Fabrication of Covalently Linked Ruthenium Complex Onto Carbon Nitride Nanotubes For The Photocatalytic Degradation of Tetracycline Antibiotic

Due to the problem of direct disposal of euents contains antibiotics to the environment and the emergence of resistant bacterial pathogens, the wastewater treatment of pharmaceutical industry has known as an importance research background. In this study, the renement and photodegradation ability of one of the most widely used antibiotics, “tetracycline” was investigated by ruthenium complex immobilized on the modied graphitic carbon nitride nanotubes. For this purpose, graphitic carbon nitride nanotubes (g-C 3 N 4 NTs) were successfully synthesized by the hydrothermal method and functionalized with 1,10-Phenantroline-5,6-dione ligand during another step. Then, the functionalized g-C 3 N 4 NTs were reinforced with immobilization of dichloro(p-cymene)ruthenium(II) dimer. The structure and morphology of the prepared photocatalyst was studied by X-ray diffraction (XRD), fourier transform infrared (FT-IR), scanning, and transmission electron microscopy (SEM & TEM) analyses. In the following, the photocatalyst's ability to optically degrade the tetracycline antibiotics was performed in a suspension reactor equipped with a LED lamp (60 W) and effective parameters such as the amount of catalyst, irradiation time, temperature, and pH were optimized. The results showed that the immobilization of Ru complex onto functionalized g-C 3 N 4 NTs improved the photocatalytic activity and increased the degradation eciencies to amount 43%. Furthermore, COD analysis was used for the determination of the amount of mineralization and results showed that the mineralization of 10 mg/L tetracycline solution of about 90% can be performed using 20 mg of Ru (II) complex/ g-C 3 N 4 NTs at pH=7 after 480 min without any additive oxidant.


Introduction
Tetracyclines are known as a group of broad-spectrum antibiotics that have a common basic structure and one of the most widely used these antibiotics is tetracycline. This antibiotic is used to treat humans and stimulate the growth of livestock and birds (Chopra and Roberts, 2001). According to the World Health Organization (WHO), the presence of this substance in nature is dangerous (Conde-Cid et al., 2020). Tetracycline nds its way to surface and groundwater through domestic and hospital e uents so livestock and poultry farms that it enters the human food cycle by drinking water and consuming fruits and vegetables (Fiaz et al., 2021). This antibiotic causes genetic resistance and accumulates in the skeletal tissues of the human body through the formation of a stable complex with calcium (Smith et al., 2015;D'Costa et al., 2011). Tetracycline is not biodegradable and disrupts in the biological treatment units, and also produces more toxic products in many chemical treatment processes such as chlorination and ozonation (Borghi et al., 2014;Park et al., 2007). On the other hand, the use of the surface adsorption technique only transfers tetracycline from one phase to another and does not eliminate its contamination (Homem and Santos, 2011). Therefore, it can be said that the common removal methods of organic contaminations are ine cient and troublesome for this antibiotic (Rodriguez-Narvaez et al., 2017).
Instead, the advanced oxidation processes, especially photocatalytic systems, are a good way to remove these compounds and are based on the production of hydroxyl radicals (Hong et  . These radicals are able to oxidize all organic compounds to the stage of carbon dioxide and water production . Most of the photocatalysts presented so far are e cient only in the ultraviolet range (Vasei et al., 2019;Yadav et al., 2018), and since only a small percentage of sunlight is made of ultraviolet light and most of that is visible light, so it needs to design and fabrication of the new generation of active photocatalysts in the visible range. In recent years, a lot of research has been done on semiconductors that have a suitable gap and can be used in the visible range Yu et al., 2020;de-Moraes et al., 2021). Graphite nitride carbon is one of the semiconductors that has a smaller bandgap (∼2.7 eV) than many other common photocatalysts such as titanium oxide and zinc oxide, and as a result, it is more e cient at absorbing sunlight (Dong et al., 2014; Zhang et al., 2021). This substance is non-toxic, cheap and easy to prepare (Cui et al., 2018). However, the most important weakness of g-C 3 N 4 , which reduces its performance in optical degradation, is the rapid recombination of electron-hole pairs (Shi et al., 2021;Liu et al., 2021). Increasing e ciency in the optical degradation operations depends on the degree of separation of electron-hole pairs. To produce active oxygen species, it is necessary that the electrons and produced holes can react separately with the water or oxygen molecules around them (Chen et al., 2020). There are two major ways to solve this g-C 3 N 4 problem and achieve maximum performance; First, changes in the morphology and particle size of the photocatalyst: The particle size of a photocatalyst has a large effect on the energy gap. As the particle size decreases along with increasing the width of the energy gap, and so its oxidation-reduction strength will increase (Singh et al., 2018). Second, the decoration of g-C 3 N 4 with other conductive and semiconductor metals; these metals increase the absorption of visible light and the e ciency of the photocatalyst by producing hot electrons and injecting them into the semiconductor .
In the present project, the above two techniques were used to fabricate an e cient photocatalyst in the visible range. In this regard, the g-C 3 N 4 nanotubes were synthesized instead of g-C 3 N 4 bulk and it cause to improve the energy band gap and reduce the recombination rate of the electron-hole pair. In addition, the use of g-C 3 N 4 nanotubes increases the surface area and has a positive effect on photocatalyst performance. To implement the second approach and further enhance the photocatalytic properties, the surface of the nanotubes was decorated with a ruthenium complex. The ruthenium complexes have the ability to absorb light in the wavelength range of 400-800 nm and cause the formation and injection of electrons into the photocatalytic system. In the end, the photocatalytic ability of Ru(II)complex/g-C 3 N 4 NTs was investigated in the decomposition of tetracycline antibiotic under visible light and the effect of factors such as radiation time, amount of photocatalyst, pH, and temperature on degradation e ciency was investigated. The advantages of this method are maintaining the photocatalyst stability and reusability in the heterogeneous system.

