The adsorption of naproxen on adsorbents obtained from pepper stalk extract by green synthesis

: This study presented the adsorption of naproxen on adsorbents ﬁ lled with Cu, Fe, and Cu/Fe nanoparticles (NPs) obtained by the green synthesis method from pepper stem waste. The resulting adsorbents were characterized by Fourier transform infrared spectroscopy, scanning electron microscopy-energy dispersive X-ray spectroscopy, and thermal gravimetric-di ﬀ erential thermal analysis. The amount of naproxen not adsorbed in the solution was determined from the cyclic voltammetry method, which is one of the electrochemical methods. The equation of the calibration curve used in the conversion of current to concentration and the R 2 value were


Introduction
Naproxen, one of the most commonly used active ingredients of the nonsteroidal anti-inflammatory drug group, is a powerful painkiller that contains aryl acetic acid in its structure and is frequently recommended by doctors due to its analgesic and anti-inflammatory properties.Owing to the recent pandemic (such as , drug supply and demand have increased.This mobility brought about the contamination of natural water resources, drinking water, and wastewater with such chemicals [1]. When naproxen is exposed to factors such as temperature, pH, and light, it can decompose into 2-acetyl-6-methoxynaphthalene (AMN) and 2-methoxy-6-ethylnaphthalene (MEN) (Scheme 1).It has been reported that if they mix with water, they will pose a fatal threat to the creatures living there [2,3] and even pose a risk of damaging their genetics [4].Another study reported that aquatic creatures cause deterioration in the tissues of organs such as kidneys and liver and changes in gill structures [5] For this reason, it is of great importance to remove naproxen from the environment.
Techniques such as photodegradation [6], biodegradation [7], biofiltration [8], electrocoagulation-flotation [9], chlorination [10], and adsorption [11] have been used to separate such drugs from the environment.Adsorption [12], which is among these techniques and has been successfully used in the removal of many metals, paints, and dirt, stands out as a practical, cheap, fast, and easy application technique, whereas other techniques are time-consuming, expensive, have limited conditions, and difficult.The success of adsorption also depends on the good surface properties of the adsorbent used.Agricultural wastes have been frequently used in adsorbent production for many years because they have the potential to exhibit high surface area due to their cellulosic structure and are abundant in nature.
In recent years, researchers have developed nanoparticulate adsorbents using natural and non-toxic raw materials such as plant extracts [13] and used them in various fields such as catalysis [14], biosensors, and medicine.The production method is called Green Synthesis because it is more environmentally friendly, has less CO 2 emissions, is cheaper, more practical, and uses fewer chemicals than traditional adsorbent production methods.In this method, phytochemicals such as polyols, polyphenols, and terpenoids found in the plant extracts act as both trapping and reducing agents, and owing to the nanostructures obtained, the unit properties of the material are moved to the nano level.In this way, it is possible to exert more properties than the larger-sized structures and to improve the physical and chemical modification and properties of the material at the nanoscale.
In the literature, metal-doped nanoparticles (NPs) were prepared from aqueous extracts of plants such as the Tilia plant [15] and peanut Vera L. shell [16] using the green synthesis method.These materials were used both as catalysts and as adsorbents in the removal of heavy metals such as Cd(II), Pb(II), Zn(II), Co(II), and Cu(II) [17].In addition, NP structures can be easily synthesized by the green synthesis method using solutions of elements such as gold [18], silver [19], zinc oxide [20], iron [21], and copper [22].
Agricultural wastes, plants, and their roots have cellulosic, hemicellulose, and lignin structures.These structures contain hydroxyl groups, and the plant roots turned into biosorbents by these groups will have a certain affinity for metal ions [23].It is known that hydrogen ions in functional groups such as amide, amine, and carboxyl play an important role in the displacement of metal ions [24].The ability of such biomass groups to remove heavy metals is due to the binding groups in the structure.
There are a few studies in the literature on naproxen adsorption using commercially activated carbon [25], agricultural waste apricot peels [26], and NPs/nanocomposites [27].There are no studies in which an adsorbent with Cu and Fe NPs was developed by green synthesis from BS.The adsorbed naproxen, and the amount of naproxen remaining in the environment after adsorption was determined by the cyclic voltammetry (CV) method.Thus, this study is thought to be the first.
The stems (BS) of red-hot pepper (Capsicum annum L.), popularly known as ISOT, are grown and the dried pepper powder is produced from the Sanlıurfa stems (BS) of this pepper.They threaten the environment as both domestic and industrial waste.According to the statement by Şanlıurfa Vegetable Fruit Brokers Association, 10% of the estimated 250,000 tons of pepper used in the production of dry powdered pepper in this region is released into the environment as waste.For this reason, it was preferred as an adsorbent raw material because it was thought to contribute to the bioeconomy.
In this study, BS was used as plant extracts to produce copper (Cu-NP), iron (Fe-NP), or copper/iron (Cu/Fe-NP) adsorbents.NPs were used because they provide high surface area, homogeneity, and high stability to the adsorbent.Naproxen adsorption performance of the obtained adsorbents was examined under conditions such as solution pH, temperature, adsorbent amount, and concentration.Structural characterization of the adsorbents was performed by Fourier transform infrared spectroscopy (FT-IR), X-ray diffraction, thermal characterization by thermal gravimetric-differential thermal analysis (TG-DTA), and morphological characterization by scanning electron microscopy-energy dispersive X-ray spectroscopy (SEM-EDX) and Brauner-Emmet-Teller (BET) surface analyzer.Naproxen adsorption was monitored using the CV method, which is one of the electroanalytical methods.

