Introduction

Water pollution is one of the most important issues around the world because clean water is essential for both life and economic development. In recent years, by increasing the activity of agricultural, industrial, and urban, the contamination level has been enhanced in the environment and especially in aquatic bodies. Dyes, pesticides, fertilizers, pharmaceuticals, organic material, nutrients, oils, and heavy metals are common contaminants (Castro et al. 2021). Wherever there are humans, drugs enter environments such as aqueous media, soil, and sediments because they can be resistant to degradation at wastewater treatment plants (WWTPs) (Mrozik and Stefanska, 2014). Also, part of the drugs enter the wastewater stream after human consumption (Angeles et al. 2020). Low concentrations of drugs are dangerous to humans, animals, and aquatic organisms (Liakos et al. 2021).

One of the pharmaceutical contaminants is Sitagliptin (STG) (Fig. 1a), which was the first dipeptidyl peptidase-4 (DPP-4) inhibitor for treating diabetes mellitus (DM). By the inhibition of DPP-4 and increasing insulin secretion, a decrease of glucagon-like peptide-1 (GLP-1) breakdown occurs, which leads to a decrease in blood sugar level (Kit Loh et al. 2020; Gumieniczek et al. 2019). Some studies showed its chemical stability in stressed conditions such as high temperature, acidic and basic pH (Krakko et al. 2022). Owing to its high persistence and increasing consumption, it is discharged in significant quantities by wastewater treatment plants (WWTPs) with concentrations up to 1.0 μg/L in WWTP effluents (Hermes et al. 2020).

Fig. 1
figure 1

Chemical structure of a STG and b FMP

Full size image

The other drug as contaminant in aqueous media is Fampridine (FMP) (Fig. 1b), which is prescribed to improve walking disorders in patients suffering from multiple sclerosis (MS) (Suneetha and Raja Rajeswari 2016; Achanti and Rajeswari Katta 2016). It is a selective potassium channel blocker, which increases the signal transfer in demyelinated nerve fibers. Therapy with prolonged-release (PR) FMP increases walking speed and stability, as well as it can create a better balance (Weller et al. 2020). The existence of this drug in the environment can lead to the degradation of surface and subsurface water quality, consequently impacting the health of living organisms. Also, its effects on aquatic populations can be observed (Bankole et al. 2023).

The presence of pharmaceuticals in aquatic environments for a long time has harmful effects on human health such as cancer, mutation, allergic reactions, and drug resistance (Mahmoud et al. 2020). Therefore, it is important to remove drugs from aqueous ecosystems.

Several procedures, including ozonation (Hermes et al. 2020), advanced oxidation (Almomani et al. 2016; Kanakaraju et al. 2018), coagulation (Gou et al. 2017), of coagulation/flocculation (Kooijman et al. 2020), nanofiltration (Gomes et al. 2020), ultrafiltration (Baresel et al. 2019), and adsorption (Dhiman et al., 2019) have been reported to eliminate types of drugs from aqueous solutions. Among the mentioned methods, adsorption is simple, affordable, high efficiency, and environmentally friendly, which has been extensively used to remove contaminants (Abolhasani et al. 2019; Singh et al. 2017). Bioadsorbents possess some advantages such as inexpensive, easy access, high performance, good adsorption capacity, easy regeneration, and eco-friendly (Mallampati et al. 2015; Pathak et al. 2015). Lemon peel (Bhattacharyya et al. 2019), potato peels (Kyzasa and Deliyanni 2015), rice husk ash (Swarnalakshmi et al. 2018), tea waste (Patil et al. 2019), and others are various bioadsorbents, which have been used to eliminate pharmaceutical contaminants.