Synthesis of g-C 3 N 4 NTs
The preparation of g-C 3 N 4 NTs was done in two separately steps. First, white crystals of melaminecyanuric acid were prepared by a hydrothermal method brie y, 0.5 mmol of melamine and 0.5 mmol of cyanuric chloride were dissolved in 75 mL deionized water and kept stirring for 30 min. The resultant transport solution was transferred to an autoclave ask and maintained at 180°C for 6 h. After this time, the autoclave was cooled until to ambient temperature and the obtained white needle-like crystals were centrifuged and washed several times with deionized water to remove organic impurities and freezer dried for 24 hours.
The second step was the calcination process as follows the melamine-cyanuric chloride crystals were calcined at 450°C for 5 h with a heating rate of 1°C min −1 under nitrogen inert atmosphere.

Functionalization of g-C 3 N 4 NTs with phenantroline-5,6dione
In order to functionalize graphitic carbon nitride nanotubes, 250 mg of g-C 3 N 4 NTs along with 1 mmol of phenantroline-5,6-dione ligand were dispersed in 50 ml of acetonitrile and after equipping the reaction vessel with the condenser it was re uxed for 60 hours at 60 ° C. At the end of the reaction time, the functionalized g-C 3 N 4 NTs were collected by centrifugation and dried at 50 ° C after washing several times with acetonitrile.

