Co3O4@CdS Hollow Spheres Derived from ZIF-67 with a High Phenol and Dye Photodegradation Activity.

The Co3O4@CdS double-layered hollow spheres were first prepared by the template-removal method with the assistance of the ZIF-67 material; the structure has been proved by transmission electron microscopy (TEM). The Co3O4@CdS hollow spheres calcinated at 400 °C exhibited the highest photodegradation activity. Nearly 90% phenol was degraded after 2 h of visible-light irradiation. More than 80% rhodamine-B (RhB) was degraded within the first 30 min and nearly eliminated after 1 h of irradiation. The mechanism of the photodegradation reaction was investigated. Based on the analysis of electron spin resonance (ESR) spectra and radical trapping test, it was found that superoxide radicals are the major oxidative species for dye degradation and holes and hydroxyl radicals are the major oxidative species for phenol degradation. These results may be used in industrial wastewater treatment. The reaction obeys first-order reaction kinetics, and the rate constant of the Co3O4@CdS hollow sphere in dye degradation is 0.05 min-1 and that in phenol degradation is 0.02 min-1, which is three times higher than that of CdS nanoparticles. These results indicated the high oxidizing ability of the samples.

Introduction

With the rapid growth of the economy and increasing demand for environment-friendly technologies, photocatalysis has attracted a lot of attention since their discovery by Honda and Fujishima.[1−3] After discovering that titanium dioxide can generate charge carriers under ultraviolet light irradiation, which are able to split water into hydrogen and oxygen, various efforts[4−9] have been taken to achieve a more efficient photocatalyst manufacturing. The charge carriers are able to not only produce hydrogen as fuel but also provide holes to oxidize organic compounds, which has promising applications in environmental pollution treatment.[2,10] The reactive intermediates that the carriers excited play an important role. Hydroxyl radical[11,12] has the advantages of strong oxidizing ability and rapid reaction. Superoxide radicals[13,14] can oxidize arsenic(III) into arsenic(V), which will significantly reduce the toxicity of industrial sewage. A lot of progress[15−18] has been made after decades of research, but there are still some holdbacks including low surface area[19,20] that restrict its charge carrier transportation; usage of rare metals[21,22] as a cocatalyst makes it economically inefficient and photocorrosion[23] and low-visible-light photocatalytic activity severely restrict its large-scale applications.

To overcome these shortcomings, numerous efforts have been made. Zhang et al.[24] successfully fabricated CdS hollow spheres under gas bubbling, which increased the surface area of the nanoparticles and resulted in a higher photodegradation activity. Engineered heterojunction[25] has been prepared not only to separate photogenerated electron–hole pairs spatially but also to reduce the usage of rare metals. Appropriate band structure[26] prepared using ions doping or other methods can significantly change the light absorption ability and suppress the occurrence of photocorrosion. Therefore, a properly designed heterojunction photocatalyst[27−30] with a unique structure[26,31−34] could benefit not only high surface area but also spatially separated photogenerated electron–hole pairs, contributing to the suppression of photocorrosion.

Recent years, metal–organic framework[35−38] (MOF) materials have become a researching hot spot due to their excellent size selectivity and manually engineered properties. Among them, zeolitic imidazolate frameworks (ZIFs),[39,40] especially ZIF-67,[4,29,41] have attracted significant attention due to their high concentration of active cobalt sites; a convenient synthetic process also makes it one of the most widely used MOF materials. Taking advantage of this property of MOF materials, the CdS/Co3O4 heterojunction can be obtained by simply applying ZIF-67 on the surface of a SiO2 sphere as a template to form a coating layer of Co3O4 after annealing. This preparation method allows the composite to become thinner and thus be used in the fabrication of double-layered hollow spheres.

