Enhancement of Pervaporation Performance of Composite Membranes Through in Situ Generation of Silver Nanoparticles in Poly (Vinyl Alcohol) Matrix
Enhancement of Pervaporation Performance of Composite Membranes Through in Situ Generation of Silver Nanoparticles in Poly (Vinyl Alcohol) Matrix
Enhancement of Pervaporation Performance of Composite Membranes Through in Situ Generation of Silver Nanoparticles in Poly (Vinyl Alcohol) Matrix
ABSTRACT: Composite membranes were prepared from an aqueous solution of poly(vinyl alcohol) (PVA) and silver sulphate. The silver nanoparticles were generated in situ before crosslinking PVA matrix by reduction of silver ions using sodium borohydride.
Physico-chemical properties of the resulting composite membranes were studied using Fourier transform infrared spectroscopy
(FTIR), UVvis spectroscopy (UVvis), thermogravimetric analysis (TGA), Wide-angle X-ray diffraction (WAXD), scanning electron
microscopy (SEM), and universal testing machine (UTM). The UVvis spectrum shows a single peak at 410 nm due to surface plasmon absorption of silver nanoparticles. This surely specified that silver nanoparticles are generated in PVA matrix. The membranes
were under go pervaporation (PV) for separation of water from isopropanol at different temperatures. The results indicated that
hydrophilicity and amorphous nature of the membranes were increased with increasing silver nanoparticles in PVA matrix. The swelling and separation performance of the membranes were studied. Both permeation flux and separation factor were increased with
increasing silver nanoparticles in PVA matrix. The results showed that the membrane containing 2.5 mass% of Ag salt exhibited excellent PV performance. The values of total flux and flux of water are almost closed to each other, indicating that membranes could be
effectively used to break the azeotropic point of water-isopropanol. The long-term test was performed at room temperature and ascertained that membranes were durable up to 30 days for the dehydration of IPA. On the basis of the estimate Arrhenius activation
energy values, the efficiency of the membranes was discussed. The calculated DHs values are negative for all the membranes, indicatC 2014 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015, 132, 41248.
ing that Longmuirs mode of sorption is predominant. V
INTRODUCTION
WWW.MATERIALSVIEWS.COM
41248 (1 of 11)
ARTICLE
WILEYONLINELIBRARY.COM/APP
Materials
w 125,000), glutaraldehyde, isopropanol, silver sulPVA (M
fate, and sodium borohydride (NaBH4) were purchased from
WWW.MATERIALSVIEWS.COM
41248 (2 of 11)
ARTICLE
WILEYONLINELIBRARY.COM/APP
DS%5
Ws 2Wd
3100
Wd
(1)
WWW.MATERIALSVIEWS.COM
Pervaporation Experiments
PV experiments were carried out using the in-house designed
set up as reported in our previous articles.32,33 The capacity
of the feed chamber was about 250 cm3 and the effective
membrane surface area in contact with the feed mixture was
34.23 cm2. The vacuum in the downstream side of the apparatus was maintained [1.333224 3 103 Pa (10 Torr)] with a
two-stage vacuum pump (Toshniwal, Chennai, India). The
composition of water in the feed mixture was varied from 5
to 25 mass%. Before performing PV experiments, the test
membrane was allowed to reach equilibrium for about 2 h
in the feed compartment with a known volume of feed mixture. After a steady state was attained, the permeate was collected in a trap immersed in liquid nitrogen jar on the
downstream side at a regular intervals of time. The experiments were carried out at 30, 40, and 50 C. The flux was
calculated by weighing the permeate on a digital microbalance. The composition of water and isopropanol in the
permeate was calculated by measuring the refractive index of
the permeate using an Abbes refractometer (sensitivity 6
0.0001, Atago-3T, Japan), and by comparing it with a standard plot that was previously established with the known
compositions of water-isopropanol mixtures. All the experiments were performed at least three times, and the average
results were reported. The results of permeation for waterisopropanol mixtures during the PV were reproducible
within the range of 65%.