Synthesis of Ru (II) complex/ g-C 3 N 4 NTs
A 250mL ask was charged with the g-C 3 N 4 NTS bearing the phenantroline-5,6-dione ligand (250 mg), [RuCl 2 (p-cymene)] 2 (0.2 mmol) and 200 mL anhydrous dichloromethane. The obtained suspension was re uxed for 2 h at 50°C. After that, the resultant product was separated through decanting of the solvent and the nanotubes were washed several times with dried dichloromethane to remove extra values of ruthenium moieties. according to ICP-OES analysis amount of Ru was 10 % in the obtained photocatalyst.
2.5. Study of the photocatalytic activity of Ru (II) complex/ g-C 3 N 4 NTs All photocatalytic reactions were performed in a 50 ml beaker containing 25 ml of tetracycline solution (10 mg/L). During irradiation of visible photons, the sample containing the nanocomposite was stirred by a magnetic stirrer to homogenize the system and prevent photocatalytic precipitation. The 60-watt visible light source was placed directly on top of the beaker at a distance of 15 cm from and at a distance of 15 cm from it, and despite the heat generated by the lamp, no cover was used. At the beginning of the test, from the moment that the samples were immersed in the solution, placed in the dark for 30 minutes to reach the adsorption /desorption equilibrium. The concentration of initial and residual tetracycline in the samples that irradiated by UV-Vis spectrometer at different times were measured after separating of nanocomposite by centrifuge and then the percentage of drug removal was calculated according to Equation 1.
Photocatalytic destruction (%) = (C 0 -C)/C 0 × 100 (1) In this equation, C 0 is the initial concentration of tetracycline and C is the concentration of the tetracycline remaining in the solution, after the photodegradation reaction. It should be noted that all experiments were repeated twice and the average value was reported. In addition, the COD of the samples was measured every hour by FAS titration with potassium permanganate (Goh and Lim, 2008).

Adsorption kinetics study
The kinetic of photocatalytic degradation of pesticides in the heterogeneous oxidation systems with visible light follows the Langmuir-Hinslowwood kinetic model  Where r, C, t, k and K are the oxidation rate (mg/L min), herbicide concentration (mg/L), irradiation time (min), reaction rate constant (1/min) and reaction adsorption coe cient (L/mg), respectively. At low initial concentrations of herbicides, Equation 2 is changed to the quasi-rst-order equation (Equation 3).
By plotting ln (C/C 0 ) versus t, the slope of the diagram shows the apparent rate of photocatalytic degradation. . In this method, the mentioned precursors are converted to g-C 3 N 4 at 550 ° C. But the main challenge of this method is the variable condensation e ciency, which leads to the production of a product with a small surface area and, of course, limits its application to photocatalytic processes (Ismael and Wu, 2019). However, the surface area of g-C 3 N 4 depends on the type of precursor and synthesis conditions, and by changing the above factors, a high surface area can be achieved. Typically, the use of melamine precursors results in the production of g-C 3 N 4 with a high N/C ratio (Yan et al., 2009). The higher the N/C ratio, the lower the energy bandgap and the higher the photocatalytic property.

Results
If melamine is used in combination with an oxygen-containing precursor to making g-C 3 N 4 , the product has more porosity and a higher surface area ( which is done by polymerizing the precursors in the created space of the template (such as silica) and then removing the template by reagents that dissolve the silica. Although the g-C 3 N 4 NTs obtained from this method has a high surface area, its crystallinity has decreased. Crystallinity is an effective parameter against the separation of excited charges in the photocatalytic process (Wanget al., 2014). The second method is the soft templating (Chen et al., 2021), which its general mechanism is based on the polymerization of nitrogen-rich precursors in the space that has been created by soft template such as trithion X10, P12, and F124 (Rastogiet al., 2008). The disadvantage of this method is the high polymerization temperature, which leads to the loss of the template before the complete formation of the nanotube. In the present project, g-C 3 N 4 NTs were synthesized by the joint polymerization of nitrogenous (melamine) and chlorinated (cyanuric chloride) precursors through the hydrothermal process without the use of any template. The cyanuric chloride is unstable under aqueous conditions therefore; it was converted to cyanuric acid as an oxygenated precursor created melamine-cyanurate crystals through a series of hydrogen bonds with melamine. Finally, the growth of these crystals within 6 hours and calcination at 450°C, it cause to produce g-C 3 N 4 NTs. Because of incomplete polymerization of melamine, there are free amine groups at the outer edges of the prepared g-C 3 N 4 NTs and these defects in the structure cause to balance in the energy bandgap and also it facilitates the functionalization of the surface. So here, the functionalization was done in the through formation imine band between carbonyls of phenantroline-5,6-dione ligand and free amines of g-C 3 N 4 NTs. In the end, the coordination of immobilized bidentate ligands with dichloro(p-cymene)ruthenium(II) led to the stabilization of the relevant complex on the outer edges of g-C 3 N 4 NTs. Scheme1 depicted the steps of the synthesis of the photocatalyst.

