula
Figure 2

Diagrammatic representation of a filtration unit used for the removal of HA.

Figure 2

Diagrammatic representation of a filtration unit used for the removal of HA.

where and are the permeate and the feed concentrations of HA, respectively.

RESULTS AND DISCUSSION

Figure 3 shows the SEM images of the cross-section of the ceramic hollow fiber membranes with the ratio of PESf/ceramic powder ranging from 13 to 20%. Upon increasing the PESf/ceramic powder ratio from 13 to 20%, the shape of the inner profile gradually became asymmetrical. It also can be seen from Figure 3 that finger-like macrovoids formed within the membrane, which could serve as channels during membrane filtration. As shown in Figure 3(d), when the ratio of PESf/ceramic powder reached 20%, the inner profile of the ceramic hollow fiber membrane was not circular and the inner wall displayed an inward bulge.
Figure 3

SEM images of the cross-section of ceramic hollow fiber membranes with a PESf/ceramic powder ratio of 13% (a), 14% (b), 17% (c) and 20% (d).

Figure 3

SEM images of the cross-section of ceramic hollow fiber membranes with a PESf/ceramic powder ratio of 13% (a), 14% (b), 17% (c) and 20% (d).

Figure 4 shows the porosity of the ceramic hollow fiber membrane with different ratios of PESf/ceramic powder. The porosity was approximately 50%, 57%, 63%, and 80% for ceramic hollow fiber membranes with PESf/ceramic powder ratios of 13%, 14%, 17%, and 20%, respectively. This result is consistent with the morphologies shown in Figure 3 in that the porosity increased with increasing of finger-like macrovoids.
Figure 4

The porosity of ceramic hollow fiber membranes with different ratios of PESf/ceramic powder.

Figure 4

The porosity of ceramic hollow fiber membranes with different ratios of PESf/ceramic powder.

The permeate flux of the ceramic hollow fiber membranes was evaluated using pure water. As shown in Figure 5, the pure water flux increased as the PESf/ceramic powder ratio increased. The pure water flux was 4,600 L/(m2·h) (75 kPa) at a PESf/ceramic powder ratio of 13%, and it increased to 7,400 L/(m2·h) (75 kPa) at the ratio of 14%, followed by additional increases to 10,200 and 31,900 L/(m2·h) (75 kPa) at the ratios of 17% and 20%, respectively.
Figure 5

The pure water flux of ceramic hollow fiber membranes with different ratios of PESf/ceramic powder at a testing pressure of 75 kPa and temperature of 25 °C.

Figure 5

The pure water flux of ceramic hollow fiber membranes with different ratios of PESf/ceramic powder at a testing pressure of 75 kPa and temperature of 25 °C.

Figure 6 presents the maximal bending strength of the ceramic hollow fiber membrane as a function of the PESf/ceramic powder ratio. The PESf/ceramic powder ratio had a significant influence on the mechanical strength. Figure 6 shows that the mechanical strength increases as the PESf/ceramic powder ratio decreases. The mechanical strengths were 490, 590, 720, and 1,170 kPa for ceramic hollow fiber membranes with PESf/ceramic powder ratios of 20%, 17%, 14%, and 13%, respectively. Meanwhile, as shown in Figure 5, the pure water flux decreased with the decreasing PESf/ceramic powder ratio. Therefore, although the ceramic hollow fiber membrane with a PESf/ceramic powder ratio of 13% had excellent mechanical strength, the pure water flux was relatively low. Compared to the membranes with PESf/ceramic powder ratios of 17 and 20%, the membrane of 14% has better mechanical strength. Considering both permeate flux and mechanical strength, the ceramic hollow fiber membrane with a PESf/ceramic powder ratio of 14% was adopted as the substrate for the fabrication of TiO2 hollow fiber membranes.
Figure 6

The maximal bending strength of ceramic hollow fiber membranes with different PESf/ceramic powder ratios.

Figure 6

The maximal bending strength of ceramic hollow fiber membranes with different PESf/ceramic powder ratios.

During the TiO2 dip-coating process, the dip time in the TiO2 solution was a key parameter influencing the mass/area ratio of the TiO2 hollow fiber membrane. The TiO2 mass/area ratio as a function of TiO2 dipping time is shown in Figure 7. As can be seen from Figure 7, the TiO2 mass/area ratio increased from 16 to 21 g/m2 as the dipping time increased from 1 to 15 min, and then leveled off from 15 min to 25 min. Therefore, the dip time was fixed at 15 min for this study.
Figure 7

TiO2 mass/area ratios for various TiO2 dipping times.