Kiwifruit belongs to the Actinidia genus which includes diverse species and cultivars with several combinations. This small meaty green fruit contains small black seeds, and a brown skin cover the fruit. Kiwi peel has mainly carbohydrates, as well as the minerals with the maximum value are Mg (8200 ppm), K (2300 ppm), Ca (2300 ppm), Na (900 ppm), P (600 ppm), Fe (82.26 ppm), Cu (6.64 ppm), Zn (9.26 ppm), and Mn (14.83 ppm). There is phenolic content along with chlorogenic acid, protocatechuic acid, p-hydroxybenzoic acid, caffeic acid, quercetin, and rutin in the kiwi peel (Sanz et al. 2021). Therefore, its peel has diverse functional groups, which may be applied as an adsorbent for the removal of drugs from aqueous media. Gubitosa et al., used Kiwi peels as an adsorbent for the removal of propranolol from water (Gubitosa et al. 2022a, b). In the other study, Kiwi peels were used as a long-lasting and recyclable adsorbent to eliminate emerging pollutants such as ciprofloxacin (Gubitosa et al. 2022a, b). Owing to its low specific gravity and trend to agglomerate, the separation of this bioadsorbent is an important problem. These are limitations for wide usage for the batch and column processes. A proper method is a magnetic separation that solid particles can be separated using an external magnetic field (Jiang et al. 2015). The most magnetic substance in nature is Fe3O4, which is very beneficial in the adsorption approach (Akpomie and Conradie 2020). The magnetic feature makes it easy to separate the adsorbent through an external magnet (Abbaszadeh et al. 2017).

In this study, the magnetic kiwi peel was prepared as a new, simple, environmentally friendly, easy availability, and inexpensive (without using any sophisticated equipment) bioadsorbent for the removal of STG and FMP from aqueous solutions. This is the first paper to use the magnetic kiwi peel to evaluate the STG and FMP removal from aqueous solutions. The use of cheap materials and simple laboratory conditions in this synthesis makes it significant and affordable for mass production. The characterization of this adsorbent was assessed by scanning electron microscope (SEM), transmission electron microscopy (TEM), Fourier-transform infrared spectroscopy (FTIR), X-ray diffraction analysis (XRD), and vibrating-sample magnetometer (VSM). Statistical optimization of empirical parameters, which are effective on the peak area values was surveyed with central composite design (CCD) as a subset of response surface methodology (RSM). Moreover, the isotherm (Langmuir and Freundlich) and kinetic models (pseudo-first-order, pseudo-second-order) were studied.

Experimental

Materials

STG and FMP with a purity of 99.4% were prepared from Stereokem and Procos companies, respectively. Potassium dihydrogen phosphate (KH2PO4), ammonium acetate (NH4CH2CO2), and sodium 1-octanesulfonate (C8H17NaO3S) were purchased from Merck.

Apparatus

The surface morphology of the magnetic kiwi peel was evaluated using Sigma VP-500 SEM (TSCAN company, Czech). In the SEM method, kiwi peel powder was fixed on a carbon layer. Then, a gold layer (with an amount of Angstrom) was placed on it, and the photo was taken. In order to determine the size of the adsorbent, TEM (Zeiss company, Germany, EM 10C–1000 kV) was applied. FTIR (PerkinElmer, U.S.A) at a wavenumber range from 4000 to 400 cm−1 was used to study the functional groups of the adsorbent. The XRD (Malvern Panalytical company, United Kingdom) was utilized to record the XRD pattern of Fe3O4 and magnetic kiwi peel at the 2θ range from 10° to 80°.

Preparation of magnetic adsorbent

6.1 g iron(III) chloride hexahydrate (FeCl3.6H2O) and 4.2 g iron(II) sulfate were dissolved in 100 mL of distilled water. This solution was transferred to a round-bottom flask and its temperature increased to 90 °C using a heater stirrer. Then, 1 g of kiwi peel powder was dispersed in 200 mL distilled water. Afterward, 10 mL NaOH (0.5 M) and dispersed kiwi peel were poured into the round-bottom flask, respectively. Using NaOH (0.5 M), the pH was set to 10 and exposed to a temperature of 80 °C for 30 min. After cooling, it was transferred to a beaker and the precipitate was collected using a magnet, as well as the supernatant was discarded. The precipitate was dried in an oven at 40 °C for 24 h and was powdered in a mortar.