Chemicals used
BSs used in the study were collected from the local manufacturer.Copper (II) sulfate pentahydrate (CuSO 4 •5H 2 O) and iron (III) chloride (FeCl 3 ), used as a source of NPs, were purchased from Merck.Naproxen was obtained from the drug tablet named Naprosyn, produced by Abdi İbrahim Company, a local pharmaceutical company.Acetonitrile (ACN; Merck) and purified water were used to prepare the solution.All solutions were kept in a cool place protected from light.

Preparation of pepper stem (BS) extract
(BS-ext.) The collected BS was washed thoroughly with distilled water and dried in an oven at 70°C.About 50 g of the dried BS was weighed and transferred to a 250 mL flask and distilled water was added to the level above it.It was boiled under a reflux system for 10 h to obtain the BS-ext.The resulting dark-brown extract was cooled and centrifuged at 1,100 rpm to separate very small particles.It was filtered with Whatman no. 1 filter paper for a clearer extract.The obtained clear extract was used immediately for green synthesis of the adsorbent-containing NPs.

Green synthesis of Cu-, Fe-, and Cu/Fe-NP-doped adsorbents
The synthesis of Cu-, Fe-, and Cu/Fe-NP-doped adsorbent is presented in Figure 1.A clear and fresh BS-ext.was added to 0.01 M CuSO 4 •5H 2 O solution at 70-80°C at a ratio of 4:1 by volume.It was kept under constant stirring for 2 h.The color of the solution, which was initially light blue, changed to yellowish-green and then to dark brownish-black over time (Figure 1).This color change was considered as an indication of NP formation [17].The mixture was left in the dark for 24 h.The mixture was centrifuged at 4,000 rpm for 6 min to separate the formed particles.The precipitated particles were washed with ethanol to avoid any residue.They were dried in an oven at 90°C for 3 h.The obtained adsorbent was labeled as Cu-NPs.Fe-NPs and Cu/Fe-NPs were obtained using the same method as that used for Cu-NPs.Here, 0.01 M Fe(III) chloride (FeCl 3 ) solution was used for the iron NP content: 0.01 M CuSO 4 •5H 2 O solution(10 mL) and 0.01 M FeCl 3 (10 mL) were used to prepare Cu/Fe-NPs.
Previously, it has been reported that antioxidants and phytochemicals in the structures of the extract captured and reduced Cu 2+ ions and, in turn, oxidized the alcohol functional groups to ketones [28].In light of this information, a reaction in which the alcohol groups in the capsaicin structure captured Cu 2+ ions and turned into ketones was proposed.The possible mechanism is given in reaction (1).
Accordingly, owing to the -OH group in the capsaicin structure in the BS-ext., it reduced Cu 2+ ions and resulted in the formation of abundant Cu NPs.

Characterization
Structural characterization of NP-doped adsorbents obtained from the BS-ext.by green synthesis was carried out by FT-IR spectroscopy using a Bruker Vertex 70 FTIR device in the wavenumber range of 4,000-400 cm −1 .The obtained data were given as wavenumber vs % transmittance.Thermal characterization by thermal gravimetric-differential thermal analysis (TG-DTA-Shimadzu DTG-60H Simultaneous DTA-TG apparatus), morphological characterization by scanning electron microscopy-energy dispersive X-ray spectroscopy (SEM-EDX) (ZEISS Evo/LS 10 electron microscope), and BET (Nova 2200e surface area and pore size analyzer) were performed.

