DESALINATION Desalination 177 (2005) 187-199
ELSEVIER
www.elsevier.conglocate/desal
Characterisation of nanofiltration membranes using atomic force microscopy N. Hilal a*, H. AI-Zoubi a, N.A. Darwish b, A.W.
Mohammad
c
"Centrefor Clean Water Technologies, School of Chemical, Environmental and Mining Engineering, The University of Nottingham, NG7 2RD, United Kingdom Tel. +44 (115) 9514168; Fax +44 (115) 9514115; email:
[email protected] hDepartment of Chemical Engineering, College of Engineering, Jordan University of Science and Technology, P.O. Box 3030 lrbid, Jordan "Department of Chemical and Proce:'s Engineering, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor Darul Ehsan, Malaysia Received 18 June 2004; accepted 6 December 2004
Abstract Atomic force microscopy (AFM) has been used to characterize five commercial nanofiltration (NF) membranes from three companies. High-resolution 3D images of the membranes were obtained without preparative treatment that may affect the membrane surface. Obtained images have been filtered to overcome the effect of tip convolution and the noises. Two sizes of the images were obtained 2 I~m x 2 gm and small sizes. The first images were used to find the surface morphology data such as average roughness, mean height, root mean square (RMS), and maximum peak-to-valley. The small size images showing visible pores were used to determine the pore size and pore size distributions, which were used to calculate porosity of membranes. A fitted line using lognormal distributions was used to represent the pore size distribution of the nanofiltration membranes. The results show that the Iognormal distribution is fitted well with AFM experimental data.
Keywords: Atomic force microscopy; Nanofiltration membranes; Surface morphology; Pore size distribution; Tip convolution
1. Introduction Atomic force microscopy (AFM) [1 ] has been considered as a powerful addition to microscopy for its ability to generate high-resolution 3D images *Corresponding author.
at subnanometer range with no sample pretreatment. AFM is being used to solve processing and material problems in a wide range o f technologies affecting the electronics, telecommunications, biological chemical, automotive, aerospace, and energy industries [2]. The materials being
0011-9164/05/$- See front matter © 2005 Elsevier B.V. All rights reserved doi: 10.1016/j.desal.2004.12.008
188
N. Hilal et al. / Desalination 177 (2005) 187-199
investigated include thin and thick film coatings [3], ceramic, composites, glasses [4], synthetic [5,6] and biological membranes [7], metals, polymers, semiconductors, proteins [8], lipids [9] ar,d DNA [10]. In addition to its imaging capabilities of the surfaces in atomic resolution, AFM can measure the force of interaction at nano to pico-Newton scale. Therefore, it has attracted the interest of a number of researchers interested in the surface properties of membranes where surfaces can be visualised and forces of interactions can be measured directly in air and aqueous solutions. It is for reasons such as these that the use of AFM is so tantalising to membrane technologists, since it produces images of membrane surfaces in both air and liquid environments without any alterations to surfaces. Hence, it may be used to determine three key parameters that influence membrane separation process [ 11]. These parameters are (i) pore size distribution and surface morphology, (ii) surface electrical properties, and (iii) surface adhesion-membrane fouling. AFM can be applied to the study of all types of membranes including microfiltration, ultrafltration and nanofiltration. Nanofiltration membranes are a relatively new type of membranes, which have properties between those of reverse osmosis (RO) membranes and ultrafiitration (UF) membranes, they have a pore size of the order of 1 nm diameter and are made from a variety of polymers. As a consequence in the differences of their manufacture, nanofiltration membranes have different surface characteristics including roughness. They have many advantages in the field of water treatment such as low operation pressure, high flux, high retention ofmultivalent anion salt and organic molecular, relatively low investment and low operation and maintenance cost. In addition, nanofiltration membranes are recently being used as pre-treatment for desalination process [12]. They can remove turbidity, microorganisms and hardness, as well as a fraction of the dissolved salts form seawater. A recent