Figure 7

TiO2 mass/area ratios for various TiO2 dipping times.

The XRD pattern of a TiO2 hollow fiber membrane is shown in Figure 8. Five distinctive peaks appeared at 2θ angles of 25.31°, 37.91°, 48.01°, 54.61°, and 62.81°, which correspond to the reflections of crystalline anatase-TiO2 planes (JCPDS 21-1272); this result is consistent with a previous study (Goei & Lim 2014).
Figure 8

XRD pattern of a TiO2 hollow fiber membrane.

Figure 8

XRD pattern of a TiO2 hollow fiber membrane.

Figure 9 shows the variation of pure water flux as a function of different numbers of impregnation cycles of the TiO2 hollow fiber membranes; the flux decreased slightly as the number of coating cycles increased. When the impregnation was repeated 5, 7, and 9 times, the pure water flux was 3,175 L/(m2·h), 2,955 L/(m2·h), and 2,790 L/(m2·h), respectively. It can be concluded that the pure water flux decreases as a result of the increased thickness of the TiO2 functional layer.
Figure 9

The pure water flux of TiO2 hollow fiber membrane with different numbers of impregnation cycles.

Figure 9

The pure water flux of TiO2 hollow fiber membrane with different numbers of impregnation cycles.

Figure 10 presents the cross-section images of TiO2 hollow fiber membranes for various numbers of impregnation cycles; the thickness of the TiO2 layer gradually increased as the number of coating cycles increased. The thickness of TiO2 was about 480, 730, and 1,110 nm, corresponding to repetitions of 5, 7, and 9, respectively. Based on the results of Figures 9 and 10, the hollow fiber membrane dipped 7 times in the TiO2 sol was employed for the rest of this study.
Figure 10

SEM images of cross-sections of TiO2 hollow fiber membranes with 5 (a), 7(b) and 9 (c) impregnation cycles.

Figure 10

SEM images of cross-sections of TiO2 hollow fiber membranes with 5 (a), 7(b) and 9 (c) impregnation cycles.

As shown in Figure 11, the pure water flux of the TiO2 hollow fiber membranes was different with or without UV irradiation. Without UV irradiation, the pure water flux of the TiO2 hollow fiber membrane was 2,955 L/(m2·h), while in the presence of UV (at UV light intensity of 1.5 mW/cm2), the pure water flux of TiO2 hollow fiber membranes reached 4,450 L/(m2·h), which is approximately 1.5 times greater. The influence of light intensity on the pure water flux was also investigated. As expected, the pure water flux increased as the UV light intensity increased. This could be attributed to the hydrophilicity of the TiO2, which has a UV-sensitive hydrophilic surface conversion agent. The hydrophilicity of TiO2 was created by the formation of a layer of adsorbed hydroxyl and multiple layers of water molecules (Lee et al. 2003). It has been reported that the photo-induced hydrophilic reaction of the TiO2 surface was enhanced by the UV illumination increasing (Banerjee et al. 2015; Kim et al. 2015). This indicates that the pure water flux of the TiO2 hollow fiber membranes could be improved with the increasing of UV light intensity.
Figure 11

The pure water flux of TiO2 hollow fiber membrane under different UV light intensities.

Figure 11

The pure water flux of TiO2 hollow fiber membrane under different UV light intensities.

The permeate flux of the TiO2 hollow fiber membrane during the HA treatment (initial concentration 10.5 mg/L) with or without UV irradiation was investigated and the result is shown in Figure 12. In the absence of UV irradiation, the permeate flux declined from 290 to 85 L/(m2·h) in 30 min, and remained constant at 80 L/(m2·h) for an additional 30 min. In the presence of UV irradiation, the permeate flux reached 410 L/(m2·h) after 10 min, and then stabilized at 250 L/(m2·h) after an additional 30 min. In other words, the average permeate flux increased from 80 to 250 L/(m2·h) with UV irradiation, and the improvement in HA permeate flux was approximately 210%. Here, it is obvious that HA molecules could be adsorbed on the membrane surface and rejected, followed by decrease of the permeate flux of the membrane. Meanwhile, under UV illumination, TiO2 could absorb UV light to generate hydroxyl radicals and other reactive oxygen species, which further accelerate the process of photocatalytic decomposition of HA. Generally, the photocatalytic degradation of HA could produce some intermediates that ultimately could be converted to CO2 and H2O (Ma et al. 2010; Xue et al. 2011). Since the HA adsorbed on the membrane surface could be destroyed by TiO2, the permeate flux would be enhanced.
Figure 12

The changes of HA permeate flux without and with UV irradiation.

