Sonocatalytic degradation of tetracycline hydrochloride with CoFe₂O₄/g-C₃N₄ composite

Abstract

In this work, different mass percent ratios of CoFe₂O₄ coupled g-C₃N₄ (w%-CoFe₂O₄/g-C₃N₄, CFO/CN) nanocomposites were integrated through a hydrothermal process for the sonocatalytic eradication of tetracycline hydrochloride (TCH) from aqueous media. The prepared sonocatalysts were subjected to various techniques to investigate their morphology, crystallinity, ultrasound wave capturing activity and charge conductivity. From the investigated activity of the composite materials, it has been registered that the best sonocatalytic degradation efficiency of 26.71 % in 10 min was delivered when the amount of CoFe₂O₄ was 25% in the nanocomposite. The delivered efficiency was higher than that of bare CoFe₂O₄ and g-C₃N₄. This enriched sonocatalytic efficiency was credited to the accelerated charge transfer and separation of e⁻-h⁺ pair through the S-scheme heterojunctional interface. The trapping experiments confirmed that all the three species i.e. ^•OH, h⁺ and ^•O₂⁻ were involved in the eradication of antibiotics. A strong interaction was shown up between CoFe₂O₄ and g-C₃N₄ in the FTIR study to support charge transfer as confirmed from the photoluminescence and photocurrent analysis of the samples. This work will provide an easy approach for fabricating highly efficient low-cost magnetic sonocatalysts for the eradication of hazardous materials present in our environment.

1. Introduction

During the last few decades, the economy of the world and the living standard of the people have been advanced significantly. However, the establishment of new industries to improve life standard has also introduced to human beings huge number of different toxic and hazardous materials including dyes, antibiotics, insecticides, heavy metals, etc [1]. The presence of these substances in the environment has threatened human health and life badly. Particularly, many antibiotics have been widely used in the treatment of fatal diseases, development of agricultural products and animal husbandry. Since these antibiotics cannot be completely digested in human or animal bodies, many antibiotics are introduced to the aqueous environment through the release of excretory products of animals which not only damage the ecosystem but also produce compound toxicity to aquatic organisms and human beings in combination with secondary pollutants in water. Tetracycline hydrochloride (TCH) antibiotic is one of the most widely-used antibiotics in livestock production and clinics because of its low price [2]. TCH is widely used as a growth promoter in livestock and poultry industries in low doses while its high doses are recommended only for curing different diseases. However, due to the low assimilation in animals, a large amount of TCH is discharged to the external environment and therefore found in ground water, river water to generate drug resistance strain [3]. To avoid its rapid accumulation in water bodies and prevent its hazardous effect, it is urgently needed to eradicate TCH for the protection of aqueous lives.

Different technologies used for the eradication of TCH mainly utilize physical [4], biological [5] and chemical treatment methods [6]. In recent years, a low-cost ultrasound technology is developed significantly to decompose a large number of different pollutants present in various water bodies [7]. This process of waste treatment has many advantages including non-selectivity, rapidity, mild degradation conditions and easy operating instrumentation to decompose structurally stable and inhibitory microbial pollutants into simple non-hazardous compounds using sound waves of frequency higher than 20 kHz [8], [9]. In 1920s, the cavitation effect of ultrasound waves were found to destroy microorganisms [10]. Crum and Fowlkes verified that ultrasound waves could split water into hydroxyl free radica (^•OH), singlet oxygen (¹O₂) and superoxide anion free radical (^•O²⁻) [11]. However, suitable catalysts that can effectively convey these oxidizing species produced during ultrasonic excitation are primarily needed.

CoFe₂O₄ (CFO) is an iron-based semiconductor oxide with a high chemical stability, low toxicity, and medium magnetization [12]. Compared with other semiconductor catalysts, CoFe₂O₄ has advantages of magnetism, i.e. to be separated easily after completion of the process [13]. However, its catalytic efficiency is badly reduced through the fast recombination of sonogenerated charge-pairs and therefore, more advanced catalysts must be investigated to overcome the wastage of excited charge-pairs during sonocatalysis.

