Influence of Co3O4 Nanostructure Morphology on the Catalytic Degradation of p-Nitrophenol

The design and fabrication of nanomaterials with controllable morphology and size is of critical importance to achieve excellent catalytic performance in heterogeneous catalysis. In this work, cobalt oxide (Co3O4) nanostructures with different morphologies (nanoplates, microflowers, nanorods and nanocubes) were successfully constructed in order to establish the morphology–property–performance relationship of the catalysts. The morphology and structure of the nanostructured Co3O4 were characterized by various techniques, and the catalytic performance of the as-prepared nanostructures was studied by monitoring the reduction of p-nitrophenol to p-aminophenol in the presence of excess NaBH4. The catalytic performance was found to be strongly dependent on their morphologies. The experimental results show that the pseudo-first-order reaction rate constants for Co3O4 nanostructures with various shapes are, respectively, 1.49 min−1 (nanoplates), 1.40 min−1 (microflowers), 0.78 min−1 (nanorods) and 0.23 min−1 (nanocubes). The Co3O4 nanoplates exhibited the highest catalytic activity among the four nanostructures, due to their largest specific surface area, relatively high total pore volume, best redox properties and abundance of defect sites. The established correlation between morphology, property and catalytic performance in this work will offer valuable insight into the design and application of nanostructured Co3O4 as a potential non-noble metal catalyst for p-nitrophenol reduction.


Introduction
Nanomaterials have been overwhelmingly developed in the fields of electronics, optics, biology, medicine, environmental science and catalysis in recent decades, due to their unique physical, chemical and biological properties that differ from those of bulk materials [1].Among these, transition metal oxide nanomaterials have attracted particular research interest for potential applications in catalysis because of their abundant reserves, low cost, high stability and strong redox properties.However, the catalytic activity of transition metal oxides needs to be further improved compared with the excellent performance of noble metal catalysts [2].Various strategies have been adopted to improve the catalytic performance of transition metal oxides, such as size control [3], crystal plane or morphology modulation [4], element doping, surface defect fabrication [5], pore introduction [6] and compound recombination [7].For example, atomically dispersed Co catalysts anchored in porous nitrogen-doped carbon (Co SA -N-C) showed outstanding performance in removing low concentrations of HCHO at room temperature [8].The catalytic activity of Ce-doped CuMn 8 O x (2.5% Ce; molar percentage) was found to be higher than that of CuMn 8 O x for the oxidation of CO at a lower temperature (65 • C) [9].Due to the synergistic effect of ZIF-8 and TiO 2 , the ZIF-8@TiO 2 micron composite has been used as a highly efficient photocatalyst for the degradation of tetracycline [10].Therefore, with the rapid development of nanotechnology and the use of various regulation methods, transition metal oxides as catalysts are expected to have potential in partially substituting noble metal catalysts.
As a typical transition metal oxide, Co 3 O 4 has been extensively used as a heterogeneous catalyst in many important chemical processes, such as low temperature oxidation of CO [11], CH 4 combustion [12], oxygen evolution reaction [13], decomposition of H 2 O 2 [14], oxidation of ethylene [15] and reduction of p-nitrophenol (p-NP) [16].In general, the physicochemical properties of Co 3 O 4 are significantly influenced by its morphology [17]; thus, it is of great significance to develop suitable preparation methods to obtain Co 3 O 4 nanomaterials with controllable morphology and size in order to upgrade the properties of Co 3 O 4 and further improve its catalytic performance.To date, many Co 3 O 4 nanomaterials with different shapes have been successfully fabricated.For example, Yan et al. [18] synthesized Co 3 O 4 nanoflower clusters by a simple low-temperature hydrothermal method.The catalytic activity of the as-prepared Co 3 O 4 nanoflower clusters for the degradation of gaseous toluene appeared to be far superior to that of Co 3 O 4 blocks under the same reaction conditions.Xie et al. [19] reported the synthesis of Co 3 O 4 nanorods by a precipitation-calcination method.The catalyst exhibited surprisingly high catalytic activity for CO oxidation at temperatures as low as −77 • C. Ultrathin Co 3 O 4 nanosheets were also synthesized by a facile solvothermal method for the catalytic oxidation of formaldehyde at room temperature [20].
p-Nitrophenol (p-NP), which is toxic, mutagenic and non-biodegradable, is one of the most common organic pollutants in industrial and agricultural wastewater.Due to its ability to bioaccumulate and its stability, effluents containing p-NP can cause serious contamination to soil and groundwater if discharged directly into the environment without treatment [21].In addition, its accumulation and residue in the environment may irritate eyes and skin, damage the liver, kidney and nervous system, and even cause chromosomal aberrations or blood disorders in humans [22].Many strategies have been proposed for the degradation of p-NP, such as adsorption [23], microbial degradation [24], photocatalytic degradation [25], electrochemical methods [26] and chemical reduction [27].Compared with other techniques, the catalytic reduction of p-NP to p-aminophenol (p-AP) by an excess NaBH 4 aqueous solution is one of the most feasible and simple methods due to its high efficiency and facile operating conditions [28].Moreover, the only reduction product, p-AP, is not only of low toxicity, but also an important intermediate in the production of various pharmaceuticals, agrochemicals, dyes, chelating agents and photographic developers [29].
The transition metal oxides, such as Co 3 O 4 , CuO, Fe 2 O 3 and NiO, were found to be catalytically active for the reduction of p-NP to p-AP [30,31].In addition, several studies have shown that the shape and size of the catalyst have a significant effect on the catalytic activity for the reduction of p-NP.For example, Konar et al. [32] synthesized rod-, spherical-, star-, and flower-shaped CuO nanostructures.The star-shaped CuO showed the maximum catalytic activity for the reduction of p-NP in the presence of NaBH 4 , which is correlated to their higher specific surface area, positive surface charge and the presence of a high index facet.Che et al. [33] obtained leaf-like, dumbbell-like, and flower-like CuO nanostructures via a simple oil bath route by introducing different additives.The order of catalytic activity for the reduction of p-NP to p-AP is leaf-like nanosheets > dumbbell-like architectures > flower-like nanostructures.In contrast, there are only a few reports available in the literature showing the effect of Co  [34].However, the morphologies of these Co 3 O 4 catalysts are all microballs consisting of granules, sheets, nanoneedles or petals.The morphology, structure and physicochemical properties of different metal oxides vary greatly, and their catalytic performance can be markedly different.Therefore, it would be essential to investigate Co 3 O 4 with different morphologies in order to explore the morphology-property-performance relationships and to provide insights into the design of Co 3 O 4 catalysts to optimize their catalytic performance for p-NP reduction.
In this work, Co(OH) 2 nanoplates, CoC 2 O 4 nanorods and Co(OH) 2 microflowers as catalyst precursors were synthesized by a simple precipitation method.CoC 2 O 4 nanocubes were obtained by a facile hydrothermal method.Then, Co 3 O 4 nanostructures with different morphologies were successfully constructed by calcining each precursor, and the as-prepared nanostructures retained the morphologies of the precursors.The composition, size, morphology and textural properties of the obtained Co 3 O 4 nanostructures were characterized by various techniques, such as XRD, TEM, SEM, BET, TPR and XPS.Furthermore, the catalytic performance of Co 3 O 4 with different morphologies was studied by monitoring the reduction of p-NP to p-AP in the presence of excess NaBH 4 .Then, the effect of the morphology of Co 3 O 4 nanostructures on the catalytic performance was investigated, and a strong dependence was found.