comprehensive review on the use ofnanofiltration membranes in water treatment is shown elsewhere [13]. For the above reasons nanofiltration membrane has given rise to worldwide interest. The studying ofnanofiltration membranes (NF) using AFM has been reported briefly in the literature [14-16]. So, a detailed AFM study for different types NF commercial membranes will be addressed in this paper. The atomic force microscope (AFM) hardware consists of a sharp tip located at the free end of a cantilever systematically probes a surface of interest in order to generate a topographical image. As the tip scans the sample, the forces between the tip and the surface cause the cantilever to bend. A detector such as an optical lever measures this deflection and allows a computer to generate a map of the surface topography [11]. Number of AFM imaging modes can be used to scan surfaces of interest; each can be applied in different situations. Comprehensive reviews of AFM operating modes for membranes are shown elsewhere in the literature [10,17]. Three modes are normally in use; contact mode, non-contact mode, and tapping mode. In contact mode the cantilever is less than a nanometer from the surface, the inter-atomic force between the cantilever and the sample is repulsive, and the dominant force component is Born Repulsion. In the non-contact mode, the cantilever is held in the order of several nanometers (~5 nm) from the sample surface, and the inter-atomic force between the cantilever and the sample is dominated by long-range attractive van der Waals interactions. The third mode, tapping mode, is a hybrid of the contact and noncontact system, in which the cantilever vibrates at a tip-sample distance closer to the region of contact imaging. This technique overcomes some of the limitations of both contact and non-contact modes and has been widely used for imaging biological molecules and cells. In the past, scanning electronic microscopy (SEM) was used to study the membrane surface. But this technique has a number of disadvantages
N. Hilal et al. / Desalination 177 (2005) 187-199
such as; (i) SEM is a vacuum technique, which results in extensive drying of the sample, (ii) extensive sample preparation that can result in damage of the original sample thereby producing imaging artefacts, and (iii) the SEM electron beam may cause damage to the sample [18]. On the contrary, AFM can image all types of membranes in both liquid and air environmentswithout special sample preparation. The only membrane preparation required is attaching it onto a steel disk. This provides a great advantage as it avoids distortion of surface structure through specialised sample preparation procedures, such as those needed in high vacuum techniques. In membrane imaging, "contact mode" is the most often used AFM method when imaging in air and "double layer mode" is preferred when imaging in liquid [ 11]. The instrument software stores the image as a data matrix of co-ordinates and can quantify the surface characteristics of the membranes, such as surface roughness, peak-to-valley distance, and mean height through region analysis of the surface. These parameters are very important for the properties of membrane separation [19]. Many authors studied the effect of these parameters on the protein-protein interaction [20] and on choosing the range of silica surface in silica adhesion study [21 ]. In addition, Surface roughness has an important effect in membrane permeability and fouling behaviour [17,22,23]. Pore size and pore size distribution are the main parameters in membrane characterisation. In order to determine pore dimensions for membranes with small pores using the AFM, it is necessary to image at a higher resolution, normally nmz. Such images in conjunction with digital stored line profiles allow the measurement of individual pores. The line profiles provide a quantification of depth in the surface; helpingto distinguish pores from other depressions. As the entrance of the pores is determined visually, the quality of the high-resolution image is most important. Factors affecting data acquisition, such as tip convolution, surface distortion and surface roughness, probe sample
189
interaction, and piezo scanner non-linearity, could contribute to imaging artefacts. Some of these problems, such as white thermal noise, causing spurious unwanted oscillation of the cantilever, can be overcome by using fast Fourier transform filtering (FFT) [24]. Optimisation of operating conditions is necessary in order to obtain high quality image and therefore more realistic surface morphology. Scan speed can play important role in image quality, recent research showed that scan speeds higher than 2 Hz lead to distortion of the membrane pores and it is very important to choose a correct cantilever and scan speed for the best image [i 7]. The cantilever should be sensitive enough to obtain a high-resolution image and also has to be stable. Cantilevers with very low spring constant and high aspect ration tip have been used throughout this work. The objective of this paper is to characterise a number of nanofiltration membranes manufactured by different companies using an AFM. Characterisation will include: pore size and pore size distribution, porosity of the membranes, and the surface analysis of a number of commercial nanofiltration membranes mainly used for pretreatment in the desalination industry.