Adsorption experiments
Naprosyn tablets were thoroughly ground, and 500 mg of the powder was weighed and dissolved in 10 mL of methanol.Then, the total solution volume was made up to 500 mL by adding ACN to prepare a stock solution of 1,000 ppm.The volume of the adsorption experiment solution was 50 mL.The studied concentration was prepared by diluting with ACN from the stock solution.For example, to prepare a 50 mL solution of 100 ppm, 5 mL of the stock solution was taken by a pipet and made up to 50 mL with ACN.
Naproxen adsorption was investigated based on the concentration (12.5, 25, 50, 75, 100, and 150 ppm), pH (2, 4.5, 7, 8.5, and 10), temperature (25, 30, 35, 40, and 45°C), and adsorbent dosage amount (0.01, 0.03, 0.05, 0.07, and 0.09 g) factors.The pH effect was investigated at acidic, neutral, and basic.For this aim, while pH values of 2, 4.5, and 7 for the neutral medium were selected in the acidic medium, pH values of 8.5 and 10 were selected in order to better observe the effect in the basic medium.In experiments, while each factor was examined within a certain range, the others were kept constant and the time-dependent change was monitored electrochemically.After adsorption with NPs from the 100 ppm naproxen solution at different pH values (2, 4.5, 7, 8.5, and 10) at 30°C, naproxen was recovered by using 50 mL of distilled water at the adsorption pH at 35 K for 2 h in a batch system [29].After adsorption, the adsorbents were filtered and the filtrate was subjected to electrochemical measurements to determine the naproxen concentration [30].

Electrochemical studies
Electrochemistry measurements were made using a triple electrode system.In this system, the Pt button electrode was used as a working electrode, the Ag wire electrode was used as a reference, and the Pt wire was used as a counter electrode.The reference electrode is set against Ce/ Ce 2+ .KCl solution (0.1 M) was used as the supporting electrolyte solution.For electrochemical measurements, after adding the adsorbent to the adsorption experiment solution, 20 µL of the solution was taken from this solution at 5,10,15,25,35,45,60,75,95,120,150, and 180 min time intervals with the help of a micropipette.To determine the current values depending on the amount of naproxen remaining in the solution, it was transferred to a CV cell containing 5 mL of the supporting electrolyte solution and mixed for 30 s.The potential range used for CV was −0.6 and 1.5 V, and the sweep rate was 100 mV/s.In CV measurements, two oxidation peaks of naproxen, in 0.2-0.4V and 1.3-1.5 V ranges, and two reduction peaks, 0.96 and −0.12 V, were observed.First, the highest current values at 1.5 V were chosen to determine the concentration.Then, a calibration curve was drawn between the current and concentration to determine the concentration of naproxen remaining in the solution after adsorption.The concentration was calculated using the equation of this curve.

Characterization of Cu-, Fe-, and Cu/Fe-NP-doped adsorbents
The FT-IR, EDX, and TGA-DTA graphs of the adsorbents doped with NPs are shown in Figure 2. The BS-ext shows special IR peaks due to its organic nature.A sharp O-H stretch peak (3,400-3,500 cm −1 ) and C-H asymmetric stretch (2,900-2,700 cm −1 ) peaks were also observed.In addition, C]C stresses, C]O stresses, and C-O stress peaks symbolized the BS-ext structure.In the structural analysis of Cu-, Fe-, and Cu/Fe-NPs, in addition to the peaks representing the organic structure, the O-H peak became weaker and wider, unlike the BS-ext, while a distinct Cu 2 O peak was observed at 621 cm −1 [31,32].These findings supported that the NP-filled adsorbent was formed.Moreover, the NP contents of adsorbents were verified by EDX images (Figure 2b, c, d, and e).Accordingly, 55.04% Cu (Figure 2c) was detected at 8 keV levels in the Cu-NP structure and 61.5% Fe was detected at 6 keV levels in the Fe-NP (Figure 2d) structure.For Cu/Fe-NP, both NPs (Cu and Fe) were detected (Figure 2e).In addition, C, K, Na, and O originating from the BS-ext.were also detected (Figure 2b).The TG-DTA curves of all adsorbents are shown in Figure 2f.A 60% mass loss was observed at 205°C in the BS-ex.This proves that there is no thermal structure degradation in the green synthesis process.However, overall, the percentage of thermal degradation remained lower for Cu-NP, Fe-NP, and Cu/Fe-NP at the end of 1,000°C due to the NP content.The mass losses were approximately calculated to be 98, 79, 90, and 94% for BS-ext., Cu-NP, Fe-NP, and Cu/Fe-NP, respectively.
The SEM images of all adsorbents are shown in Figure 3.
The surface morphological structure of the BS-ext. is given in Figure 3a.It was clearly seen that the surface of the material obtained as a result of green synthesis consisted of regular geometric shapes (Figure 3b and c).These changes in the morphological structure of the BE-ext.occurred in the presence of Cu and/or Fe particles and homogeneously distributed on the surface.It was determined that Cu causes more cubic geometric shapes in the structure, while Fe causes more acicular structures.The effects of both NPs can be observed together in Cu/Fe-NP (Figure 3d).

