eprints.binadarma.ac.ideprints.binadarma.ac.id/2658/1/Full paper ICRAMST2012 (E... · Web viewRefinery Produced Wastewater Treatment by PVDF Composite Hollow Fiber Membrane E. Yuliwatia,b,*

ICRAMST2013

Refinery Produced Wastewater Treatment by PVDF Composite Hollow Fiber Membrane

E. Yuliwatia,b,* , A.F. Ismailb,c

aDepartment of Industrial Engineering, Faculty of Engineering, Universitas Bina Darma, 30251 Palembang, Indonesia

Tel:+62(711)515-579; Fax: +62(711)515-581bAdvanced Membrane Technology Research Centre Universiti Teknologi Malaysia

cFaculty of Petroleum and Renewable Energy Engineering,Universiti Teknologi Malaysia, 81310 UTM, Skudai Johor, Malaysia

Tel. +60 (7) 553-5592; Fax: +60 (7) 558-1463

*Corresponding author: [email protected]

Abstract

The aim of this study is to investigate the effect of surface modified of PVDF membranes by adding the hydrophilic additives for refinery produced wastewater treatment. This paper presents the results of a research on direct clean water treatment using hollow fiber ultrafiltration. The source of water is the synthetic refinery wastewater with mixed liquor suspended solids (MLSS) concentration of 3 g/l. All experiments were conducted at 25oC and 0.5 bar absolute. The morphological and performance tests were conducted on PVDF ultrafiltration membranes prepared from different additives concentrations. The cross sectional area of the hollow fiber membranes was observed using a field emission scanning electron microscope (FESEM). The surface wettability of porous membranes was determined by measurement of contact angle. Mean pore size and surface porosity were calculated based on the permeate flux. The results also indicated that the PVDF composite membranes with lower additives concentration loading possessed smaller mean pore size, more apertures inside the membranes with enhanced membrane hydrophilicity. The flux and rejection of refinery wastewater using PVDF composite membrane achieved 140.82 L/m2h and 98.8 % , respectively.

Keywords: surface modified membrane; composite; inorganic additives; average pore size; hydrophilicity.

1. Introduction

Waterborne outbreaks of enteric diseases are a major public health concern, yet monitoring

and identifying the disease-causing microorganism from water samples remain difficult.

1

ICRAMST2013

Produced water is by far the largest contaminated stream resulting from thermal heavy oil

recovery operations and its treatment and reuse is essential for the sustainability of oil sands

processing [1]. Organic contaminants in produced waters are toxic and corrosive leading to

environmental and operational problems. From an environmental sustainability and perspective,

it is necessary to recycle produced water and thus it must undergo proper treatment in order to

avoid potentially negative impacts on drinking water supplies and aquatic organisms [2,3]. From

an industrial standpoint, the different contaminants in the produced water may adversely affect

equipment leading to scaling and corrosion [4-6].

Many studies have been documented on the use of UF membranes for treating oily

wastewater [7-13]. These membranes were prepared from polymeric materials such as cellulose

acetate (CA), polysulfone (PSf), polyethersulfone (PES) incorporated with inorganic material

such as alumina (Al2O3) and titanium dioxide (TiO2). As these membranes were quite

hydrophilic and displayed relatively smaller pores, water which was free of oil or with reduced

oil content were recovered as permeate.

Polyvinylidene fluoride (PVDF) is one of the most extensively applied membrane material

in UF system due to its outstanding antioxidation activity, excellent chemical resistance and

thermal stability, highly organic selectivity, as well as good mechanical and membrane forming

properties. However, its hydrophobic nature, which often resulted in severe membrane fouling

and declined permeability, has been a barrier to its application in water and wastewater treatment

[14]. In general, PVDF shows a good solubility in many common organic solvents such as N,N-

dimethylacetamide (DMAc), N,N-dimethylformamide (DMF), N-methyl pyrrolydone (NMP)

and dimethylsulfoxide (DMSO). As a semi-crystalline polymer, PVDF generally exhibits more

complicated phase separation behaviour than amorphous polymer [15]. These advantageous

