3.1. Morphological Properties
The morphological property of the as-constructed CuOSC nanocomposites was characterized by FE-SEM; illustrative micrographs are shown in
Figure 1. The micrographs clearly show that the chemical combination of CuO and SWCNTs resulted in hierarchical growth of cauliflower-shaped composite particles with diverse sizes between 30 nm and 1 μm. All of the CuO crystals are nanodimensional; however, the sizes of the CuOSC nanocomposites appear to rely on the numbers of agglomerating crystals surrounding the SWCNTs. CuO nanocrystals encased the SWCNTs and resulted in the formation of nanocauliflowers. The space between consecutive nanocauliflowers reveals separation between the SWCNTs. Large-sized nanocauliflowers grew in areas that contained more SWCNTs, whereas voids were left in the areas without SWCNTs. SWCNTs are not distinct in the composites in
Figure 1a–e due to hierarchical growth of the CuO encapsulating SWCNTs, however, the SWCNTs are apparently observable in
Figure 1f (SWCNTs are revealed distinctly in the HR-TEM images, later in this section). The
Figure 1f shows how CuO nanocrystals surrounded the SWCNTs to form chemically bonded nanocomposites. The size of the CuOSC nanocauliflowers appears to depend on the calcination time also. Bigger sized nanocomposites were observed in FE-SEM analysis of CuOSC-3 (
Figure 1e).
The nanostructure of and chemical attachments between the constituents of the CuOSC composites were investigated by field emission energy-filtered TEM (FE-EF-TEM); representative micrographs are displayed in
Figure 2. CuO nanocrystals are conspicuously attached to the walls of the SWCNTs; however, the nanocauliflower architecture observed by FE-SEM was dismantled by the bath sonication procedure used to prepare the TEM samples. CuO nanocrystals and SWCNTs combined chemically to generate an interconnected network of constituents. The chemical combination between the SWCNT wall and CuO nanoparticles gave rise to a permanent heterojunction between them; this heterojunction is a pivotal factor in the photocatalytic performance of the composite by promoting the split of photo-generated electron–hole pairs and preventing their recombination [
5]. The CuO nanoparticles in the nanocomposites appear to possess monoclinic crystalline geometry. The well-distributed small nanocrystals exhibit sizes (as measured by FE-EF-TEM analysis) ranging from 2 to 15 nm. Since the TEM images of all the composite samples were almost similar, the representative micrographs of only CuOSC-3 nanocomposite are presented in
Figure 2.
3.2. Structural Characterization
The crystallinity and the crystal structure of pure CuO, SWCNTs, and the CuOSC heterojunction-composites were examined by XRD analysis; the corresponding diffraction patterns are shown in
Figure 3. The XRD profiles of the pure CuO (
Figure 3a) include robust peaks at 2
θ values of 43.34°, 50.53°, and 74.21°, which are attributable to the (111), (202), and (222) crystalline facets of monoclinic CuO [
6]. All of the aforementioned peaks were also observed in the patterns of the CuOSC composite samples, along with a new peak due to pure SWCNTs. The XRD pattern of CuOSC-1 (
Figure 3a) matches that of pure CuO except for the peak attributed to the (002) plane of the pure SWCNTs at 26.03°. However, two more peaks emerged at 36.42° and 61.46° in the pattern of CuOSC-2 (
Figure 3c), corresponding to the (−111) and (−113) planes of monoclinic CuO, respectively [
6]. Those two feeble peaks of CuOSC-2 became more prominent in the pattern of CuOSC-3 (
Figure 3d), likely because of the longer calcination time of the CuOSC-3 nanocomposite.
The inset in
Figure 3 shows the XRD pattern of untreated SWCNTs; the crests at 25.64° and 43.13° are attributed to the (002) and (100) crystal facets of graphitic carbon, respectively. The peak at 25.64° corresponds to the
d-spacing of CNTs, confirming their crystalline form. The peak at 43.13° corresponds to disordered carbon present on the sample [
20]. The peak corresponding to the (002) facet of crystalline carbon is observed in the patterns of all of the CuOSC nanocomposites, although it is shifted slightly toward higher 2
θ values (approximately at 26°) compared with its position in the pattern of pure SWCNTs. A peak associated with the (100) plane of SWCNTs is not observable in the spectra of the nanocomposites because of its superimposition with the peak associated with the (111) plane of the monoclinic CuO. The presence of carbon peaks in the spectra of the composites suggests that the crystallinity of the carbon was unaffected by the nanocomposite formation process.