Adsorption experiments
Experiments were carried out at six different concentrations in the range of 12.5-150 ppm, including 12.5, 25, 50, 75, 100, and 150 ppm.Each concentration was prepared by diluting the Naproxen stock solution with ACN.Adsorption was carried out at 25°C with a 200 rpm shaking speed using 50 mL of the experiment solution in 200 mL flasks.About 20 µL of this solution was taken with the help of a micropipette (NICHIRYO brand Nichipet EXII "10-100 µL volume range") at the aforementioned time intervals and added to 5 mL of KCl electrolyte solution for electrochemical measurements.The measurements were repeated three times and the average value was used in the calculations.The calibration curve was drawn according to the Randles-Sevcik equation (equation ( 2)), which expresses the linear relationship between the current and concentration [33]: where I p is the peak current value (A), n is number of electrons is the electrode area (cm 2 ), D is the diffusion coefficient (cm 2 /s), C is the concentration (mol/cm 3 ), and ϑ is the scanning speed (V/s).For accurate calibration, the calibration curve was plotted by using the oxidation CV curves (0-1.5 V range) obtained at nine different values (150, 100, 75, 50, 25, and 12.5 ppm) in the 150-12.5 ppm concentration range of naproxen.The curves are shown in Figure 4a.A calibration chart was created by using the peak current values in these curves (Figure 4a, inset 1).The calibration curve of the graph in Figure 4 was obtained from equation (3), and an R 2 value of 0.999 (Figure 4a, inset 1).By using equation (2), concentration data were obtained, which corresponds to the current values read (Figure 4a, inset 2) from the solutions taken at different conditions and times.This value was used as the concentration of naproxen (C t ) remaining in the medium.Using these data, CV graphs of each adsorbent in each factor were plotted over time.Since the electrode type and solution content did not change, the characteristics of the CV plots were the same.Only the change in the solution pH caused the characteristics of   the CV curves to change as well.All these graphics are given in the Supplementary Information (SI).Equation ( 4) was used to reach the adsorption equilibrium values: where q e is the amount of substance in equilibrium (mg/g), C 0 is the initial concentration of naproxen (mg/L), and C t is the equilibrium concentration of naproxen in solution (mg/L) at time t of adsorption.V represents the adsorption solution volume (mL) and m represents the amount of adsorbent used (mg).
Basically, the breakdown of a drug molecule through oxidation depends on the presence of OH radicals in the environment.While the oxidant in the environment (this can sometimes be a chemical, sometimes a ray, or the oxidizing voltage) is adsorbed by the NP, it takes electrons from there and forms OH radicals.Meanwhile, the OH − ion is oxidized.The radicals formed to cause the decomposition of the molecule adsorbed on the surface.As a result of the reaction, CO 2 and H 2 O are formed.The rate of this reaction depends on the electron transfer of the NPs.The possible degradation mechanism occurs with the formation of hydroxyl (HO˙) radicals in the presence of H 2 O, OH − , and oxidants in the environment.[34,35] This situation is illustrated in Figure 4b.The possible electrochemical degradation mechanism and the reaction of NAP are shown in reactions ( 5) and ( 6): [36] 2 2 ˙(6)