TEM and SEM
The morphology and structure of the as-prepared Co3O4 nanostructures were investigated by scanning electron microscopy (SEM) and transmission electron microscopy (TEM) (Figure 2).As shown in Figure 2a,c, the size distribution of the as-prepared Co3O4 nanoplates is nearly uniform and they all consist of well-defined hexagonal plates with an average side length of about 100 nm and an average thickness of about 20 nm.Numerous

TEM and SEM
The morphology and structure of the as-prepared Co 3 O 4 nanostructures were investigated by scanning electron microscopy (SEM) and transmission electron microscopy (TEM) (Figure 2).As shown in Figure 2a,c, the size distribution of the as-prepared Co 3 O 4 nanoplates is nearly uniform and they all consist of well-defined hexagonal plates with an average side length of about 100 nm and an average thickness of about 20 nm.Numerous mesopores with a diameter of 5-10 nm were observed on the surface of the Co 3 O 4 nanoplates, which were formed by the dehydration of Co(OH) 2 during the hightemperature calcination process.No fragment was observed in the TEM and SEM images, indicating that all the particles were assembled into hexagonal nanoplates.The Co 3 O 4 microflowers, assembled by a large number of intertwined nanorods and some nanosheets, have a uniform particle size distribution, and the diameter of the Co 3 O 4 microflowers ranges from 0.8 µm to 1.2 µm.Since the intersection of nanorods can form various pores, there are many interstices in the microflowers.No cracked microcrystalline flower was observed in the SEM or TEM images.In addition to the flower-like structure, some of the nanorod pieces did not form a flower-like structure (Figure 2d,f).As shown in Figure 2g,i, the Co 3 O 4 nanorod samples have an average length of about 1 µm and an average width of about 100 nm.Numerous pores and gullies were clearly observed on the surface of the nanorods, which can be attributed to the release of gas from CoC 2 O 4 during the high-temperature decomposition process (described in Section 3.4).It is obvious from the images that some particles have not been assembled into nanorods.As shown in Figure 2j,l, the Co 3 O 4 nanocube samples all have a typical cubic shape with some very small pores on the smooth surface and an average edge length of 40 nm.Furthermore, the size distribution of the Co 3 O 4 nanocubes is regular, and no fragments were observed in the images.Figure 2b,e,h,k shows the HRTEM images of these Co 3 O 4 nanostructures.The lattice fringes of the Co 3 O 4 nanoplates, Co 3 O 4 nanorods and Co 3 O 4 nanocubes with an interplanar spacing between adjacent planes of 0.28-0.29 nm are attributed to the (2 2 0) facets of the cubic Co 3 O 4 (JCPDS file value, 0.29 nm).The lattice space of the Co 3 O 4 microflowers is 0.4680 nm, which is comparable to the (1 1 1) plane of Co 3 O 4 (JCPDS file value, 0.467 nm) and, thus, is in good accordance with the XRD result.