2. Materials and methods
2.1. Atomic force microscopy The AFM used in the present study was an Explorer (TMX 2000), a commercial device from Vecco Instruments (USA). Silicon cantilevers (Ultralevers Park Scientific Instruments) with a high aspect ratio tip were used to scan the membranes. Contact mode in air medium was used to characterise all NF membranes at room temperature of 25°C. Nanofiitration membrane samples were imaged without preparative procedure that may affect the membrane structure. The only preparation of the membranes for AFM was their attachment to steel discs with double-side-scotch tape. In operation of AFM in contact mode, the
190
N. Hilal et al. / Desalination 177 (2005) 187-199
sharp tip is moved vertically above the samples while a Z-feedback loop was operated to maintain a constant force on the cantilever tip by adjusting the tip-to-sample distance. The vertical motion of the tip due to inter-atomic forces, is detected by sensing, with a four segment photodiode, the deflection of a laser beam reflected off the back of the cantilever. The Z voltage applied to the (piezoelectric) scanner to maintain constant repulsive force was used to map out the surface topography. Low scan rates were used with 300300 pixel resolution. Once a clear image was obtained, all parameters of surface morphology as well as pore size distribution were obtained directly from the image using software analysis associated with the Explorer. Two types of images were obtained for each membrane; one over an area of 2 pm x 2 !am and another one with high resolution over small area where pores are clearly visible. The 2 gm x 2 lam images were used to determine surface morphology data such as average roughness, mean height, root mean square (RMS), and maximum peak-tovalley. The small size images were used to determine the pore size and pore size distributions, which were used to calculate porosity of membranes. All images were filtered to overcome the effect of tip convolution and the noises in the images (if any) using imaging processing in the software of AFM with two choices; (i) the first one is using the convolution with 5x5 filtering matrix (Kernel) and smooth choice, (ii) the second choice is using filter with low or medium Zvalue.
0,2 n m
AFM in conjunction with surface analysis software associated with the Explorer can give full details about features that might be on the surface and also surface analysis including peakto-valley, average roughness and root mean square roughness. Fig. 1 shows typical line and surface analysis of one of the nanofiltration membrane (NF 90). The obtained average roughness (Ra) from the software is defined as the arithmetic mean of the deviation in height from the image to the mean value as shown in Eq. (1).
Ro
1
N
0.1
0
72
Ra: Rp: Rprn: RI: Rim:
(1)
The root mean square roughness (RMS) is defined as the square root of the mean value of 144A
72.&
OA (a)
Standard Roughness 1
z,-2
0.0174 0.9474 0,0352 00977 02165']
nm nm r~rn r~m nm
~-i'..:tar u:: e
H~ ~:1h:
4 i~!?.1. . . rl 01 ,,o.,c,nm
144
Fig. 1. Line profile of (a) 2-dimentional NF90 membrane, (b) line profile data and surface analysis.
b)
N. Hilal et al. / Desalination 177 (2005) 187-199 the squares o f the distance of the points from the image mean value - - Eq. (2).
RMS =
Z, - ~)2
(2)
Average mean height (.4 Vmean) is an arithmetic mean defined as the sum o f all height values divided by the number of data points - - Eq. (3). 1
N
tz[=-~=~ Z,
(3)
where Z is the current Z value, while Z and Nare the average of Z values and the number of points within the area, respectively. Surface porosity of membranes was calculated using the calculated mean pore size and number of pores in an image using Eq. (4).
nl~l
×100% (4)
porosity % =
ma.o
where n is number o f pores, dp is the mean diameter, A ~mage is the surface area o f the typical image.
