CTAB Surfactant Assisted and High pH Nano-Formulations of CuO Nanoparticles Pose Greater Cytotoxic and Genotoxic Effects

Toxicity of synthesized nanoparticles is the area of concern to all the researchers due to their possible health implications. Here we synthesized copper oxide nanoparticles (CuO NPs) without surfactant at pH value of 2, 7, 10 and with cetyletrimethylammoniumbromide (CTAB) surfactant at pH 7. Synthesized nanoparticles were characterized for various structural parameters including crystallite size, lattice parameters, strain, phase analysis using X-ray diffraction analysis, and morphological aspects have been analyzed using FESEM and HRTEM imaging. All the four nano-formulations were analyzed for their toxic potential using Allium cepa L. at three different concentrations (0.1, 0.01 and 0.001 g/100 ml). Cytological and genetic parameters including mitotic index, mitotic inhibition, aberrant cells, binucleated cells, micronucleated cells, chromosomal bridges, fragmentation, stickiness, laggards, vagrants, c-mitosis and disturbed spindle were analyzed. Our results revealed a dose dependent increase in cytotoxic parameters including decreased total dividing cells, mitotic index, and increased mitotic inhibition. Genotoxic parameters also increased at higher treatment concentrations including chromosomal aberrations and percent aberrant cells. The pH value at the time of particle synthesis has significant influence on the crystallite size and agglomeration as assessed by XRD, FESEM and HRTEM analysis. The NPs synthesized at pH 2 and 10 were found to be of smaller size and posed more toxic effects as compared to particles synthesized at neutral pH. On the other hand, CTAB assisted CuO NPs synthesized at pH 7 revealed even smaller crystallite sizes and thus boost the toxicity in all the parameters as compared to NPs synthesized without CTAB. The present study suggested an increase in toxic parameters of synthesized CuO NPs with respect to crystallite size which is pH dependent. Addition of CTAB at pH 7 decreased the crystallite as well as particle size and enhanced the toxic potential. Further studies are recommended to analyze the effect of surfactant addition in toxicological studies on CuO NPs.

Copper is one of indispensable elements for maintaining homeostasis in various types of organisms 15 and in its ionic form, it may lead to a toxic situation once they exceed the physiological tolerance range in vivo 16,17 . Number of investigations were performed to explore the environmental issue of copper compounds in water reservoirs. Thus the studies related to assessment of toxicity of CuO NPs is of keen interest among researchers to explore their health effects, genotoxicity and carcinogenic effects [18][19][20][21][22][23][24][25][26][27][28] . As CuO NPs can enter via respiratory or gastrointestinal pathways, they are known to cause acute inflammations, oxidative stress, DNA damage and cytotoxicity. Histopathological assessment of CuO particles in nano-domain has shown to result in DNA damage and cytotoxicity 29 . Oxidation-sensitive fluorescent probe was used to measure oxidative lesions by monitoring intracellular production of reactive oxygen species (ROS). Previous studies have proved that toxic effects of CuO NPs may involve oxidative stress as a major role player 30 . Cells exposed to CuO NPs were reported to supress the catalase and glutathione reductase activity in comparison to cells cultured in normal medium. The upsurge of the ratio of oxidation to total glutathione revealed that CuO NPs not only generated ROS but they also blocked cellular antioxidant defences 31 .
Several higher plants and bio-assays with their roots have provided an economical, simpler and sensitive method for the determination of the hazardous effects of various environmental pollutants. Plants have been used for the evaluation of environmental pollutants as they are direct recipients of agrotoxics. Tradescantia paludosa, Vicia faba and Allium cepa are generally chosen as test materials for environmental mutagenesis analysis as they have large monocentric chromosomes in reduced numbers. Allium cepa has been used in number of studies with an aim to find cytotoxicity and genotoxicity as it a cheap source and is available throughout the year. Allium cepa root test has been employed for finding genotoxic effects of different compounds like mitotic activity and chromosomal aberrations 32,33 . Genotoxicity of Thermopsis turcica on Allium cepa L. roots was tested by random amplified polymorphic DNA and alkaline comet assays 34 . Similarly, Allium cepa has been used in many other studies involving genotoxic evaluation of different metal compounds including copper [35][36][37] . Here in this paper, Allium cepa L. has been used to investigate the effect of pH of precursor solution and CTAB coating on toxicological aspects of synthesized CuO NPs.