N 2 Physisorption
In general, the catalytic activity of metal oxide nanomaterials is highly dependent on their surface area and pore volume; therefore, it is indispensable to investigate the relationship between the specific surface area and porosity of these Co 3 O 4 nanostructures for a better elucidation of their catalytic activity.The nitrogen adsorption-desorption isotherms and BJH adsorption pore size distributions of these Co 3 O 4 nanostructures with different morphologies are plotted in Figure 3.As depicted in Figure 3a, all the as-prepared Co 3 O 4 samples exhibit a type IV adsorption isotherm pattern and an H1 hysteresis loop according to the IUPAC classification, demonstrating the presence of a regular mesoporous structure [35].The specific surface areas and pore volumes of these nanostructures based on the BET results are summarized in Table 1.The Co 3 O 4 nanoplates have the largest specific surface area of 106.3 m 2 •g −1 .The specific surface areas of the Co 3 O 4 microflowers, nanorods and nanocubes are 69.3 m 2 •g −1 , 51.2 m 2 •g −1 and 31.7 m 2 •g −1 , respectively.The total pore volumes of the Co 3 O 4 nanoplates and microflowers are relatively similar, being 0.27 cc.g −1 and 0.30 cc.g −1 , respectively.In contrast, the Co 3 O 4 nanorods and nanocubes have identical pore volumes, but the value is relatively small at only 0.21 cc.g −1 .The pore size distribution derived from the adsorption branch of the Co 3 O 4 nanoplates shows a broad distribution, ranging from 3.3 nm to 59.9 nm, while those of the other three Co 3 O 4 samples are even broader, ranging from 1.9 nm to 78.2 nm (Figure 3b), which can be attributed to the intra-aggregated pores within the agglomerated particles [36].
The total pore volumes of the Co3O4 nanoplates and microflowers are relatively similar, being 0.27 cc.g −1 and 0.30 cc.g −1 , respectively.In contrast, the Co3O4 nanorods and nanocubes have identical pore volumes, but the value is relatively small at only 0.21 cc.g −1 .
The pore size distribution derived from the adsorption branch of the Co3O4 nanoplates shows a broad distribution, ranging from 3.3 nm to 59.9 nm, while those of the other three Co3O4 samples are even broader, ranging from 1.9 nm to 78.2 nm (Figure 3b), which can be attributed to the intra-aggregated pores within the agglomerated particles [36].The catalytic activity of a catalyst for redox reactions depends largely on its reducibility [37].The H 2 -TPR profiles were determined to investigate the oxidation-reduction properties of the Co 3 O 4 samples with different morphologies, as illustrated in Figure 4.All four Co 3 O 4 nanostructures exhibit two distinct reduction peaks in the range of 200-500 • C, corresponding to an overall two-step reduction process with CoO as the intermediate (i.e., the reduction of Co 3+ to Co 2+ at a relatively low temperature and then Co 2+ to Co at a relatively high temperature), similar to that reported in the literature [38].Curvefitting of the TPR profiles of the Co 3 O 4 nanoplates, microflowers and nanorods resulted in the resolution of three reduction sub-processes, while four reduction sub-processes were identified in the TPR curve of the Co 3 O 4 nanocubes.The peak located at a relative high temperature can be deconvoluted into two adjacent peaks, indicating that the reduction of Co 2+ to Co over all four catalysts involved the overlapping of two events, and that Co 2+ was not completely reduced to metallic Co in one step.The peaks of the Co 3 O 4 nanoplates, microflowers and nanorods at a relatively low temperature can only be deconvoluted into one peak, indicating that Co 3+ was reduced to Co 2+ in one step.However, the peak of the Co 3 O 4 nanocubes centered around such a low temperature can be fitted with two peaks at temperatures of 287 • C and 333 • C (area ratio 1:4.2).The weak peak at 287 • C may be attributed to the reduction of a small amount of Co 3+ to Co 2+ [39].The relative area of the two peaks located at the low-and high-temperature regions can be determined by means of the deconvolution above.The area ratio of the two peaks is calculated to be 1:3.2.1.5.XPS X-ray photoelectron spectroscopy was carried out to qualitatively characterize the chemical composition and surface chemical states of these Co 3 O 4 nanostructures [40].Figure 5a shows the XPS full spectra of these nanostructures.It is evident that each peak in the spectra can be attributed to Co, O or C elements, indicating the absence of other metallic or inorganic impurities.In general, the weak C1s peak can be attributed to the hydrocarbon contaminations inherent in the XPS analysis; therefore, these Co   2+ and more oxygen vacancies were formed [16].Therefore, the amount of oxygen vacancies on the surface of the Co 3 O 4 nanostructures is ordered as follows: Co 3 O 4 nanoplates > Co 3 O 4 nanorods > Co 3 O 4 microflowers > Co 3 O 4 nanocubes.Furthermore, the high-resolution O 1s spectra of these Co 3 O 4 nanostructures observed at 528.9~529.2eV can be deconvoluted into two typical peaks with binding energies of 528.9~529.2eV and 530.0~530.4eV, which are attributed to the lattice oxygen (O lat ) and the surface adsorption oxygen species (O sur ), respectively [41].No characteristic peaks assigned to oxygen vacancies were observed, indicating that the amount of oxygen vacancies formed was very small.

