Effect of vertically aligned carbon nanotube density on the water flux and salt rejection in desalination membranes
© The Author(s) 2016
Received: 7 April 2016
Accepted: 6 July 2016
Published: 22 July 2016
In this paper, vertically aligned carbon nanotube (VACNT) membranes of different densities are developed and their performances are investigated. VACNT arrays of densities 5 × 109, 1010, 5 × 1010 and 1011 tubes cm−2, are initially grown on 1 cm × 1 cm silicon substrates using chemical vapour deposition. A VACNT membrane is realised by attaching a 300 μm-thick 1 cm × 1 cm VACNT array on silicon to a 4″ glass substrate, applying polydimethylsiloxane (PDMS) through spin coating to fill the gaps between the VACNTs, and using a microtome to slice the VACNT–PDMS composite into 25-μm-thick membranes. Experimental results show that the permeability of the developed VACNT membranes increases with the density of the VACNTs, while the salt rejection is almost independent of the VACNT density. The best measured permeance is attained with a VACNT membrane having a CNT density of 1011 tubes cm−2 is 1203 LMH at 1 bar.
Within the last one and half decades, many researchers have worked on different types of CNT based membranes. Vertically aligned carbon nanotubes (VACNT) embedded in a polymer matrix have been developed and tested for gas and liquid transport and filtration. Hinds et al. (2004) have pioneered the multiwall carbon nanotubes (MWCNTs) sealed membrane and observed that liquid transportation was much faster than that predicted by the hydrodynamic theory. Holt et al. (2006) have adopted the same concept and developed a membrane using a chemical vapour deposited (CVD) double-wall carbon nanotube (DWCNT) matrix in silicon nitride. Gas transportation was more than one order rapid than predicted by the Knudsen diffusion model. Kim et al. (2007) have used single-wall carbon nanotubes and incorporated them into existing membranes. The space between CNTs was filled with polymer and the permeance of the membrane for various gases was investigated, demonstrating a reduction in permeability, mainly caused by the polymer layer.
In all above-mentioned works, the total flux (of liquid or gas) was typically dependent on the type of the used CNTs and their densities. As each membrane structure was prepared for a specific application, with different polymer materials being used to fill the space between CNTs, no definite conclusion has so far confirmed the effect of CNT density on the membrane’s performance. Conventionally, VACNT membranes have been fabricated using either compression and rolling techniques (Yu et al. 2009), with the main aim of research being to improve the membrane permeability without affecting the salt rejection property.
Recently, Wang et al. have reported wafer-scale transfer of VACNT arrays (Wang et al. 2014), demonstrating that after a short time of weak oxidation, VACNTs can be easily detached from the native growth substrates, and thus, a freestanding VACNT film can be obtained. This demonstration opens the way for the development of large-size VACNT-based membranes by transferring multiple VACNT films onto large-scale membranes (or substrates) for commercial applications.
In this paper, the permeance and salt rejection properties of four membranes of different VACNT densities are experimentally investigated. The developed VACNT membranes display adequate permeability and salt rejection in comparison with previously reported membranes (Hinds et al. 2004; Holt et al. 2006; Kim et al. 2006, 2007; Yu et al. 2009; Sharma et al. 2010).
Drops of liquid PDMS were added to and spread over the VACNTs using a spin coater operating at 2500 rpm. It is typically possible that some PDMS enters into the CNTs from the opening area, however, due to its high dynamic viscosity 3500 Centipoise (obtained from product data sheet of Sylgard 184, Dow Corning), the PDMS does not enter deeply into the CNTs. Thus, by slicing the VACNT–PDMS composite block into 25 µm thick slices and discarding the top slice, CNT blockage by PDMS is minimised.
PDMS was diluted using xylene and a sample was placed in a vacuum desiccator to remove any air trapped by the PDMS. SEM images were taken after every fabrication step and SEM images of the final samples are reported in the manuscript. The various VACNT membranes were purchased from DK Nanomaterials Co. Ltd. Company, which also measured the dimensions of CNTs using TEM and their densities using SEM.
The fabrication method is summarised as follows: VACNTs on silicon wafers of different VACNT densities were purchased from DK Nanomaterials Co. Ltd (China). A two-step fabrication process was used to develop the membranes. First, PDMS was deposited onto the purchased VACNTs using spin coating; second, 25 µm thick membranes were sliced out of the VACNTs + PDMS block using a microtome machine. Figure 2 shows the fabrication steps used for the development of the VACNT membranes. A glass substrate was uses as a mechanical support onto which the Si wafer (which has the VACNTs) was glued. The glass support was subsequently removed before the VACNT membranes were sliced.