On the basis of the PV data, separation performance of the
membranes was evaluated in terms of total flux (J) and separation selectivity (asep). These were calculated respectively using
the following equations:
J5
asep 5
W
A:t
Pw =PIPA
Fw =FIPA
(2)
(3)
41248 (3 of 11)
ARTICLE
WILEYONLINELIBRARY.COM/APP
Figure 2. UVvisible absorption spectra of crosslinked PVA and its AgNps incorporated composite membranes: (M) 0 mass%; (M-1) 1 mass%;
(M-2) 1.5 mass%; (M-3) 2 mass%; (M-4) 2.5 mass% of Ag-Nps.
Membrane Characterization
UVVisible Analysis. The absorption spectra of crosslinked
PVA and its composite membranes are shown in Figure 2. It is
observed that no absorbance peak was found in the investigated
wavelength range for the crosslinked PVA membrane (M). However, all the composite membranes showed a broad peak and
whose maximum absorbance occurred at 410 nm. This is attributed to an occurrence of surface plasmon resonance of silver
nanoparticles (Ag-Nps) in the resulting composite membranes.34 As the content of silver nanoparticles was increased in
the membranes, the absorbance at 410 nm was increased. This
ensures the formation of Ag-Nps as well as their increased concentration in the membrane matrix. This is very well demonstrated in the actual photographs of the crosslinked PVA and its
Ag-Nps incorporated composite membrane films presented in
Figure 3.
FTIR Analysis. The spectra of crosslinked PVA and its composite membranes were recorded using FTIR and the spectra thus
obtained are presented in Figure 4. The crosslinked PVA (M)
exhibits a characteristic band at around 3400 cm21 and multiple bands between 1000 and 1150 cm21, which correspond
to AOH and CAO stretching vibrations, respectively. Upon
Figure 3. Photographs of the crosslinked PVA and its Ag-Nps incorporated composite membranes. [Color figure can be viewed in the online issue, which
is available at wileyonlinelibrary.com.]
WWW.MATERIALSVIEWS.COM
41248 (4 of 11)
ARTICLE
WILEYONLINELIBRARY.COM/APP
through the membrane and thereby enhances the selective permeants in the composite membranes.
Thermogravimetric Analysis (TGA). Thermal stability of the
crosslinked PVA and its composite membranes was analyzed
using TGA under nitrogen atmosphere, and the resulting thermograms are shown in Figure 6. All the membranes exhibited
three consecutive steps for weight loss and the first weight loss
is around 5%, which occurred between 60 and 110 C. This is
attributed to desorption of physically absorbed water molecules
in the membrane. Although second weight loss was started at
around 230 C for all the membranes, the percentage weight loss
was about 60% varied from 230 to 380 C, which corresponds
to the decomposition of PVA chain. This weight loss was
decreased correspondingly with increasing the amount of AgNps in crosslinked PVA matrix. The third weight loss occurred
between 380 and 480 C, and is attributed to decomposition of
glutaraldehyde. All these clearly suggest that Ag-Nps incorporated membranes demonstrated better thermal stability than
that of crosslinked PVA membrane.
Figure 5. Wide-angle X-ray diffraction patterns of crosslinked PVA and its
Ag-Nps incorporated composite membranes: (M) 0 mass%; (M-1) 1
mass%; (M-2) 1.5 mass%; (M-3) 2 mass%; (M-4) 2.5 mass% of Ag-Nps.
WWW.MATERIALSVIEWS.COM
Figure 7. SEM micrographs of crosslinked PVA and its Ag-Nps incorporated composite membranes: (A) surface view and (B) cross-sectional view.