Changes in adsorption with time in the adsorbent obtained using different NPs
The adsorption results using the BS-ext., Cu-NP, Fe-NP, and Cu/Fe-NP are shown in Figure 5, and adsorbents containing different NPs are compared to observe the adsorption efficiency obtained.The BS-ext.showed the lowest adsorption efficiency while Cu/Fe-NP showed the highest adsorption efficiency.Since the BS-ext.has a completely rigid, tight, and non-porous structure, and low adsorption efficiency was an expected result.Since there may be interaction with pi electrons between Cu NPs and functional groups in the naproxen structure, it was likely to be adsorbed on the NP-filled adsorbent surface [37].The same approach can be observed for Fe-NP.The related graphics are shown in Figure 5c.In Figure 5d, the comparison of the equilibrium values is shown.While the adsorption efficiency was 3.3% with BS-ext., it was 46.142% with Cu-NP, 57.622% with Fe-NP, and 82.406% with Cu/Fe-NP.It was clear that Cu and Fe NPs BS-ext.possessed active sites suitable for adsorption, and the two NPs exerted a synergistic effect together and formed more active sites.

Effect of solution pH
It was expected that the presence of NPs increased the adsorption efficiency.However, as excess ions such as H + and OH − in the environment will affect the electrostatic attraction between the adsorbent and the adsorbate, the adsorption process proceeds depending on the pH of the solution.Thus, the adsorption of the naproxen molecule was investigated by adjusting the ambient pH to 2, 4.5, 7, 8.5, and 10.The obtained data are shown in Figure 6a for each pH value as percent separation and adsorption capacity.As expected, higher adsorption occurred in the acidic and neutral media compared to the basic medium.Considering that the pH of the naproxen solution is around 4.5-5, it was observed that the best adsorption pH was its solution medium.The highest adsorption efficiency and adsorbent capacity were observed at pH 4.5 for all adsorbents.At this pH, 82.4% removal and a capacity of 165 mg/g were achieved with Cu/Fe-NP.However, these values were 57,622% and 116 mg/g for Fe-NP and 46% and 92 mg/g values for Cu-NP.The removal percentage and adsorption capacity at higher pH values were 50% and 100 mg/g for Cu/Fe-NP, respectively, and were 35% and 70 mg/g for Fe-NP and 28% and 56 mg/g for Cu-NP.
A decrease was observed from each adsorbent in both adsorption capacity and percent removal at high pH.The reason for this could be that the excess OH − ions in the adsorption medium prevented the interaction between the naproxen molecules and the NPs on the adsorbent.In addition, at high pH values, NPs may cause leaching from the surface.Thus, the surface could become unstable and inefficient [38].
The variations in adsorption capacity and removal percentage at different concentrations (12.5, 25, 50, 75, 100, and 150 ppm) are shown in Figure 6b.It was expected that the adsorption capacity and removal percentage would increase with increasing concentration because concentration played a driving force in the movement of ions in the liquid phase toward the solid surface.In addition, the structure of the molecule to be adsorbed, the presence of active sites of the surface to be adsorbed, and the electrostatic attraction increase the adsorption efficiency and adsorption capacity.
According to the data obtained, the removal of naproxen with Cu/Fe-NP increased from 47 to 88%, while the adsorption capacity increased from 12 to 264 mg/g.These values increased from 33 to 63% and from 8.22 to 184 mg/g for Fe-NP; and from 26.25 to 49% and from 7 to 147 mg/g for Cu-NP, respectively.In the literature, it was observed that adsorption decreases with the concentration in adsorption processes with naproxen.In this study, an increase was observed between increasing concentrations but this increase was low due to the increase rate.The reason for this could be due to the lack of active sites on the surface and the electrostatic interaction between these sites and the molecules.
The type of adsorption with activated carbon and its derivatives was observed to be physical adsorption.Since heat is released to the environment during the adsorption of the molecule on the surface by the active sites, the increase in the temperature did not affect the adsorption in physical adsorption, since mass transfer was prevented at high temperatures [39].This effect can be clearly observed in Figure 6c.It showed an adsorption capacity of 164.73 mg molecules per gram of Cu/Fe-NP with a removal of 82.36% at 25°C.The removal reached 82.58% with a 0.22 increase at 45°C and the adsorption capacity reached 165.16 mg/g with an increase of 0.43 mg/g.The amount of increase observed in Cu/Fe-NP due to the NP difference was determined to be 0.32 mg/g for 0.16% adsorption capacity for the removal with Fe-NP, and the increase in removal for Cu-NP was 0.03% for an adsorption capacity 0.06 mg/g.This behavior was compatible with physical adsorption.