Graphite-phase carbon nitride (CN) is a potential metal-free green catalyst with many advantages such as narrow band gap (2.7 eV), non-toxicity, cheap raw materials and simple synthetic procedure [14], [15], [16]. Due to its suitable electronic structure, it can maintain a high stability even in acid or alkali conditions. However, g-C₃N₄ has obvious disadvantages, e.g. low surface area, effortless recombination of excited carrier-pairs and reduced quantum productivity [17], [18], [19], [20]. Therefore, g-C₃N₄ must be properly treated to obtain high efficiency sonocatalysis. When two or more semiconductors are coupled to form heterojunctions, the resultant composite delivers highly accelerated sonocatalysis due to the efficient separation of sonogenerated charge-pairs. The CB of g-C₃N₄ is above the CB of CoFe₂O₄ while the VB of CoFe₂O₄ is below the VB of g-C₃N₄, therefore the excited electrons in the CB of g-C₃N₄ entertain high reduction potential while the excited holes in the VB of CoFe₂O₄ possess high oxidation potential. Thus, heterojunctional combination of CoFe₂O₄ with g-C₃N₄ is expected to present better sonocatalytic degradation efficiency under ultrasonic irradiation due to the enhanced excited charge-pairs separation through S-scheme charge transfer mechanism.

In this work, CoFe₂O₄ coupled g-C₃N₄ nanocomposites of different mass percent ratios were prepared through the simple hydrothermal and polymerization methods. The sonocatalytic efficiency of the samples was checked by degrading TCH under ultrasonic irradiation. The conduct of exciton separation and migration of the samples were probed through different analytical approaches. Based on the results of free radical trapping experiments and the Semiconductor Energy Band Theory, a charge transfer mechanism has been proposed to elucidate the excitation and separation of charge-pairs for the degradation of TCH antibiotic.

2. Experimental section

2.1. Materials

All the reagents used in this work were analytical reagent (AR) and used without secondary treatment. The water used in the experiment was ultrapure water. Melamine (C₃H₆N₆, 99 %) was purchased from Aladdin Reagent Co., ltd. Cobalt (II) nitrate hexahydrate (Co(NO₃)₂·6H₂O, 99 %), iron(III) chloride hexahydrate (FeCl₃·6H₂O, 99 %) and tetracycline hydrochloride (C₂₂H₂₄N₂O₈·HCl) were purchased from McLean Reagent Co., ltd. NaOH (AR, 96 %), sodium sulfate (Na₂SO₄, 99 %) and barium sulfate (BaSO₄, AR) were purchased from Sinopharm Chemical Reagent Co., ltd. Ascorbic acid (AA, AR), methanol (MeOH, AR) and isopropanol (IPA, AR) were obtained from Sigma Co.

2.2. Preparation

g-C₃N₄ was prepared through thermal polymerization using melamine as the raw material. First, 20 g melamine was put in a crucible with a lid and heated at 550 °C for 4 h at a heating rate of 2 °C/min. After the completion of the polymerization reaction, the crucible was cooled to room temperate, the content was ground finely and stored for further experimental work.

0.5820 g of cobalt nitrate hexahydrate and 1.0812 g of ferric chloride hexahydrate were dissolved in 20 ml deionized water to form a homogenous mixture. 0.1 M NaOH solution was introduced dropwise to the ferric chloride solution while stirring, and the precipitates formed were separated through centrifugation. The materials were rinsed thoroughly with water and then more NaOH solution was introduced for a hydrothermal process, which was supervised by keeping the mixture in an oven at 160 °C for 12 h. The product obtained was rinsed thoroughly with deionized water and then dried in an oven overnight. The dried sample was ground completely and then calcined at 600 °C for 3 h. The final product was grounded and stored for further experimental work.

Different nanocomposites containing both g-C₃N₄ and CoFe₂O₄ were prepared by selecting their appropriate masses and dispersing them in a mixed solution of water and methanol. The mixture was heated at 80 °C while stirring to vaporize the whole solvent. The obtained residue was dried at 80 °C overnight and represented by x%-CFO/CN, where ‘x%’ represents the mass percentage (5, 10, 15, 20, 25 and 30 %) of CoFe₂O₄ in the prepared nanocomposites. The whole synthetic process is schematically shown in Scheme 1.

Scheme 1.

Synthesis process of CFO/CN composite.