Materials and Methods
preparation of nano-crystalline Cuo Nps. Nano-crystalline CuO NPs in the form of powder has been synthesized using hydrated cupric nitrate (Cu (NO 3 ) 2 .3H 2 O) and monohydrate citric acid (C 6 H 8 O 7 .H 2 O) as precursors. A 100 mL precursor solution in doubly distilled water was prepared with copper nitrate to citric acid (CN:CA) ratio of 1:1. The acidity (pH) of the solution was monitored using Naina make microprocessor-based pH meter. The pH of the precursor solution was attained to value of 2, 7 and 10 by adding dropwise 25% liquid ammonia solution. In an another set of experiment, cetyletrimethylammoniumbromide (CTAB) was selected as cationic surfactant and 10 mL of 0.5 M surfactant solution was mixed in 100 mL precursor solution at pH value of 7. The precursor solution in different sets of experiment was thermally dehydrated in a hot air oven maintained at temperature of 80 ± 5 °C. The viscous fluid was heated to higher temperature using hot plate and gel underwent auto catalytic combustion. The reaction ended with foamy, blackish decomposed residue. The residue was further calcined at temperature of 400 °C in muffle furnace of Macro Scientific make for 4 hours in order to get rid of leftover organic matter. The heating rate was kept at 5 °C/min and the detailed mechanism has been reported earlier 38 . The calcination temperature was selected on the basis of the thermal analysis for the similar type of samples reported earlier 4,38,39 . The calcined CuO powder samples were designated as C1, C2, C3, and C4 corresponding to sample without CTAB at pH value 2, 7 and 10 and with CTAB at pH7 respectively, of the precursor solution.
X-Ray diffraction analysis. The various calcined powder samples were tested for pure CuO phase identification using X'Pert Panlytical X-ray diffractometer with Cu K α radiation having wavelength of 1.5405 Å, and other operating conditions were 30 mA, 40 kV. The samples were scanned in an angle of 2θ ranging from 30-80°.
Field emission seM and High Resolution teM analysis. The various CuO samples were scanned for surface topography using FESEM and HRTEM techniques. The scanning electron micrographs were taken using JEOL JSM-6700F with a beam voltage of 30 KV whereas TEM images were taken using transmission electron microscope system (HRTEM, model FEI Technai 30) operated at 300 kV. For HRTEM images, CuO NPs were ultrasonically dispersed in deionized water and dispersion dropped out onto the copper grid which was air dried and scanned in TEM chamber. treatment sample preparation. Three concentration groups per calcined CuO powder samples, C1, C2, C3 and C4 were made and named as per the Table 1.
Allium cepa root test. Test Material. Onion bulbs (Allium cepa L.) of average size (15-20 mm diameter) were procured from the local market. The onion bulbs were sun-dried for 5 weeks. The roots of dried bulbs were shaved off from the base with a sharp blade. This exposed the fresh meristematic tissues and the bulbs were kept in distilled water to shield the primordials from drying up.
Treatment of test material. After removing excess water with a blotting paper, the bases of the onion bulbs were dipped in solutions of all the CuO-NP test solutions as described in Table 1. A series of seven onion bulbs were used for each sample concentration and control (tap water). The experiment was run for 14 days in dark. After the exposure time is over, out of the seven exposed onion bulbs, best five onions in terms of root length development were chosen for analysis. Rest of the two were not considered in the experiment.
Root length measurement. After the exposure period, five selected onion bulbs were taken for the root length measurement. The root length (cm) of all onion bulbs was measured on 3 rd , 7 th and 14 th day using a calibrated www.nature.com/scientificreports www.nature.com/scientificreports/ ruler. After taking the root lengths, the mean was calculated for each concentration treatment. The mean root length of the control samples was also calculated.