Isotherm studies
Isotherms are very useful equations used to understand and explain the connection between the mass transfer and transition of the adsorbed molecules to the adsorbent active surface in the adsorption event.Important constants reflecting these isotherms were calculated by using isotherms such as Freundlich [40], Langmuir [41], Halsey [42], Jovanovic [43], and Dubinin-Radushkevich [44], which are frequently used to explain the characteristics of the adsorption phenomenon on the surface of the adsorbent.The purpose of using the Langmuir isotherm is to know whether the adsorption process on the adsorbent surface is in a single layer if the results obtained comply with this isotherm, and whether the adsorbent surface has a homogeneous structure.The equation used was where C e is the equilibrium concentration (mg/L), q e is the adsorption capacity (mg/g), and K L is an adsorption energy constant (L/mg).These values were reached with the help of the graphs drawn from the relevant data and are given in Table 1.In addition, R L , which is the equilibrium parameter depending on the initial concentration (C o ) and the K L value, expresses the applicability of the isotherm, and is given as follows: This gives information that the applied isotherm and adsorption conditions are unfavorable if the R L value is greater than 1, linear if it is equal to 1, favorable if it is between 1 and 0, and irreversible if it is equal to 0 [45].It is accepted that the adsorbent surface used in the adsorption events to which the Freundlich isotherm fits is heterogeneous and the molecule is adsorbed on these scattered active sites.The equations representing the Freundlich isotherm are as follows: and These equations give the amount of equilibrium substance (q e , mg/g) adsorbed on the adsorbent surface used in adsorption given its mass.C e (mg/g) depends on the equilibrium concentration and from the graph drawn by taking the logarithm, the Freundlich constant (K F ) expressing the adsorption capacity and the data expressing the degree of adsorption efficiency (n) are obtained.Naproxen adsorption on adsorbents obtained from pepper stalk extract  11 The Halsey isotherm is used for multi-layer adsorption events on the heterosporous adsorbent surface, and is expressed as follows: k H and n are the Halsey isotherm constants and are obtained from the graph drawn using the equilibrium adsorption capacity (q e , mg/g) and the final concentration (C e , mg/L) of the adsorbent.
However, Jovanovic's isotherm suggests that some mechanical interactions between the adsorbents and molecules should be taken into account in addition to some of the assumptions of the Langmuir model.The expression for this isotherm is as follows: This equation relates the equilibrium adsorption capacity (q e , mg/g) to the final concentration (C e , mg/g).The Jovanovic constant K J can be found in the graph of C e drawn against ln q e .The mean adsorption energy is defined as the free energy required to adsorb one mole of adsorbent from the solution onto the solid adsorbent surface, and these data provide information about whether the adsorption behavior is chemical or physical [44].
In this study, the DR isotherm was used to calculate the E.This isotherm is expressed by the following equation : where q e and q s are the adsorption capacities (mg/g) at equilibrium and the saturated state, respectively: Ɛ is a value called the Polanyi potential (kJ 2 /mol 2 ) [46], R is the gas constant (J/mol K), T is the temperature (K), and C e is the equilibrium concentration (mg/L).If equation ( 10) is linearized, K is found from the slope of the resulting equation.K is a constant that depends on the adsorption energy and is used in the following equation to find the mean energy: The graphs obtained from all three adsorbent data in this study are shown in Figure 7.The DR characteristic data and E obtained from the calculations are given in Table 1.Accordingly, the calculated energy values are 0.105 kJ/mol for Cu/Fe-NP, 0.215 kJ/mol for Fe-NP, and 0.317 kJ/mol for Cu-NP.Since it is known that the parameter E < 8 kJ/mol in physical adsorptions and E > 8 kJ/mol in chemical adsorptions [41], NAP adsorption occurs physically with these adsorbents in this study.
When the R 2 values of the curves were examined and the adsorption of the Naproxen molecules with Cu/Fe-NP, Fe-NP, and Cu-NP was compared, it was seen that the most compatible isotherm was Freundlich and, another version of it, Halsey.According to the calculations, R 2 values for each adsorbent in Freundlich and Halsey were in the range of 0.998-0.997.Accordingly, the adhesion of the naproxen molecules on the adsorbent surface under these conditions was realized by a multi-layer adsorption process and the adsorbent surface had a heterogeneous structure.In fact, there was a homogeneous distribution of NPs on the surface, as observed from SEM mapping images, but this harmony can be interpreted as the clustering of some NPs caused the surface to behave as if it was a heterogeneous surface, and the adsorption took place in a multilayered manner.