Standard solution preparation

1000 mL solution of STG and FMP with concentration of 30 μg mL−1 was prepared separately (stock solutions). Then, the standard solutions with concentrations of 25, 20, 15, 10, and 5 μg mL−1 were individually prepared from stock solutions. 5 mg of adsorbent was added to one of the parts. The solutions were injected into the HPLC instrument for plotting calibration curve.

Preparation of real sample

STG and FMP products were produced in the form of tablets at Alhavi Pharmaceutical Company. After cleaning and washing the production line, sampling of rinse water obtained from washing was done at 3 points. Then, drug removal was done by magnetic NPs.

Chromatographic conditions

Chromatographic analysis was accomplished with an HPLC system (Agilent 1200 series) with a C18 column (25 cm × 4.6 mm, 5 μm). The mobile phase related to the STG and FMP was phosphate buffer, methanol, acetonitrile (60:30:10 v/v/v) and phosphate buffer, methanol (70:30 v/v), respectively. The mobile phase flow rate was adjusted to 1 ml/min for both components with the injection volume of 20 μL. The STG and FMP wavelength was set at 268 and 240 nm, respectively.

Removal experiments

The removal of STG and FMP was assessed by magnetic kiwi peel, and estimation of removal percentage was accomplished at various concentrations of pollutant (1–1500 μg/mL) with a volume of 25 mL, time between 1 and 45 min, different peak areas, the adsorbent dosage of 15 mg and pH of 9.

Central composite design (CCD)

A single-factor experimental design is commonly performed by researchers, which is named as one-factor-at-a-time (OFAT) method. In this approach, only one variable is changed at a time and the other variables are constant. The experimental ranges can be obtained for each parameter without eliminating any conditions, which is the advantage of the OFAT method. However, this technique is time-consuming and it does not represent the interaction effects among the factors, which makes it difficult to achieve optimal safety conditions. Therefore, RSM can be proposed as a statistical method (Amirah Mohd Zahri et al. 2021). The CCD is a powerful form of RSM, which investigates the factors affecting the experiment (Rao Amarachinta et al. 2021). The benefit of this approach is time-saving through screening of more factors (Elshafei et al. 2021).

In this study, the interaction between pH, contact time, and nano dosage was assessed using CCD and 17 experiments were obtained using Eq. (1).

$$N = 2^{n} + 2n + n_{c}$$
(1)

where n and nC denote the number of independent factors and number of center points. Also, the total of experiments is shown by N (Bayuo et al., 2020). The factor levels of STG and FMP are given in Table 1. The predicted values were determined using a quadratic model as represented in Eq. (2).

$$Y = \beta_{0} + \mathop \sum \limits_{i = 1}^{k} \beta_{i} X_{i} + \mathop \sum \limits_{i = 1}^{k} \beta_{ii} X_{i}^{2} + \mathop \sum \limits_{i = 1}^{k} \mathop \sum \limits_{i \ne j = 1}^{k} \beta_{ij} X_{i} X_{j} + \varepsilon$$
(2)

where Y demonstrates the response value. β0, βi, βii, and βij depict the constant coefficient, linear, quadratic, and interaction effects, respectively. The independent variables are shown by Xi and Xj, as well as ε and k indicate the error and the number of factors, respectively (Farhadi et al. 2021).

Table 1 The levels of pH, adsorbent dosage, and time in CCD model for STG and FMP
Full size table

Results and discussion

Characterization

SEM analysis

SEM image of powder kiwi peel (Fig. 2a) displays spherical shape. Figure 2b reveals the SEM image of the kiwi peel after modification with Fe3O4. The magnetic Fe3O4 is dispersed on the peel surface and the agglomerated morphology with a spherical structure is observed. This image signifies the homogeneous distribution of spherical nanoparticles.

Fig. 2
figure 2

SEM image of a powder kiwi peel before adding Fe3O4 and b magnetic kiwi peel

Full size image

TEM analysis

The TEM image of kiwi peel before adding Fe3O4 is exhibited in Fig. 3a. No specific agglomeration is observed. On the other hand, the prepared magnetic kiwi peel has an agglomerated spherical morphology, which indicated that Fe3O4 particles are located on the peel surface (Fig. 3b). The aggregation of NPs is due to their superparamagnetic nature.