Cytological analysis. For the analysis of chromosomal aberrations, the tips of the emerged roots from the onion bulbs exposed to different sample concentrations were cut and fixed in ethanol:glacial acetic acid (3:1, v/v). This procedure was done for all the five selected onion bulbs in each category of treated and control samples after 3 rd , 7 th and 14 th day. After fixation, the root tips were hydrolyzed in 1 N HCl at 60 °C for five minutes and were washed with double distilled water. Root tips were squashed on a microscopic slide and stained with acetocarmine for 10-15 minutes. The stained slides were covered with cover slips and were sealed to prevent moisture loss. Slides were analyzed at 1000X magnification on a trinocular microscope (Olympus, CX21) fitted with a digital camera (Olympus, E520). statistical analysis. The difference between various cytological parameters for control and exposed groups was analysed using Mann-Whitney U-test. Mean and standard error values were found using descriptive analysis and p < 0.05 was considered as the significant level of the statistical analysis. The data was statistically analysed using the Minitab software version 16.1.0 (Minitab Inc.) for windows.

Results and Discussion
X-ray diffraction analysis. Figure   where β is the full width at half maximum (FWHM), λ = 1.5405 Å is the wavelength of Cu K α radiations, and θ is the Bragg's angle. The crystallite size was calculated using diffraction data of most prominent (002) peak and values obtained were recorded in Table 2. The results show that the crystallite size was found to be minimum (27.73 nm) in case of C3 sample (pH 10 precursor solution), followed by C1 (29.36 nm) and C2 (63.40 nm) samples. Moreover, the addition of CTAB reduced the crystallite size of CuO NPs (C4) to 18.12 nm and was found to be smallest among the four samples. The various lattice parameter (a, b, c, β) and the unit cell volume (V) of CuO lattice in all samples have been calculated using following relations (6) where d corresponds to spacing between adjacent planes, h, k, l are Miller indices of the respective crystal plane. The values obtained for various cell parameters are tabulated in Table 2. The values are found to be in match with results reported in ICDD card. The variation of cell parameter values from respective standard ones reveals the existence of strain as well as imperfections in the lattice structure. The C1 and C2 samples shows tensile strain in crystal structure while the samples C3 and C4 possess compressive type strain. The variation of unit cell volume with pH also indicated various types of strain in the crystal structure 40 .
Morphological characterization of samples. SEM images of various samples recorded at magnifications ranging from 500X to 12000X are shown in Fig. 2. The variation of porosity in agglomerates of particles was influenced by changing the reaction conditions in terms of pH and addition of surfactant. The pH variation affects the reaction rate as well as the liberation of gaseous byproducts in the auto combustion of viscous liquid. The particles in samples C1-C4 appeared to be bound together into agglomerates of various sizes which strongly   www.nature.com/scientificreports www.nature.com/scientificreports/ depend on the rate of combustion reaction. The disintegration of agglomerates as revealed by SEM images with rise in pH of precursor solution is attributed to the increase in gaseous byproducts 40 . The combustion process depends upon the reaction conditions and the rate of various intermediate steps (pre-hydrolysis, hydrolysis and poly-condensation). It has been noticed that variation in pH affects the rate of reaction as well as ionisation of citric acid 40 . In low pH conditions, ionisation of citric acid (CA) is weak as described follows The incomplete ionization of CA results in the formation of weak copper-citrate complex. Further increase in pH value to about 7 increases the ionisation of citric acid as given by following equation 2 2 In case, the pH value is further increased beyond 7, citric acid ionizes completely and results in the formation of strong copper-citrate complex Thus alkaline medium promotes the formation of homogenized viscous gel. In low pH conditions, the rates of hydrolysis and poly-condensation reactions are retarded due to the formation of hydronium ions (H 3 O + ) whereas the respective rates are accelerated in precursor solution having high pH value. The burning of branched copper-oxygen polymeric network in high pH conditions results in the formation of porous CuO powder as a solid product. The differential thermal analysis being conducted for CuO samples were reported earlier 40 which revealed that addition of ammonia not only controls pH but also forms NH 4 NO 3 , which breaks into NO x and O 2 molecules in the gaseous by-products during combustion process.
The addition of CTAB surfactant in reaction mixture further helps to increase porosity in the product as observed in Fig. 2(C4). CTAB consists of long chain of carbon atoms that acts as an additional fuel besides citric acid and its addition retards the reaction 4,38-40 . Increasing the fuel content would further results in more gas liberation which helps to disintegrate the agglomerates into nanoparticles. On the other hand, the increased exothermicity 40 raises the internal reaction temperature that further calcined the formed particles. The particles grow as the reaction propagates even during the post-thermal treatment applied to remove the organics. The long chain of carbon atoms in surfactant acts as space filling secondary material that leaves gap in the calcination process and maintains porosity in residual product as observed in SEM images of sample C4.