Kinetic studies
Figure 8 shows the adsorption capacity (q t ) and percent removal graphs of the naproxen molecules at time t in comparison with four different adsorbents.Adsorption varied with time and increased rapidly in about 20 min, and then equilibrium was reached in the range of about 30-35 min.However, it is observed that adsorption continues in very small percentages due to the presence of NPs, the presence of active sites formed on the surface, and multilayer adsorption.
In the adsorption event, the molecules were adsorbed on the adsorbent surface by mass transfer.In order to better understand the mechanism of this phenomenon, adsorption kinetics were examined.The Lagergren [47] approximation for pseudo first order, .
The Ho-McKay [48] approach for pseudo-second-order, the Elovich [49] approach, and the Weber-Moris [50] approach for the intraparticle diffusion mechanism, were used in calculations as adsorption kinetics.
In electrochemical analyses, whether the process that occurs during the interaction of the electrode and the substance in the medium is diffusion-controlled or not was determined by the Randles-Sevcik equation.Here, the oxidation and reduction current values corresponding to the different scan rates were plotted against the square root of the scan rate.As the R 2 value of the curves approached 1, it was inferred that the electrochemical analysis was diffusion-controlled. Accordingly, the graphs obtained are shown in Figure 9.
When the graphs drawn from the obtained data were examined, it was found that the adsorption process from the R 2 values of the curves complied with the pseudosecond-order kinetics (Figure 9b).These kinetic curves had an R 2 value of 1 for all adsorbents.It is observed from the graph curves (Figure 9d) that the adsorption phenomenon between adsorbents and naproxen does not comply with intraparticle diffusion.However, when examined in detail, it was found that it followed intraparticle diffusion in two stages (Figure 9e), the first 20 min and the next 100 min.
The parameters obtained from the curves and equations used in kinetic studies and the kinetic mechanism are given in Table 2.
Thermodynamic data were determined by the following equation, depending on the change in enthalpy (ΔH), entropy (ΔS), and Gibbs free energy (ΔG): K c was found as the ratio of the naproxen concentration in the medium to the equilibrium concentration.When the equation, also known as the Van't Hoff equation, was plotted for different temperatures, enthalpy and entropy changes were obtained from the slope and the cut of the equation.From here, the Gibbs free energy was calculated.The values obtained by using five different temperatures in this study are shown in Figure 10.
When the curves of the graphs were examined, the enthalpy, entropy, and free energy values of the naproxen adsorption process were calculated for each adsorbent from the equations of the curves, whose R 2 values were determined to be 0.9984, 0.9979, and 0.999 for each Cu/Fe-NP, Fe-NP, and Cu-NP adsorbent, respectively.These data are given in Table 3.
The enthalpy value for Cu/Fe-NP was calculated to be 9.92 kJ/mol, the entropy value was 9.11 J/mol-K, and the free energy was −7.21 kJ/mol.These values are 10.56 kJ/mol, 9.96 J/mol-K, and −7.58 kJ/mol for Fe-NP, and were 7.39 kJ/mol, 6.78 J/mol, and −5.37 kJ/mol for Cu-NP, respectively.Since the Gibbs free energy is in the range of 0 and −20 kJ/mol, the Naproxen adsorption by NPs is physical adsorption.In addition, with the enthalpy change below 40 kJ/mol, it also revealed that the adsorption was physical.

Recovery experiment
The reuse of the adsorbent used in the technique of separating substances such as heavy metals, dye, or drugs from the medium is very important in terms of both economy and practicality.For the recovery experiment, naproxenadsorbed NPs were regenerated in the same pH solutions at room temperature for 2 h with continuous stirring.In addition, the amount of naproxen recovered was calculated.Figure 11 shows the naproxen concentration (Figure 11a) and percent recovery (Figure 11b) obtained as a result of three reuse cycles of three different adsorbents at different pH values.While obtaining the percentage recovery, the amount of naproxen adsorbed at the same pH value was taken as the basis.
According to the data obtained, all NPs (except for Cu-NP) had high recovery percentages in three cycles, while this rate dropped to 80% at high pH.For Cu-NP, the recovery decreased at pH 8.5 and 7, but surprisingly it increased at pH 10.However, the performance of Cu-NP  was weaker than the other two NPs at about 77%.Although desorption is described, it was actually thought that the desorbed substance was adsorbed in the first place.Thus, the recovery process of naproxen in a sense ensured the regeneration of the adsorbents.As a result, NP plays a good role in adsorbent development, of which Fe and Cu NP-doped BS adsorbents had better effect on naproxen adsorption and were considered suitable adsorbents for several times use.