191
2.2. Nanofiltration membranes Five nanofiltration (NF) membranes from three different companies were characterised using AFM in this study. Table 1 shows the commercial names of these NF membranes and their companies as well as the main polymers used in manufacture of the membranes. All membranes were made from polyamide except for NF-PES-10 which is made from polyethersulfone. Information has been taken from the manufacturers' websites. Other details such as the maximum pressure, maximum temperature, water permeability flow, and salt rejection percentage are shown in Table 2. These details are provided by the manufactures. The maximum operating pressure for all the membranes has the range of 40-40 bar while the maximum operating temperature has a range of 45-95°C. All Dow's membranes have a high rejection ofmultivalent ions except NF200 which has relatively low rejection around 35-50% of Table 1 Properties of NF membranes used in this work (so far) and their manufacturers No 1 2 3 4 5
Membrane NF90 NF200 NF270 NF-PES-10 AFC40
Company Dow Dow Dow NADIR PCI membrane
Polymer Polyamide Polyamide Polyamide Polyethersulfone Polyamide
Table 2 Manufacturers' characterisation of NF membranes Membrane
Maximumoperated Maximumoperating CaCI 2 rejection % (product MgSO4rejection% (product pressure, bar temperature, °C water flow rate, m3/d) water flow rate, m3/d)
NF90 NF200 NF270 NF-PES- I 0 AFC40
41.0 41.0 4 i .0 40.0 60
+ Water product flow (L/m2h) ++Rejection of NaCI A Retention of lactose solution
45 45 45 95 60
85-95 (28.4) 35-50(30.3) 4(L60(55.6) 5-15++(200-400)+ 60 (20-80)+
>97 (36.0) 97 (25.7) >97 (47.3) 25-45 ^(200-400) + >95^(20-80)+
192
N. Hilal et al. / Desalination 177 (2005) 187-199
calcium chloride. The values of rejection of Nadir's membrane NF-PES-10 was given as the rejection of NaC! and as the rejection of lactose (uncharged solutes). It is clear that this membrane has relatively low rejection value with high pure water flux. Although PCI membrane worked at high conditions, it claims to reject salt with medium percent.
i
15r~
3. Results and discussion
(a)
3.1. Visualisation of membrane surfaces In this study, the images were obtained over two different areas; an area of 2 ~tm × 2 lam and small area. The large size was imaged to measure surface morphology of the membranes while the small size images were used to determine the pore size and pore size distribution. Figs. 2 and 3 show high resolution AFM images over small areas. The colour intensity shows the vertical profile of the membrane surface, with light regions being the highest points and the dark points being the depressions and pores. The pores are very clearly visible in all small size images. The diameters of the pores were measured in conjunction with line profiles stored by AFM as shown in section 2. I. The profiles aid in assessing the size of pore entrance on membrane surface. Great care should be taken when measuring the diameters of the pores by AFM as the results depend on a convolution of the macroscopic dimensions of tip and the pore diameter, therefore the effect of tip convolution was taken into account by applying convolution and filter treatments on the obtained images of membranes. Dow's membranes (NF90, NF200, NF270) have small pore size. Fig. 2 shows high resolution images of the surfaces of these membranes with different sizes using contact mode. Images are given over of small sizes with the light regions being the highest points, and the dark regions the depressions (pores). NF images show that the overall surface structure is not similar for the three membranes and they have very different numbers
7r~
~
(b)
}24 tnn
0A
(c)
Fig. 2. AFM images for Dow's membranes with small sizes; (a)NF90, (b) NF200, and (¢) NF270 membranes.
N. Hilal el aL / Desalination 177 (2005) 187-199
o f pores distributed non-uniformly over the scanned area. The resulting pore size distributions were found to be: for the NF90 membrane with a mean of 0.55 nm and a standard deviation o f 0.126 nm; for the NF200 membrane the range was 0.2-0.48 nm with a mean of 0.31 nm and a standard deviation of 0.066 nm; for the NF270 membrane the range was 0.47-0.99 nm with a mean of 0.71 nm and a standard deviation of 0.14 nm. These data show that the membranes have a narrow size distribution with low values of standard deviation. In addition, NF200 has the highest number of pores -200 over an area of 10 nm x I 0 nm, but with small size diameter, while NF270 has the smallest number of pores -125 over an area of 20 nm × 20 nm. Nadir membrane NF-PES-10 and PCI membrane AFC40 are shown in Fig. 3. Again, the light regions being the highest points and the dark points being the depressions and pores. The pores are very clear in the two images. NF-PES-10 has around 75 pores over area of 9 nm × 9 nm with mean diameter of 0.33 nm and very low value o f standard deviation of 0.059 nm, while AFC40 has 64 pores over area of12.2 nm x 12.2 nm with mean diameter o f 0.48 nm and standard deviation of 0.069 nm.