Catalytic Performance of Co 3 O 4 Nanostructures
The catalytic reduction of p-NP to p-AP in the presence of excess NaBH 4 was chosen as a probe reaction to evaluate the catalytic performance of Co 3 O 4 nanostructures with different morphologies.The reaction is considered as a thermodynamically favorable process according to the standard electrode potential (p-NP/p-AP = −0.76V, H 3 BO 3 /BH 4 − = −1.33V), but it is difficult for electrons to transfer from BH 4 − to the p-NP substrate due to the large difference in the redox potential and the high kinetic barrier between the mutually exclusive negative donor (BH 4 − ions) and acceptor (p-nitrophenolate ions) [42].Therefore, the NaBH 4 solution itself cannot reduce p-NP to p-AP in the absence of a highly efficient catalyst.Co 3 O 4 has been proven as an effective catalyst for the reduction of p-NP, which can reduce the kinetic barrier [43].The catalytic reduction of p-NP to p-AP using Co 3 O 4 nanostructures as a catalyst was conducted in a standard quartz cell and monitored by an in situ UV-vis spectrophotometer.The p-NP solution was first homogeneously mixed with a NaBH 4 solution, and then different shapes of Co 3 O 4 nanostructures were added to initiate the reduction reaction.The pale-yellow aqueous solution of p-NP exhibited a strong absorption peak centered at 317 nm, as shown in Figure 6.When the NaBH 4 aqueous solution was added, the absorption peak was red-shifted to 400 nm, accompanied by a color change to green-yellow, which could be attributed to the formation of p-nitrophenolate ions (i.e., p-NP anions) under alkaline conditions [44].When the Co 3 O 4 nanostructures were introduced into the mixed solution of p-NP and NaBH 4 , the absorption peak of p-NP at 400 nm gradually decreased with the fading of the yellow color and the simultaneous appearance of a new absorption peak at 300 nm (which can be attributed to the formation of p-AP).The absorption peak at 400 nm was significantly stronger than that at 300 nm; therefore, the progress or kinetics of the reaction was measured based on the concentration of p-NP anions in the reaction mixture by monitoring the change in the absorbance at 400 nm at different time intervals.
tion peak was red-shifted to 400 nm, accompanied by a color change to green-yellow, which could be attributed to the formation of p-nitrophenolate ions (i.e., p-NP anions) under alkaline conditions [44].When the Co3O4 nanostructures were introduced into the mixed solution of p-NP and NaBH4, the absorption peak of p-NP at 400 nm gradually decreased with the fading of the yellow color and the simultaneous appearance of a new absorption peak at 300 nm (which can be attributed to the formation of p-AP).The absorption peak at 400 nm was significantly stronger than that at 300 nm; therefore, the progress or kinetics of the reaction was measured based on the concentration of p-NP anions in the reaction mixture by monitoring the change in the absorbance at 400 nm at different time intervals.

Catalytic Activity Test
Figure 7 shows the extinction UV-vis absorption spectra of p-NP after reaction for different durations in the presence of 0.2 mg of Co3O4 nanostructures with different morphologies.It can be seen that the absorption peak of p-NP at 400 nm decreased with time and a new peak was simultaneously generated at a wavelength of 300 nm, indicating the