Recently, owing to the excellent light absorption ability and rational band structure, the CdS/Co3O4 heterojunction structures[42−44] have been studied in the literature. To pursue higher loadings of Co3O4, CdS nanorods were chosen due to the higher specific surface area. Cha et al.[44] electrostatically assembled Co3O4 on the CdS nanorods to achieve higher efficiency of water oxidation. However, solely applying Co3O4 nanoparticles onto the surface of the CdS nanorods is insufficient for enhancing the photocatalytic activity of CdS, thus a method to prepare CdS/Co3O4 photocatalysts with a core–shell structures is badly needed. Hu et al.[45] reported a CoO@CdS nanorod core–shell structure using an impregnation–calcination method. The hydrogen evolution rate is 43-fold higher than that of CdS nanorods, which indicated that the core–shell structure might contribute to the charge carrier transportation, but the surface area is low since it is largely determined by the size of the CdS nanorods. Therefore, to achieve a high photocatalytic activity and a high surface area at the same time, the Co3O4@CdS double-layer hollow spheres are worth exploring. Nevertheless, the mechanism and reactive intermediate radicals of CdS/Co3O4 composites have never been investigated. Herein, a novel template-removal method was proposed to fabricate Co3O4@CdS composite hollow spheres that are derived from ZIF-67. The morphology of the Co3O4@CdS composite hollow sphere was proved by transmission electron microscopy (TEM). The photodegradation activity was tested using organic dyes and phenol, the calculated reaction rate constant of the Co3O4@CdS composite hollow spheres in dye degradation is 7 times higher than that of CdS nanoparticles and the calculated reaction rate constant of the Co3O4@CdS composite hollow spheres in phenol degradation is 3 times higher than that of the CdS nanoparticles. Electron spin resonance (ESR) was brought to investigate the reaction intermediate, the results indicate that the high degradation efficiency of the Co3O4@CdS composite hollow sphere is derived from the higher production of the photogenerated holes and hydroxyl radical as well as the higher production of superoxide radicals. These results revealed that the Co3O4@CdS composite hollow spheres may have an excellent application prospect in environmental wastewater treatment and ESR measurement would help to better understand the relationship between intermediate and reactions.

Results and Discussion

The morphology of the sample was characterized by transmission electron microscopy. Figure a is a typical TEM image of a Co3O4@CdS hollow sphere. It can be seen that the thickness of both CdS and Co3O4 layers is around 50 nm, while the total diameter of 250 nm matches perfectly with the diameter of the SiO2 template (shown in Figure S1). Figure b–d is the element mapping of the hollow sphere samples. It illustrates that cadmium sulfide was uniformly covered on the surface of Co3O4. Co3O4 was also completely covered on the surface of the SiO2 template, but the concentrations of Co in specific parts are significantly higher than those in other parts due to the MOF structure. The morphology of the ZIF-67@SiO2 spheres are not perfectly spherical (shown in Figure S2), and the annealing process of MOF materials tends to result in the formation of tiny particles, which stick to the surface of the SiO2 template. According to the high-resolution TEM (HRTEM) images, the characteristic lattice fringes of 0.45 and 0.24 nm correspond to the (111) plane of Co3O4 and the (311) plane of CdS, respectively, verifying that the hollow sphere samples contain cobaltous oxide and cadmium sulfide.

Figure 1

HRTEM images of the CdS@Co3O4 hollow spheres (a, e, f) and element mapping of Co (b), S (c), and C (d).

Figure displays the X-ray diffraction (XRD) patterns of the CdS@Co3O4 hollow sphere samples derived at different calcination temperatures. The diffraction peaks at 36.85 and 31.27° could be attributed to the (311) and (220) crystal planes of Co3O4 (JPCDS No. 42-1467) with a space group of Fd3̅m (227). The XRD results show that with the rising temperature during the calcination of ZIF-67, the degree of crystallinity of Co3O4 increases accordingly. When the calcination temperature reaches 500 °C, a new phase with a different chemical composition is generated. According to the XRD peaks at 34.15 and 39.64°, the component is identified as CoO (JPCDS No. 42-1300).

Figure 2

XRD pattern of the CdS@Co3O4 hollow spheres at different calcination temperatures.