2

ICRAMST2013

properties, coupled with its hydrophobicity, make it an outstanding membrane material

particularly for industrial wastewater treatment applications involving oily emulsion [16],

organic/water separations [17,18], gas absorption and stripping [19,20], and membrane

distillation [21,22]. Although its hydrophobicity is favorable in promoting the transport of the

organic component of an organic/water feed solution, the neat PVDF membrane is liable to be

contaminated and resulted in a dramatic decreased of membrane water flux [23]. Many attempts

have been carried out to improve the hydrophilicity of PVDF membranes through various

methods for instance such as physical blending, chemical grafting, and surface modifications

[24]. The effect of hydrophilic additives, i.e. LiCl and PVP, on the thermodynamic/kinetic

relations during the phase inversion process in the preparation of PVDF-based membranes was

investigated by Fontananova et al. [25].

In this study, the PVDF UF membranes have been fabricated by addition of LiCl.H 2O and

TiO2 in various concentrations to modify membrane surface properties and filtration

performance. Surface wettability of membranes is usually expressed in terms of contact angle for

a liquid drop on the membrane surface to measure the tendency for liquid to wet of the

membrane surface. Pore size, porosity, and elemental composition analysis of the PVDF UF

membranes were investigated. The surface and inner structures of the sample membranes were

studied using field emission scanning electron microscope (FESEM) and energy dispersive x-ray

(EDX) apparatus. The pretreatment used phytoremediation is an emerging cleanup technology

for contaminated wastewater. The performance for refinery wastewater treatment was

characterized by pure water flux and rejection efficiency of refinery wastewater.

3

ICRAMST2013

2. Experimental

2.1. Materials

Ultrafiltration membranes have been prepared using Kynar®740 PVDF polymer pellets were

purchased from Arkema Inc. Philadelphia, USA. The solvent N,N-dimethylacetamide (DMAc,

Aldrich Chemical) (Synthesis Grade, Merck, >99%) was used as polymer solvent without further

purification. Lithium chloride monohydrate (LiOH.H2O) and nanoparticles titanium dioxide

(TiO2) were used as inorganic additives. Both chemical additives were purchased from Sigma-

Aldrich and used as received. Glycerol was purchased from MERCK (Germany) and used as non

solvent for post treatment of membrane. In all cases, tap water was used as the external

coagulation bath medium in the spinning process.

2.2. Preparation of PVDF spinning dopes

An amount of pre-dried (24 h oven dried at 50 oC) PVDF pellets was weighed and poured

into pre-weighed DMAc solvent. The mixture was stirred to ensure thorough wetting of polymer

pellets, prior to the addition of appropriate amounts of LiCl.H2O at 50 oC. TiO2 was then added

to the polymer dope mixtures which were continuously stirred for 48 h (IKA-20-W) at 500 rpm

until a homogenous solution was formed. The polymer solution was kept in a glass bottle and air

bubbles formed in the dope were removed using water aspirator for several hours. The fully

dissolved polymer solution was transferred to a stainless steel reservoir, allowed to stand and

degassed for 24 h at room temperature prior to spinning process. Solution viscosity was

measured using rheometer (Bohlin Instrument Ltd.) at various temperatures between 25 and 50

oC.

4

ICRAMST2013

2.3. Membrane preparation

PVDF hollow fiber UF membranes were spun at room temperature by a dry-jet wet

spinning method. The spinning solutions were divided into two batches. First batch consisted of

different PVDF concentration ranging from 16 to 22 wt.%. The second one was prepared from

19 wt.% PVDF in DMAc at different TiO2 concentration (0, 5, 10, 15, 20 wt.%) and LiCl.H2O

was maintained at 5.2 wt % of the weight of PVDF, as shown in Table 1 respectively.

Table 1. Membrane composition

Sample PVDF wt. % TiO2 wt. % LiCl.H2O wt. %

PTL-0 19 0 0.98

PTL-5 19 1 0.98

PTL-10 19 1.95 0.98

PTL-15 19 2.85 0.98

PTL-20 19 3.8 0.98

The hollow fiber spinning process by dry-jet wet phase inversion was explained elsewhere.

The detailed spinning parameters are listed in Table 2.

