Preparation of ZnO/BaTiO3 adsorbent using Elaeagnus Angustifolia L. leaf extract and its evaluation for ciprofloxacin removal from aqueous solutions: an optimization study.

Especially, most papers have reported an increase in antibiotic resistance (AR) bacterial infections during the COVID-19 pandemic. Because of the outbreak of the SARS-CoV-2, antimicrobial resistance (AMR) should be controlled and reduced. Researchers have reported that the adsorption technique is an sufficient procedure for separating drugs such as antibiotics from aqueous solutions. The prepared of ZnO/BaTiO3 nanocomposite using Elaeagnus Angustifolia L. leaf extract was successfully obtained using green route. The synthesized nanocomposite was interacted with ciprofloxacin hydrochloride (CPF) to aim at eliminating the antibiotic from aqueous solutions. The incorporation of Elaeagnus Angustifolia leaf extract onto ZnO/BaTiO3 proved a sustainable chemistry study. Hence, this study indicated that green nanoparticles include neither the use of hazardous chemicals nor toxic chemicals. FTIR, XRD, and SEM-EDX analyses were applied to give information about the structural properties of the green nanocomposite. Box-Behnken design (BBD) was executed by response surface methodology (RSM) to gain optimal conditions. The effect of pH, initial concentration of CPF, and nanocomposite dose on CPF-nanocomposite interaction was examined. The experimental findings of adsorption study revealed that the optimal adsorption capacity of CPF onto ZnO/BaTiO3 was found as 125.29 mgg-1 under optimal conditions (adsorbent dose: 3.00 mg, pH value of solution: 9.88, initial concentration CPF: 49.63 mgL-1).

Introduction

Nowadays, antibiotic resistance (AR) is a severe topic that affects global health. Antibiotics are designed to defeat bacteria and microorganisms. These antibiotics, taken uncontrollably and unconsciously, cause the bacteria and microorganisms to gain resistance and create bacterial resistance. There are many reasons that accelerate the spread of AR. AR arises from discharging antibiotics into agricultural land, drinking water and sewage [1].

Ciprofloxacin (CPF), a second-generation fluoroquinolone, has broad-spectrum antibiotic owing to its widespread-uses in treatment for both gram-positive and gram-negative bacteria [2]. It can be used therapeutically for the lower respiratory tract, urinary tract, especially bladder infection, bone infection such as osteomyelitis, intra-abdominal infection urinary tract, skin-related infections, neutropenic patients [3].

One of the solution ways against antibiotic resistance is the process of removing antibiotics from aqueous solutions. Many techniques have been progressed to eliminate antibiotics from the aqueous phase. These are membrane technology [4], Fenton oxidation [5], photocatalytic degradation [6], electrocoagulation [7] and adsorption [8]. Among these techniques, adsorption stands out because it is the easiest, cheapest and most applicable process.

Zinc oxide (ZnO) has drawn great attention owing to its nature such as low cost, non-toxic, thermal stability [9]. Nagpal and Kakkar (2019) underlined in their review that ZnO NPs are used in wastewater treatment due to the reasons aforementioned above [10]. Nowadays, green chemistry processes in the preparation of nanoparticles are becoming more and more popular thanks to environmentally benign manufacturing. Especially, the synthesis of nanoparticles via plant extracts makes them more biocompatible [11]. Published papers related to ZnO nanoparticles exemplified the promising application of the green synthesis method [12-14]. BaTiO3 is a common photocatalyst owing to the semiconductor material. When the studies in the literature were examined, it was seen that BaTiO3 was scarcely used in the adsorption process. Kumari et al. (2015) prepared BaTiO3-mesoporous silica nanocomposite for the extract of Cr6+ from water [15]. Therefore, ZnO/BaTiO3 has been used for the first time in CPF adsorption.

Phytochemicals are compounds such as polyphenols, phenolic acids, and flavonoids. They contribute significantly in the reduction of metallic ions [16]. Elaeagnus Angustifolia known as the Russian olive has numerous phytochemicals such as flavonoids, polyphenols, phenolic acids and vitamins [17]. It is used for the treatment of ulcer, sore throat, asthma treatment, and bronchial and lung diseases [18, 19] due to its antioxidant, anticonvulsant, analgesic, and anti-inflammatory properties [17].

Response surface methodology (RSM) is widely chosen to figure out the impacts of the inputs on outcomes. RSM is a helpful statistical technique. It indicates how the responses change in the selected input range. Besides, it helps to optimize the response [20].