The Scherrer equation (Equation (1)) [
17] was applied to obtain the mean crystallite size from full-width at half-maximum (FWHM) value determined from the XRD patterns of the nanocomposites:
where
D is the mean diameter of the crystallites,
k denotes the Scherrer constant (
k ≈ 0.9),
λ is the wavelength of the X-rays (
λ = 0.154 nm for Cu Kα radiation),
β is the FWHM in radians, and
θ represents the diffraction angles in degrees. The peak at 43.34° (2
θ) was chosen for reckoning the mean size, which was calculated to be 23.95 nm.
The charge states of the constituting elements in the samples were explored using high-performance equipped X-ray photoelectron spectroscopy (HP XPS); the corresponding spectra are depicted in
Figure 4, and
Figures S1 and S2.
The survey spectrum of the as-synthesized CuOSC nanocomposites is demonstrated by
Figure 4a, where the presence of chemically combined Cu, O, and C is explicitly observed. The Cu-2
p core-level profiles are displayed in
Figure 4b. Deconvolution of the broad and asymmetric range reveals the coexistence of Cu
2+ ions in neighboring different chemical environments. The peaks at 931.88 and 951.85 eV are indexed to Cu-2p
3/2 and Cu-2p
1/2 states, respectively and indicate the existence of Cu
2+ in the CuOSC nanocomposite samples. The binding-energy gap of 20.03 eV, as obtained from a straightforward calculation using these values, coincides with the value reported in the literature, further confirming the Cu
2+ state in the nanocomposites [
21]. The satellite peaks at 943.70 and 962.17 eV confirm the purity of CuO in the composite samples [
17].
The O-1
s core-level details of the CuOSC-3 nanocomposite are illustrated by
Figure 4c, where the broad asymmetric spectra have been deconvoluted into three discrete plots, revealing the existence of nonequivalent chemical atmosphere surrounding the O
2− ions. The profile with a peak at a binding energy of 530.26 eV affirms that the O
2− ions were connected chemically to Cu
2+ ions in the monoclinic CuO, whereas the other element with a peak at 531.42 eV corresponds to vacant oxygen sites prevalent in the composite [
17]. The XPS spectra reveal a substantial number of oxygen vacancies. The third and robust component at a relatively higher binding energy (533.13 eV) corresponds to chemisorbed O on the surface of the nanocomposite in the form of O
2, H
2O, or OH; such groups represent loosely attached oxygen species on the surface of the nanocrystalline CuO [
10].
Deconvolution of the core-level C-1
s profiles (
Figure 4d) confirms the existence of dissimilar chemical environments around the carbon atoms in the CuOSC composite samples. The most intense peak at 283.98 eV indicates the dominance of
sp2-hybridized graphitic carbon in the nanocomposites and that at 284.64 specifies defects in the form of sp
3-hybridized carbon, whereas the peaks at 285.89 and 289.82 eV ascribe to C–O and O–C=O species, respectively [
20]. The bonding states and chemical environments of the constituent elements, as previously discussed (bonding states of Cu, O and C discussed above), led us to conclude that heterojunctions formed between the CuO and SWCNTs via strong covalent bonds (explicitly, CuO–SC or Cu–OOC–SWCNT) or through linkage by means of van der Waals force.
3.3. Thermal Properties
The effect of thermal energy on the as-fabricated composite samples was evaluated using a thermal analyzer, in which the samples were heat-treated gradually from 27 °C to 800 °C sustaining a ramp speed at 10 °C min
−1 under a nitrogen atmosphere. Thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) curves were constructed on the basis of the thermal analysis results; the corresponding spectra are displayed in
Figure S3.
As expected for a stable metal oxide, pure CuO (
Figure S3a) exhibits a trivial mass loss when heated to 800 °C; the remnant mass of CuO after thermal action is 98.43%. A meager mass loss near 100 °C in the TGA trace of pure CuO is assigned to the evaporation of physically attached water molecules. The gradual decrease in mass between 545 °C and 785 °C is attributable to the thermal decomposition of residual acetate groups [
7]. By contrast, the TGA trace of the SWCNTs (
Figure S3a) shows an enormous mass loss due to thermal decomposition; when the sample was heated to 555 °C, only 3% residual carbon remained; the remnant mass did not decrease further when the sample was heated to 800 °C. The slight weight loss observed in the temperature span of 100 to 390 °C is ascribed to the elimination of moisture content and to the deterioration of –OH groups attached to the SWCNTs. Afterwards, the fast loss between 400 and 550 °C is ascribed to the breakdown of SWCNTs [
14].