5

ICRAMST2013

Table 2. The detailed spinning condition

Dope extrusion rate (ml/min) 4.20Bore fluid H2O

Bore fluid flow rate (ml/min) 1.40External coagulant Tap water

Air gap distance (cm) 1 cmSpinneret o.d./i.d. (mm) 1.10/0.55

Coagulation temperature (oC) 25

In general, the polymer solution was pressurized through spinneret with controlled extrusion

rate, while internal coagulant was adjusted at 1.4 ml/min. The hollow fiber emerged from the tip

of the spinneret was guided through the two water baths at a take up velocity 13.7 cm/s, carefully

adjusted to match free falling velocity before landed in a final collection bath to complete the

solidification process. The spun hollow fibers were immersed in the water bath for a period of 3

days, with daily change of the water, to remove the residual DMAc and the additives. The

hollow fibers were then post-treated using 10 wt.% glycerol aqueous solution as non solvent

exchange for 1 day in order to minimize fiber shrinkage and pores collapse. After the fibers were

dried for 3 days, they were ready for making hollow fiber test modules.

2.4. Membrane characterizations

The morphology of the membrane was observed by field emission scanning electron

microscope (FESEM) (JEOL JSM-6700F). The FESEM micrographs were taken at certain

magnifications. It produced photographs at the analytical working distance of 10 nm. Surface

composition analysis was carried out on energy dispersive x-ray (EDX) (JEOL JSM-6380LA).

6

ICRAMST2013

The static contact angle of membrane was measured by the sessile drop method using a

DropMeter A-100 contact angle system (Maist Vision Inspection & Measurement Co. Ltd.) to

characterize the membrane wetting behaviour. A water droplet at 3 µL was deposited on the dry

membrane using a microsyringe. A microscope with a long working distance 6.5x objectives was

used to capture micrographs.

2.5 Permeation flux and rejection of refinery wastewater measurements

The permeation flux and rejection of PVDF hollow fiber membranes were measured by

submerged ultrafiltration experimental equipment as shown in Fig. 1. An in-house produced U-

shape hollow fiber bundle, with a filtration area of 11.23 dm2, was submerged in prepared

suspension in membrane reservoir with volume of 14 L. A cross-flow stream was produced by

air bubbling generated by a diffuser situated underneath the submerged membrane module for

mechanical cleaning of the membrane bundle. The air bubbling flow rates per unit projection

membrane area was set constantly at 1.8 L/min in order to maintain proper turbulence. The

filtration pressure was supplied by a vacuum pump. Permeate flow rates were continually

recorded using flow meter respectively.

The rejection test was carried out with distilled water and synthetic refinery wastewater

with mixed liquor suspended solid (MLSS) concentration of 3 g/L. All experiments were

conducted at 25 OC. Firstly, the pure water permeation flux (Jw) was measured using prepared

PVDF submerged membrane under reduced pressure (0.5 bar absolute) on the permeate side.

Finally, the permeation measurement with refinery wastewater (JR) and rejection (R) were

measured under reduced pressure on the permeate side.

7

ICRAMST2013

Figure 1. Schematic diagram of submerged ultrafiltration

Pure water flux was measured after the flux was steady, then calculated as

F = VAt (1)

where F is the pure water flux (L/m2 h), V is the permeate volume (L), A is the membrane surface

area (m2), and t is the time (h).

Rejection (R) was characterized with a synthetic refinery wastewater after the membrane

was previously filtered with pure water until the flux became steady. The synthetic refinery

wastewater was an in-house synthesized and consisted of fresh water, hydraulic oil, diesel fuel,

surfactant, and carbon black in proper composition, based on mixed liquor suspended solid

(MLSS) measurement 3 g/L and UV wavelength 2.6 cm-1. It was calculated as

R = (1-

c p

c f ) x 100 (2)

8

ICRAMST2013

where R is the rejection ultrafiltration process (%), cp is the concentration of the permeate (%)

and cf is the concentration of the feed (%).

3. Results and discussions

3.1. Effect of additives concentration on the structural and physical properties of PVDF

membranes

3.1.1. Morphological studies of PVDF membranes

Fig. 2 shows the FESEM micrographs of the PVDF ultrafiltration membranes prepared

using different concentrations of TiO2 at a constant concentration of LiCl.H2O. Improvement of

membrane morphology occurs with small amount addition of TiO2 nanoparticles. TiO2

nanoparticles have high specific areas and good hydrophilicity, which will affect the mass

transfer during the spinning process.