In this context, the ultrasound-assisted extraction method was preferred to prepare Eleagnus Angustifolia L. leaf extract. The prepared ZnO/BaTiO3 with the help of Eleagnus Angustifolia L. leaf extract was used for the elimination of CPF from aqueous solutions. The secondary aim was to construct adsorption conditions containing three factors such as pH, adsorption dose, and initial concentration of CPF using RSM integrated into Box-Behnken design (BBD). Therefore, RSM was predicted to be the most suitable optimal CPF adsorption conditions, which gave the highest adsorption capacity onto ZnO/BaTiO3. Both variance analysis (ANOVA) and Pareto chart were applied in order to show the impacts of inputs on the outputs. Besides, the changes of the adsorbent surface after adsorption were described using by ATR-FTIR spectroscopy. Hence, the surface morphology of the adsorbent and adsorption mechanism were illuminated.

Materials and methods

Materials

ZnO (99.99 %, 18 nm) and BaTiO3 (99.95 %, 90 nm, cubic) were purchased from Nanografi (ODTU Teknopark, Turkey). Elaeagnus Angustifolia L. leaves were obtained from İzmir, Turkey in 2020.

Extraction of Elaeagnus Angustifolia leaves and synthesis of ZnO/BaTiO3

First, Elaeagnus Angustifolia L. leaves were ground by using a grinder. Then, a homogenizer (Bandelin, Sonoplus HD 2200.2) with 13 nm titanium probe was used to extract Elaeagnus Angustifolia L. leaf. Briefly, 30 mL of distilled water was poured into 0.75 g of grinding leaves. The sonicater was operated with a 70% amplitude (50% nominal) during 10 min. Then, the extracts were filtered using RC filter, 20 μm. 20 mL of extracts were added to the beaker and BaTiO3 by taking 50 % of the weight of ZnO and ZnO were added. The solution was mixed by magnetic stirrer at a speed of 400 rpm. Absolute ethanol was used as the washing solution. After the nanoparticle was washed three times and was dried at 60 oC in a vacuum dryer until it dries. The dried material was calcinated at 550 °C during 5 h at a heating rate of 10 °C. Figure 1 represented the preparation of ZnO/BaTiO3 using Elaeagnus Angustifolia leaf extract as a solvent.

Fig. 1

The representation of the preparation of ZnO/BaTiO3 using Elaeagnus Angustifolia leaf extract as a solvent

Characterization of the prepared ZnO/BaTiO3

Spectrum Two FT-IR Spectrometer (FTIR) (Perkin Elmer) was applied to describe the functional groups of the prepared ZnO/BaTiO3 nanocomposite. Scanning electron microscopy (SEM) with energy dispersive X-ray analysis (XL-30 SFEG, Philips, Eindhoven, Holland) was performed to illuminate the surface analysis of the samples. X-ray diffraction (XRD) analyses were done using (Bruker D8 Advance device) via Cu Kα radiation over a 2θ range from 2-90° with the scanning rate of 3° min-1.

Adsorption experiments and statistical analysis

The design of adsorption runs were constructed using Design Expert 12 software ((Trial version 12, Stat-Ease, USA). The desired adsorbent dose was added to Erlenmayer and the specified initial antibiotic solution was poured into the Erlenmayer. Then, shaker was run at the speed of 100 rpm for two hours. Two hours later, the initial and final solution were read at 275 nm by using a UV-Vis Spectrophotometer (Pekin Elmer, Lambda 365, USA). The adsorption capacity was calculated in Eq. (1) as follows;

In Eq. (1), qe is the adsorption capacity (mg g-1), Co is the initial CPF concentration in solution (mgL-1), Cf is the final concentration in 2 h later (mgL-1), m is the mass of ZnO/BaTiO3 (g), and V is the solution volume (L).

Response surface methodology (RSM) was preferred in order to optimize the adsorption process conditions. RSM suggested a mathematical model indicating the process conditions. By using this mathematical equation in coded factor in Eq. (2).

where α0 is the intercept, αi is a single factor, αij is the interaction of factors and ε is a pure error.

The evaluation of the percentage effect (P %) of each variable on the adsorption capacity could be drawn Pareto chart by following Eq. (3) [21].

In Eq. (3), αi is the coefficient of variables in Eq (2) in accordance with coded factor, Pi is the percentage of each variable, respectively.

In order to assess the impacts of the inputs (pH, CPF initial concentration and adsorbent dose) on the response (adsorption capacity), BBD design was selected. Box and Behnken devised Box-Behnken design in 1960 [22] BBD is an alternative method instead of a full factorial design serving labor efficiency with fewer essential experiments [23]. Herein, 3 factors with three levels of Box-Behnken design were applied to maximize the response (adsorption capacity) under the optimal conditions such as pH (A), adsorbent dose (B), and CPF initial concentration (C). Table 1 represented the coded factors.