193
Surface analysis o f membranes was investigated to the membranes over an area o f 2 ~tm × 2 ~tm. Fig. 4 shows a typical o f selected image o f the NF200 and NF270 membranes over an area of 2 ~tm × 2 ~tm.As shown in the figure, the images have light and dark regions. The light region represents the highest points in the membranes, while the pores are mainly founded in the dark points. Characterisation of the surfaces was achieved by measuring the roughness, average roughness, mean height, and peak-valley. Table 3 shows these parameters for all studied NF membranes. These parameters were defined in section 2.1. Roughness is one of the most important surface properties as it has a strong influence on adhesion (fouling) and also on local mass transfer [15]. Adhesive force has been shown to be larger for membranes with high roughness compared to the smooth membranes [25,26]. It is clear from Table 3 that NF90 has the highest value o f roughness among the investigated NF membranes. This means that this membrane is expected to have relatively high adhesive force and as the result high fouling on its surface. NF200, NF270, and AFC40 have low value of roughness while NF-PES-10 membrane has the lowest value of roughness. So, NF-PES-I 0
~ d a ~ o B t t * , l T t l o n m a r n h t . a ~ o . M F - P F . ~ . IO
P C I m,
J
122A~
0.24 nr0 0,26ran
0A
(a)
uA
Fig. 3, AFM images for NF membranes with small sizes; (b)NF-PES-IO, (b)AFC40.
(b)
194
N. Hilal et al. f Desalination 177 (2005) 187-199
~5 n m nrfl
9r~
0r~
(a)
0nr6"0nm
(b)
Fig. 4. Typical AFM images to the Dow's membranes with size to of 2 txm × 2 i~m; (a) NF200, (b) NF270 membranes.
Table 3 Surface characteristics ofsmall size NFmembranesasmeasured by AFM Membrane
Avemge+roughn~s ~,nm
Roughness + RMS, nm
Mean height+, nm
Peak-valley +, nm
% porosi~
NF90 NF200 NF270 NF-PES-10 AFC40
22.7632 2.7098 3.3611 1.76365 2.5142
27.7515 3.6750 4.3797 2.33985 3.2660
43.0776 5.3582 6.5646 3.1614 4.3702
142.8548 30.2165 32.8695 16.1155 20.2515
17.1 15.5 11.7 7.9 7.8
+These values were taken the averageto three different images was considered to be the smoothest membrane among the all studied membranes. Surface roughness has been shown to relate to initial membrane fouling behaviour [27]. 3.2. Determination of pore size distribution Pore size and size distributions of each membrane, which are shown in Figs. 5 and 6, have been determined using line analysis software described in section 2. The pore size distribution identifies the pore size width of the membrane. Fifty pores were measured in each membrane. In order to represent the pore size of all membranes with the suitable fitted equation, Iognormal distributions has been chosen for fitting the data. This was found to give a good fitting to the pore size
distribution data. The Iognormal distribution shown in equation (5) was represented by the percentage frequencies (% f) [28,29].
(5) where %f is the percentage of pore density, %fm~ is the maximum of %f, la is the value of pore diameter d that makes %fmax maximum (most probable d ), and ~ is the standard deviation or width of the distribution. The fitted line shown in the pore size distribution in Figs. 5, 6 corresponds well with the pore size distribution, providing a good justification for the use of such a distribution in calculations. The diameters (la) obtained by the
N. Hilal et aL / Desalination 177 (2005) 187-199
(a)
AFM data fitted data
® /.//:C" o
fitted equation %[= %jfmax exp[-O.5(In(d~/a)! o) 2] W~0"513-+0"0t9 nm er= 0.280_+0.042 nm
B.!
°i N
{}i 0.2
............
-
0,3
0.4
0.5
0,7
0.6
0.8
0,9
~0
Poresize (rim) 35~
(b)
..............
fitted equation %f= %fm~x exp[-0.5(In(d~,/~t), o) ~] it= 0.2886_+0,0l 7 nm ~ ~ ~..-¢°=0"271 +0.060 nm
30 25
20 15 10 5
0 0.10
0,15
020
{},25
0,;30 0.35
0.40
0,45
0 . 5 { } 0.55
0.60
Pore size (nm)
'J6-.