Fourier transform-infrared spectroscopy (FT-IR)
The functional groups of melamine-cyanuric acid, g-C 3 N 4 NTs, Ru(II)complex/g-C 3 N 4 NTs samples were con rmed by FTIR spectroscopy. The melamine-cyanuric acid spectrum (Fig. 1a)  of melamine-cyanuric acid crystals. After pyrolysis treatment, the arrangement of the absorption bands was changed (Fig. 1b). For example, the intensity of absorption bands attributed to N-H groups dropped sharply in region >3000 cm −1 , and also the assigned peak of carbonyl at 1731 cm −1 disappeared which con rmed the release of NH 2 and H 2 O during pyrolysis treatment to the formation of g-C 3 N 4 . Figure 1c again shows the decreasing of the vibrations of the amino groups, which, together with the new appeared band at 1650 cm −1 , proves the chemical modi cation of the nanotubes through the formation of the imine band with phenantroline-5,6-dione ligand.

Ultraviolet and visible absorption spectroscopy (UV-Vis)
To get the UV-Vis absorption spectrum of g-C 3 N 4 Nts, this material was dispersed in ethanol by ultrasonic bath for 20 minutes and then its adsorption was measured. As shown in Figure 2, g-C 3 N 4 NTs have maximum adsorption at 368 nm, which has a blue shift (the transition to a higher energy level or shorter wavelength) relative to the bulk sample (Xin and Meng, 2013). In nanomaterials, the distance between the electron-hole is controlled by the particle size, so that as the particle size shrinks to the nanometer scale, the movement of excitons are limited and a transition in the optical spectrum is observed (Ren et al., 2016). In general, the transfer of optical spectrum to higher energies due to the decreasing of particle size means an increase in the forbidden band energy. The energy difference (in units of electron volts) between the highest valence band and the lowest conduction band is called the energy bandgap (Kumar et al., 2016). To calculate the bandgap energy of a material, the absorption coe cient parameter of the material must rst be obtained. The absorber coe cient is an important parameter in optical applications and is calculated for a transparent thin lm from the following relation: Here, d is the thickness of the thin lm in nanometers and T is the percentage of light transmission. Since g-C 3 N 4 has a direct energy gap, so by direct electron transfer between the capacitance band and the conduction band, the band gap energy (Eg) is obtained from the Tuac equation (5).
Here, A is constant, α is the absorption coe cient, Eg is the width of the energy gap, and hν is the energy of the emitting photon with respect to the absorption spectrum of the synthesized g-C 3 N 4 NTs. Also, n for g-C 3 N 4 NTs with a straight and permissible bandgap is equal to 1/2. According to Figure 2, the energy gap of g-C 3 N 4 NTs can be calculated by drawing (αhν) 2 (eV.cm −1 ) 2 in versus of hν (eV) and extrapolation its linear part. This energy gap was obtained 2.96 eV.

X-ray diffraction (XRD)
The prepared melamine-cyanuric acid sample and its Ru(II)complex/g-C 3 N 4 NTs have been characterized by X-ray powder diffraction technique ( gure 3). The presence of sharp diffraction peaks in the XRD of the synthesized melamine-cyanuric acid sample during the hydrothermal process con rms that the obtained product is highly crystalline (Liu et al., 2017). The X-ray diffraction pattern changed completely after the pyrolysis process and the formation of g-C 3 N 4 NTs nanotubes, followed by chemical modi cation.
New peaks have appeared around 2θ = 13.1 and 27.2 and con rm the formation of graphite nitride carbon phases in the sample (Pattnaik et al., 2019). An important point that can be seen in this image is the decrease in the intensity of the peaks after heat treatment, which indicates the percentage of crystallinity of the sample is reduced, or in other words, the long-range order of the structure is reduced.
During the heat treatment, the carbon nitride surfaces appear to be oxidized and the layers separated.
Also, due to the oxidation process, the structure is defective and its effects in reducing the peak intensity of this material have appeared. Another reason for the low intensity of peaks is the modi cation process when an element enters the structure of the compound through chemical bonding, causing the corresponding peaks to atten or shift and due to the stabilization conditions of the complex, there is no possibility of ruthenium phases in the nal structure.