Figure 12

The changes of HA permeate flux without and with UV irradiation.

Although the membrane shows attractive prospects for water treatment, the fouling often restricts its practical application. The degree of membrane fouling could be investigated by monitoring the permeate flux variation. As shown in Figure 12, the fouling of the TiO2 hollow fiber membrane could be demonstrated by the changes in permeate flux of HA without and with UV irradiation. In the case of without UV irradiation, the permeate flux of HA declined from 290 initially to 85 L/(m2·h) after 30 min, suggesting the cake layer of HA formed due to surface adsorption by filtration alone. However, the permeate flux reached 250 L/(m2·h) with UV irradiation, indicating the coating of TiO2 was beneficial to resist fouling, due to its distinguished photocatalytic ability to mineralize HA, ensuring the attractive performance of the TiO2 hollow fiber membrane.

Figure 13 shows the TOC of HA before and after treatment with the TiO2 hollow fiber membrane with and without UV irradiation. The initial TOC value of HA was set at 10.5 mg/L; it decreased to 7.2 mg/L under TiO2 hollow fiber membrane treatment in absence of UV irradiation and decreased to 1.7 mg/L in presence of UV irradiation. This indicated that HA could be partially rejected by filtration alone, whereas HA removal by filtration and concurrent photocatalysis was remarkable. It could be calculated that the 31.4% rejection of HA was observed by filtration alone, whereas 83.8% of HA was removed by filtration and concurrent photocatalysis, suggesting a typical synergistic effect by integration. Consequently, compared with the filtration alone, the HA removal efficiency was dramatically increased by the contribution of both rejection and photocatalysis.
Figure 13

TOC variation during HA treatment without and under UV irradiation (UV light intensity of 1.5 mw/cm2).

Figure 13

TOC variation during HA treatment without and under UV irradiation (UV light intensity of 1.5 mw/cm2).

CONCLUSIONS

A TiO2 hollow fiber membrane was fabricated by a controlled wet-spinning process, followed by a subsequent dip-coating process. It is shown that the optimized fabrication process is very effective in controlling membrane performance. In this study, using the TiO2 hollow fiber membrane, the pure water flux reached 4,450 L/(m2·h). The permeate flux was significantly enhanced in the presence of UV irradiation, and increasing the UV light intensity resulted in increased pure water flux. Meanwhile, the TiO2 hollow fiber membrane was applied for treating an aqueous HA solution. Compared with the 31.4% rejection of HA by filtration alone, 83.8% of HA removal efficiency was achieved. The HA removal was improved under UV irradiation, likely due to enhanced performance of filtration along with photocatalytic degradation of HA simultaneously. In addition, the use of TiO2 hollow fiber membranes provided a better reduction of membrane fouling in the presence of UV light irradiation. Considering the high permeate flux and enhanced performance, the TiO2 hollow fiber membranes may find various applications in water treatment.

ACKNOWLEDGEMENTS

This work was supported by the National Natural Science Foundation of China (51478075) and the Natural Science Foundation of Liaoning Province of China (2014020149).