2.3. Characterization

X-ray diffractometric (XRD) inspections were purchased with Bruker D8 Advance Diffractometric Instrument (Germany) fortified with a Cu-Kα radiation source. FTIR investigations were pursued with a Nicolet Magna 560 spectrophotometer made in the US with the detection range between 400 and 4000 cm⁻¹. The binding energy and elemental state of the constituent elements were defined using an ESCALAB MKII X-ray photoelectron spectrometric (XPS) instrument (from UK) with Mg-Kα radiation. The microscopic photos of the models were compiled with the help of a JEM-2010 apparatus (Japan) while photoelectrochemical inspections were defined with the aid of an IVIUM V13806 electrochemical workstation invented in Netherland.

2.4. Sonocatalytic activity test

The sonocatalytic activities of the prepared samples were measured by oxidizing TCH antibiotic in water. The precise tentative measures were as follows: Before the sonocatalytic degradation experiment, 50 ml TCH solution (10 M) and 0.2 g sonocatalyst were stirred in dark for some time to adhere as many antibiotic particles on the catalyst surface as possible. The mixture was then put in an ultrasonic instrument to conduct the catalytic degradation reaction. Throughout the sonolysis, a content of 5 ml was isolated from the reaction system after every 5 min. After centrifugation and filtration, the absorbance of the solution was measured with a UV visible spectrophotometer at the maximum absorption wavelength of 357 nm.

3. Results and discussions

3.1. XRD

The crystal-phase structure of x%-CFO/CN was analyzed through X-ray diffractometry (XRD) as shown in Fig. 1. Evidently, g-C₃N₄ has two XRD diffraction peaks centered at 13.1° and 27.5° (JCPDS 87–1526). The peak found at 13.1° corresponds to crystal plane (1 0 0) and is attributed to the layered stacking structure. The sharp diffraction peak at 27.5° communicates to crystal plane (0 0 2) formed by aromatic stacking units [21]. The diffraction peaks of CoFe₂O₄ are available at 30.0°, 35.4°, 37.0°, 43.0°, 53.4°, 56.9° and 62.5° which respectively match the crystal planes (2 2 0), (3 1 1), (2 2 2), (4 0 0), (4 2 2), (5 1 1) and (4 4 0). All these diffraction peaks are assigned to the spinel CoFe₂O₄ (JCPDF 22-1086) [22]. The characteristic peaks of g-C₃N₄ and CoFe₂O₄ can be found in the diffraction peaks of x%-CFO/CN composites. As the mass of CoFe₂O₄ increases, its diffraction peaks are gradually becoming obvious. The XRD patterns of 30%-CFO/CN are analogous to those of CoFe₂O₄ and g-C₃N₄, signifying the existence of CoFe₂O₄ in the composites. There is no impurity XRD peak signifying that the product prepared has a superior purity.

Fig. 1.

XRD patterns of g-C₃N₄, CoFe₂O₄ and x%-CFO/CN samples.

3.2. FTIR

The FTIR study was conducted to find the nature of different functionalities in the nanocomposites, and the data has been summarized in Fig. 2. The FTIR peak centered at 3167 cm⁻¹ goes to the stretching mode of the O−H group of water molecules adsorbed on the surface of g-C₃N₄, or may be related to the stretching mode of the N−H group in an uncondensed state [23]. The peak available at 1637 cm⁻¹ is provided by the bending vibration of hydroxyl of water attached to the surface of g-C₃N₄. Some other peaks found in the range of 1545–1242 cm⁻¹ are shown in by either NH-C₂ or N-C₃ groups of the polymerized samples g-C₃N₄ [24]. A relatively strong peak at 805 cm⁻¹ is provided by the breathing-mode of s-triazine constituent units. In case of CoFe₂O₄, a strong characteristic peak seen at 485–690 cm⁻¹ is delivered by the metal–oxygen (Co-O and Fe-O) stretching vibration [25]. After coupling, the FTIR peaks responsible for the adsorbed water particles or uncondensed N−H groups show a highly diminished intensity. While, the metal–oxygen peak at 485–690 cm⁻¹ show reduced intensities due to the strong interactions between CoFe₂O₄ and g-C₃N₄ in the nanocomposites which play a significant role in charge-pairs separation and the accelerated catalysis.

Fig. 2.

FTIR spectra of g-C₃N₄, CoFe₂O₄ and x%-CFO/CN samples.