Catalytic Activity Test
Figure 7 shows the extinction UV-vis absorption spectra of p-NP after reaction for different durations in the presence of 0.2 mg of Co 3 O 4 nanostructures with different morphologies.It can be seen that the absorption peak of p-NP at 400 nm decreased with time and a new peak was simultaneously generated at a wavelength of 300 nm, indicating the reduction of p-NP to p-AP. Figure 7a-d shows that with the prolongation of the reaction, the decrease rate of the peak intensity at 400 nm varies greatly among these Co 3 O 4 nanostructures, implying that different shapes of Co 3 O 4 have different catalytic activities for the p-NP reduction reaction.When the Co 3 O 4 nanoplates were used as catalyst, the time to complete the reduction process was less than 3.9 min (Figure 7a).The complete conversion of p-NP to p-AP in the presence of Co 3 O 4 microflowers took 5.4-6.0 min (Figure 7b).These are in comparison with 10.7 min to complete the reaction over Co 3 O 4 nanorods (Figure 7c) and approximately 30.0 min over Co 3 O 4 nanocubes (Figure 7d).Thus, Co 3 O 4 nanoplates render the best catalytic activity, followed by Co 3 O 4 microflowers, then nanorods and, finally, nanocubes.
In addition to the different time required for the complete conversion of p-NP to p-AP over these Co 3 O 4 nanostructures, an induction time (t 0 ) is present, during which no reaction occurred initially.The induction time of these Co 3 O 4 nanostructures differs significantly, as summarized in Table 3.The order of induction time required for the reduction of p-NP catalyzed by these Co 3 O 4 catalysts is as follows: nanoplates < microflowers < nanorods < nanocubes, which is consistent with the order of the catalytic activity, that is, the shorter the induction time, the higher the catalytic activity.There have been many reports on the existence and happenstance of the induction time.Three main hypotheses have been found in the literature, namely, the spontaneous reconstruction time of the catalyst surface, the activation time of the catalyst active site, and the time required for reactants to diffuse onto the catalyst surface; however, the existence of the induction time is still controversial, and a concrete reason needs to be further verified [45].
nanorods < nanocubes, which is consistent with the order of the catalytic activity, that is, the shorter the induction time, the higher the catalytic activity.There have been many reports on the existence and happenstance of the induction time.Three main hypotheses have been found in the literature, namely, the spontaneous reconstruction time of the catalyst surface, the activation time of the catalyst active site, and the time required for reactants to diffuse onto the catalyst surface; however, the existence of the induction time is still controversial, and a concrete reason needs to be further verified [45].7 and 8.
As a typical model reaction, the kinetics of the catalytic reduction of p-NP to p-AP by NaBH 4 has been extensively investigated.In this work, the initial molar ratio of NaBH 4 to p-NP was 100, i.e., there was a significant excess of NaBH 4 .In this case, the concentration of BH 4 − remained essentially constant throughout the reaction, and the reaction rate was independent of the concentration of NaBH 4 .Then, it is reasonable to assume that the reaction follows the pseudo-first-order kinetics with respect to p-NP.The kinetic equation for the reduction reaction can be described as follows: where A t and A 0 are the absorbance values of p-NP (λ = 400 nm) at time t and at the initial time, respectively.C t and C 0 represent the equivalent concentrations of the p-nitrophenolate ions obtained from A t and A 0 , respectively.K app and t represent the apparent rate constant and reaction time, respectively.Therefore, the apparent rate constant values of these Co p-NP was 100, i.e., there was a significant excess of NaBH4.In this case, the concentration of BH4 − remained essentially constant throughout the reaction, and the reaction rate was independent of the concentration of NaBH4.Then, it is reasonable to assume that the reaction follows the pseudo-first-order kinetics with respect to p-NP.The kinetic equation for the reduction reaction can be described as follows: where At and A0 are the absorbance values of p-NP (λ = 400 nm) at time t and at the initial time, respectively.Ct and C0 represent the equivalent concentrations of the p-nitrophenolate ions obtained from At and A0, respectively.kapp and t represent the apparent rate constant and reaction time, respectively.Therefore, the apparent rate constant values of these Co3O4 nanostructures are determined by the slopes of the respective linear curves of t versus −ln(At/A0).As shown in Figure 8 The literature has demonstrated that the reduction of p-NP to p-AP is a surfacecontrolled process that can be analyzed by the Langmuir-Hinshelwood mechanism [35].The model reaction assumes that p-NP ions and BH 4 − are competitively adsorbed on the surface of Co 3 O 4 .The adsorbed BH 4 − reacts with the Co 3 O 4 surface to transfer active hydrogen species and electrons to the surface.These Co 3 O 4 catalysts with different morphologies then serve as catalysts to relay active hydrogen species and electrons to the adsorbed p-NP anions to form p-AP, which is considered as the rate-determining step.The generated p-AP is then desorbed from the catalyst surface to expose the active sites for the next reduction cycle.The adsorption and desorption processes are supposed to be fast; therefore, the efficient adsorption of p-NP and BH 4 − and the efficient desorption of p-AP, as well as the superior electron transport capacity are key factors for the excellent activity of a catalyst [46].It is well known that the surface properties of the catalyst can be tailored by controlling the morphology, size, composition and porosity of the catalyst, thereby improving its catalytic performance [47,48].Different morphologies of Co 3 O 4 are associated with different exposed crystalline planes, specific surface area, surface defects, pore structure and reducibility, which can, in turn, have a significant impact on their catalytic performance [49].
In general, the large specific surface area of the catalyst is conducive to the adsorption of reactant molecules (i.e., BH 4 − and p-NP) on its surface, and its surface can expose more active sites, which largely ensures its excellent catalytic performance.According to the BET results in Figure 3, the specific surface areas of the Co 3 O 4 nanoplates, microflow-ers, nanorods and nanocubes are 106.3,69.3, 51.2 and 31.7 m 2 /g, respectively, which is consistent with the order of their catalytic activity [33].Pore structure is also one of the most important factors affecting catalyst performance, as the abundance of interparticle mesopores and high pore volume can provide more active surface sites for a heterogeneous catalytic reaction.It can be clearly seen from the TEM and SEM results (Figure 2) that there are many small pores on the surface of the Co 3 O 4 nanoplates, and the surface of the Co 3 O 4 microflowers is rough.The BET results (Figure 3 and Table 1) show that the total pore volumes of the Co 3 O 4 nanoplates and microflowers are 0.27 and 0.30 cc/g, respectively, which is significantly higher than those of the Co 3 O 4 nanorods and nanocubes.Despite the rough surface of the Co 3 O 4 nanorods, the total pore volume is only 0.21 cc/g, which is the same as that of the Co 3 O 4 nanocubes with a smooth surface.As a result, the catalytic activity of the Co 3 O 4 nanoplates and microflowers is much higher than that of the Co 3 O 4 nanorods and nanocubes [50].Oxygen vacancies, as a common inherent surface defect in metal oxides, have been proven to significantly affect or even modify the physical and chemical properties of materials and are an effective means to regulate the structure and catalytic performance of catalysts.In addition, our previous results have demonstrated that oxygen vacancies can act as active sites for the reduction of p-NP [16].The XPS results show that the amount of oxygen vacancies on the surface of Co − to the acceptor p-NP.Therefore, the reducibility of Co 3 O 4 also greatly affects its catalytic performance.According to the TPR results (Figure 4), the order of the reducibility of these different Co 3 O 4 catalysts is as follows: Co 3 O 4 nanoplates > Co 3 O 4 microflowers > Co 3 O 4 nanorods > Co 3 O 4 nanocubes, which is consistent with the order of their catalytic activity.In summary, different morphologies of Co 3 O 4 render different properties, such as different specific surface area, surface defects, porosity and reducibility, which, in turn, significantly affect their catalytic performance.
Among these factors, the specific surface area (S BET ) seems to have the greatest influence on the catalytic performance, given that it is much higher for the Co 3 O 4 nanoplates than that for the other shapes (Table 1).To better elucidate the role of other factors, the reaction rate constants per unit specific surface area (k app /S BET ) were adopted to compare the activity of Co 3 O 4 with different morphologies.The values of k app /S BET are 0.020 g•min −1 •m −2 for the Co 3 O 4 microflowers, 0.015 g•min −1 •m −2 for the Co 3 O 4 nanorods, 0.014 g•min −1 •m −2 for the Co 3 O 4 nanoplates and 0.007 g•min −1 •m −2 for the Co 3 O 4 nanocubes (Figure 8b, Table 4).The sequence of the catalytic activity per unit-specific surface area of these Co 3 O 4 nanostructures is as follows: Co 3 O 4 microflowers > Co 3 O 4 nanorods > Co 3 O 4 nanoplates > Co 3 O 4 nanocubes.The Co 3 O 4 microflowers exhibited the best catalytic activity per unit specific surface area, which can be attributed to its largest total pore volume.Therefore, compared with the specific surface area and total pore volume, the effect of reducibility and surface defects is believed to be relatively small in determining the current catalyst performance.

