Introduction

With the rapid development of industrialization, certain hazardous effects on the environment and human survival have emerged beside its benefit. Effluents from the use of pesticides, the textile, petrochemical, dyeing, plastic, and paper industries are highly toxic, carcinogenic and recalcitrant [1,2], and yet not readily degradable. Chlorophenols are toxic chemicals which are used in many industrial applications such as petrochemicals, pesticide, dye intermediates and paint [3]. Especially, 2,4-dichlorophenol (2,4-DCP) is an important chemical precursor for the manufacture of a widely used herbicide, 2,4-dichlorophenoxy acetic acid (2,4-D) [4]. However, 2,4-DCP may cause some pathological symptoms and changes to human endocrine systems. Their mode of exposure is through the skin and gastrointestinal tract. In recent years, concerns have been raised because of chlorophenols persistence and bioaccumulation both in animals and in humans [5-7].Therefore, it is important to find innovative and effective ways to minimize the harm of chlorophenols in the environment.  Heterogeneous semiconductor photocatalysts have received significant attention owing to their potential application in a wide range of photoinduced reactions, notably photocatalytic hydrogen production, removal of organic contaminants and air pollutants, and electricity production using solar cells [8–11]. TiO₂ mediated photocatalytic degradation is a successful and convenient alternative to the conventional methods for the treatment of wastewater containing organic pollutants. TiO₂ has the advantage of good chemical stability, the absence of toxicity and relatively low cost, but a serious disadvantage is its wide band gap (Eg = 3.2 eV) that requires that UV radiation is used to trigger this attractive photocatalyst, which would greatly hinder the commercialization of TiO₂ photocatalysis. Photocatalytic degradation of organic contaminants using abundant natural solar radiation can be highly economical compared with the processes using artificial UV radiation, which require substantial electrical power input. In regard to this, various attempts have been made to extend the spectral response of TiO₂ into the visible region of the solar spectrum and enhance its photocatalytic activity [12]. Lots of attempts were tried to improve the photocatalytic performance of TiO₂ and doping with transition metal ions, such as silver, nickel, and iron was found to be a useful method [13-18]. Among all the available transition metals, cobalt was proved to be one of the most effective dopants to enhance the light response and photoactivity of TiO₂. Ebrahimian and et al  [19] prepared cobalt doped TiO₂ nanoparticles which shows a wide absorption range extended into the visible region. Iwasaki [20] synthesized cobalt doped TiO₂ and found that the introduction of Co²⁺‏ could apparently shift the light absorption edge of anatase TiO₂ to the visible region and enhance photoactivity under both UV and visible light irradiation. Also, researchers found that the addition of co-sorbent carbon materials can enhance the photocatalytic efficiency of TiO₂ [21-23]. As a new member of the carbon family, carbon nanotubes (CNTs) with one-dimensional and hollow structure have received considerable interest since their discovery [24] due to their outstanding structural characters, e.g., mechanical strength [25], excellent thermal conductivity, unique electronic properties [26] and thermal stability [27]. CNTs can be used as a promising material for environmental cleaning. The collection process for MWCNTs with TiO₂ in new composite depends on many causes: one of them suggests that carbon nanotubes behave as a semiconductor supports because of their combination of physiochemical properties which include excellent conductance, high abilities for adsorption [28]. A similar literature has shown that the activities of TiO₂ increased due to abilities of MWCNTs to decrease TiO₂ crystalline grain and particle sizes [29], or increased in the activity of the particles because the direct interaction between MWCNTs and TiO₂ reduces the recombination of electron and hole ( h⁺/e^-) [30]. Generally, It is believed that the change on activities can be related to the TiO₂–CNTs bonding which can be formed through some physic/chemical interactions such as Vander Walls interaction. The above finding stimulated the advance improvement of an efficient photocatalyst in visible light region. To the best of our knowledge, MWCNTs /Co-TiO₂ nanocomposites prepared using modified sol-gel process have not yet been reported. In the present work, these nanocomposites were prepared using titanium isopropoxide (TIP) as titanium precursor. The performance of the resultant photocatalysts was evaluated by photocatalytic treatment of synthetic wastewater containing 2,4 dichlorophenol  (2,4-DCP) under visible light. The textural properties of the resulting photocatalysts were investigated using XRD, FTIR and SEM/EDX. The reaction intermediates were identified by gas chromatography-mass spectrometry (GC–MS) technique.