Fig. 3
figure 3

TEM image of a powder kiwi peel before adding Fe3O4 and b magnetic kiwi peel

Full size image

FTIR analysis

The FTIR spectrum of kiwi peel is shown in Fig. 4a. The band at 3444 cm−1 corresponds to hydroxyl group. The absorption peak at 2917 cm−1 indicates the functional group C–H. The peak at 1733 cm−1 is stretching vibration of carboxyl (COOH) group. The peaks at 1247 cm−1 and 1045 cm−1 are assigned to the C–C and C–H bending, respectively (Sartape et al. 2017). As illustrated in Fig. 4, the broad peak at 3424 cm−1 is related to the hydroxyl groups on the magnetic adsorbent surface. The peak at 2923 cm−1 represents the stretching vibration of the C–H (Jiang et al. 2015; Weng et al. 2018). The absorption band at 2899 cm−1 confirms the existence of the –CH2 group (Singh et al. 2017). The peak at 1747 cm−1 belongs to the stretching vibration of C = O of ester or carboxylic acid functional groups (Mallampati et al. 2015). The absorption peak at 1639 cm−1 shows the stretching vibration of C = O for carboxylic groups. The observed peak at 1384 cm−1 is attributed to –CH3, which is specific. The sharp peak at 582 cm−1 is due to the presence of the Fe–O–Fe bond of Fe3O4 (Abbaszadeh et al. 2017; Pashazadeh-Panahi and Hasanzadeh 2020).

Fig. 4
figure 4

FTIR spectrum of a kiwi peel powder before adding Fe3O4 and b magnetic kiwi peel

Full size image

XRD analysis

The observed peaks at 2θ = 29.89°, 35.36°, 43.19°, 53.43°, 57.14°, and 62.82° (Fig. 5a) are assigned to the (220), (311), (400), (422), (511), and (440) planes. These peaks are similar to the peaks in the standard pattern of magnetite Fe3O4 (JCPDS 19–0629) (Chang Lin and Yen Lee 2020; Liu et al. 2020). The XRD spectra of kiwi peel before adding Fe3O4 (Fig. 5b) showed a sharp peak at 2θ value around 25º. In Fig. 5c, a similar pattern of Fe3O4 can be observed, where the crystallinity corresponding to the magnetic kiwi peel remained the same. It can be stated that the formation of this adsorbent does not indicate any effect on the magnetic property of Fe3O4.

Fig. 5
figure 5

XRD pattern of a Fe3O4, b kiwi peel before adding Fe3O4, and c magnetic kiwi peel

Full size image

VSM analysis

The magnetic feature of Fe3O4 and magnetic kiwi peel was assessed by VSM (Fig. 6). The magnetization of Fe3O4 and magnetic kiwi peel was found to be 65.58 and 45.45 emu/g, respectively. These results reveal that the kiwi peel was successfully magnetized by coating with Fe3O4 NPs. The decline in the magnetization of the kiwi peel may be assigned to the decreased ferrite content in the kiwi peel and the formation of an outer layer of metal oxides over the surface of the Fe3O4 core.

Fig. 6
figure 6

VSM analysis of Fe3O4 and magnetic kiwi peel

Full size image

Removal results

The removal percentage of STG and FMP was calculated (Table 2). As shown, the maximum removal of STG (69.04%) was assigned to the concentration of 1 mg/mL and a peak area of 164.93. By increasing the concentration of STG from 1 to 1500 μg/mL, the removal percentage using the adsorbent was decreased. The diminish in the removal efficiency can be assigned to the less active sites in high concentrations, which prevent the drug from binding completely to the adsorbent surface (Rani and Shanker 2018). Also, the FMP concentration of 50 mg/mL with the peak area of 1714.85 represented the maximum removal percentage (63.76%). At lower concentrations of FMP up to 50 μg/mL, the specific surface area and abundant reactive surface center are high, which are more available to the target molecules. On the other hand, a decrease in the mass transfer occurs at concentrations higher than 50 μg/mL. Hence, there is a lack of available unoccupied sites of adsorbent.