(c)
~?~ i f ~ ':~
~4 i
o=
10
&
6
fitted equation %]'= % / ~ x exp[-O.5(ln(dp/la)/o) 2] B = 0.681+_0.016 nm o = 0.223+0.026 nm
!i!~i
......
NN®
6 4
2 0
........
0,3
0.4
0.5
0.6
N® 0.7
o.6
o.~
~.o
~,~
Pore size (nm)
Fig. 5. Pore size distribution of (a) NF90, (b) NF200, (c) NF270 Dow membranes with fitted Eq. (5).
195
196
N. Hilal et aL / Desalination 177 (2005) 187-199
(a)
30-~
25
2O
®NI' s
0~30
0.35
0,,40
0,45
O,C~
0,55
0,60
0.65
0,70
Pore Size (rim)
fitted ~]~tion %1= 0'~I,~ ex4-0.s0n(~@)/o)-]":' it= 0.307_+0.0023nm o=0.2(D~.008 nm
(b) 36
30
26
x~
o
n
1
0.20
0.25
0.30
0,~
0.40
0.45
0.50
0,,5~5
Pore s~e (nm)
Fig. 6. Pore size distribution of(a) AFC40, (b) NF-PES-10 membranes with fitted Eq. (5).
fitting Eq. (5) is very close to the mean diameter obtained by AFM while the values of represent the width of the pore size of the membranes. The fitted of equation with values of ~tand were shown in the plots o f pore size distribution in Figs. 5, 6, while all the constants and parameters of Eq. (5) including the value of R that is obtained from the fitting calculation are shown in Table 4. Fig. 5 shows the pore size distribution with the fitted equations for three Dow's membrane (NF90, NF200, and NF270). It is clear that the fitted line
in the three membranes is very close to the data obtained by AFM with very small error and the values of Ia are very close to the values of the diameter obtained by AFM. Moreover, the figures confirm, as mentioned above, that pore size of NF200 membrane is the smallest size among all studied NF membranes. Fig. 6 shows that the pore size distribution of AFC40 and NF-PES- 10 membranes. According to the values shown in Table 4, the fitted lines are very close to the AFM data.
N. Filial et al. / Desalination 177 (2005) 187-199
197
Table 4 Summary of all obtained constants from fitted curves (Iognormaldistributions) in Eq. (5) Membrane
R-value
%./m~
~t, nm
c, nm
NF90 NF200 NF270 NF-PES-10 AFC40
0.857 0.864 0.891 0.997 0.967
14.59a:1.61 26.01±4.71 13.74±1.27 33.25±1.03 25.23±1.983
0.513±0.019 0.2886±0.017 0.681±0.016 0.307±0.0023 0.460-~:0.0075
0.280-~0.042 0.271~0.060 0.223±0.026 0.20001-0.008 0.178±0.017
3.3. Surface porosity Surface porosity was calculated from the ratio of the area of the pores to the total area in the typical image as shown in Eq. (4). The surface porosity gives a good idea about the percentage area of membrane that the fluid may pass through it. Table 3 shows values of calculated porosities of the investigated membranes. As shown in Table 3, NF90 has a relatively high value of porosity. This is because NF90 has more pores with larger pore diameter. NF200 and NF270 have lower value of porosity. Finally, NF-PES-10 and AFC40 have the lowest values of porosity. Again, the reason for that these membranes have low number of pores and relatively low pore size diameter. In summary, the mean pore size of all membranes is shown in Fig. 7 in which the diameter of the membranes has a range of 0.31-0.71 nm. It is clear that NF200 has the smallest mean pore diameter (0.31 nm), while NF270 has the largest mean diameter (0.71 nm). It should be noted that the pore size alone dose not govern the efficiency of NF membrane in separation process in water treatment [30]. In addition, NF90 is the roughest membrane with high value of porosity while the NF-PES-10 is the smoothest membrane with the lowest value of porosity. Finally, all investigated membranes have porosity in the range of 7.817. 1%. Li and Elimelech [31 ] studied organic fouling and chemical cleaning of nanoflitration membranes, they measured the adhesion force between foulant and and NF270 membrane surface using carboxylate modified AFM colloid probe as a surrogate for humic acid.