The nanocomposite samples (
Figure S3a) demonstrated remarkably higher thermal stability than the pure SWCNTs. Almost no noticeable loss is observed in the TGA traces of the heterojunction composites heated to 500 °C; however, a slight mass loss occurred between 500 and 800 °C. Such a mass-loss trend, which is comparable to that of pure CuO but substantially greater than that of the SWCNTs, is ascribed to the complete removal of physisorbed water fragment, the degradation of bonded hydroxyl groups, and the decomposition of remnant acetates [
14]. Among the nanocomposite samples, CuOSC-3 exhibited the greatest thermal stability, which is ascribed to its longer calcination time (3 h) compared with that of the other samples. The longer calcination time resulted in extensive removal of surface hydroxyl groups and unconsumed acetates. The results confirm that the thermal stability of the SWCNTs in the as-synthesized composites was dramatically enhanced by their chemical combination with CuO nanocrystals via strong chemical bonds. As validated by XPS analysis, the TGA results confirm that the attachment of CuO nanocrystals to the SWCNTs, as observed by FE-SEM and HR-TEM, occurs through chemical bonding between them. Such chemical bonds result in a permanent heterojunctions between their surfaces. The DSC profiles (
Figure S3b) affirm that the samples’ thermal events were all endothermic processes.
3.6. Photocatalytic Activity
The as-synthesized nanocomposites were tested for their photocatalytic proficiency by evaluating the catalyzed decomposition of MB under irradiation by natural sunlight, and their photocatalytic efficiency under identical conditions was compared with that of pristine CuO nanocrystals and SWCNTs. The obtained results are depicted in
Figure 6. To investigate the possibility of self-degradation of MB alone under exposure to sunlight, we conducted a blank test by subjecting an MB solution under solar light exposure with no photocatalyst; the results are displayed in
Figure S5. The results of the blank test show that natural sunlight alone (when no catalyst was used) caused negligible deterioration of the MB solution.
The photocatalytic performance of CuOSC-3, which reveals the highest photocatalytic performance among the investigated nanocomposite samples, is shown in
Figure 6a; it exhibited 96% photo-decomposition of an MB solution through 2 h of sunlight exposure. The other samples (viz., CuOSC-2 and CuOSC-1) resulted in 93% and 90% photodecomposition at the same exposure time, respectively. By contrast, pure CuO and pure SWCNTs decomposed 56% and 11% of the MB solution, respectively, (
Figure 6d and
Figure S6a, correspondingly) under the same experimental conditions and within the same time period.
Figure S6 demonstrates the photocatalytic action of pristine SWCNTs and the dose optimization of composite photocatalysts. The optimized doze (
Figure S6b) of photocatalyst, i.e., 100 mg photocatalyst per 100 mL substrate solution (1 g L
−1) was employed in all experiment sets. These findings affirm that the establishment of heterojunctions between the crystalline CuO and the SWCNTs remarkably enhanced the photocatalytic ability of the CuO nanocrystals. Such an enhancement in photocatalytic performance is ascribed to the generation of an abundant number of charge carriers and their segregation due to synergistic interaction between the CuO and SWCNTs [
16].
The progress of the photocatalytic decomposition reaction was monitored by plotting ln (
C/
C0) against the time of irradiation and the corresponding values are displayed in
Figure 7a. The figure demonstrates that our as-synthesized photocatalysts are significantly better compared to the action of pristine CuO. We accessed the adsorption of dye on the surface of the photocatalyst by stirring the dye solution magnetically with the photocatalysts (1 h) and keeping undisturbed in a dark chamber (additional 1 h) for adsorption–desorption equilibrium. We found that MB was notably adsorbed on the surface of the photocatalyst during the adsorption–desorption equilibrium period. It is evident from 7a, where the concentration at 0 min is notably less than the initial concentration. There is the best adsorption of MB molecules by CuOSC-3 photocatalyst.
The kinetics of the photon-induced decomposition of MB via the photocatalytic action of as-synthesized catalysts was studied by calculating the decomposition rate constant with a pseudo-first order rate equation (Equation (2)):
where
k is the first-order rate constant of the photodeterioration,
t is the irradiation time in minutes,
C0 is the initial concentration of the solution, and
C is the concentration at the evaluated time [
22,
23]. The degradation rate constants calculated for CuOSC-3, CuOSC-2, CuOSC-1, and pure CuO samples were 0.0248, 0.0184, 0.0165, and 0.0082 min
−1, respectively. Among the composite samples, the best photocatalyst, i.e., CuOSC-3, was examined for reusability and exhibited almost constant catalytic performance until the fifth reuse cycle; the photocatalyst decomposed 93% of the MB in the solution in its fifth recycle (
Figure 7b). However, some changes in morphology due to repeated handling were observed (
Figure S7a,b). The recycling performances of CuOSC-1 and CuOSC-2 were not examined assuming that their abilities are comparable with that of CuOSC-3.