3.3. Energy band structure

The UV–vis absorption survey of g-C₃N₄, CoFe₂O₄ and 25%-CFO/CN samples was completed to understand the photoabsorption property of the catalysts as demonstrated in Fig. 3a. g-C₃N₄ has a strong absorption in the visible light region estimated up to 450 nm. CoFe₂O₄ has a wide absorption range between 200 and 800 nm which leads to the black color of the sample. The absorption intensity of 25%-CFO/CN composite is significantly increased in the visible range compared with pure g-C₃N₄ indicating that the introduction of CoFe₂O₄ is indeed conducive to the absorption of visible light. Excess CoFe₂O₄ may inhibit the optical utilization of the composite due to the light absorption by aggregating and shielding of other g-C₃N₄ particles. Based on Kubelka Munk formula [26], A(h$\mathrm{\nu }$ − E_g)^n/2 = αh $\mathrm{\nu }$ (α, h, $\mathrm{\nu }$, A and E_g are absorption coefficient, Planck constant, optical frequency, constant and energy band gap respectively), the band gaps of g-C₃N₄, CoFe₂O₄ and 25%-CFO/CN samples calculated are 2.76, 1.33 and 1.60 eV respectively as is shown in Fig. 3b.

Fig. 3.

UV–vis absorption spectra (a), band gaps (b), Mott Schottky curves (c) and energy band structure diagram (d) of g-C₃N₄, CoFe₂O₄ and 25%-CFO/CN samples.

The Mott Schottky curves of the samples selected are shown Fig. 3c. All the samples studied show positive slopes signifying that all the semiconductors under investigation are n-type. By generalizing the cures and cutting x-axis, the V_fb of g-C₃N₄, CoFe₂O₄ and 25%-CFO/CN are −0.58, −0.62 and −0.54 V (vs SCE) respectively. Since the flat band potential of n-type semiconductors is almost equal to CB potential [27], the CB potentials of g-C₃N₄, CoFe₂O₄ and 25%-CFO/CN are −0.58, −0.62 and −0.54 V (vs SCE) respectively. The VB potentials were calculated according to the empirical formula given in equation (3).

$${\mathrm{E}}_{\mathrm{V}\mathrm{B}}={\mathrm{E}}_{\mathrm{C}\mathrm{B}}+{\mathrm{E}}_{\mathrm{g}}$$ (3)

here E_VB, E_CB and E_g respectively represent the VB potential, CB potential and band gap energy. Thereby, the calculated valence band potential (vs SCE) of g-C₃N₄, CoFe₂O₄ and 25%-CFO/CN is 2.18, 0.71 and 1.06 V (vs SCE) respectively as is shown in Fig. 3d.

3.4. TEM

A morphological study of the 25%-CFO/CN nanocomposite was conducted by getting TEM and HRTEM images as demonstrated in Fig. 4. CoFe₂O₄ nanoparticles with about 50 nm size are regularly distributed on the surface of g-C₃N₄ nanosheets in Fig. 4a. Interestingly the nanoparticles show no sign of aggregation, and they are well dispersed on the surface of g-C₃N₄. In Fig. 4b, the lattice fringes are obviously clear. The 0.148, 0.132 and 0.205 nm lattice fringes are respectively devoted to the planes (4 4 0), (6 2 0) and (4 0 0) of CoFe₂O₄ lattice. From Fig. 4c, the EDX analysis reveals that the prepared sample is composed of C, N, Co, Fe and O respectively in 40.36, 46.74, 2.45, 4.93 and 5.51 weight percentages in 25%-CFO/CN sample. And, the elemental mappings of 25%-CFO/CN demonstrate that the 25%-CFO/CN sample is comprised by well-spread C, N, Co, Fe and O elements as shown in Fig. 4d.

Fig. 4.

TEM (a) and HRTEM images of 25%-CFO/CN (b), EDS spectrum (c) and elemental distribution mapping of C, N, Co, Fe and O in 25%-CFO/CN sample (d).