Vertically aligned carbon nanotubes (VACNT) membrane parameters used to calculate the permeability and enhancement factor
VACNTs density (tubes cm−2)
CNT diameter (nm)
Dynamic viscosity µ (Pa S) at 20 °C
Pressure difference Δp (torr)
Membrane thickness (µm)
5 × 109–1 × 1011
1.002 × 10−3
Result and discussion
The results shown in Fig. 6 also indicate that the average diameter of the CNT is less than 10 nm (average iron oxide nanoparticle size) and that the gaps between the VACNT were completely occupied by PDMS (Vatanpour et al. 2011; Zhang et al. 2015; Zhao et al. 2009a).
Mi group (Mi et al. 2007)
Hinds group (Hinds et al. 2004)
Holt group (Holt et al. 2006)
Kim group (Kim et al. 2007)
VACNT + PDMS composite
Porous aluminium support
Average outer diameter (nm)
Average inner diameter (nm)
7.5 ± 2.5
1.6 ± 0.4
Thickness of CNT layer (μm)
CNT density (tubes cm−2)
5 × 109, 1010, 5 × 1010, 1011
1.87 × 109
6 × 1010
2.5 × 1011
(7.0 ± 1.75) × 1010
Maximum permeance (LMHBar)
917, 1007, 1111, 1203
It is obvious from Table 2 that the maximum fluxes (rounded to nearest integer) are 917, 1007, 1111 and 1203 LMH for the VACNT densities of 5 × 109, 1 × 1010, 5 × 1010 and 1 × 1011 tubes cm−2, respectively. Note that in order to confirm the accuracy of our experimental results, the performance of the VACNT membrane were compared with that reported by Hinds et al., which has a CNT density (6 × 1010 tubes cm−2) that is slightly less than that of the third membrane developed in this work (of density 5 × 1010 tubes cm−2). Table 2 shows that the water flux achieved using our membrane is slightly higher than that achieved by Hinds et al., who used Polystyrene as the filler material.
Salt rejection performance achieved by key reported CNT membrane types
CNT inner diameter (nm)
Salt rejection (%)
2.5 × 1011 (Corry 2008)
Vertically aligned (VA)
2.5 × 1011 (Corry 2008)
Vertically aligned (VA)
5 × 109 (this paper)
1 × 1010 (this paper)
5 × 1010 (this paper)
1 × 1011 (this paper)
20 wt% CNT (Thomas and Corry 2015)
Mixed matrix (MM)
0.05 wt% CNT (Ocvirk et al. 2000)
Mixed matrix (MM)
It is important to notice from Fig. 7a, b that, for both NaCl solutions and DI water, the increase in flux is not directly proportional to the VACNT density. This is because when the density of VACNT increases, the number of CNT walls also increases, while the active inner diameter of CNT remains the same. Therefore, the slight increase in flow rate is attributed to additional small volumes of water flowing between walls of the MWCNTs. Note also that, the main advantage of increasing the VACNT density is the prevention of membrane biofouling, while achieving a slight increase in flow rate, with negligible impact on the salt rejection (Youngbin et al. 2014). Figure 7c compares the flux attained with DI water and NaCl solutions for different VACNT densities.
Note that the charge-based filtering mechanism, exhibited in the proposed VACNT filters, allows a relatively high CNT diameter to achieve better salt rejection than size-based filtering counterparts. This is due to the electrical and surface properties (Zeta potential and surface roughness, respectively) of PDMS, which are the key factors affecting ion transportation through CNTs (Schrott et al. 2009). Note also that the energy barrier of the CNT pores for Na+ ions depends on the pressure, temperature and concentration of the ions in the feed (Schrott et al. 2009; Corry 2008).
The salt ion rejection depends on two main factors, (1) inner diameter of the carbon nanotubes (the average inner diameter of MWCNT is 5 nm) and (2) the surface charge of the material used to fabricate the membrane. Salt rejection reduces with increasing the diameter of the CNTs (Thomas and Corry 2015). A native PDMS surface is typically negatively charged as demonstrated by Ocvirk et al. (2000). Therefore, the Na+ ions are trapped by the PDMS surface, hence increasing the salt rejection of the PDMS–CNT membrane. During the experiments, initially, the surface charge of the membrane was high, since both the low CNT diameter and high surface charge of the membrane contributed to the salt rejection. After 60 min of filtration, salt ions accumulated on the surface of the membrane, thus reducing the salt rejection contributed by the surface charge of the membrane, as shown in Fig. 7, wherein the results are in agreement with the investigation reported by Schrott et al. (2009).
Note that, the concentration polarisation (due to the accumulation of rejected salt particles at the membrane surface) typically reduces the salt rejection capability of the VACNT membranes and negatively influences mass transfer, thus increasing the osmotic pressure and reducing the water flux at the feed side. Concentration polarisation can be overcome by osmotic backwash, which is typically induced when the feed-side osmotic pressure exceeds the applied hydraulic pressure across the membrane (Chen et al. 2004; Juang et al. 2008).