Scanning Electron Microscopy (SEM)
Scanning electron microscopy provides useful information from topography and morphology of the prepared photocatalyst. Scanning electron microscope images of the sample are shown in Figure 5. These images were taken with different magni cations to determine the overall morphology of the sample and how the nanotubes are arranged. As you can see, the study sample contains clusters of orderly nanotubes and the amount of unwanted particles in that is very low.
To determine the composition of the elements in the prepared photocatalyst, the distribution of active sites, and to determine whether the ruthenium complex is embedded in the surface of the nanotube matrix or in the inner part of the matrix, two complementary analyzes of energy-dispersive X-ray pattern and element distribution map (EDS & Map) were used. The EDS spectrum con rmed the presence of the elements carbon, nitrogen, ruthenium, and chlorine with the appearance of the corresponding peaks and the mapping analysis of polymer shows uniform distribution and non-accumulation of elements that enhance the properties of the photocatalyst.

Transmission electron microscopy (TEM)
Transmission electron microscopy (TEM) images showed that the structure of g-C 3 N 4 is in the form of hollow curved tubes with an outer diameter of less than 10 nm and a length of several hundred nanometers ( Figure 5). It appears that after chemical functionalization of nanotubes and loading of ruthenium (II) complex, failure has occurred in areas where defects such as pentagons and heptagons. In addition, the modi cation process has resulted further distance between the nanotubes, which has a constructive effect on increasing the surface area of the photocatalyst. The Ru(II) complex/g-C 3 N 4 NTs show accumulations in some areas, which possibly due to the agglomeration of graphitic nanotubes. This re-stacking of graphitic nanotubes is dependent on strong van der Waals forces and π-π interactions of sp 2 structures and hydrogen bonding.

Investigation of effective factors in optical degradation of tetracycline
3.3.1. In uence of amount of photocatalyst As shown in Figure 6, the degradation e ciency of tetracycline increased with increasing photocatalyst concentration to 30 mg, after which the degradation e ciency remained almost constant. This can be attributed to the fact that when all the antibiotic molecules landed on the nanotubes, the excess amount of photocatalyst due to the lack of antibiotic molecules had no effect on the percentage of degradation and even slightly reduced the reaction rate. The decreasing in reaction rate at a dose of 40 mg was caused by two main factors. First; g-C 3 N 4 nanotubes tend to agglomeration due to their nanometric size and high surface energy, and when the concentration of these nanotubes exceeds a certain limit, the activated nanotubes will be deactivated by contact with the base catalyst, and to follow it, the catalytic e ciency decreases. Second, with the increasing amount of photocatalyst, the turbidity increased and the light scattering occurs due to the collision of optical rays with the catalyst particles scattered in the solution and a number of light photons lose their energy and thus the e ciency of photocatalytic processes decreases.

In uence of irradiation time
To investigate the effect of irradiation time on degradation e ciency, a certain concentration of tetracycline was exposed to an optimized amount of photocatalyst (30 mg) and a frequency of 60 watts at different times. The results are shown in Figure 7. As you can see, the percentage of antibiotic degradation increased with increasing reaction time, so that in 90 minutes, more than 99% of the tetracycline molecule was broken down into smaller components. This step was followed by monitoring the UV spectrum. But the degradation of the small organic components until producing carbon dioxide and water takes longer so that after 6 hours, 30% of the smaller molecules are still present in the solution. Chemical Oxygen Demand (COD) technique was used to determine the concentration of organic components in the second step. In this method, the sample is strongly re uxed by acidic solutions and a certain amount of potassium dichromate (K 2 Cr 2 O 7 ). After digestion of the sample to determine COD, the residual and unoxidized amount of potassium dichromate is titrated with ammonium sulfate to determine the amount of potassium dichromate consumed and the oxidized material is calculated in the oxygen equations.