Conclusion
Naproxen adsorption was carried out with adsorbents filled with Cu, Fe, and Cu/Fe NPs obtained by the green synthesis method from pepper stem waste.Cu-NP formation was supported by the decrease and shift in the intensity of the broader and other functional group peaks in the BS-Cu spectrum, the sharp OH peak observed in the extract spectrum, and the characteristic Cu 2 O vibration peak at 621 cm −1 .NP contents of adsorbents were verified by EDX images.Accordingly, 8.44% Cu was detected at 8 keV levels in the Cu-NP structure, and 7.05% Fe was detected at 6 keV levels in the Fe-NP structure.The NPs of Cu-NP and Fe NP adsorbents were visualized at 200 nm size and dispersed on the surface with different regular geometrical formations, as determined by SEM.The amount of naproxen not adsorbed in the medium was determined by electrochemical analysis methods such as CV and square wave.To determine the amount of unknown concentration, a calibration curve was drawn from the current values corresponding to the specific concentration.The calibration curve of the graph was obtained with the equation = − y x 2.6165 288.22 and an R 2 value of 0.999.While the adsorption efficiency was 3.3% with BS-ext., it was 46.142% with Cu-NP, 57.622% with Fe-NP, and 82.406% with Cu/Fe-NP.The highest adsorption efficiency and adsorbent capacity were observed at pH 4.5 for all adsorbents.At this pH, 82.4% removal with Cu/Fe-NP and a capacity of 165 mg/g were achieved.These values were 57.622% and 116 mg/g for Fe-NP and 46% and 92 mg/g values for Cu-NP, respectively.
For Cu/Fe-NP, when the concentration increased from 12.5 to 150 ppm, the removal increased from 47% to 88%, and the adsorption capacity from 12 to 264 mg/g.For Fe-NP, it increased from 33 to 63% and from 8.22 to 184 mg/g, while for Cu-NP it increased from 26.25 to 49% and from 7 to 147 mg/g.By contrast, temperature did not affect the adsorption process since physical adsorption took place, and the adsorption data remained constant.
As the R 2 values for each adsorbent were in the range of 0.998-0.997 in the isotherm calculations, it was found that the Freundlich and Halsey isotherms were compatible.Accordingly, the adhesion of the naproxen molecules on the adsorbent surface under these conditions occurred with a multi-layer adsorption process and it was found that the adsorbent surface has a heterogeneous structure.
When the graphs drawn from the kinetic data were examined, it was determined that the adsorption process from the R 2 values of the curves complied with the pseudosecond-order kinetics.The R 2 values of these kinetic curves were 1 for all adsorbents.
Thermodynamic data were determined using Van't Hoff curves with R 2 values in the range of 0.998-0.999.The enthalpy value for Cu/Fe-NP was calculated to be 9.92 kJ/mol, the entropy value was 9.11 J/mol-K, and the free energy was −7.21 kJ/mol.These values were 10.56 kJ/ mol, 9.96 J/mol-K, and −7.58 kJ/mol for Fe-NP; and 7.39 kJ/ mol, 6.78 J/mol-K, and −5.37 kJ/mol for Cu-NP, respectively.Since the Gibbs free energy is in the range of 0 to −20 kJ/ mol, Naproxen adsorption by NPs is physical adsorption.In addition, with the enthalpy change below 40 kJ/mol, it was also revealed that the adsorption was physical [51].
In this study, BS-ext shows a low percentage of adsorption and can reach high adsorption values with Cu and/or Fe NPs.It was observed that it is possible to reach a higher adsorption percentage by increasing the amount of adsorbent.
Funding information: This research received no external funding.
Author contributions: The author confirms sole responsibility for the following: study conception and design, data collection, analysis and interpretation of results, and manuscript preparation.
Linear fit of Cu-NP Linear fit of Fe-NP Linear fit of Cu/Fe-NP

Figure 11 :
Figure 11: (a) The naproxen concentration and (b) the percent desorption obtained as a result of three reuse of three different adsorbents at different pH values (adsorption conditions: 30°C, 100 ppm, and 0.05 g adsorbent; CV conditions: 100 mV/s, −0.6 to 1.5 V, and 0.1 M KCl/ACN electrolyte solution).

Table 3 :
Gibbs free energy data of the adsorbents