It is obvious from Fig. 9 that all developed VACNT membranes displayed similar salt rejection properties. The experimental results shown in Figs. 7 and 8 demonstrate the ability of the developed VACNT membranes to achieve RO filtration water and high fluxes, in addition to preventing biofouling (Youngbin et al. 2014). The ability of VACNT membranes to prevent biofouling has been reported by Youngbin et al. (2014). This manuscript mainly focuses on comparing the water flux and salt rejection VACNT-based membranes of different densities. A comparison between the biofouling properties of the various developed VACNT membranes will be addressed in detail along with different types of CNTs in future publications.
The performance of VACNT membranes of densities 5 × 109, 1010, 5 × 1010 and 1011 tubes cm−2 have been developed and their performances investigated. The VACNT membrane development process has been described in detail. Experimental results have confirmed that the permeability of VACNT membranes increases with the density of the VACNT, while the salt rejection is almost independent of the VACNT density. A permeance of 1203 LMHBar and salt rejection exceeding 96.5 % have been experimentally achieved using a VACNT membrane of VACNT density around 1011 tubes cm−2.
ST performed the experiments and collected the data included in the manuscript. The manuscript was written and edited by ST. Prof. KA checked and approved the manuscript. Both authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
- Ackerman DM, Skoulidas AI, Sholl DS, Johnson JK (2003) Diffusivities of Ar and Ne in carbon nanotubes. Mol Simul 29:677View ArticleGoogle Scholar
- Anh CH, Baek Y, Lee C, Kim SO, Kim S, Lee S, Kim S-H, Bae SS, Park J, Yoon J (2012) Carbon nanotube-based membranes: fabrication and application to desalination. J Ind Eng Chem 18:1551View ArticleGoogle Scholar
- Chen JC, Li Q, Elimelech M (2004) In situ monitoring techniques for concentration polarization and fouling phenomena in membrane filtration. Adv Colloid Interface Sci 107:83View ArticleGoogle Scholar
- Cooper SM, Chuang HF, Cinke M, Cruden BA, Meyyappan M (2003) Gas permeability of a buckypaper membrane. Nano Lett 3:189View ArticleGoogle Scholar
- Corry B (2008) Designing carbon nanotube membranes for efficient water desalination. J Phys Chem B 112:1427View ArticleGoogle Scholar
- Du F, Qu L, Xia Z, Feng L, Dai L (2011) Membrane of vertically aligned super long carbon nanotubes. Langmuir 27:8437View ArticleGoogle Scholar
- Futaba DN, Hata K, Yamada T, Hiraoka T, Hayamizu Y, Kakudate Y, Tanaike O, Hatori H, Yumura M, Iijima S (2006) Shape-engineerable and highly density packed single-walled carbon nanotubes and their application as super-capacitor electrodes. Nat Mater 5:987View ArticleGoogle Scholar
- Ge L, Wang L, Du A, Hou M, Rudolph V, Zhu Z (2012) Vertically-aligned carbon nanotube membranes for hydrogen separation. RSC Adv 2:5329View ArticleGoogle Scholar
- Gilani N, Towfighi J, Rashidi A, Mohammadi T, Omidkhah MR, Sadeghian A (2013) Investigation of H2S separation from H2S/CH4 mixtures using functionalized and non-functionalized vertically aligned carbon nanotube membranes. Appl Surf Sci 270:115View ArticleGoogle Scholar
- Hinds BJ, Chopra N, Rantell T, Andrews R, Gavalas V, Bachas LG (2004) Aligned multiwalled carbon nanotube membranes. Science 303:62View ArticleGoogle Scholar
- Holt JK, Park HG, Wang Y, Stadermann M, Artyukhin AB, Grigoropoulos CP, Noy A, Bakajin O (2006) Fast mass transport through Sub-nanometre carbon nanotubes. Science 12:1034View ArticleGoogle Scholar
- Juang RS, Chen HL, Chen YS (2008) Membrane fouling and resistance analysis in dead-end ultrafiltration of Bacillus subtilis fermentation broths. Sep Purif Technol 63:531View ArticleGoogle Scholar
- Kim S, Pechar TW, Marand E (2006) Poly(imide siloxane) and carbon nanotube mixed matrix membranes for gas separation. Desalination 192:330View ArticleGoogle Scholar
- Kim S, Jinschek JR, Chen H, Sholl DS, Marand E (2007) Scalable fabrication of carbon nanotube/polymer nanocomposite membranes for high flux gas transport. Nano Lett 7:2806View ArticleGoogle Scholar