The cross-section morphology of PVDF UF membranes indicated that the finger-like

macrovoids extended from both inside and outside of the membranes and caused suppression in

an intermediate spongy substructure at lower TiO2 concentration, as illustrated in Fig. 2(b-c).

However, with further increasing TiO2 concentration it was observed that the intermediate and

the outer and inner membrane layer have been changed significantly. The long finger-like

structure became shorter at the both outer and inner layer of membrane and the intermediate

layer presented a thicker sponge-like structure, whereas in Fig. 2(d-e). These results can be

explained on the basis of the delayed liquid-liquid demixing process, which could be attributed

to the higher viscosity and lower phase-inversion rate of the spinning dope. Therefore, the kinetic

9

ICRAMST2013

hindrance due to viscosity overcomes the thermodynamic factor and thus resulted in the

formation of thick sponge-like layer. Moreover, formation of few drop cavities in the sponge-like

layer was also associated to the slow solidification process during phase inversion.

The higher TiO2 concentration induced also an aggregate phenomenon and absorbed into

the substructure of PVDF UF membrane. Those aggregates blocked the pores and caused the

decreased of the average pore size. This result was attributed to the porous structure and possible

hydrophilicity of the TiO2 nanoparticles. It indicates that hydrophilicity of nanoparticles TiO2

was directly correlated with porosity and might be responsible for the higher liquid uptake.

(a)

(b)

10

ICRAMST2013

(c)

(d)

(e)

Figure 2. FESEM images of prepared PVDF membranes with maintained LiCl.H2O of 0.98 wt.% and TiO2 of a) 0 b) 1c)1.95 d) 2.85 e) 3.8 wt.%.

3.1.2. Porosity and surface wettability studies of PVDF UF membranes

The membranes were characterized in terms of surface wettability measurement. The

results are shown in Fig. 3. Surface wettability is one of the important membranes properties

which could affect the flux and antifouling ability of membranes. As presented in Fig. 3, it was

11

ICRAMST2013

found that contact angle of PVDF UF membranes decreased significantly with increasing TiO2

concentration above 1.95 wt.%, then increased with further increasing TiO2 content. The

decreased contact angle indicates the decrease in effective hydrophilic area and hydroxyl group

number. The hydrophilic TiO2 particles, which contained hydroxyl groups and adsorbed on the

membrane surface, were responsible for increased hydrophilicity. Thus, it might be considered

that hydrophilicity was the most important factor among the membrane performances [26].

0

20

40

60

80

100

0 1 1.95 2.85 3.8

TiO2 content (wt. %)

Contac

t ang

le (o

)

Figure 3. Contact angle with water of the PVDF composite membrane and standard

deviation on measurement in different regions of the membrane surface

3.2. Effect of additives concentration on the performance of PVDF membranes

As shown in Fig. 4, PTL-10 membranes showed the flux peak value of 140.82 L/m2 h

when TiO2 concentration were at 1.95 wt.% and decreased with further increasing TiO2

concentration. The values of rejection demonstrated the similar trend to the flux, which increased

12

ICRAMST2013

to the peak value 98.8 % at 1.95 wt.% TiO2 concentration then decreased with further increasing

TiO2 concentration. The hydrophilicity TiO2 particles on the membrane surface reduced the

interaction between contaminants and the membrane surface. The increased membrane

hydrophilicity and membrane pore size with lower TiO2 concentration (≤1.95 wt.%) could attract

water molecules inside the composite membrane; facilitated their penetration through the

membrane, enhancing the flux and rejection. However, higher TiO2 concentration (> 1.95 wt.%)

resulted in the formation of a highly viscous dope. This slowed down the formation process of

PVDF UF membranes and produced a compact sublayer, as shown in Fig. 4. Moreover, the

enhanced flux and rejection values were also developed due to air bubbling flow rate per unit

projection membrane area. This was set constantly in order to maintain turbulence, increase mass

transfer coefficient, decrease oil droplets and suspended solid concentrating on the membrane

surface, and weakens the effect of polarization.