HRTEM images of various samples are also embedded in Fig. 3. These images also show the agglomerated CuO NPs. The bigger aggregates, consisting of primary particles were observed in the samples C2. The sample formed with pH 10 (Fig. 3C3) of the precursor solution exhibited a sharp particle distribution and low agglomeration. The C1-C3 samples show comparatively dense agglomerated structure composed of non-uniform spherically shaped particles whereas C4 sample shows improvement in porosity with the CTAB addition as shown in the Fig. 3(C4). CuO NPs synthesized with the addition of CTAB surfactant appeared as almost spherical in shape. TEM analysis of the samples shows broad particle size distribution in C1-C3 samples whereas in C4 sample, crystallite size is reduced as well the particle size distribution is confined in the range of 5-10 nm. The uniformity in the size of CTAB assisted CuO NPs may be assigned to the formation of spherical micelles in copper-citrate complex. The formation of micelles inhibits the crystallite growth as well as controls agglomeration 4,40 . The prominent role played by porosity in improving the adsorption of ammonia gas on the CuO NPs has been already discussed 4,40 . The higher ammonia gas adsorption rate constant indicating the fastest reaction has been noticed in C1 and C3 samples as compared to C2.
The size and stability of CuO NPs can also be investigated by performing Dynamic light scattering experiment to obtain hydrodynamic size (diameter) and zeta potential 41,42 . The hydrodynamic size describing behaviour of the particle in a fluid will be different from the particle size as measured by HRTEM if the particles are coated with protective surfactant/stabilizing agent in post synthesis treatment. In such cases the observed size includes the centre core and the protective layer of surfactant. The techniques like electron microscopy or SAXS separate out the core from protective layer and provides more accurate resultant particle size. In our case, the protective coating of CTAB surfactant controlling particle size and agglomeration in the copper-citrate complex burnt in the auto combustion reaction and any leftover surfactant traces are further eliminated in the calcination process. Thus, average size of CuO NPs as revealed by HRTEM images is approximately the exact size.

Root length measurements.
Mean root lengths were recorded for all the four treatment samples in each concentration group of 0.1, 0.01 and 0.001 g/100 ml of the synthesized CuO nanoparticles (Figs 4 and 5). Figure 4 reveals that at lowest concentration 0.001 g/ml, maximum mean root length at 14 th day of exposure was found in the treatment group C2 (2.8 ± 0.03 cm) followed by C1 (2.16 ± 0.09 cm), C3 (1.66 ± 0.05 cm) and C4 (0.54 ± 0.02 cm) as compared to control group (3.78 ± 0.06 cm; p < 0.05). Similar trend was seen at highest concentration viz. C2: 1.8 ± 0.08 cm, followed by C1: 1.14 ± 0.06 cm, C3: 0.9 ± 0.05 cm and C4: 0.32 ± 0.02 cm (C2,C1,C3: p < 0.05; C4: p < 0.01). Figure 5 shows the relative root lengths at 3 rd , 7 th and 14 th exposure day and reveals the increasing trend among all concentration groups in all the four types of samples with respect to exposure period. Though the mean root length in CuO synthesized using CTAB group (C4) was found to be increasing with exposure period but its rate of growth is found to be lowest among all the groups. The decrease in growth rate in C4 sample might be due to its comparatively smaller size and uniformly distributed crystallite size. Thus, our results suggest a decreased root length with the exposure to CuO NPs at various concentrations. On the same line, Shaymurat et al. 43 reported a concentration-dependent inhibition of root length by ZnO NPs. Treatment with 50 mg/L ZnO NPs for 24 h blocked the root growth of garlic completely. Similarly, Manesh et al. 44 reported that germination index and root length were affected by TiO 2 nanoparticle (n-TiO 2 ) exposure. A significant reduction in root elongation and germination percentage was observed in seeds with co-exposure to n-TiO2 and CdCl2 at the highest concentrations (1000 mg/L and 250 mg/L, respectively) as compared to co-exposure at lower concentrations (1 mg/L and 1 mg/L, respectively) and controls (p < 0.05). Reduced root and shoot lengths in seedlings of CuO and ZnO NP treated plants are also presented 45 . Fe 2 O 3 -NPs hindered the seed germination and root length www.nature.com/scientificreports www.nature.com/scientificreports/ in radish 46 . In another study, 16, 12 and 18 percent reduction in root length; and 22, 16 and 27 percent reduction in shoot length at concentration of 1000 mg/L was found for CuO NPs, ZnO NPs and binary mixture of NPs respectively 47 . Similarly, gamma-Fe 2 O 3 NP concentration of 50 and 100 mg/L remarkably reduced the root length by 13.5 and 12.5 percent, respectively in Zea mays L. 48 .  On the contrary, use of copper nanoparticles with chitosan-PVA hydrogels (Cs-PVA-nCu) was shown to improve the root length in grafted water melon. Cs-PVA-nCu application was found to upsurge primary root and stem length by 14% and 8%, respectively 49 . Similarly, Ag-NP treated seedlings showed an enhancement in root length and production of phytochemical diosgenin to a level of 214.06 ± 17.07 µ/mL as compared to control (164.44 ± 7.67 µ/mL) 50 .