- Krishnakumar P, Tiwari PB, Staples S, Luo T, Darici Y, He J, Lindsay S (2012) Mass transport through vertically aligned large diameter MWCNTs embedded in parylene. Nanotechnology 23:4551011View ArticleGoogle Scholar
- Kumar S, Srivastava S, Vijay YK (2012) Study of gas transport properties of multi-walled carbon nanotubes/polystyrene composite membranes. Int J Hydrogen Energy 37:3914View ArticleGoogle Scholar
- Majumder M, Chopra N, Andrews R, Hinds BJ (2005) Nanoscale hydrodynamics: enhanced flow in carbon nanotubes. Nature 438:44View ArticleGoogle Scholar
- Majumder M, Chopra N, Hinds BJ (2011) Mass transport through carbon nanotube membranes in three different regimes: ionic diffusion and gas and liquid flow. ACS Nano 5:3867View ArticleGoogle Scholar
- Mi W, Lin YS, Li Y (2007) Vertically aligned carbon nanotube membranes own macroporous alumina supports. J Membr Sci 304:1View ArticleGoogle Scholar
- Ocvirk G, Munroe M, Tang T, Oleschuk R, Westra K, Harrison DJ (2000) Electrokinetic control of fluid flow in native poly(dimethylsiloxane) capillary electrophoresis devices. Electrophoresis 21:107View ArticleGoogle Scholar
- Schrott W, Slouka Z, Červenka P, Ston J, Nebyla M, Přibyl M, Šnita D (2009) Study on surface properties of PDMS microfluidic chips treated with albumin. Biomicrofluidics 3:044101View ArticleGoogle Scholar
- Sears K, Dumée L, Schütz J, She M, Huynh C, Hawkins S, Duke M, Gray S (2010) Recent developments in carbon nanotube membranes for water purification and gas separation. Materials 3:127View ArticleGoogle Scholar
- Sharma A, Tripathi B, Vijay YK (2010) Dramatic improvement in properties of magnetically aligned CNT/polymer nanocomposites. J Membr Sci 361:89View ArticleGoogle Scholar
- Skoulidas AI, Ackerman DM, Johnson JK, Sholl DS (2002) Rapid transport of gases in carbon nanotubes. Phys Rev Lett 89:1859011View ArticleGoogle Scholar
- Srivastava A, Srivastava ON, Talapatra S, Vajtai R, Ajayan PM (2004) Carbon nanotube filters. Nat Mater 3:610View ArticleGoogle Scholar
- Thomas M, Corry B (2015) A computational assessment of the permeability and salt rejection of carbon nanotube membranes and their application to water desalination. Philos Trans R Soc A 374:20150020View ArticleGoogle Scholar
- Vatanpour V, Madaeni SS, Moradian R, Zinadini S, Astinchap B (2011) Fabrication and characterization of novel antifouling nanofiltration membrane prepared from oxidized multiwalled carbon nanotube/polyethersulfone nanocomposite. J Membr Sci 375:284View ArticleGoogle Scholar
- Verweij H, Schillo MC, Li J (2007) Fast mass transport through carbon nanotube membranes. Small 12:1996View ArticleGoogle Scholar
- Wang M, Li T, Yao Y, Lu H, Li Q, Chen M, Li Q (2014) Wafer-scale transfer of vertically aligned carbon nanotube arrays. J Am Chem Soc 136:18156View ArticleGoogle Scholar
- Youngbin B, Kim C, Seo DK, Kim T, Lee JS, Kim YH, Ahn KH, Bae SS, Lee SC, Lim J, Lee K, Yoon J (2014) High performance and antifouling VACNT membrane for water purification. J Membr Sci 460:171View ArticleGoogle Scholar
- Yu M, Funke HH, Falconer JL, Noble RD (2009) High density, vertically-aligned carbon nanotube membranes. Nano Lett 9:225View ArticleGoogle Scholar
- Zhang L, Zhao B, Wang X, Liang Y, Qiu H, Zheng G, Yang J (2014) Gas transport in vertically-aligned carbon nanotube/parylene composite membranes. Carbon 66:11View ArticleGoogle Scholar
- Zhang L, Jiang C, Yang J, Zheng G (2015) Preparation and transport performaces of high-density, aligned carbon nanotube membranes. Nanoscale Res Lett. doi:10.1186/s11671-015-0970-8 Google Scholar
- Zhao B, Futaba DN, Yasuda S, Akoshima M, Yamada T, Hata K (2009a) Exploring advantages of diverse carbon nanotube forests with tailored structures synthesized by supergrowth from engineered catalysts. ACS Nano 3:108View ArticleGoogle Scholar
- Zhao B, Song Z, Yang J (2009b) Tunable field emission properties of carbon nanotube arrays by engineering Fe catalysts. Mater Lett 63:2556View ArticleGoogle Scholar