In uence of temperature
The in uence of temperature on tetracycline degradation was shown in gure 8. Changes in process velocities in the range of 0 to 75 ° C indicate that degradation rate is directly related to reaction temperature ( gure 8). In fact, the reaction temperature, on the one hand, contributes to the production of dissolved O 2 and the degradation of H 2 O 2 , which ultimately leads to the production of active hydroxyl radicals, and on the other hand, it provides the activation energy for the reaction. Based on this, it can be said that the catalytic process is exothermic and by drawing ln k (rate constant) versus the inverse of the temperature (kelvin) and in accordance with Arrhenius equation (Jensen, 1985) (Eq. 6), the activation energy of the reaction was obtained from the slope of the straight line. The value of the slope is equal to -Ea/R where R is a constant equal to 8.314 J/mol K. (Fig. 9).
The activation energy (Ea) of tetracycline degradation is estimated to be 0.94 kJ/mol.

In uence of pH
In optical degradation processes, pH can affect the desired decomposition rate of the contaminant. Previous studies have shown that pH plays an important role in the breakdown and elimination of antibiotics. As shown in Figure 5, there is a signi cant difference between the percentage of tetracycline degradation at different pHs and the rate of decomposition at alkaline pH is signi cantly higher and optimized at pH=9. The pH variable affects the adsorption and dissociation capacity of the target compounds, the electric charge distribution on the surface of the catalysts, and the oxidation potential of the conduction band.
Since the isoelectric point of g-C 3 N 4 based materials is around pH=9 [Zhu et al., 2015] and the tetracycline antibiotic acidic, the effect of pH on the photocatalytic process can be justi ed by the presence of electrostatic forces between the surface of g-C 3 N 4 NTs and tetracycline. Thus, at pHs > 9, the g-C 3 N 4 NTs surface has a negative charge and the tetracycline molecules also have a negative charge, therefore, the force between them is repulsive and thay don't have any reluctant to react and this cause to decrease resulting in the yield of photodegradation decreased at the pH=11. At pHs < 9, the g-C 3 N 4 NTs surface has a positive charge and the tetracycline molecules also have a positive charge due to protonation. Therefore, the electrostatic force between the photocatalyst and the tetracycline is the repulsive force, which leads to a decrease in degradation e ciency. Although, the highest percentage of pesticide degradation was obtained at an alkaline pH but, due to the economic aspect and ease of operation treatment at the neutral pH, investigation of other parameters in photocatalytic degradation of the mentioned pesticides was performed at pH = 7

Investigation of synergistic effect
In order to study the synergistic effect, tetracycline degradation reaction in the presence of g-C 3 N 4 NTs and Ru(II) complex/g-C 3 N 4 NTs was investigated separately and the results are reported in Table1.
According to the results, g-C 3 N 4 nanotubes alone are able to destroy 62% of tetracycline in 90 minutes.
After loading the ruthenium complex, the degradation percentage was higher than 99% for the same period. It seems combining two photon-active substances; g-C 3 N 4 NTs and Ru(II) complex has been able to optimize the prepared photocatalyst bandgap. Therefore, the synergistic effect of Ru(II) complex and g-C 3 N 4 NTs enhances the photocatalytic properties.

Conclusions
In this study, Ru (II) comples/g-C 3 N 4 nanocatalyst was synthesized and its photocatalytic properties for the degradation of tetracycline antibiotic was investigated. The synthesized photocatalyst were evaluated by various methods such as Infrared spectroscopy, X-ray diffraction, scanning and transmission electron microscopy, and visible ultraviolet spectroscopy. The results showed the successful synthesis of graphite nitride carbon nanotubes reinforced with ruthenium (II) complex. In general, one of the most important properties of a photocatalyst or a catalystis its speci c surface area. Therefore, the synthesis of g-C 3 N 4 in the form of nanotubes had a signi cant performance in photocatalytic properties. In addition to increasing the speci c surface area, the accessibility surface for light also increases automatically. Finally, since the bulk graphite carbon nitride alone is a suitable option for photocatalytic degradation of pollutants, the use of improved its, which has a higher speci c surface area and better activity, is not farfetched.

Figure 2
The spectrum of a) UV-Vis spectrum b) curve of (αhν)2 versus hν of g-C3N4 NTs to bandgap estimation.  Plot of Ln k versus 1/T to estimation of Ea.