Stability Test
Considering the economic benefits and environmental factors in the practical applications, recyclability is another criterion to evaluate the performance of catalysts in addition to their catalytic activity.The recyclability of these Co 3 O 4 nanostructures was evaluated by continuously adding a concentrated p-NP aqueous solution after each cycle to adjust the concentration of p-NP to its initial value of 0.125 mmol/L before the reaction.The NaBH 4 concentration in the first cycle was 12.5 mmol/L, which is 100 times the concentration of p-NP, and no NaBH 4 was added in the subsequent cycles.As can be seen from Figure 9, although the apparent rate constants of all the Co 3 O 4 nanostructures are somewhat reduced, they show no significant deactivation and still maintain satisfactory stability over multiple consecutive cycles.After five consecutive cycles, the apparent rate constants could still reach 1.37 min −1 for the Co 3 O 4 nanoplates, 1.16 min −1 for the Co 3 O 4 microflowers, 0.61 min −1 for the Co 3 O 4 nanorods and 0.16 min −1 for the Co 3 O 4 nanocubes.To better evaluate the performance of the current catalysts, the activity of some common (supported) noble metal catalysts and other cobalt-based catalysts are summarized in Table 4.As can be seen from the table, the catalytic activity (k nor ) of the as-prepared Co 3 O 4 nanoplates and Co 3 O 4 microflowers in this work is comparable or even better than that of some noble metal catalysts and supported ones (e.g., Au, Pd, Au-Cu x O y in HPSNs and Ag-CeO 2 ) and is significantly higher than that of other cobalt-based catalysts (e.g., PdO-Co 3 O 4 , NiCo 2 O 4 , and Co 3 O 4 @C).The results here clearly demonstrate the relatively high efficiency of the Co 3 O 4 nanoplates and microflowers for the catalytic reduction of p-NP.
Table 4.As can be seen from the table, the catalytic activity (knor) of the as-prepared Co nanoplates and Co3O4 microflowers in this work is comparable or even better than that some noble metal catalysts and supported ones (e.g., Au, Pd, Au-CuxOy in HPSNs and A CeO2) and is significantly higher than that of other cobalt-based catalysts (e.g., PdO-Co3 NiCo2O4, and Co3O4@C).The results here clearly demonstrate the relatively high efficien of the Co3O4 nanoplates and microflowers for the catalytic reduction of p-NP.

Preparation of Co(OH)2 Nanoplates
The Co(OH)2 nanoplates were synthesized according to our previous work [

Materials
Cobalt nitrate hexahydrate (Co , polyvinyl pyrrolidone K90 ((C 6 H 9 NO) n , PVP), sodium dodecyl benzene sulfonate (C 18 H 29 NaO 3 S, SDBS), sodium borohydride (NaBH 4 ) and p-nitrophenol (C 6 H 5 NO 3 ) of analytical grade were purchased from Aladdin Industrial Corporation (Shanghai, China).All the chemicals were used as received without further treatment.Deionized water was used in all the experiments.

Preparation of Co(OH) 2 Nanoplates
The Co(OH) 2 nanoplates were synthesized according to our previous work [16].Briefly, CoCl 2 •6H 2 O (5 mmol) and NaOH (50 mmol) were dissolved separately in 50 mL deionized water.The precipitation was started by the dropwise addition of CoCl 2 and NaOH aqueous solutions into a three-necked flask containing 50 mL deionized water at 50 • C under N 2 atmosphere.The obtained pink suspension was stirred at 50 • C for 1 h.The N 2 flow was then stopped.The product was cooled to room temperature, centrifuged, washed several times with deionized water and dried in the oven (60 • C) for 8 h.

Preparation of Co(OH) 2 Microflowers
The Co(OH) 2 microflowers were synthesized by a simple precipitation method.In a typical experiment, Co(NO 3 ) 2 •6H 2 O (4 mmol) and CO(NH 2 ) 2 (0.5 mol) were dissolved in 50 mL deionized water.Then, polyvinyl pyrrolidone (3.85 µmol) was added to the solution described above.The mixture was ultrasonicated for 1 h to form a homogeneous solution.The solution was then transferred to a three-necked flask and heated in a water bath at 80 • C for 6 h with magnetic stirring.The flask was semi-sealed during the reaction process.Finally, a purple precipitate was obtained.After cooling to room temperature, the purple precipitate was centrifuged, washed several times with deionized water and dried in the oven (60 • C) for 8 h.

Preparation of CoC 2 O 4 Nanorods
The CoC 2 O 4 nanorods were synthesized by a simple precipitation method.Briefly, Co(CH 3 COO) 2 •4H 2 O (5.9 mmol) and NaOH (10 mmol) were dissolved into 50 mL and 10 mL deionized water, respectively.Then, 0.6 g of H 2 C 2 O 4 was added to the Co(CH 3 COO) 2 aqueous solution.The precipitation process was started by the dropwise addition of a NaOH aqueous solution into the Co(CH 3 COO) 2 aqueous solution under continuous stirring.The obtained pink suspension was stirred at 85 • C for 12 min.The product was cooled to room temperature, centrifuged, washed several times with deionized water and dried in the oven (60 • C) for 8 h.

Preparation of CoC 2 O 4 Nanocubes
The CoC 2 O 4 nanocubes were synthesized by a facile hydrothermal method.In a typical synthesis, 4 mmol Co(CH 3 COO) 2 •4H 2 O and 5 mmol NaOH were dissolved in 20 mL deionized water, respectively.The precipitation was initiated at room temperature by the dropwise addition of a NaOH aqueous solution to the Co(CH 3 COO) 2 aqueous solution with continuous stirring.The obtained suspension was stirred for another 0.5 h at room temperature.Then, 1.2 g SDBS was added to the suspension described above.The suspension was transferred to a 100 mL Teflon-lined stainless-steel autoclave, and then the autoclave was maintained at 160 • C for 36 h.The temperature was programmed to increase from room temperature to 160 • C at a heating rate of 2.0 • C/min.After cooling to room temperature, the black precipitate was centrifuged, washed several times with deionized water and dried in the oven (60 • C) for 8 h.