Experimental

Materials and reagents

Cobalt (II) chloride hexahydrate (CoCl₂.6H₂O) was supplied by (Merck, No.102539). Titanium isopropoxide (TIP), (Merck No. 8.21895), ethanol (Merck No. 818760), deionized water and multi-walled carbon nanotubes functionalized by carboxylic groups (MWCNTs) were provided by Neutrino Corporation (Iran), (The average diameter of the MWCNTs was 10-20 nm, and the length was 0.5-2μm), were used for photocatalyst synthesis. High-purity 2,4-DCP, 98%,  (Merck No. 803774) was used as a probe molecule for photocatalytic tests.

Preparation of MWCNTs/Co-TiO₂ nanocomposite

MWCNTs/Co-TiO₂ nanocomposite was prepared by a modified sol-gel method. An appropriate amount of CoCl₂.6H₂O (Table 1), 10 mL TIP and 30 mL ethanol were stirred for 2 h (Solution A). Then, solution B, 20 mL ethanol, 5 mL deionized water and 2 mL hydrochloric acid and an amount of MWCNTs (Table 1) was added into the Solution A and stirred for 12 h at room temperature. The sol was formed after 12 h of stirring followed by aging at room temperature for 24 h and evaporated at 80 °C for 8 h. Finally, the dried powder was calcined at 450 °C under air for 2 h to get an MWCNTs/Co-TiO₂ sample. For comparison, four MWCNTs/Co-TiO₂ samples with different amounts of MWCNTs, pure TiO₂, Co-TiO₂ and MWCNTs/TiO₂ samples were synthesized by the same route. From now on, the prepared samples will be shown according to Table 1.

Characterization

Fourier transforms infrared (FTIR) analysis was applied to determine the surface, functional groups, using FTIR spectroscopy (FTIR-2000, Bruker), where the spectra were recorded from 4000 to 400 cm^-1. The XRD patterns were recorded on a Siemens, D5000 (Germany). X-ray diffractometer using Cu K_α radiation as the X-ray source. The diffractograms were recorded in the 2θrange of 20-80°. The morphology of the prepared samples was characterized using scanning electron microscope (SEM) (Vegall-Tescan Company) equipped with an energy dispersive X-ray (EDX).

Photocatalytic degradation of 2,4-DCP

In a typical run, the suspension containing 10 mg photocatalyst and 100 mL aqueous solution of 2,4-DCP (40 mg/L) was stirred first in the dark for 10 min to establish adsorption/desorption equilibrium. Irradiation experiments were carried out in a self-built reactor. A visible (Halogen, ECO OSRAM, 500W) lamp was used as irradiation source (its emitting wavelength ranges from 350 nm to 800 nm with the predominant peak at 575 nm).At certain intervals, small aliquots (2 mL) were withdrawn and filtered to remove the photocatalyst particles. These aliquots were used for monitoring the degradation progress, with Rayleigh UV-2601 UV/VIS spectrophotometer (λ_max= 227nm).

Statistical analysis

All experiments were performed in triplicate and the average values were presented. The data were analyzed by one-way analysis of variance (ANOVA) using SPSS 11.5 for Windows. The data were considered statistically different from control at P < 0.05.