41248 (5 of 11)
ARTICLE
WILEYONLINELIBRARY.COM/APP
Membrane
Tensile strength
(MPa)
Elongation at
break (%)
76
86
M-1
63
56
M-2
54
50
M-3
51
36
M-4
47
25
Scanning Electron Microscopy (SEM) Analysis. Figure 7 illustrates the SEM photographs of the surface and cross-sectional
views of crosslinked PVA and its composite membranes. At
lower concentration, Ag-Nps were distributed evenly throughout
the crosslinked PVA matrix. However, at higher concentration,
we can notice Ag-Nps on the surface as well as in the bulk of
the material. Further the photographs explicitly show that silver
particles implanted in the membrane matrix have no voids
around them. This ensures that Ag-Nps incorporated membranes obtained here are free from the possible defects and can
be employed for PV process without any difficulty.
Mechanical Properties
Tensile strength and percent elongation of a membrane often
tell its suitability for PV applications. The effect of Ag-Nps on
the tensile strength and percent elongation of the crosslinked
membranes were studied and the data thus obtained are summarized in Table I. It could be seen from the data that both
tensile strength and percent elongation were respectively
decreased from 76 to 47 MPa and 86 to 25% with increasing
the content of Ag-Nps. This was expected due to an establishment of ionic clusters in the membranes and this becomes
prominent as the content of Ag-Nps was increased in the membrane. Despite of the fact, the resulting membranes are suitable
for the PV separation.
WWW.MATERIALSVIEWS.COM
41248 (6 of 11)
ARTICLE
WILEYONLINELIBRARY.COM/APP
WWW.MATERIALSVIEWS.COM
Figure 12. Variation of flux and selectivity with different mass% of AgNps at 10 mass% of water in the feed.
Figure 13. Variation of total flux, fluxes of water, and isopropanol with
different mass% of Ag-Nps at 10 mass% of water in the feed.
41248 (7 of 11)
ARTICLE
WILEYONLINELIBRARY.COM/APP
Table II. Diffusion Coefficients of Water and Isopropanol for All Membranes at Different Mass% of Water in the Feed
Dw 3 108 (cm2/s)
Mass%
of water
M-1
M-2
M-3
5.97
7.82
8.04
9.62
12.20
5.97
5.81
5.58
5.45
4.61
10
6.20
8.47
9.55
11.63
14.41
10.73
10.42
10.10
8.52
7.43
15
8.06
9.68
10.80
11.91
15.21
22.90
20.91
20.51
19.30
18.70
20
9.94
11.20
12.11
13.01
15.52
43.70
41.80
41.71
39.31
38.14
25
12.12
12.52
13.12
14.22
17.22
78.12
72.72
69.82
69.20
68.54
M-4
plasticized upstream membrane surface allows some of isopropanol molecules along with selective permeants, which causes a negative impact on the membranes selectivity.
On the contrary, the selectivity was increased remarkably from
membrane M-1 to M-4, upon increasing the Ag-Nps in the
membrane. This was assigned to increased selective interaction
between membrane and the water molecules owing to the presence of Ag-Nps in the crosslinked matrix. This was further
reveals from the Figure 12, in which flux and selectivity were
plotted as a function of silver content in the membranes at 10
mass% of water in the feed. Because of crosslinking or incorporation of fillers into the membrane matrix packing density of
the membranes increases, as a result the permeation flux
decreases and selectivity increases.18,37 However, in the present
study both permeation flux and selectivity were increased with
increasing Ag-NPs in the membrane. This is due to a trade-off
phenomenon existing between flux and selectivity in PV process, a significant enhancement of hydrophilicity and establishment of greater interaction by the incorporation of Ag-Nps in
the crosslinked PVA matrix overcome this phenomenon.
The efficiency of membranes was assessed by plotting the individual fluxes as a function of mass% of silver at 10 mass% of water
in the feed are shown in Figure 13. It is clearly indicated that the
total flux and flux of water are almost closed to each other,
whereas the flux of isopropanol is negligibly small, indicating that
the membranes developed in the present study by the incorporation of Ag-Nps are highly selective towards water with a wonderful
improvement in the flux as compared to membrane M. On the
basis of this plot, it is further concluded that membrane containing 2.5 mass% of Ag-Nps demonstrated excellent flux and selectivity among the membranes developed in the present study.