Root elongation has been found to be minimum in case of CTAB assisted CuO NPs. This toxic effect may be attributed to the smaller particulate size of the C4 nanoparticle group. Here we suggest that addition of CTAB at pH7 during the synthesis of CuO NPs decreased the particle size as revealed by XRD analysis. The decrease in particle size with CTAB addition is also reported by Varghese et al. 51 . XRD and Williamson-Hall plot of CuO nanoparticles, with and without CTAB, has revealed the sizes as 11 and 22 nm, respectively. The CTAB appeared to influence the properties and morphology of CuO powder 4 . Variation in CTAB concentration has been shown not to significantly affect the size of Cu nanoparticles 52 . Some studies have also reported the nanoparticles to be better stabilized using CTAB concentration above 1 mM 53 .
Cytological analysis. All the four synthesized CuO NP samples were checked for their cytotoxic potential using Allium cepa. Allium cepa samples were analyzed following treatments with the synthesized CuO nanoparticles at different concentrations as mentioned in the material and methods section 2.5.2.

Number of dividing cells and mitotic index. Number of dividing cells (DC) per 2000 cells at different applied
concentrations of all the four nanoparticles types was recorded. At highest concentration (V3), DC was found to be 312 for C1, 330 for C2, 290 for C3 and 264 in case of C4 as compared to 396 for the control samples (Tap water treatment) ( Table 3). Maximum values for DC was found to be 374 per 2000 cells in the lowest concentration of 0.001 g/100 ml in C2V1 sample. A decreasing trend among DC with increasing concentration was found among all the four treatments groups. On the basis of number of dividing cells, after group-wise clubbing of DC, the order of samples came out to be Control > C2 > C1 > C3 > C4 (Fig. 6). Thus, maximum division reduction was seen in CuO synthesized with the addition of CTAB surfactant exposure group (C4) at all the applied concentrations. The addition of CTAB decreased the particle size of NPs as described in Table 2. This addition may be the reason for reduced DC values in C4 samples.
Mitotic Index (MI) was calculated as described previously in the methods section (Eq. 1). At highest concentration group (0.1 g/100 ml), a decreasing trend was found in MI viz. Control (19.8%) > C2V3 (16.5%) > C1V3 (15.6%) > C3V3 (14.5%) > C4V3 (13.2%). Same trend was found at each applied concentration for all four treatment groups (Table 3). In line with our results, a concentration-and time-dependent decrease in mitosis index has been reported 43 . Similarly, Ag NPs have been shown to induce a decrease in mitotic index using Allium cepa 54 .   www.nature.com/scientificreports www.nature.com/scientificreports/ Anti-proliferative and antimitotic activities were found to be amplified on cellular treatment with plumbagin -AgNPs 55 . A dose dependent reduction in the mitotic index from 69 to 21 was also found in Allium cepa root test when exposed to TiO 2 NPs 56 . Decrease in MI was also reported in many other studies following exposure to different NPs like AgNPs 54,57 ; chromium (III) oxide nanoparticles 58 and aluminum oxide nanoparticles 59 . On the contrary, exposure to Bismuth (III) oxide nanoparticles was found to significantly increase MI in Allium cepa 60 .