Identification of degradation intermediates

The photocatalytic reaction intermediates were identified by GC–MS in an Agilent 190915-433 instrument equipped with an HP-5MS capillary column (30 m × 0.25 mm). The column temperature was programmed at 50 °C for 2 min, and from 50 to 250 °C at a rate of 10 °C min^-1. The sample used for GC–MS analysis was prepared according to the following procedure: The obtained degradation product was acidified to pH 1 and subsequently extracted with dichloromethane. After dichloromethane was evaporated to dryness under vacuum, 10 mL methanol was added to dissolve the residue. Then, 1 mL concentrated sulfuric acid was added and the combined solution was refluxed for about 3 h. The solution was further extracted with dichloromethane followed by concentrating to about 1 mL under reduced pressure. The released chloride ions originating from the degradation of 2,4-DCP were identified and determined by the AgNO₃ method.

Result and discussion

X-ray diffraction analysis

Fig. 1a shows the XRD pattern of the prepared TiO₂. The diffractions found at 2θ= 27.4º, 36.1º, and 41.2º belonged to the rutile crystalline phase of TiO₂ [31] and the diffractions at 2θ= 25.28º, 37.80º, 48.18º, and 54.09 are the main diffractions for the anatase crystalline phase of TiO₂ (JCPDS 21-1272). The XRD pattern revealed the prepared TiO₂ containing predominant anatase crystalline phase and a few of rutile crystalline phase. The XRD pattern of MWCNTs (Fig. 1b) shows a broad crystalline diffraction around 2θ=25.5°, which represents the characteristic diffraction of MWCNTs [32]. In XRD pattern of MWCNTs/TiO₂ (Fig. 1c), we didn’t observe the strong and main diffraction of MWCNTs at 2θ=25.5°, which was overlapped with the main diffraction of anatase TiO₂ at 2θ=25.3º and the relatively large difference between the mass percent of MWCNTs and TiO₂ and low crystallinity of MWCNTs could be the reasons for the MWCNTs diffraction not to be detectable [33]. Also, the XRD pattern of Co-TiO₂ sample (Fig. 1d) didn’t show any cobalt phase indicating that cobalt ions uniformly dispersed among the anatase crystallites and showed only pure anatase phase for TiO₂ [34-37]. The XRD patterns of the ternary MWCNTs/Co-TiO₂ nanocomposites (Fig. 2) revealed the main diffractions of anatase and rutile crystalline phase for TiO₂. Also, the XRD patterns (Fig. 2), showed, increasing of cobalt doping inhibited the phase transformation from anatase to rutile in the ternary MWCNTs/Co-TiO₂ nanocomposites [38] and we observed pure anatase crystalline phase for the sample MWCNTs/Co-TiO₂ (4.14) (containing the highest cobalt doping in our synthesized samples) (Fig. 2d).

The diffraction patterns of the prepared samples show considerable line width, indicating the samples containing small crystal. The average crystal size of each sample is calculated from the full width at half maximum (FWHM) of the (101) diffraction peak using Scherrer’s equation [39].

                                                  (1)

Where D is the average crystal size of the sample, λthe X-ray wavelength (1.54056 Å), βthe full width at half maximum (FWHM) of the diffraction peak (radian), K is a coefficient (0.89) and θis the diffraction angle at the peak maximum. All the prepared samples are in nano-size range (Table. 2), from 10.11 to 18.32 nm, and all the samples showed smaller crystal size compared to pure TiO₂. In the case of ternary MWCNTs/Co-TiO₂ nanocomposites, it can be concluded that the addition of cobalt to titania hinders the growth of TiO₂ nanoparticles. This may be due to the formation of Co–O bond on the surface of the doped TiO₂, which restricted the crystallite growth of TiO₂ [40].