Diffusion Coefficient
A proper understanding of the PV, the solution-diffusion mechanism is accepted by several researchers. According to this
M-1
M-2
M-3
M-4
mechanism, PV process consists of three consecutive steps: sorption of the permeant from the feed mixture to the membrane,
diffusion of the permeant in the membrane and desorption of
the permeant to the vapor phase on the downstream side of the
membrane.38 Thus, rate of permeation and selectivity are governed by the solubility and diffusivity of each component of the
feed liquid to be separated. In the PV process, the diffusion step
controls the transport of penetrants, because of the establishment
of fast equilibrium distribution between bulk feed and the
upstream surface of a membrane.39 It is therefore, to understand
the mechanism of transport of molecules important to estimate
the diffusion coefficient, Di of penetrating molecules.
From Ficks law of diffusion, the diffusion coefficient can be
expressed as:40
Ji 52Di
dCi
dx
(4)
Ji d
Ci
(5)
Table III. Pervaporation Flux and Separation Selectivity of All Different Membranes at Different Temperatures for 10 Mass% of Water in the Feed
J 3 102 (kg/m2h)
asep
Temp.
( C)
M-1
M-2
M-3
M-4
M-1
M-2
M-3
M-4
30
3.18
4.31
4.84
5.85
7.16
244
271
293
361
634
40
6.25
6.82
7.08
8.24
10.19
197
213
271
293
329
50
10.30
10.87
12.92
13.23
14.19
144
152
184
213
244
WWW.MATERIALSVIEWS.COM
41248 (8 of 11)
ARTICLE
WILEYONLINELIBRARY.COM/APP
Figure 14. Variation of log J with temperature for crosslinked PVA and its
Ag-Nps incorporated composite membranes at 10 mass% of water in the
feed.
WWW.MATERIALSVIEWS.COM
Figure 15. Variation of log D with temperature for crosslinked PVA and
its Ag-Nps incorporated composite membranes at 10 mass% of water in
the feed.
M-1
M-2
M-3
M-4
Ep
47.92
39.85
37.64
33.19
27.87
ED
48.13
40.81
37.91
33.81
28.14
Epw
46.93
39.14
36.63
32.51
26.97
EpIPA
68.32
60.14
58.018
53.95
52.40
EDw
47.87
40.23
37.49
33.51
27.94
DHs
20.21
20.96
20.27
20.62
20.27
41248 (9 of 11)
ARTICLE
WILEYONLINELIBRARY.COM/APP
(7)
ACKNOWLEDGMENTS
26. Zhang, Q. G.; Liu, Q. L.; Chen, Y.; Chen, J. H. Ind. Eng.
Chem. Res. 2007, 46, 913.
REFERENCES
1. Jeffrey, R.; Long, J.; Yaghi, O. Chem. Soc. Rev. 2009, 38,
1213.
28. Alghezawi, N.; Sanli, O.; Aras, L.; Asman, G. Chem. Eng.
Process. 2005, 44, 51.
29. Srinivasa Rao, P.; Smitha, B.; Sridhar, S.; Krishnaiah, A. Sep.
Purif. Technol. 2005, 44, 244.
30. Gautam, A.; Ram, S. Mater. Chem. Phys. 2010, 119, 266.
31. Calinescu, I.; Patrascu, M.; Gavrila, A. I.; Trifan, A.;
Boscornea, C. U. P. B. Sci. Bull. Series B, 2011, 73, 4.
WWW.MATERIALSVIEWS.COM
M.;
ARTICLE
WILEYONLINELIBRARY.COM/APP
36. Gonzalez-Campos, J. B.; Prokhorov, E.; Sanchez, I. C.; LunaBarcenas, J. G.; Manzano-Ramirez, A.; Gonzalez-Hernandez,
J.; opez-Castro, Y. L.; del Rio, R. E. J. Nanomater. 2012,
2012, 11.
WWW.MATERIALSVIEWS.COM