During the sol-gel process, TEOS was first hydrolyzed to silicic acid (Equation 2). Then, condensation reactions led to the formation of Si-O-Si bounds (Equation 3) and colloidal silica nanoparticles would be appeared by the emergence of a milky silica sol [37
]. In this stage, synthesized nanoparticles of colloidal silica had the mean size of 80 nm (Figure ).
Particle size analysis. (a) Undoped sol. (b) 0.5% Cu just doped sol.
After drying and curing, the solvent was evaporated and the agglomeration of silica nanoparticles fabricated silicon nanostructures on cotton fabrics. Since the presence of hydroxyl groups (Si-(OH)3) in silicon nanostructures, the surface fabricated still remained hydrophilic (Figure ). However, due to its covering effect on cotton fabrics, a water droplet could not easily penetrate into the fabric as in pristine fabric. The interaction of a long-chain alkylsilane agent like HDTMS with silanol groups produced a hydrophobic surface which in turn would increase the contact angle (Figure ). Figure represented SEM images of untreated sample and cotton fabrics treated with alkylsilane and SiO2 nanoparticles. Alkylsilane-treated SiO2 surface (Figure ) apparently showed higher roughness than untreated one (Figure ). Its SWC and WSA were 151.1° and 30°, respectively (Table Figure ). The contact angles of pristine fabric and samples only treated with alkylsilane or silica were not measurable, due to rapid absorbance of falling water droplets.
Schematic drawings. (a) A colloidal silica nanoparticle. (b) Silica nanostructured surface containing hydroxyl groups. (c) Silica substrate treated with alkylsilane.
SEM micrographs of different samples. (a) Untreated. (b) Treated with SiO2.
Static water and water shedding angles of fabricated surface on cotton fabric samples
When silica sol was doped with Cu nanoparticles and then cotton fabrics were immersed in it, as expected, the energy dispersive X-ray (EDX) analysis confirmed the presence of Cu nanoparticles on the sample surface (Figure ).
EDX analysis for treated cotton fabrics containing different doped sols. (a) 0.5% Cu. (b) 2% Cu.
Cu nanoparticles were introduced into the silica sols when the colloidal silica nanoparticles had been previously formed by the sol-gel process. Hence, they would be settled on the surface of colloidal SiO2 nanoparticles. This was confirmed by TEM images (Figure ). Dissolution in alkaline silica sol may result to various cuprous and cupric complexes like Cu(OH)2, Cu2CO3(OH)2, Cu(NH3)2+, and Cu(NH3)2+, indicating a tendency towards colloidal silica nanoparticles.
TEM micrographs of the synthesized nanoparticles. (a) Silica. (b) 0.5% Cu-doped silica.
The addition of 0.5% wt/wt Cu into silica sol caused the flocculation of colloidal silica nanoparticles (Figure ). The emersion of two peaks and the broadening of silica peaks in a size distribution graph just 5 min after introducing Cu nanoparticles may be attributed to the gradual agglomeration of silica and Cu particles (Figure ).
Such agglomeration would produce more grape-like clusters on the final fabricated surface. Compared with ordinary SiO2 nanostructured surface, this morphology showed higher air trapping capability and SWC(Table ).
The valleys generated in (0.5%) Cu-doped treated samples were also obvious in SPM micrographs (Figure ).
SEM and SPM micrographs of fabricated surface. Fabricated by: (a and b) (left) untreated, (middle) silica sol, (right) 0.5% Cu-doped silica, and (c) water drops on surface fabricated by 0.5% Cu-doped silica sol and its contact angle.
Based on the fundamental theories on motion of liquid droplets on the rough surfaces [41
], there are two important models about the wetting behavior of these surfaces: Wenzel and Cassie-Baxter [42
]. The major difference between the two is the existence of air packets trapped in the valleys between liquid droplets and the solid substrates. Regarding the below equation [41
], if the air pockets fraction (fLA
) is high, then the value of cosθ
is decreased which may be followed by the enhanced superhydrophobic effect on the roughened surfaces:
where Rf denotes the roughness factor, and θ and θ0 are the contact angles of liquid droplets on rough and flat surfaces, respectively. The WSA value for such sample was decreased and reached to 24°, and also, the slippery of treated surface was increased.
Increasing the amount of Cu nanoparticles (2% owf) in silica sol may probably disintegrate the agglomerated clusters of silica nanoparticles and furthermore fill in the valleys of fabricated surfaces. Therefore, a homogeneous silica-copper hybrid nanocomposite would be formed on the cotton fabric samples (Figures and ). Creating a fairly flat level of roughness with filled-in valleys may result to a decreased SWC, comparing to silica networks with low Cu content (Table ). In contrast with low Cu content silica network, however, a water droplet showed less tendency (or "petal effect") to adhere to surfaces of high Cu content silica network. This was consistent with WSA values, showing lower water shedding angle (Table ).
SEM micrograph of 2% Cu-doped silica surface.
Schematic drawings of the possible effect of Cu content on the morphology of fabricated surface. Fabricated by: (left) 0.5% Cu-doped silica sol, (right) 2% Cu-doped silica sol.
Comparing data sets of static contacts angles "SWC" for different samples through analysis of variance demonstrated that differences between silica-alkylsilane, silica-Cu (0.5%) and silica-Cu (2%) were significantly important (significant value < 0.5). Standard errors for the measured static contact angles were increased by introducing Cu nanoparticles (Table ). Post hoc test (Duncan's multiple range test) showed that (homogenous subsets of) means of all three abovementioned samples were different (Table ). The effect of Cu nanoparticle on superhydrophobic property of the surface was even more than treating the surface with an alkylsilane agent like HDTMS. It should be noted that the silica-Cu (0.5%) can be considered as a hierarchical structure. In the case, the fabricated surface may have a self-cleaning capability, like Hygroryza aristata leaves.
Post post hoc test (Duncan's multiple range test) for three samples
All fabricated surface containing Cu nanoparticles displayed acceptable antibacterial properties against E. coli and S. aureus bacteria (Table ). The criterion for passing the test or evaluating them was the percentage of bacteria growth reduction. Approximately, the total numbers of bacteria for samples were 22,800 (for E. coli bacteria) and 17,440 (for S. aureus bacteria) CFU/ml at zero contact time. These amounts reduced considerably (more than 70% for E. coli and 90% for S. aureus bacteria) for doped silica treated samples with Cu nano particles but increased for control samples (undoped silica treated). It may be attributed, first, to the antibacterial activities of Cu nanoparticles and, second, to the prohibition of Cu nanoparticles agglomeration resulted from their settlements on silica nanoparticles (Figure ). The latter is considered as an important parameter because it has been known that the antibacterial activity of metallic nanoparticles has a strong relationship with their sizes. The samples containing 2% Cu showed less antibacterial activity, especially against E. coli bacteria. This may result from the flocculation of Cu nanoparticles of high concentration.
Percent reduction of bacteria on the fabricated control and doped silica surfaces