Table 2 The obtained removal percentage related to the STG and FMP
Full size table

Results of CCD

The implementation of the CCD approach was performed to evaluate the effect of pH, nano dosage, and time on the peak area value of STG and FMP (Table 3 and Table 4). As shown in these tables, the actual and predicted peak area values were compared with each other. The maximum peak area of the STG (actual = 2170.1, predicted = 2170.21) was assigned to the pH = 7.8, nano adsorbent of 15.5 (mg), and contact time of 33.5 min. On the other hand, the highest peak area of FMP (actual = 1932.5, predicted = 1901.11) was related to the obtained values at the pH = 7.8, nano adsorbent of 11.8 (mg), and contact time of 33.5 min. In order to determine peak area, a quadratic equation was applied for STG and FMP using Eqs. 3 and 4, respectively.

$$\begin{aligned} {\text{R1}} = & {1946}.0{7155}.{\text{32 A}} + {58}.{\text{63 B}} + {5}0.{\text{48 C}} + {126}.{\text{97 AB}} \\ \quad + {57}.00{\text{ AC}} + {33}.{\text{36 BC125}}.{\text{88 A}}^{{2}} + {42}.{\text{31 B}}^{{2}} {56}.{\text{42 C}}^{{2}} \\ \end{aligned}$$
(3)
$${\text{R1}} = 1901.11 + 3.48 {\text{A}}16.{\text{19 B1}}0.{\text{19 C1}}00.80{\text{ A}}^{{2}} 19.{\text{86 B}}^{{2}} 71.90{\text{C}}^{{2}}$$
(4)

where A is the pH, B is the nano dosage (mg), and C is the time (min).

Table 3 CCD results and interaction between factors related to the STG
Full size table
Table 4 CCD results and interaction between factors related to the FMP
Full size table

Statistical analysis

Analysis of variance (ANOVA) results related to the STG and FMP are given in Table 5 and Table 6. The results revealed the significance of the model where their F-value and P-value are 287.46 = STG, 55.17 = FMP and < 0.0001, respectively. Furthermore, the “Prob > F” value is less than 0.05 indicating the significance of the model terms. Based on the p-values of STG ( p < 0.05), all the linear coefficients, interaction coefficients, and quadratic coefficients were significant. On the other hand, the values higher than 0.05 show the insignificance of model terms. In the current case, the nano dosage (B) is significant model terms. “Adeq Precision” is known as one parameter to study signal-to-noise ratio. This value is proper where the ratio is higher than 4. In current models, the ratio values of STG and FMP were found to be 60.817 and 21.548, respectively, which represent the signal sufficiency and the model dependency. The high values of R2 (STG = 0.9973, FMP = 0.9707) were obtained for the prediction of peak area under various conditions. In addition, the closeness adjusted R2 value (STG = 0.9938, FMP = 0.9531) to the value of R2 indicates the predicted values are near to the actual (experimental) values using the proposed model.

Table 5 ANOVA results of CCD for the prediction of peak area of STG
Full size table
Table 6 ANOVA results of CCD for the prediction of peak area of FMP
Full size table

The experimental values (actual) and CCD predicted values of STG were compared with each other (Fig. 7a). This plot represented that the values are very near to each other. Hence, the experiments possess excellent accuracy and less error. The normality of data was studied by a normal probability plot (Fig. 7b), indicating data points are located either on a straight line or in close proximity to the straight line. This propagation of data points confirms that the data have a normal distribution.

Fig. 7
figure 7

Plot of predicted values versus actual values for a STG and c FMT, as well as a normal plot of residuals for b STG and d FMT related to the peak area values

Full size image

The obtained peak area values versus experimental values of the FMP are shown in Fig. 7c. The data is almost scattered around the straight line. These results suggest that the two data sets have a high correlation with each other (R2 = 0.9707). The diagram of the normal probability (%) against studentized residuals is exhibited in Fig. 7d. As shown, data possess a linear normal distribution without a certain variance.