3.8. Mechanism of Photocatalytic Action
The mechanism of photocatalytic action of the nanocomposite samples can be explained based on the production of electron–hole pairs on the exposed surface of the photocatalyst being irradiated by sunlight comprising energy higher than or equivalent to the bandgap of the photocatalyst. In the nanocomposite photocatalyst, the energy of the radiation excites its filled valence-band (VB) electrons into the empty conduction band (CB), making photogenerated electrons (e
−) available in the CB and generating positively charged holes (h
+) in the VB [
20]. In the hetero-combined CuO and SWCNTs, synergistic migration of electrons and holes is enhanced by the difference in their edge potentials; the VB edge potential of the CuO is lower than that of the SWCNTs, whereas it is exceedingly positive in comparison to the lowest unoccupied molecular orbital of the SWCNT [
14]. Hence, the photogenerated (CB) electrons of the SWCNTs spontaneously drift towards the CuO CB and the photogenerated holes in the CuO VB correspondingly migrate to the SWCNTs through the heterojunction. Such an exchange of charge carriers between CuO and SWCNTs is responsible for their effective separation, resulting in a sufficiently large excess of photon-induced electrons and holes for initiating redox reactions that lead to the decomposition of dye solutions [
6].
Oxygen vacancies (V
o in CuO) and the surface defects (SDs) between the CB and VB of SWCNTs (as observed via XPS analysis) in the hetero-composites further inhibit the recombination of photon-generated electrons and holes by ensuring their precise separation. The flow of photogenerated electrons from the VBs to the SDs and V
o regions and from these locations to the CB results in the substantial separation of charge carriers and promotes the generation of additional electrons and holes [
17]. The photoinduced electrons attack dissolved O
2 to generate superoxide radicals (˙O
2−), which in turn produce powerful hydroxyl radicals (˙OH) that degrade MB molecules. The ˙OH radicals are primarily the prominent oxidizing agents that invade MB adsorbed onto the surface of the photocatalyst and generate transition intermediates [
6]. The intermediates are repeatedly attacked by the ˙OH, ˙O
2−, or h
+ until they are converted into innocuous inorganic compounds such as H
2O and CO
2. The schematic diagram of the generation of reactive species is presented in
Figure S9 and the details of mechanism is provided thereafter in the
supplementary information.
3.10. Detection of the Degradation Products
MB solutions before and after photocatalytic degradation were analyzed by LC–MS to characterize their photodecomposition products; the corresponding spectra are disclosed in
Figure 8.
Figure 8a shows the LC–MS spectrum of an MB solution, which includes a robust peak at
m/
z 284.10; the detailed spectrum of the degradation products is shown in
Figure 8b. Since the photocatalytic reactions with all as-synthesized samples were carried out under same experimental conditions and we expected the similar degraded products, we selected the degraded sample of MB by CuOSC-3 (sample with the best activity) for LC–MS analysis. The degradation intermediates or byproducts resulting from the photocatalytic action of our as-synthesized photocatalysts on MB were observed at
m/
z values of 145, 171, 203, 337, 365, 398, 421, and 463. Most of these
m/
z values of the degradation byproducts or intermediates are similar to those previously reported [
25,
26,
27,
28]. Because 96% degradation of the MB solution occurred during 2 h of catalyst exposure, some intact MB molecules were present in the solution, as evident in the LC–MS spectra. In addition, the intensity of the prominent MB peak in the LC–MS spectrum (
Figure 8a) was substantially diminished in the spectrum of the photodegraded sample (
Figure 8b). The structural formulae of the degradation products can be deduced on the basis of their
m/
z ratios. During the photocatalytic reaction, strongly oxidizing ˙OH radicals exfoliate the benzene rings of MB via demethylation or hydroxylation accompanied by oxidation [
25,
27]. A series of reactions occurs until the innocuous inorganic end products such as CO
2, H
2O, sulfates, and nitrates are produced from their organic parent molecules. Moreover, some nontoxic organic acids such as acetic acid and oxalic acid may also be produced. The LC–MS spectra confirm the photodegradation of MB and the formation of less-harmful organic or harmless inorganic products.
Based on the LC–MS analysis, the degradation mechanism of MB is proposed as revealed by
Figure S11. Some of the additional intermediates (that are supposed to be vanished completely due to degradation) are also shown in the mechanism based on the literature for better understanding of the photodegradation mechanism.