SEM/EDX analysis

SEM images of the ternary MWCNTs/Co-TiO₂ nanocomposites are shown in Fig. 3. Some aggregation can be observed that probably happened during the synthesis process. The high viscosity of the sol might be one of the reasons to induce this phenomenon [41]. The EDX patterns of the ternary MWCNTs/Co-TiO₂ nanocomposites in Fig. 4 show two peaks around 0.2 and 4.5 keV. The intense peak is assigned to the bulk TiO₂ and the less intense one to the surface TiO₂. The peaks of cobalt are distinct in Fig. 4 at 0.6, 6.9 and 7.5 keV. The less intense peak is assigned to cobalt in the TiO₂ lattice [42, 43]. These results confirmed the existence of cobalt atoms in the ternary MWCNTs/Co-TiO₂ nanocomposites but the XRD patterns do not show any diffractions related to cobalt. Therefore, it may be concluded that cobalt ions are uniformly dispersed among the TiO₂ lattice during the synthesis process. EDX results are given in Table 3. Fig. 5 shows elemental mapping images of the ternary MWCNTs/Co-TiO₂ nanocomposites. From the elemental mapping mode, highly and uniformly dispersion of cobalt was observed in the TiO₂ lattice especially for the sample MWCNTs/ Co-TiO₂ (3.13). This implies a good interaction between cobalt and TiO₂ in the preparation process using the sol-gel method.

FTIR analysis

The FTIR spectra of the ternary MWCNTs/Co-TiO₂ nanocomposites are shown in Fig. 6. The vibrations observed at ∼3400, 2930 and 2850 cm^-1 are attributed to the Ti – OH bond [44]. The spectra show the relatively strong band at ∼ 1630 cm^-1 observed for all the samples which are due to the OH bending vibration of chemisorbed and/or physisorbed water molecule on the surface of the catalysts. The strong vibration in the range of 700-500 cm^-1 is attributed to stretching vibrations of Ti –O–Ti bond [44]. The weak peak at about 514 cm⁻¹ assigned to stretching vibrations of Co–O emerged a little [45], the Co–O vibration is not strong because of the broad spectrum of TiO₂ and a small amount of Co dopant. FT-IR results reminded the formation of a small part of Co–O bond. It was probably the existence of Co–O bond that hindered the recombination of generated photo holes and photoelectrons [40]. FTIR results showed the ternary MWCNTs/Co-TiO₂ nanocomposites contain MWCNTs, cobalt, and TiO₂.

Photocatalytic degradation of 2,4-DCP

The photocatalytic degradation efficiency of 2,4-DCP using the prepared samples under visible light showed in Fig.7 and Table 4. Among the ternary nanocomposites, the photocatalytic activity of the MWCNTs/Co-TiO₂ (3.13) sample was the highest and  82% degradation of 2,4-DCP obtained after 180min irradiation under visible light. We obtained the degradation percent of 2,4-DCP in the presence of pure TiO₂, MWCNTs/TiO₂ and Co-TiO₂ samples, 44%, 71% and 67%, respectively during 180min under visible light. The degradation percent of 2,4-DCP in the presence of the ternary MWCNTs/Co-TiO₂ nanocomposites is higher than pure  TiO₂, it can be noticed that the introduction of MWCNTs and cobalt obviously caused a synergetic effect on 2,4-DCP degradation and led to a higher photocatalytic activity (Table 4). The higher photocatalytic activity of MWCNTs and Co co-modified TiO₂ in the ternary nanocomposites may be explained as firstly, an appropriate amount of the doped Co in TiO₂ could effectively capture the photo-induced electrons and holes, which inhibited the combination of photoinduced carriers and improved the photocatalytic activity. Secondly, MWCNTs/Co-TiO₂ samples had more surface hydroxyl groups than the pure TiO₂ sample which would be beneficial for the adsorption of 2,4-DCP. The abundant hydroxyl groups adsorbed on the surface of the catalyst could facilitate the formation of hydroxyl radicals which could optimize the degradation process of the adsorbed 2,4-DCP on the surface [46]. Thirdly, because of presence MWCNTs, the surface area of MWCNTs/Co-TiO₂ samples were slightly larger than that of pure TiO₂ which might favor the adsorption of 2,4-DCP and provide more possibly accessible active sites. Also, there is a synergetic effect between MWCNTs and TiO₂ and MWCNTs acting as a photosensitizer. MWCNTs can trap the photo-induced electrons and form superoxide radical ion and/or hydroxyl radical on the surface of TiO₂, which are responsible for the degradation of the organic compound. Due to the introduction of the MWCNTs, an increase of surface charge on TiO₂ in the hybrid catalysts can be suggested. The surface charge may lead to modifications of the fundamental process of electron/hole pair formation while applying visible irradiation [47]. Consequently, it may be the unique interaction between TiO₂ and the MWCNTs that endows the ternary nanocomposite with a higher catalytic activity in the photocatalytic removal of MO compares to pure TiO₂.