3.5. XPs

The XPS data was obatined to indentify elemental states and the interations between the constituent g-C₃N₄ and CoFe₂O₄ nanopaticles as is shown in Fig. 5. Fig. 5a reveals the survey spectra, and the substances detected are C, N, Fe, Co and O. The rest of plots are XPS study for individual electron orbital in detail. In Fig. 5b, C1s signifies that it possesses two energy peaks spotted at 284.9 and 288.4 eV. These peaks are respectively acknowledged from sp² C-atom covalently attracted by N-atom of the ring (C–N–C) and sp² C-atom covalently attracted by NH₂ (C–(N)₃) and responsible to connect different s-triazine units in the polymeric structure. From the spectra in Fig. 5c, it is accepted that N1s owns three energy peaks centralized at 398.7, 399.6 and 400.9 eV. These peaks are respectively acknowledged from either the N-atom covalently attracted sp² C-atom or the terminal N-atom covalently attracted to carbon atoms, i.e. N-(C)₃. The binding energy peaks sighted at 780.0, 795.7 and 803.5 eV, carrying shake-up satellite peaks at 782.8, 786.4 and 790.1 eV, are donated respectively to Co 2p_3/2 and Co 2p_1/2 of Co²⁺ as provided in Fig. 5d [28]. For Fe, the two characteristic binding energy peaks at 710.5 and 723.8 eV are respectively donated to Fe 2p_3/2 and Fe 2p_1/2 as revealed in Fig. 5e. In addition, another characteristic peak at 718.6 eV is assigned to the Fe³⁺ satellite transitions. There are also weak energy peaks at 713.3, 726.6 and 732.4 eV indicating that the some iron atoms are also present in Fe²⁺ [29]. The spectra of O1s are provided in Fig. 5f, the peaks at 529.8 and 533.4 eV belong to the metal–oxygen chemical bond existing in the catalyst (O₂⁻), and the peaks at 531.8 is related to the water molecules adsorbed on the surface of the catalysts.

Fig. 5.

XPS survey spectrum (a), high-resolution spectra of C 1 s (b), N 1 s (c), Co 2p (d), Fe 2p (e) and O 1 s (f) of 25%-CFO/CN.

3.6. Sonocatalytic decomposition of pollutants

The materials prepared were subjected to be investigated in the degradation of TCH under ultrasonic conditions. The removal rates of the antibiotic with 5%-CFO/CN, 10%-CFO/CN, 15%-CFO/CN, 20%-CFO/CN, 25%-CFO/CN, and 30%-CFO/CN are respectively 15.11, 19.83, 19.94, 22.61, 26.71 and 19.93% as is shown in Fig. 6a. Obviously, all the nanocomposite samples show enhanced degradation activity. Further, as the amount of CoFe₂O₄ increases, the pollutant eradication performance of the composites is first enhanced and then reduced. When the loading amount of CoFe₂O₄ is 25%, the pollutant removal performance of the composite is the highest. The rate constant for eradicating the antibiotic with 25%-CFO/CN is 0.01317 min⁻¹ which is 2.5 times higher than that of 0.00523 min⁻¹ with g-C₃N₄ as is shown in Fig. 6b. The reason for this may be the efficient charge-pairs transfer and lower recombination rate, and the larger number of active sites provided by the loading of CoFe₂O₄. However, as its amount is further increased, the particles of CoFe₂O₄ are gradually agglomerated which may decrease the number of active sites for TCH attachment and shield the sonoluminescence so as to reduce the degradation rate.

Fig. 6.

Sonocatalytic degradation of TCH with g-C₃N₄ and x%-CFO/CN nanocomposites under ultrasonic irradiation for in 5 min (a) and reaction rates with g-C₃N₄ and 25%-CFO/CN (b). Here x% represents the weight amount of CoFe₂O₄ in the nanocomposites.

3.7. Fluorescence spectra and photoelectric property

Fluorescence spectra (PL) are normally applied to understand the charge carrier transport efficiency of semiconductor materials. Generally, higher photogenerated charge carrier-pairs recombination rates are reflected by stronger fluorescence intensity. Therefore, a PL study was conducted as is shown in Fig. 7a. Compared with g-C₃N₄, the PL intensity of the 25%-CFO/CN composite is significantly weaker signaling that the separation of exciton-pairs is facilitated by loading CoFe₂O₄ over g-C₃N₄.

Fig. 7.

PL spectra (a) and PC curve (b) of g-C₃N₄, CoFe₂O₄ and 25%-CFO/CN samples.