Figure shows the X-ray photoelectron spectra (XPS), which demonstrate the chemical state and chemical composition of the sample. The X-ray photoelectron spectra confirmed the presence of Co, O, Cd, and S elements in the Co3O4@CdS hollow sphere samples. The binding energies of 797.6 and 780 eV are assigned to Co 2p1/2 and Co 2p3/2, respectively, corresponding to the Co element in Co3O4. The results analyzed from XPS and XRD spectra further confirmed that the hollow sphere samples contained CdS and Co3O4 based on the inference of the preparation process and XRD results.

Figure 3

XPS spectra of the CdS@Co3O4 hollow sphere samples: (a) Co 2p, (b) O 1s, (c) Cd 3d, and (d) S 2p.

The ultraviolet–infrared light absorption spectra were used to investigate the light absorption abilities of the samples. To eliminate the influence of the nanostructure on light absorption, CdS hollow spheres were used as the reference object. The difference in the light absorption ability between the CdS nanoparticles and CdS hollow spheres are shown in Figure S3. Figure displays the UV–vis absorption spectra of the samples. From Figure , it can be easily seen that the light absorption ability was significantly enhanced with the attachment of the Co3O4 layer. It can also be inferred that the CdS hollow spheres are a direct band gap semiconductor whose band gap is around 2.3 eV, according to the UV–vis absorption spectra (shown in Figure S3). With the increasing calcination temperature, the light absorption rose in the first stage and then decreased. The spectra indicated that 400 °C is a more favorable calcination temperature that would endow the sample with a stronger light-absorbing ability.

Figure 4

Ultraviolet–infrared light absorption spectra of the CdS@Co3O4 hollow spheres at different calcination temperatures and the samples of HS-300, HS-400, and HS-500, respectively.

The efficiency of the photodegradation of organic dyes (rhodamine-B, RhB) under visible-light irradiation is shown in Figure . According to the absorption spectra of RhB, the maximum absorption peak located at 554 nm (shown in Figure S4). The blank test was brought to indicate that photolysis can be ignored since the concentration of RhB negligibly decreased under visible-light irradiation (shown in Figure S4). RhB degraded slowly solely due to the presence of pure CdS nanoparticles. Less than 60% RhB was degraded for 3 h, and the efficiency of the catalyst was around 20% in the first half-hour under visible-light irradiation. With the presence of the Co3O4 cocatalyst, the efficiency rose remarkably, which indicates that cobalt oxide would be beneficial for the degradation of RhB under visible-light irradiation. Among them, the Co3O4@CdS hollow spheres calcinated at 400 °C exhibited the highest photodegradation activity. More than 80% RhB was degraded within the first 30 min, and then nearly eliminated after 1 h of visible-light irradiation. The Co3O4@CdS hollow sphere calcinated at 500 °C showed a similar photodegradation activity; the efficiencies for the first half hour and for the first hour are approximately 60 and 80%, respectively. To obtain the photodegradation ability of these photocatalysts, the mechanism of the first-order reaction is applied to evaluate the photocatalytic performance.[11] The calculated rate constants are listed in Table . The rate constants of the Co3O4@CdS hollow sphere samples calcined at temperatures of 400, 300, and 500 °C were 0.05, 0.009, and 0.03 min–1, respectively, which are 7, 1.3, and 4.2 times higher than those of the CdS nanoparticles. The results indicated that the Co3O4@CdS hollow sphere calcined at 400 °C possessed a superior photocatalytic performance among all of the samples, which may be attributed to the morphology and chemical composition of the heterostructure.

Figure 5

Photodegradation efficiency of RhB over CdS and CdS@Co3O4 hollow spheres (left) and the corresponding calculated rate constants (right).