Also from Table 4, while Co dopant content increased from 3.03 wt.% to 3.13 wt.%, the dominant Co²⁺ captured electrons which then moved to the absorbed O₂ to reach a higher photocatalytic reaction. As the literature reported [48], with the substitution for Ti⁴⁺by Co²⁺in crystal structure of TiO₂, the catalyst can introduce a new impurity level to the conduction band of TiO₂ and the electrons can be promoted from the valence band to these impurity levels, resulting in a narrowing of the band gap. This fact indicates that there are more photogenerated electrons and holes which can be introduced to participate in the photocatalytic reactions [49]. However, recombination of photogenerated electrons and holes is one of the most significant factors that deteriorate the photoactivity of the TiO₂ catalyst. Any factor that suppresses the electron-hole recombination will, therefore, enhance the photocatalytic activity[50,51]. In general, if the size of doping metal ion is very similar toTi⁴⁺, it is very likely that metal ion enters into the interstitial site of the TiO₂ crystal. The doping metal ion located mainly on the shallow surface of TiO₂ can induce defects. The defects can become the centers of shallow electrons or holes traps, which would efficiently improve separation of an electron-hole pair. Hence the photocatalyst will have a high photocatalytic activity [52]. The radius of Co²⁺(0.074 nm) is slightly bigger than Ti⁴⁺(0.068 nm) [52]. When calcination was performed at high-temperature Co²⁺ions may enter the interstitial site of TiO₂ crystal and can create crystal defect. In thispresent work, FT-IR spectra revealed that there were Co–O bonds in the TiO₂/Co nanoparticles. The formation of the Co–O hindered the recombination of photogenerated electrons and holes, so the cobalt doping increased the photocatalytic activity of naked TiO₂. According to Fig. 7 and Table 4, 3.13 wt% was the optimal doping content. When the doping ratio is 3.52 or 4.14 wt%, the degradation rate became slower. This is because, at a low doping level of metal, photogenerated holes and electrons are well separated, increasing the efficiency of the photocatalyst. However, at high doping level, there is a considerable chance for multiple trapping which will reduce the efficiency of the catalyst. Moreover, the detrimental activity at high cobalt loading may be attributed to the blockage of many TiO₂ active sites due to the large amounts [40].

The activity of the samples for degradation of 2,4-DCP may even be better by addition of hydrogen peroxide to the reactor (Fig. 8). Almost 82% of 2,4-DCP is removed within 180 min in the presence of MWCNTs/Co-TiO₂ (3.13), when there is no H₂O₂ added. In the presence of H₂O₂, more hydroxyl radicals are produced compared to the conditions without using H₂O₂ and we obtained 91% removal for 2,4-DCP. Hydroxyl radicals may be generated by direct photolysis of hydrogen peroxide [53,54], or by reaction of hydrogen peroxide with superoxide radical [55, 56]. Since in our study a source of visible light is employed, it is unlikely that direct photolysis of hydrogen peroxide is significant. Basically, wavelengths shorter than 300 nm provide enough energy for photocleavage of the H₂O₂ molecules. This can be explained according to the below reaction:

               (2)

The faster hydroxyl radical formation is associated with the higher degradation rate. This photocatalyst did not exhibit any photoactivity in the dark, neither in the absence nor in the presence of H₂O₂, suggesting that it is necessary to photoexcite both TiO₂ and the cobalt in order to a obtain substantial improvement of the degradation of 2,4-DCP.