=E
0.8, i 0.6 i 0.4
O D.
0.2 0 NF90
NF200
NF270
NF-PES-10
AFC40
NF membranes
Fig. 7. The mean pore diameter of NF membranes measured by AFM.
4. Conclusions
The present paper has shown that atomic force microscopy is a valuable tool for the characterisation of surface morphology ofnanofiltration membranes. AFM can generate high-resolution 3D images of membrane surfaces without any previous treatment of the membrane surface that can alter the membrane structure. Two different sizes of images of five NF membranes form three companies were characterised using AFM. Surface properties were estimated using images over an area of 2 ~m × 2 p,m, while the small size images were used to measure pore size, pore size distribution, and calculate porosity. The imaging processing in the software of AFM was used to overcome the effect of tip convolution and the noises in the original images. A good fitting line using iognormai distributions was used to represent the pore size distribution.
i 98
N. Hilal et al. / Desalination 177 (2005) 187-199
NF200 was found to have the lowest diameter among investigated NF membranes in this paper while NF270 has the largest diameter. In addition, NF90 was the roughest membrane with high value o f porosity while NF-PES-10 was the smoothest one with the lowest value of porosity.
Acknowledgements We thank the Middle East Desalination Research Centre (MEDRC) for funding this work.
References [I] G. Binnig, C.F. Quate and C. Gerber, Atomic force microscope, Phys. Rev. Lett., 56 (1986) 930. [2] S.A.C. Gould, The atomic force microscope: a tool for science and industry, UItramicroscopy,33 (1990) 93-98. [3] D. Raghavan, X. Gu, T. Nguyen and M. Vanlandingham, Characterization of chemical heterogeneity in polymer systems using hydrolysis and tapping-mode atomic force microscopy, J. Polymer Sci.: Part B, Polymer Physics, 39(13) (2001 ) 14601470. [4] E. Lester, N. Hilal and J. Henderson, Porosity in ancient glass from Syria (c. 800 A.D.) using gas adsorption and atomic force microscopy,J. Sur. Inter. Analysis, 2004, in press. [5] W.R Bowen, N. Hilal, R.W. Lovitt and P.M. Williams, Atomic force microscope studies of membranes: surface pore structures of Anopore and Cyclopore membranes, J. Membr. Sci., 110(2) (1996) 233-238. [6] W.R Bowen, N. Hilal, R.W. Lovitt and P.M. Williams, Atomic force microscope studies of membranes: surface pore structures of Diaflo ultrafiltration membranes. J. Sur. Inter. Analysis, 180(2) (1996) 350-359. [7] M.H. Beckmannn, A. Kolp and F. Lang, Atomic force microscopy of biological cell membranes: from cell to molecules, Microscopy and Analysis, 45 (1995) 7-9. [8] H.X. You and C.R. Lowe, AFM studies of protein adsorption, 2. Characterization of immunoglobulin adsorption by detergent washing, J. Colloid lnterf. Sci., 182 (1996) 586--601.