3D response surface plots

In order to better visualize and quickly select optimal conditions, surface plots were used. The plots of STG and FMP are exhibited in Figs. 8 and 9, in which the optimum conditions, the relationships between parameters, and their mutual dependence can be obviously assessed. The concurrent effect of two variables of time and pH on the peak area is displayed in the 3D diagram of Fig. 8a. The maximum peak area (1860) is proper at a pH value equal to 7.8 owing to the elevated electrostatic attraction force between the adsorbent and STG, because the increment of pH leads the adsorbent surface to become more negative. On the other hand, the decline of peak area with reducing pH can be attributed to the competition between STG ions and hydrogen ions for the active sites of the adsorbent (Kalantary et al. 2016).

Fig. 8
figure 8

3D response surface plots of CCD for the prediction of peak area related to the STG

Full size image
Fig. 9
figure 9

3D response surface plots of CCD for the prediction of peak area related to the FMP

Full size image

The simultaneous effect of adsorbent dosage and pH on the peak area is shown in Fig. 8b. By increasing adsorbent dosage up to 15.5, the active sites and the surface areas available for the STG enhance.

According to Fig. 8c, the highest peak area is observed at 33.5 min, which is a sufficient opportunity for interaction between the adsorbent and STG. In fact, access to active sites of the adsorbent is more. The peak area value was increased up to 33.5 min. It can be concluded that at initial times all active sites related to the adsorbent surface are available. By decreasing available sites, the system was reached an equilibrium state.

Figure 9a illustrates the 3D surface plot related to the combined effects of pH (A) and time (C) on the peak area. The peak area was enhanced (1901.11) by increasing pH value and contact time up to 7.8 and 33.5 min, respectively. The mutual effect of the nano dosage (B) and pH (A) is represented in Fig. 9b. The highest peak area was obtained by increasing the nano dosage up to 11.8 mg. At higher values of both parameters, the peak area was decreased. Moreover, it was discovered that by enhancing the time (up to 33.5 min) and nano dosage (up to 11.8), the response was improved (Fig. 9c).

Kinetic study

The kinetic data express the mechanism of transport rate related to the adsorbate onto the magnetic adsorbent. Two kinetic models, including pseudo-first-order (PFO) and pseudo-second-order (PSO) models were applied. The linear form PFO and PSO models are defined using Eqs. (5) and (6).

$$ln (q_{e} - q_{t} ) = lnq_{e} - k_{1} \left( t \right)$$
(5)
$$\frac{t}{{q_{t} }} = \frac{1}{{k_{2} q_{e}^{2} }} + \frac{t}{{q_{e} }}$$
(6)

In these equations, the amount of STG and FMP adsorbed at a certain time and equilibrium time is represented by qt and qe (mg g−1), respectively. Also, K1 (min−1) and K2 (g mg−1 min−1) show the PFO and PSO rate constant, respectively. As shown in Fig. 10a, by plotting ln (qe –qt) versus t, the values of qe and K1 can be achieved from the slope and intercept, respectively (Table 7). The calculation of qe and K2 values of PSO (Table 7) was performed using the slope and intercept of the plot of t/qt versus t, respectively (Fig. 10b) (Radoor et al., 2021). The high R2 (0.9999) of the PSO kinetic model reveals the goodness of fit of this model for the adsorption of both drugs. Hence, the control the reaction mechanism and adsorption rate of the PSO model are performed using the chemical adsorption.