Mitotic inhibition. Mitotic inhibition (MI) was calculated for all the treatment samples as per Eq. 2. MI was found to be increasing with the treatment concentration as maximum was found in C4V3 (33.33) and least was found to be in C2V1 (5.55). Table 4 shows the relative mitotic inhibition in four exposure groups of synthesized nanoparticles. Increasing trend in mitotic inhibition was found to be C2 < C1 < C3 < C4 samples (Fig. 7).
Genotoxic parameters. Chromosomal aberrations. Genotoxic parameters were assessed following exposure treatments with all the four nanoparticles samples at different concentrations. Chromosomal aberrations were observed in the slides for each treatment concentration. Various types of chromosomal aberrations per 2000 cells including binucleated cells, micronucleated cells, chromosomal bridges, fragmentation, stickiness, laggards, vagrants, c-mitosis and disturbed spindle were scored (Table 5). Maximum total aberrations were found to be in 0.1 concentration group of C4 (50) as compared to C3 (35), C1 (28) and C2 (25). Total aberrations were found to be associated with increasing sample concentrations. On the same line, ZnO NPs have been reported to induce several kinds of mitotic aberrations including chromosome stickiness, bridges, breakages and laggings 43 . Many    61 demonstrated that Ag NPs and AgNO3 induced oxidative stress that was manifested in terms of DNA degradation. In another study, a concentration dependent increase in DNA strand breaks has been reported in Fe 2 O 3 -NPs treated groups of Raphanus sativus using comet assay. Cell cycle analysis also revealed 88.4% of cells in sub-G1 apoptotic phase, signifying cell death in 2.0 mg/mL Fe 2 O 3 -NPs treated group 46 . About 2.4-fold higher DNA damage was observed by comet assay in Cobalt oxide nanoparticles treated cells of eggplant as compared to untreated control 62 . Increased chromosomal aberrations in Allium cepa were also reported following exposures to AgNPs 54,57 ; chromium (III) oxide nanoparticles 58 and TiO 2 NPs 56 .
Aberrant cells. Percentage of aberrant cells (AC) was calculated as the number of cells with any kind of aberration as per Eq. 3 (Table 5). AC in highest treatment concentration of 0.1 g/100 ml was found out to be 8.97% (C1V3); 7.57% (C2V3); 12.06% (C3V3) and 18.93% (C4V3). Out of the four treatment groups, maximum aberrant cell percentage was found to be in 0.1 concentration group of CuO NPs synthesized with CTAB surfactant (18.93%; C4V3), lowest being in 0.001 concentration treatment group of CuO synthesized at pH7 (2.67%; C2V1). AC compared at the lowest applied concentration of 0.001 g/100 ml was found to be 3.71% (C1); 2.67% (C2); 4.90% (C3) and 7.74% (C4). Again the maximum aberrant cells were found in the CTAB-CuO exposure treatments which may be attributed to smaller particle size of CuO NPs. In another study, size dependent toxicity was reported as 20 and 50 nm AuNPs did not induce obvious DNA damage at the tested concentrations whereas 5 nm NPs induced a dose-dependent increment in DNA damage in HepG2 cells 63 . Thus, smaller particles induce more toxicity in terms of genetic instability and aberrations. Higher DNA damages were also reported after exposure to different nanoparticles in Allium cepa root test 54,56,60 .   www.nature.com/scientificreports www.nature.com/scientificreports/

Conclusion
CuO nanoparticles (NPs) with high catalytic activity were synthesized using sol-gel auto combustion route at different pH values and with the addition of cationic surfactant, CTAB at pH value of 7 of the precursor solution. The increase in pH value of the precursor solution increases the rate of combustion resulting in the production of highly porous and active CuO particles. The catalytic activity of synthesized CuO particles was found to be further enhanced with the addition of surfactant CTAB. Surfactant addition not only decreased the crystallite size but also made the particle agglomerates of uniform size. This uniform and small size made the CuO particles more toxic. The toxicity of the particles is found to be a function of the pH and strong dependence on the surfactant addition. The high catalytic activity in terms of toxicity has been proved in the form of decrease in root length measurements among the tested samples for same exposure duration at different treatment concentrations. Cytological analysis with measurements of dividing cells, percentage of aberrant cells, mitotic index and mitotic inhibition proved that CTAB assisted particles are more toxic than nascent CuO NPs revealing their high catalytic activity. Further studies are recommended so as to explore the effects of CTAB and other surfactants on the crystallite structure of CuO NPs.