Table 1

Calculated Rate Constants of CdS and Co3O4@CdS Hollow Spheres

sample name	CdS	HS-300	HS-400	HS-500
k/min–1	0.007	0.009	0.05	0.03

The efficiency of the photodegradation of phenol under visible-light irradiation is shown in Figure . Less than 30% phenol was degraded due to the presence of CdS nanoparticles after 1 h of visible-light irradiation. With the addition of the Co3O4 cocatalyst, the photodegradation activity enhanced significantly. Similar to the previous conclusion, the Co3O4@CdS hollow spheres calcined at 400 °C exhibited the highest photodegradation activity. More than 70% phenol was degraded within the first hour, and nearly 90% phenol was eliminated after 2 h of visible-light irradiation. The calculated rate constants of HS-300, HS-400, and HS-500 are 0.0079, 0.0198, and 0.0127, respectively (Table ). The calculated first-order reaction rate constants of different Co3O4@CdS hollow spheres are 0.3, 2.1, and 1 times higher than those of the pure CdS nanoparticles. The photodegradation reactions indicated that the as-prepared hollow spheres have excellent organic degradation activities.

Figure 6

Photodegradation efficiency of phenol over CdS, CdS@Co3O4 hollow spheres (left), and the corresponding calculated rate constants (right).

Table 2

Calculated Rate Constants of CdS and Different Co3O4@CdS Hollow Spheres

sample name	CdS	HS-300	HS-400	HS-500
k/min–1	0.0062	0.0079	0.0198	0.0127

To evaluate the long-term stability of the Co3O4@CdS hollow spheres, the curves of the normalized concentration of RhB with time are shown in Figure . It can be observed from Figure that the high efficiency of the photodegradation activity remained after three continuous cycling tests. The results indicated that the samples are self-robust and have the potential for practical applications in wastewater treatment.

Figure 7

Long-term stability of the CdS@Co3O4 hollow spheres.

The nitrogen adsorption–desorption isotherms were introduced to investigate the specific surface area of the samples (Figure ). The Brunauer–Emmett–Teller (BET) surfaces of 300, 400, 500, and CdS nanoparticles were determined to be 62.37, 75.63, 48.14, and 2.43 m2·g–1, respectively. The absorption type of 300 and 400 agreed with Langmuir V; meanwhile, at the tail of the isotherms (high relative pressure), the absorbance increased quickly, suggesting the presence of mesopores. The nitrogen adsorption–desorption isotherm of 500 rose did not show an obvious rise, which indicated that the calcination may destroy the structure of the Co3O4@CdS hollow sphere as well as the pores. Compared with the BET surface of the CdS nanoparticles, the Co3O4@CdS hollow spheres exhibited a 24.6-, 30.1-, and 18.8-fold increase in specific surface area, respectively, which proved that a hollow nanostructure would be beneficial to the charge carrier transportation.

Figure 8

Nitrogen adsorption–desorption isotherms of CdS and the samples of HS-300, HS-400, and HS-500.

To reveal the mechanism of photodegradation, electron spin resonance[46,47] (ESR) spectra was introduced to measure the reactive intermediate. It can be easily observed that significant evolution of ESR signals at room temperature under visible-light irradiation. According to Figure , the peak intensity of the Co3O4@CdS hollow spheres is much stronger than that of Co3O4 and the CdS nanoparticles, which indicated that the concentrations of the intermediate radicals generated by the Co3O4@CdS hollow spheres were higher than normal Co3O4 or CdS. The result indicates the higher photodegradation activity of the Co3O4@CdS hollow spheres than those of Co3O4 and CdS. Figure a displays the ESR spectra of the samples tested in air under visible-light irradiation and dark condition. The peak intensity of the Co3O4@CdS hollow spheres is much stronger, which reveals that the photogenerated holes and intrinsic defects in the sample are more than those of CdS and Co3O4. More photogenerated holes indicate higher photodegradation activity, while more intrinsic defects may contribute to more rapid carrier transportation. Figure b,c exhibits the spectra of the samples tested in 5,5-dimethyl-1-pyrroline N-oxide (DMPO) under visible-light irradiation and dark condition. The pattern of the signals indicated the presence of superoxide radicals.[48,49] From Figure b, the increased intensity of the quadruple characteristic peaks of DMPO indicates that the concentration of superoxide radicals is higher than that of pure CdS or Co3O4. The concentration of hydroxyl radicals is also higher than that of pure CdS or Co3O4 according to Figure c.