Fig. 10
figure 10

a Pseudo first-order and b Pseudo second-order adsorption kinetics of STG and FMP by the magnetic kiwi peel

Full size image
Table 7 The obtained kinetic parameters for the removal of STG and FMP
Full size table

Adsorption isotherm models

Adsorption isotherm studies indicate the interaction between the adsorbent and adsorbate. In this study, the interaction between adsorbent and STG and FMP was evaluated through Langmuir and Freundlich isotherm models that their assumption is monolayer and multilayer adsorption, respectively. The linear form of Langmuir and Freundlich isotherm models can be described using Eqs. (7) and (8), respectively.

$$\frac{{C_{e} }}{{q_{e} }} = \frac{1}{{q_{max} K_{L} }} + \frac{1}{{q_{max} }}C_{e}$$
(7)
$$Log q_{e} = log \left( {K_{F} } \right) + \frac{1}{n} log \left( {C_{e} } \right)$$
(8)

where Ce and qe displays the pollutant concentration at equilibrium (mg L−1) and the amount of pollutant adsorbed (mg g−1), respectively. The values of qm (mg g−1) and KL (L/mg) reveal the maximum adsorption capacity and a Langmuir isotherm constant, respectively. The qmax and KL values (Table 8) were determined from the slope and intercept of the linear plot of Ce/qe versus Ce (Fig. 11a). As shown in Table 8, KF (mg g−1) and n denote the Freundlich constant and heterogeneity factor, respectively. These values were estimated from the linear plot of log qe against log Ce (Fig. 11b). The values of n between 1 to 10 and n < 1 indicate favorable and unfavorable adsorption, respectively (Kaur et al. 2020). According to higher values of R2 related to the Langmuir isotherm compared to the Freundlich, there is a monolayer distribution of STG and FMP on homogenous active sites available on the magnetic adsorbent. The maximum adsorption capacity was obtained 59.88 and 58.08 mg g−1 for STG and FMP, respectively. The value of n equal to 2.67 for STG and 1.9 for FMP revealed the favorable adsorption.

Table 8 The obtained parameters of isotherm models for the removal of STG and FMP
Full size table
Fig. 11
figure 11

a Langmuir and b Freundlich isotherms of STG and FMP by the magnetic kiwi peel

Full size image

Reusability of adsorbent

In order to decrease costs and waste production, the adsorbent recycling can be evaluated. The removal efficiency of adsorbent versus cycles (4 successive cycles) is illustrated in Fig. 12. It can be concluded that magnetic kiwi peel has high ability after four removal process (from 83.45 to 52.96%) for the reusability.

Fig. 12
figure 12

Reusability of magnetic kiwi peel for removing STG and MP

Full size image

Stability of adsorbent

The effect of different pH (2–12) and time (5–90 min) was studied on the stability of the adsorbent. It can be seen that the adsorption changes are insignificant for both drugs (Fig. 13).

Fig. 13
figure 13

The effect of pH and time on the stability of adsorbent

Full size image

Real sample analysis

Sampling was done after producing 3 real batches. According to the obtained results (Table 9), the method has good precision and reproducibility for both drugs.

Table 9 The results obtained from real sample analysis
Full size table

Comparison with literature

A comparison of the magnetic kiwi peel with previously reported adsorbents for the removal of various pharmaceutical pollutants was performed and given in Table 10. The removal efficiency and maximum adsorption capacity of the suggested adsorbent with a short adsorption time (35 min) were more satisfactory than other reported absorbents.

Table 10 Comparison of various adsorbent for the removal of different drugs
Full size table

Conclusion

In this study, the magnetic kiwi peel was synthesized and used to remove STG and FMP from an aqueous medium. HPLC technique was applied to calculate removal percentage values at different concentrations of drugs. The CCD experimental design was implemented to obtain the optimum conditions, including pH, adsorbent dosage, and contact time for the highest peak area values of both pharmaceutical components. The R2 values of adjusted determination and adequate precision demonstrated that the experimental data were matchable with the data predicted by the quadratic model. The pH, adsorbent dosage, contact time, the interaction of pH-adsorbent dosage, the interaction of pH-contact time, and the interaction of adsorbent dosage-contact time were considered the most efficient experimental factors for obtaining the peak area values for STG. On the other hand, the adsorbent dosage had a significant effect on the peak area values of FMP. The proposed adsorbent is simple, cost-effective, environmentally friendly, high efficiency, and reliable without needing a time-consuming synthesis process for the removal of pharmaceutical pollutants from aqueous solutions, which could be a suitable alternative to expensive and complex absorbents.