REFERENCES

REFERENCES
Banerjee
S.
Dionysiou
D. D.
Pillai
S. C.
2015
Self-cleaning applications of TiO2 by photo-induced hydrophilicity and photocatalysis
.
Applied Catalysis B: Environmental
176–177
,
396
428
.
Chong
M. N.
Jin
B.
Chow
C. W. K.
Saint
C.
2010
Recent developments in photocatalytic water treatment technology: a review
.
Water Research
44
,
2997
3027
.
Fan
Z.
Xiao
C.
Liu
H.
Huang
Q.
Zhao
J.
2015
Structure design and performance study on braid-reinforced cellulose acetate hollow fiber membranes
.
Journal of Membrane Science
486
,
248
256
.
Haworth
P. F.
Smart
S.
Serra
J. M.
Diniz da Costa
J. C.
2012
Combined investigation of bulk diffusion and surface exchange parameters of silver catalyst coated yttrium-doped BSCF membranes
.
Physical Chemistry Chemical Physics
14
,
9104
9111
.
Hoffmann
M. R.
Martin
S. T.
Choi
W. Y.
Bahnemann
D. W.
1995
Environmental applications of semiconductor photocatalysis
.
Chemical Reviews
95
,
69
96
.
Jyothi
M. S.
Nayak
V.
Padaki
M.
Geetha
B. R.
Ismail
A. F.
2014
The effect of UV irradiation on PSf/TiO2 mixed matrix membrane for chromium rejection
.
Desalination
354
,
189
199
.
Kim
H. M.
Kim
D.
Kim
B.
2015
Photoinduced hydrophilicity of TiO2/WO3 double layer films
.
Surface and Coatings Technology
271
,
18
21
.
Kingsbury
B. F. K.
Li
K.
2009
A morphological study of ceramic hollow fibre membranes
.
Journal of Membrane Science
328
,
134
140
.
Lee
Y. C.
Hong
Y. P.
Lee
H. Y.
Kim
H.
Jung
Y. J.
Ko
K. H.
Jung
H. S.
Hong
K. S.
2003
Photocatalysis and hydrophilicity of doped TiO2 thin films
.
Journal of Colloid and Interface Science
267
,
127
131
.
Leo
A.
Smart
S.
Liu
S.
Diniz da Costa
J. C.
2011
High performance perovskite hollow fibres for oxygen separation
.
Journal of Membrane Science
368
,
64
68
.
Miao
X.
Tang
Y.
Wong
C. W. Y.
Zang
H.
2015
The latent causal chain of industrial water pollution in China
.
Environmental Pollution
196
,
473
477
.
Molinari
R.
Borgese
M.
Drioli
E.
Palmisano
L.
Schiavello
M.
2002
Hybrid processes coupling photocatalysis and membranes for degradation of organic pollutants in water
.
Catalysis Today
75
,
77
85
.
Razmjou
A.
Resosudarmo
A.
Holmes
R. L.
Li
H.
Mansouri
J.
Chen
V.
2012
The effect of modified TiO2 nanoparticles on the polyethersulfone ultrafiltration hollow fiber membranes
.
Desalination
287
,
271
280
.
Schreiber
C.
Rechenburg
A.
Rind
E.
Kistemann
T.
2015
The impact of land use on microbial surface water pollution
.
International Journal of Hygiene and Environmental Health
218
,
181
187
.
Sholl
D. S.
Johnson
J. K.
2006
Making high-flux membranes with carbon nanotubes
.
Science
312
,
1003
1004
.
Sujaul
I. M.
Hossain
M. A.
Nasly
M. A.
Sobahan
M. A.
2013
Effect of industrial pollution on the spatial variation of surface water quality
.
American Journal of Environmental Sciences
9
,
120
129
.
Wu
Z.
Faiz
R.
Li
T.
Kingsbury
B. F. K.
Li
K.
2013
A controlled sintering process for more permeable ceramic hollow fibre membranes
.
Journal of Membrane Science
446
,
286
293
.
Yu
J. C.
Yu
J. G.
Ho
W. K.
Zhang
L. Z.
2001
Preparation of highly photocatalytic active nano-sized TiO2 particles via ultrasonic irradiation
.
Chemical Communications
19
,
1942
1943
.
Yu
L. Y.
Shen
H. M.
Xu
Z. L.
2009b
PVDF-TiO2 composite hollow fiber ultrafiltration membranes prepared by TiO2 sol-gel method and blending method
.
Journal of Applied Polymer Science
113
,
197
204
.
Zhang
X.
Wang
D. K.
Lopez
D. R. S.
Diniz da Costa
J. C.
2014
Fabrication of nanostructured TiO2 hollow fiber photocatalytic membrane and application for wastewater treatment
.
Chemical Engineering Journal
236
,
314
322
.
Zhang
X.
Hu
J.
Chang
Q.
Wang
Y.
Zhou
J.
Zhao
T.
Jiang
Y.
Liu
X.
2015
Influences of internal coagulant composition on microstructure and properties of porous YSZ hollow fiber membranes for water treatment
.
Separation and Purification Technology
147
,
337
345
.