Figure 9

ESR spectra of CdS, Co3O4, and CdS@Co3O4 hollow spheres under the dark condition and light illumination: (a) holes, (b) superoxide radicals, and (c) hydroxyl radicals.

To evaluate the effect of the intermediate radicals in the photodegradation reaction of RhB, the scavenging agents were introduced. Isopropanol (IPA) was used for the elimination of hydroxyl radicals; AgNO3 was used for the elimination of electrons; and ammonium oxalate (AO) was used to eliminate photogenerated holes. According to Figure , less than 60% RhB was degraded in the first half-hour while approximately 80% RhB was degraded for the first hour with the presence of IPA. The results indicated that hydroxyl radicals are effective for the photodegradation of RhB. Since the existing form of electrons is superoxide radicals and they cannot be directly tested by UV–vis spectra (shown in Figure S5), AgNO3 was introduced to eliminate photogenerated electrons. With the presence of AgNO3, the efficiency of the photodegradation reaction decreased severely, with only around 20% RhB degraded during the first half-hour, while more than 70% RhB still existing in the solution during 1 h of reaction. The low photocatalytic efficiency indicated that superoxide radicals may be the main oxidative species for the Co3O4@CdS hollow sphere samples. The results indicated that the enhanced photocatalytic activity of the Co3O4@CdS hollow spheres is mainly due to the larger amount of superoxide radicals (•O2–) (Table ).

Figure 10

Photodegradation of RhB over Co3O4@CdS hollow spheres with scavengers (left) and the corresponding calculated rate constants (right).

Table 3

Calculated Rate Constants of the Co3O4@CdS Hollow Spheres with/without Scavengers

types of scavengers	original HS (without scavenger)	IPA	AO	AgNO3
k/min–1	0.054	0.029	0.013	0.006

The mechanism of phenol degradation is studied in Figure . According to the radical trapping test, around 20% phenol was degraded in the first half-hour, while approximately 40% phenol was degraded for the first 2 h in the presence of IPA and AO. The results indicated that the hydroxyl radicals and holes may be the major reactive intermediates for the photodegradation of phenol. With the presence of AgNO3, the change in the efficiency of the photodegradation reaction decreased negligibly compared to the test without AgNO3, which indicated that electrons played a very slight role in the photodegradation reaction of phenol (Table ).

Figure 11

Photodegradation of phenol over the Co3O4@CdS hollow spheres with scavengers (left) and the corresponding calculated rate constants (right).

Table 4

Calculated Rate Constants of the Co3O4@CdS Hollow Spheres with/without Scavengers

types of scavengers	original HS (without scavenger)	IPA	AO	AgNO3
k/ min–1	0.0198	0.005	0.005	0.017

The superoxide radicals (•O2–) were derived from photogenerated electrons; from the equations, the photogenerated electrons and holes are both beneficial for the degradation of organic dyes, thus increasing the concentration of hydroxyl radicals or superoxide radicals would be favorable. According to the ESR spectra, the enhancement in the photodegradation activity was due to the increasing concentrations of hydroxyl radicals and superoxide radicals.

Co3O4 is a p-type semiconductor,[43] while CdS is an n-type material. A p–n heterojunction would be formed at the interfaces between Co3O4 and CdS as soon as two semiconductors are in contact. Scheme illustrates the band structure of the Co3O4@CdS hollow sphere heterostructures. The migration of the photogenerated electrons to the conduction band of CdS can be promoted by the built-in internal electric field, leaving sufficient photogenerated holes in the valence band of Co3O4. More rapid transportation of carriers apparently is able to reduce the recombination of carriers significantly, which will result in a big enhancement in photocatalytic activity. The p–n heterostructure have promising applications in hydrogen evolution.