[9] J. Yang, L.K. Tamm, A.P. Somlyo and Z. Shao, Promises and problems of biological atomic force microscopy, J. Microscopy, 171 (1993) 183-198. [10] M. Benzanilla, S. Manne, D.E. Laney, Y.L. Lyubchenko and H.G. Hansma, Adsorption of DNA to MICA, silylated MICA and minerals characterisation by atomic force microscopy, Langmuir, ! 1 (1995) 655-659. [11] W.R. Bowen, N. Hilal, R.W. Lovitt, P.M. Williams and C.J. Wright, in Surface Chemistry and Electrochemistry of Membrane Surfaces, Surfactant Science Series, T.S. Sorensen, Ed., Dekker, New York,, 1999, chap. 1, pp. 1-37. [12] A. Hassan, M. AI-So, A. AI-Amoudi, A. Jamaluddin, A. Farooque, A. Rowaili, A. Dalvi, N. Kither, G. Mustafa and I. A1-Tisan,A new approach to thermal seawater desalination processes using nanofiltration membranes (Part 1), Desalination, I 18 (1998) 3551. [13] N. Hilal, H. Al-Zoubi, N.A. Darwish and A.W. Mohammad, A comprehensive review ofnanofiltration membranes: treatment, pretreatment, modelling, and atomic force microscopy, Desalination, 170 (2004) 28 !-308. [14] W.R. Bowen, A. Mohammad and N. Hilal, Characterisation of nanofiltration membranes for predictive purposes use of salts, uncharged solutes and atomic force microscopy, J. Membr. Sci., 126 (1997) 91105. [15] W.R. Bowen and T.A. Doneva, Atomic force microscopy studies ofnanofiltration membranes: surface morphology, pore size distribution and adhesion, Desalination, 129 (2000) 163-172. [16] A.W. Mohammad, N. Hilal and M. Abu Seman, A study on the producing composite nanofiltration membrane with optimised properties, Desalination, 158 (2003) 73-78. [17] W.R. Bowen and T.A. Doneva, Atomic force microscopy of membranes, In Encyclopedia of Surface Colloid Science, Marcel Dekker, Inc., 2001. [18] J.A. Stoton, The application of atomic force microscopy in the characterisation of membrane surfaces, PhD thesis, University of Wales Swansea, 2001. [19] A.G. Fane and C.J.D. Fell, A review of fouling and fouling control in ultrafiltration, Desalination, 62 (1987) 117-136. [20] W.R. Bowen, N. Hilal, R.W. Lovitt and C.J. Wright, Direct measurement of interactions between adsorbed protein layers using an atomic force micro-
N. Hilal et al. / Desalination 177 (2005) 187-199 scope, J. Colloid Interf. Sci., 197(i 998) 348-352. [21 ] W.R. Bowen, N. 14ilal, R.W. Lovitt and C.J. Wright, Atomic force microscopy study of the adhesion of a silica sphere to a silica surface - - effect of surface cleaning. Colloids and Surfaces A: Physico. Engin. Aspects, 157 (1999) 117-125. [22] M. Elimelech, X. Zhu, A.E. Childress and S. Hong, Role of membrane surface morphology in colloidal fouling of cellulose acetate and composite aromatic polyamide reverse osmosis membranes, J. Membr. Sci., 127 (1997) 101-109. [23] X. Zhu and M. Elimelech, Colloidal fouling of reverse osmosis membranes: measurements and fouling mechanisms, Environ. Sci. Technol., 31 (1997) 3654-3662. [24] W.R. Bowen and T.A. Doneva, Artefacts in AFM studies of membranes: correcting pore images using fast Fourier transform filtering, J. Membr. Sci., 171 (2000) 141-147. [25] W.R. Bowen, N. Hilal, R.W. Lovitt and C.J. Wright, A new technique for membrane characterisation: direct measurement of the force of adhesion of a single particle using an atomic force microscope, J. Membr. Sci., 139 (1998) 269-274. [26] W.R. Bowen and T.A. Doneva, Atomic force microscopy studies of membranes: effect of surface
[27]
[28]
[29]
[30]
[31]
199
roughness on double-layer interactions and particle adhesion, J. Colloid interf. Sci., 229(2000) 54A, 549. E.M. Vrijenhoek, S. Hong and M. Elimelech, Influence of membrane surface properties on initial rate of colloidal fouling of reverse osmosis and nanofiltration membranes, J. Membr. Sci., 188 (2001) 115-128. N.A. Oehoa, P. Pradanos, L. Palacio, C. Pagliero, J. Marchese and A. Hernandez, Pore size distributions based on AFM imaging and retention of multidisperse polymer solutes characterization of polyethersulfone UF membranes with dopes containing different PVP, J. Membr. Sci., 187 (2001) 227-237. A.W. Mohammad, N. Ali and N. Hilal, Investigation characteristics of increasing molecular weight cutoff polyamide nanofiltration membranes using solute rejection and atomic force microscopy, Sep. Sci. Technol., 38 (2003) 1307-1327. W.R. Bowen and A.W. Mohammad, A theoretical basis for specifying nanofiltration membranes - dye/salt/water streams, Desalination, I 17 (1998) 257-264. Q. Li and M. Elimelech, Organic fouling and chemical cleaning of nanofiltration membranes: measurements and mechanisms, Environ. Sci. Technol., 38 (2004) 4683-4693.