Preparation of Reduced Co 3 O 4
The as-prepared Co(OH) 2 nanoplates, Co(OH) 2 microflowers, CoC 2 O 4 nanorods and CoC 2 O 4 nanocubes were calcined at 300 • C for 2 h in air to obtain the corresponding Co 3 O 4 nanostructures.The temperature was programmed to increase from room temperature to 300 • C at a ramp of 2 • C/min.Then, these Co 3 O 4 nanostructures were separately soaked in a NaBH 4 aqueous solution (0.05 mol/L) at 30 • C for 40 min.Afterwards, the powder was collected via centrifugation, washed several times with deionized water and used directly for the catalytic reduction of p-NP to p-AP.

Catalyst Characterization
The powder X-ray diffraction (XRD) analyses of the as-prepared samples at room temperature were recorded by a PANalytical X'pert-Pro powder X-ray diffractometer equipped with Cu Ka monochromatic radiation (λ = 0.1541 nm) with 2θ ranging from 10 • to 90 • at a scanning rate of 5 • /min.The X-ray tube was operated under 40 kV and 40 mA.Nitrogen sorption measurements of the samples were performed on a Quadrasorb SI adsorption analyzer.The specific surface areas and pore properties were calculated from the adsorption-desorption isotherms collected at 77 K. Transmission electron microscopy (TEM) were taken on a JEOL JEM-2100 at an accelerating voltage of 120 kV, and scanning electron microscopy (SEM) were taken on a JEOL JSM-7800F at an accelerating voltage of 3 kV to examine the morphologies of the as-prepared samples.Temperature-programmed reduction (H 2 -TPR) was carried out on a Micromeritics AutoChem 2920 apparatus.A 10% mixture of H 2 /Ar passed through the catalyst bed at a flow rate of 40 mL min −1 , and at the same time, the temperature was programmed to increase from 50 • C to 700 • C at a heating rate of 10 • C/min.

Catalytic Reduction of p-NP to p-AP by Excess NaBH 4
The catalytic reduction of p-NP to p-AP was conducted under controllable conditions to evaluate the catalytic performance of the obtained nanostructures.Typically, the reduction reaction was conducted in a 3.5 mL standard quartz cell at room temperature (ca.23 • C).Excess NaBH 4 aqueous solution was chosen as the reducing agent.Approximately 2.0 mL p-NP aqueous solution (0.175 mmol/L) was first mixed with 0.7 mL freshly prepared NaBH 4 solution (0.05 mol/L) in the quartz cell.Then, 0.1 mL suspension containing one of the obtained Co 3 O 4 nanostructures (2 g/L) was quickly injected into the mixture under temperature control and rapid stirring.The reduction process was monitored by detecting the extinction of p-NP via an in situ UV-vis spectrophotometer (METASH UV8000) over a scanning range of 200 nm-600 nm.The kinetic evaluation was judged by the disappearance of the p-NP adsorption peak at 400 nm as a function of the reaction time.

Conclusions
Co 3 O 4 nanoplates, microflowers, nanorods and nanocubes were successfully synthesized by facile precipitation or hydrothermal methods, followed by a thermal decomposition process.The catalytic performance of these Co 3 O 4 nanostructures for the reduction of p-NP to p-AP in the presence of excess NaBH 4 was studied.The order of the catalytic activity of these Co 3 O 4 nanostructures is nanoplates > microflowers > nanorods > nanocubes.Co 3 O 4 nanoplates exhibit the highest catalytic activity due to their largest specific surface area, relatively high total pore volume, abundant oxygen vacancies and strong reducibility.The results indicate that Co 3 O 4 catalysts with different morphologies exhibit different properties, such as different specific surface area, porosity, surface defects and reducibility, which, in turn, significantly affect their catalytic performance.Therefore, morphology control in the preparation of Co 3 O 4 nanocrystals is crucial for the preparation of highly efficient heterogeneous catalysts.The controllable synthesis of Co 3 O 4 nanostructures with different morphologies is expected to bring new opportunities for regulating the catalytic performance of metal oxide catalysts.

2. 1 . 21 Figure 1 . 4 Figure 1 .
Figure 1 shows the wide-angle XRD patterns of the as-prepared Co 3 O 4 with different morphologies.The data used in the XRD pattern are all original data without background subtraction.These Co 3 O 4 nanostructures were obtained from different precursors (the detailed materials and methods are described in Section 3).The morphology of these nanostructures is shown in Figure 2. The XRD peaks of the four samples match the JCPDS file (01-076-1802) of the cubic phase of Co 3 O 4 , indicating the complete transformation of the four precursors to Co 3 O 4 via high-temperature calcination.Furthermore, the characteristic diffraction patterns of these Co 3 O 4 nanostructures are identical except for their different intensities, implying that the same product was obtained but the crystallinity is different.For all the samples, the diffraction peaks at 19.03 • , 31.32 • , 36.90 • , 38.61 • , 55.75 • , 59.45 • and 65.34 • in the XRD patterns of the Co 3 O 4 nanostructures are well-matched with the (1 1 1), (2 2 0), (3 1 1), (2 2 2), (4 2 2), (5 1 1) and (4 4 0) facets of the cubic Co 3 O 4 , respectively.No peaks from the other phases or impurities were observed in the patterns of the Co 3 O 4 nanostructures.In addition, there was an apparent increase of counts as the 2θ angle increased in the XRD patterns, especially for the Co 3 O 4 nanocubes, nanorods and microflowers.This can be mainly caused by imperfect sample preparation or the sample not being aligned well during the scanning process.Molecules 2023, 26, x FOR PEER REVIEW 19 of 21
44 for the Co 3 O 4 nanoplates, 1:3.40 for the Co 3 O 4 microflowers, 1:3.37 for the Co 3 O 4 nanorods and 1:3.37 for the Co 3 O 4 nanocubes, which are all close to the stoichiometric ratio of Co 3+ to Co 2+ in Co 3 O 4 .Furthermore, all the Co 3 O 4 samples were completely reduced to metallic cobalt before 500 • C.However, the peak position, peak intensity and half-peak width of these samples are different, suggesting that their redox properties are, to some extent, dependent on their morphologies.The first reduction peaks of the Co 3 O 4 nanostructures are 287 • C for the Co 3 O 4 nanoplates, 306 • C for the Co 3 O 4 microflowers, 318 • C for the Co 3 O 4 nanorods and 333 • C for the Co 3 O 4 nanocubes.The results indicate that the reducibility of the Co 3 O 4 nanoplates is the strongest among the four Co 3 O 4 nanostructures, i.e., lattice oxygen species coordinated with Co 3+ cations are more easily extracted from the surface of the Co 3 O 4 nanoplates, thereby facilitating the electron transfer and the formation of more oxygen vacancies.This is followed by the Co 3 O 4 microflowers.The decrease in reducibility of the Co 3 O 4 nanorods and nanocubes can be attributed to the decrease in the specific surface areas, as observed from the BET analyses above [35].Molecules 2023, 26, x FOR PEER REVIEW 19 of 21