Scheme 1

Band Gap Schematic Illustration of the CdS@Co3O4 Hollow Spheres

Conclusions

In summary, the Co3O4@CdS hollow spheres possessed better light absorption and larger surface area, thus an enhanced photocatalytic activity can be obtained. The Co3O4@CdS hollow spheres calcinated at 400 °C exhibited the highest photodegradation activity. Nearly 90% phenol was degraded after 2 h of visible-light irradiation. More than 80% RhB was degraded within the first 30 min and then nearly eliminated after 1 h of visible-light irradiation. The mechanism of photodegradation was investigated through the ESR spectra and radical trapping test; it can be concluded that superoxide radicals are the major oxidative species for the Co3O4@CdS hollow sphere samples. Nevertheless, the Co3O4@CdS hollow spheres formed a p–n heterojunction, which may be beneficial for applications in hydrogen evolution.

Experimental Section

Synthesis of SiO2 Nanospheres

The SiO2 nanospheres were prepared using a conventional Stöber method.[50,51] The prepared SiO2 nanoparticles were purified by repeated centrifugation and dispersion in alcohol. Finally, the silica nanoparticles were dried overnight at 80 °C in an oven.

Preparation of SiO2@Co3O4 Nanospheres

The SiO2@Co3O4 nanospheres were prepared by coating ZIF-67 onto the surface of the SiO2 nanospheres. The typical preparation process is as follows: 0.3 g of SiO2 nanospheres was added into 100 mL of methanol under stirring to form a homogeneous solution; then, 2.91 g of cobaltous acetate was added into the solution and stirred for 10 min; after that, 0.82 g of methylimidazole was slowly added into the above solution under stirring. The color of the solution changed from light purple to dark purple. After stirring for 2 h, the SiO2@ZIF-67 nanospheres were centrifuged and then calcinated at different temperatures for 8 h after drying. For the sake of later convenience, the samples are marked as HS-300, HS-400, and HS-500, according to the calcination temperatures of 300, 400, and 500 °C, respectively.

Preparation of Co3O4@CdS Hollow Spheres

The preparation process for the Co3O4@CdS spheres is schematically illustrated in Scheme . First, 0.3 g of SiO2@Co3O4 nanospheres, 0.66 g of CdCl2, 0.7 g of citric acid, and 0.4 g of thiourea were added into 300 mL of ultrapure water to form a homogeneous solution; 10 mL of ammonia was then added into the solution to adjust the alkalinity. The solution was heated up to 80 °C for 3 h under stirring. Then, the SiO2@Co3O4@CdS nanoparticles were collected through centrifugation. Then, the particles were dipped into ammonium bifluoride solution (4 M) for 12 h. Finally, after distilling in ultrapure water for at least three times, the Co3O4@CdS hollow spheres were dried overnight at 80 °C in an oven.

Scheme 2

Schematic Illustration of Co3O4@CdS Hollow Sphere Fabrication Process

Photocatalytic Activities Test

The photocatalytic degradation experiments were carried out in a 250 mL of photoreduction cell. The cell was connected to a cold-water circulation system. The cell was photoilluminated using a solar simulator with a 420 nm cutoff filter. The concentrations of RhB[52] were determined at 554 nm using a UV–vis spectrometer.

In a typical photocatalytic RhB photodegradation experiment, 15 mg of photocatalyst powders was suspended in 100 mL of 6.0 M RhB aqueous solution with/without scavenger. Prior to illumination, the cell was covered with an aluminum foil and the suspension was stirred in the dark to reach an adsorption–desorption equilibrium. Then, the suspension was irradiated under 300 W Xe lamp for 180 min. The samples were taken out at regular time intervals and centrifuged before a UV–vis spectrometer analysis.

In a typical photocatalytic phenol photodegradation experiment, 15 mg of photocatalyst powders was suspended in 100 mL of 20 mg·L–1 phenol aqueous solution. Prior to illumination, the cell was covered with an aluminum foil, and the suspension was stirred in the dark to reach an adsorption–desorption equilibrium. Then, the suspension was irradiated under the 300 W Xe lamp for 180 min. The samples were taken out at regular time intervals and centrifuged before high-performance liquid chromatography (HPLC) analysis.