0

20

40

60

80

100

120

140

0 0.95 1.95 2.8 3.85

TiO2 wt %

J (L

/m2 h

)

0

20

40

60

80

100

R, (J

R/J

w) %

J w

J R

R

J R/ J w

Figure 4. Effect of the TiO2 concentration on permeation flux and rejection

13

ICRAMST2013

The antifouling properties of PVDF ultrafiltration membranes could be evaluated by the

ratio of refinery wastewater flux (JR) and pure water flux (JW). For the higher antifouling

submerged UF membrane, the feed of refinery wastewater would cause a small flux loss and the

ratio (JR/JW) would be higher. Fig. 4 also shows that initially, the ratio (JR/JW) is increased

sharply and reached the highest peak of 1.95 wt.% TiO2 concentration. However, higher TiO2

concentration, namely, 2.8 and 3.95 wt.% TiO2, resulted the decreased value of the ratio (JR/JW).

The membrane surface hydrophilicity was improved significantly and reduced the interaction

between the contaminants and the membrane surface, then effectively improved the antifouling

properties. However, the ratio (JR/JW) of PVDF UF membranes decreased slightly at >1.95 wt.%

TiO2 concentration until the fouling phenomenon disappeared at 3.85 wt.% TiO2 concentration.

As the hydrophilic additives content in the membrane increased the decrement of ratio (JR/JW)

became gradually less, but it still indicated that the antifouling properties of the PVDF UF

membranes were promising by adding of the inorganic additives LiCl.H2O and TiO2.

4. Conclusions

PVDF UF membranes were fabricated via a dry-jet wet spinning method. Various

concentrations of TiO2 at constant value of LiCl.H2O were used as inorganic additives in the

spinning dopes in order to improve the phase-inversion rate and provide porous asymmetric

membranes with advanced structure for refinery produced wastewater treatment. Several

characterizations and measurement techniques such as membrane structure, surface wettability,

porosity, average pore size, and permeability were utilized to evaluate fine structural details of

14

ICRAMST2013

the membrane and membrane performance. Refinery produced wastewater filtration was

conducted through in-house prepared PVDF hollow fiber ultrafiltration membranes. FESEM

analysis indicated that 19 wt.% PVDF concentration had suppressed both inner and outer

membrane surface finger-like macrovoids with slightly denser skin layer which decreased mass

transport resistance. Addition of 1.95 wt.% TiO2 nanoparticles resulted in smaller nanoparticles

which in turn achieved higher hydrophilicity, small pore size, and high porosity. Permeability

test results illustrated that LiCl.H2O and TiO2 nanoparticles affected the PVDF UF membranes

performance remarkably. Significantly higher flux and rejection of refinery wastewater were

observed. Furthermore, the achieved filtration performance was indicated that approximately

18.3 % flux reduction was gradually occurred within the first 18 h before became constant until

the end of the filtration operation.

References

[1] J.F. Kong, K. Li, Oil removal from oil-in-water emulsions using PVDF membranes, J.

Membr. Sci., 16 (1999) 83-93.

[2] S.P. Deshmukh, K. Li, Effect of ethanol composition in water coagulation bath on

morphology of PVDF hollow fibre membranes, J. Membr. Sci., 150 (1998) 75-85.

[3] K. Jian, P.N. Pintauro, Asymmetric PVDF hollow fiber membranes for organic/water

pervaporation separations, J. Membr. Sci., 135 (1997) 41-53.

15

ICRAMST2013

[4] A.W. Zularisam, A.F. Ismail, and M.R. Salim, Behavior of natural organic matter (NOM) in

membrane filtration for surface water treatment: A review, Desalination, 194 (1-3) (2006)

211-231.

[5] M. Tomaszewska, Preparation and properties of flat sheet membranes from

polyvinylidenefluoride for membrane distillation, Desalination, 104 (1996) 1-11.

[6] M. Khayet, T. Matsuura, Preparation and characterization of polyvinylidene fluoride

membranes for membrane distillation, Ind. Eng. Chem. Res., 40 (2001) 5710-5718.

[7] M. Khayet, C.Y. Feng, K.C. Khulbe, T. Matsuura, Study on the effect of a non-solvent

additive on the morphology and performance of ultrafiltration hollow fibre membranes,

Desalination, 148 (2002) 321-327.