The kinetic of 2,4-DCP degradation

The photocatalytic degradation of 2,4-DCP is a first-order reaction and its kinetics may be expressed as ln(C/C₀) = -k_obst (Fig. 9). In this equation k_obs (min⁻¹) is the apparent rate constant, C₀ and C are the initial concentration and concentration at reaction time t of 2,4-DCP, respectively. The k_obs are found from the slopes of the straight lines obtained by plotting ln(C/C₀) versus irradiation time (Fig. 9). The reaction rates, rate constants and half-lifes (t_1/2) at various initial concentrations of 2,4-DCP are given in Table 5. The results summarized in Table 5 show that the reaction rate of degradation of 2,4-DCP is faster at higher initial concentration. However, the rate constants decrease to some extent when the initial concentration increases.

Identification of intermediate products

The intermediate species formed during photocatalytic degradation of 2,4- DCP, were identified by GC–MS technique. The major reaction intermediates identified in an aliquot withdrawn after 200 min following a degradation condition specified as in Fig. 7. Presence of these intermediates (Fig. 10) supports our proposed mechanism which is based on OH radicals. The hydroxyl radicals attack 2,4-DCP converting it to chlorocatechol and then to chlorobenzoquinone. Subsequently, hydroxyl groups would break the aromatic rings of chlorobenzo-quinone transferring them into simple acids like oxalic acid, acetic acid, etc. as the final products [57,58]. In addition to identifying the organic intermediates, chloride ions were also detected and identified as one of the final products of the photocatalytic removal. The amount of Cl^- in the reaction media at the end of the photocatalytic experiment almost equals the amount of chlorine present in 2,4-DCP indicating essentially complete degradation.

Photocatalyst reuse

The effectiveness of photocatalyst reuse was examined for degradation of 2,4-DCP during a four cycles experiment. Each experiment was carried out under identical conditions. 100mL 2,4-DCP with an initial concentration of 40 mg/L, 10mg MWCNTs/Co-TiO₂ (3.13)  and 180 min irradiation time under visible light were used. After each experiment, the solution residue from the photocatalytic degradation was filtered, washed and the solid was dried. The dried catalyst samples were used again for the degradation of 2,4-DCP, employing similar experimental conditions. Recycling experiments showed, no obvious decrease of the photocatalytic removal efficiency (from 82% to 78%) was observed after four cycles and indicating that our MWCNTs/Co-TiO₂ (3.13) is renewable for environmental applications. Possibly deactivation of the part of the photocatalyst surface, due to permanent adsorption of intermediate species, might be involved in the reduction of its activity.

Conclusions

In summary, ternary MWCNTs /Co-TiO₂ nanocomposites were successfully prepared by the modified sol-gel process and characterized by a different analysis. It was found that the ternary MWCNTs /Co-TiO₂ nanocomposites presented enhanced photocatalytic degradation of 2,4-dichlorophenol (2,4-DCP) and exhibited expansion in spectral response range shifted to the visible region. This was probably due to the affiliation of special properties by MWCNTs and Co dopant. The presence of MWCNTs could create many active sites and increase surface area while Co doping promoted the separation of photogenerated carriers. Furthermore, the major problem of TiO₂ photocatalyst was reduced by narrowing their band gap by the Co dopant. Therefore, adding a suitable amount of MWCNTs and Co into the TiO₂ lead to a great improvement in the photocatalytic degradation of 2,4-DCP and this new photocatalyst showed remarkable activity in the visible light region. The reactions follow a pseudo-first-order kinetics and the observed rate constant values change with 2,4-DCP concentrations. Oxalic acid and maleic acid were the major intermediate species at the final stage of the degradation process as identified by gas chromatography-mass spectrometry (GC–MS) technique.

Acknowledgements

The authors wish to acknowledge the financial support of the University of Tehran for supporting this research.

CONFLICT OF INTEREST

The authors declare that there is no conflict of interests regarding the publication of this manuscript.

If you have any problem in viewing images please download the PDF version.