Figure 4 .Figure 4 .
Figure 4. H2-TPR profiles and their deconvolution for Co3O4 nanostructures: (1) nanoplates, (2) microflowers, (3) nanorods and (4) nanocubes (measurement conditions: temperature from 100 °C to 700 °C at a rate of 10 °C/min using a stream of 10% H2/Ar balance with a flow rate of 40.18 cm 3 STP/min).2.1.5.XPS X-ray photoelectron spectroscopy was carried out to qualitatively characterize the chemical composition and surface chemical states of these Co3O4 nanostructures[40].Figure5ashows the XPS full spectra of these nanostructures.It is evident that each peak in 3 O 4 nanostructures are merely constituted of Co and O elements.The high-resolution XPS spectra of the Co element of the four Co 3 O 4 samples are shown in Figure 5b to provide detailed information on the chemical oxidation states.The sharp peak at the binding energy of about 779.0 eV is characteristic of Co 2p
3 O 4 nanostructures are determined by the slopes of the respective linear curves of t versus −ln(A t /A 0 ).As shown in Figure 8, the apparent rate constant values for the p-NP reduction reaction in the presence of 0.2 mg of different shapes of Co 3 O 4 are 1.49min −1 (nanoplates), 1.40 min −1 (microflowers), 0.78 min −1 (nanorods) and 0.23 min −1 (nanocubes).The activity sequence for the reduction of p-NP catalyzed by Co 3 O 4 with different morphologies is as follows: Co 3 O 4 nanoplates > Co 3 O 4 microflowers > Co 3 O 4 nanorods > Co 3 O 4 nanocubes.
3 O 4 nanostructures is ordered as follows: Co 3 O 4 nanoplates > Co 3 O 4 nanorods > Co 3 O 4 microflowers > Co 3 O 4 nanocubes.Therefore, the catalytic activity of the Co 3 O 4 nanoplates is the highest, and that of the Co 3 O 4 nanocubes is the lowest.For the reduction of p-NP, Co 3 O 4 served as a catalyst to transfer active hydrogen species and electrons from the donor BH 4
3 O 4 morphology on the reduction of p-NP.For example, conical Co 3 O 4 , stacked Co 3 O 4 , needled Co 3 O 4 and floral Co 3 O 4 have been synthesized, which show remarkably enhanced activity in the reduction of p-NP than commercial Co 3 O 4 catalysts.Conical Co 3 O 4 exhibits the highest catalytic activity because of the more reactive surface and more superior redox properties than other Co 3 O 4 catalysts

Table 1 .
Surface properties of Co 3 O 4 nanostructures.
2.1.4.H 2 -TPR 3/2 with two curve-fitted peaks corresponding to Co 3+ (778.5~778.55eV) and Co 2+ (780.4~780.55eV).The shoulder peak with a binding energy of about 794.0 eV can be deconvoluted into two typical peaks of Co 3+ (793.6~793.7 eV) and Co 2+ (795.5~795.7 eV), which can be attributed to the corresponding Co 2p 1/2 .The shake-up satellite peaks with low intensity at ca. 8.8 eV from the main spin-orbit components of Co 2p 3/2 are also characteristic of Co 3 O 4 .Apparently, the element Co is present in the chemical state of Co 2+ and Co 3+ , and there is no metallic cobalt.The molar percentages of Co 2+ and Co 3+ in these Co 3 O 4 nanostructures are obtained according to the integrated area ratios obtained by XPS splitting, as summarized in Table 2. Our previous work has demonstrated that when Co 3 O 4 is treated with a NaBH 4 aqueous solution, some of the Co 3+ can be reduced to Co 2+ and at the same time, oxygen vacancies are formed on the Co 3 O 4 surface.More Co 2+ on the surface of Co 3 O 4 indicates that more Co 3+ was reduced to Co

Table 2 .
The molar percentages of Co 2+ and Co 3+ in Co 3 O 4 nanostructures.

Table 3 .
The induction times of various Co 3 O 4 nanostructures for the reduction of p-NP a .
a Based on the reaction results in Figures

Table 4 .
Comparison of the rate constants of different catalysts for the catalytic reduction of p-NP to p-AP by NaBH 4 in the recent literature and this work.

Table 4 .
Cont.The mass concentration of the catalyst in the solution.b The mass-normalized rate constant calculated based on the catalyst amount, including the support based on the equation: k nor = k app /m cat . a