[8] S. Chabot, C. Roy, G. Chowdhury, T. Matsuura, Development of poly(vinylidene fluoride)

hollow fiber membranes for the treatment of water/organic vapor mixtures, J. Apply.

Polym. Sci., 65 (1997) 1263-1270.

[9] W.D. Benzinger, B.S. Parekh, J.L. Eichelberger, High temperature ultrafiltration with

Kynar poly(vinylidene fluoride) membranes, Sep. Sci. Technol., 15 (4) (1980) 1193-1204.

[10] X.C. Cao, J. Ma, X.H. Shi , Z.J. Ren, Effect of TiO2 nanoparticle size on the performance of

PVDF membrane, Apply. Surf. Sci., 253 (2006) 2003-2010.

[11] J. Bush, A. Cruse, W. Marquardt, Modelling submerged hollow-fiber membrane filtration

for wastewater treatment, J. Membr. Sci., 288 (2007) 94-111.

[12] C.H. Lu, W. H. Wu, R. B. Kale, Microemulsion-mediated hydrothermal synthesis of

photocatalytic TiO2 powders, J. Hazard. Mat., 154 (2008) 649-654.

16

ICRAMST2013

[13] M.L. Hami, M.A. Al-Hasyimi, M.M. Al-Doori, Effect of activated carbon of BOD and

COD removal in a dissolved air flotation unit treating refinery wastewater, Desalination,

216 (2007) 116-122.

[14] M.X. Loukidou, A.I.Zouboulis, Comparison of two biological treatment processes using

attached growth biomass of sanitary landfill leachate treatment, Environ. Pollut., 111 (2001)

273-281.

[15] J. Busch, A.Cruse, W. Marquadt, Modelling submerged hollow fiber filtration for

wastewater treatment, J. Membr. Sci., 288 (2007) 94-111.

[16] C.J. Jou, G.C. Huang, A pilot study for oil refinery wastewater treatment using a fixed film

bioreactor, Adv. Environ. Res., 7 (2002) 463-469.

[17] Z. Yuan, X.D. Li, Porous PVDF/TPU blends asymmetric hollow fiber membranes prepared

with the use of hydrophilic additive PVP (K30), Desalination, 223 (2008) 438-447.

[18] T.H. Bae, I.C. Kim, T.M. Tak, Preparation and characterization of fouling-resistant TiO2

self-assembled nanocomposite membrane, J. Membr. Sci., 275 (2006) 1-5.

[19] A. Bottino, G. Capanelli, S. Munari, A. Turturro, High performance ultrafiltration

membranes cast from LiCl doped solution, Desalination, 68 (1998) 167-177.

[20] W.J. Lau and A.F. Ismail, Theoritical studies on the morphology and electrical properties of

blended PES/SPEEK nanofiltration membranes using different sulfonation degree of

SPEEK, J. Membr. Sci., 334 (2009) 30-42.

[21] Z. Wang, Y.W. Lu, S.M. Wang, Study on the preparation and characterizatics of PVDF/CA

blend ultrafiltration membrane, J. Membr. Sci., 22 (6) (2002) 4-8.

17

ICRAMST2013

[22] Y.J. Wang, D.J. Kim, Crystallinity, morphology, mechanical properties and conductivity

study of in situ formed PVDF/LiClO4/TiO2 nanocomposite polymer electrolytes,

Electrochimica Acta, 52 (2007) 3181-3189.

[23] A. Bottino, G. Capanelli, S. Munari, A. Turturro,Effect of coagulation medium properties of

sulfonated polyvinylidene fluoride membranes, J. Apply. Polym. Sci., 30 (1985) 3009-3022.

[24] Z.H. Xu, L. Li, F.W. Wu, S.J. Tan, Z.B. Zhang, The application of the modified PVDF

ultrafiltration membranes in further purification of Ginkgo biloba extraction, J. Membr.

Sci., 255 (2005) 125-131.

[25] E. Fontananova, J.C. Jansen, A. Cristiano, E. Curcio, E. Drioli, Effect of additives in the

casting solution on the formation of PVDF membranes, Desalination, 192 (2006) 190-197.

[26] Z.P. Fang, Y.Z. Xu, C.W. Xu, Modification mechanism of nanoparticles on polymers, J.

Mater. Sci. Eng., 21 (2) (2003) 279-282.

18