The ultrasonic cavitation induces sonoluminescence, resulting in excitation of semiconductor as shining by photons and in generation of e⁻-h⁺ pairs [30]. The separation of photogenerated charge-pars in the semiconductor can also be investigated by analyzing the photocurrent density of the samples. A stronger photocurrent intensity usually reflects a better separation magnitude of the charge carrier-pairs. Fig. 7b reveals the photocurrent response curves (PC) of g-C₃N₄, CoFe₂O₄ and 25%-CFO/CN under visible light. Clearly, the photoelectric currents are rapidly generated from all samples indicating that the samples can respond to visible light. Among them, 25%-CFO/CN composite has the highest photocurrent intensity which indicates that it has the highest photogenerated charge carrier-pairs separation efficiency.

3.8. Mechanism of sonocatalysis

To track the contribution of different oxidation species generated during sonocatalysis, radical trapping experiments were performed with 25%-CFO/CN as shown in Fig. 8a. Ascorbic acid (AA), methanol (MeOH), Na₂SO₄ and isopropanol (IPA) were directed to check the role of ^•O₂⁻, h⁺, e⁻ and ^•OH. It is interesting, but with no surprise, that the oxidation of the target antibiotic is not hindered when electrons trapper Na₂SO₄ is introduced to the reaction mixture. However, all the other three trappers, namely AA, MeOH and IPA, show a significant hindering effect on the sonocatalytic decomposition of the antibiotic. By the way, the effect of AA is dominant. These experiments indicate that ^•O₂⁻, ^•OH and h⁺ are the active species accountable for the oxidation of TCH antibiotic.

Fig. 8.

Sonocatalytic degradation activity of 25%-CFO/CN for TCH in the presence of different scavengers under ultrasonic irradiation for 5 min.

To enlighten the charge transfer in CFO/CN composite, a schematic sonocatalytic mechanism has been planned as available in Fig. 9. The Fermi levels of both g-C₃N₄ and CoFe₂O₄ are located near their corresponding conduction bands, and before hybridization takes place their electronic densities should acquire reaching a steady state. Therefore, after hybridization an appropriate heterojunction is formed, their electronic densities will be powerfully governed by the Coulomb interaction at the interface of the heterojunction. As a result, the charge density of g-C₃N₄ is rapidly transferred to CoFe₂O₄ until the Fermi levels of both the constituents are equilibrated. At the hot interface between CoFe₂O₄ and g-C₃N₄, depleted and accumulated charge density layers are shaped, resulting in different charge states in the two components. Thus, we can say that an electric field is settled at the hot interface of the two components. The generation of electrons and holes as well as their consequent separation at the hot interface can be arrived in terms of the built-in electric field formed via sonoluminescence. It maybe suggested that when exposed to ultrasound waves for interaction, the holes in the VB of g-C₃N₄ are going to oxidize or degrade or decompose the antibiotic. At the meantime, the sonoluminesence excited electrons react with the O₂ in solution forming ^•O₂⁻. Thus, both electrons and holes are transformed into oxidizing species and carry elevated thermodynamic energies to exercise the oxidation of antibiotics. The super oxide anions formed react with water to form H₂O₂. The formed H₂O₂ undergoes further reaction to form hydroxyl free radicals. The produced H₂O₂ and ^•OH species react with TCH and further accelerates the sonocatalytic performance under the stipulated conditions.

Fig. 9.

Proposed mechanism for the sonocatalytic degradation of TCH with CFO/CN.

4. Conclusion

This work presents a simple method to prepare an effective ferrite-based sonocatalyst for the ultrasonic-assisted eradication of hazardous antibiotics from aqueous environments. The prepared sonocatalyst was then coupled with g-C₃N₄ in different mass percent ratios to form nanocomposites. The nanocomposites displayed enhanced sonocatalytic activity and degraded about 26.71% antibiotic in 10 min. The extraordinary decomposition efficiency was credited to the excellent charge separation in the nanocomposite sonocatalysts. Radical trapping experiments confirmed that ^•O₂⁻, h⁺ and e⁻ were dominant oxidizing species in the eradication of TCH. Electrons from the sonoluminescence and holes in the VB of CoFe₂O₄ contributed to the highest production of super oxide anions and hydroxyl free radicals owing to their high thermodynamic energy in the given redox reactions.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

No data was used for the research described in the article.