Table 1

Selected parameters with three levels for the adsorption of CPF onto ZnO/BaTiO3

Parameters	Levels
	-1	0	1
pH	3	7	11
Adsorbent Dose (mg)	3	6	9
Initial CPF concentration (mgL-1)	20	35	50

Results and discussions

Characterization of ZnO, BaTiO3, and ZnO/BaTiO3

FTIR analysis of ZnO, BaTiO3 and ZnO/BaTiO3 was done using Perkin Elmer Spectrum Two IR in Fig. 2. In FTIR spectrum of BaTiO3, a peak appeared at 511 cm-1 corresponding to a common Ti-O absorption into BaTiO3 [24]. Additionally, a small peak at 3500 cm-1 was assigned to asymmetric and symmetric -OH bond. In ZnO FTIR’s spectrum depicted that a small peak at 687 cm-1 was attributed to the Zn-O stretching bond [25]. FTIR spectrum of the synthesis ZnO/BaTiO3 showed that the peak at around 498 cm-1 was the characteristic bond of Ti-O vibration [26]. The other small intensity peak at about 1400 cm-1 was related to the crystalline Ba-Ti-O vibration [27]. The small (C = O)p (p :pyridone) vibrations of CPF was seen at 1630 cm–1 in the spectrum of CPF-adsorbed ZnO/BaTiO3 [28]. The other small peak emerged at 1260 cm–1 owing to the vibration of C–F bond of CPF [28]. This result showed that electrostatic interactions occurs between raw CPF and ZnO/BaTiO3. Briefly, it was seen that there is a strong overlapping between the spectrum of CPF-adsorbed ZnO/BaTiO3 and the spectrum of pure ZnO/BaTiO3.

Fig. 2

FTIR spectrum of pure ZnO, pure BaTiO3, ZnO/BaTiO3, CPF-adsorbed ZnO /BaTiO3

XRD pattern of BaTiO3 was seen in Fig. 3. The peaks at 2θ = 22.27, 31.66, 39.02, 45.35, 51.03, 56.32, 65.98, and 69.12° were assigned to (1 0 0), (1 1 0), (1 1 1), (200), (2 1 0), (2 1 1), (2 2 0), (3 0 0) support the perovskite phase of barium titanate [29]. The peaks at 2θ = 31.67°, 34.58°, 36.02°, 47.68°, 56.68°, 62.91°, 68.02°, 69.12 indexes as (100), (002), (101), (102), (110), (103), (200), (112) of ZnO revealing the wurtzite structure of nanoparticles (Zincite, JCPDS 5-0664) [30]. The peaks related to BaTiO3 and ZnO were seen in the XRD pattern of ZnO/BaTiO3.

Fig. 3

XRD patterns of BaTiO3, ZnO, ZnO/BaTiO3

SEM image of pure BaTiO3 nanoparticle showed quasi-spherical shape in Fig. 4(b). Besides, the surface of BaTiO3 had less agglomerated. The agglomeration form of ZnO appeared in Fig. 4(a). In addition, the spherical shape of ZnO nanoparticles occurred. Both the spherical shape of ZnO and the quasi-spherical shape of BaTiO3 belonging to the synthesized ZnO/BaTiO3 were seen in Fig. 4(c). It might be claimed that a new surface occurred. A well-synthesized form of the BaTiO3 and ZnO was observed from the SEM-EDX result.

Fig. 4

SEM image and EDX spectra of a) ZnO, b) BaTiO3, and c) ZnO/BaTiO3

Response surface design with Box-Behnken design

Seventeen runs with five replicates at center points were shown in Table 2. Through the ANOVA analysis (Table 3) with the F and p-value indicates which parameter is more crucial than the others. The most effective variable was seen as pH (F-value: 229.19 and p-value<0.0001). It can be sorted by other effective parameters as follows; dose of nanoparticle (B), CPF initial concentration (C), and the squared dose of nanoparticle (B2). In addition to variance analysis, Pareto chart serves as a better understand which parameter is more significant than the others. Hence, Pareto chart implies the percentage values of factors that are statistically crucial.

Table 2

BBD for CPF adsorption onto ZnO/BaTiO3

No	pH	Adsorbent dose(mg)	CPF concentration (mgL-1)	Adsorption capacity (qe:mgg-1)
1	7	6	35	47.84
2	7	6	35	45.34
3	7	6	35	46.52
4	4	6	50	22.1
5	7	9	20	26.23
6	10	3	35	109.21
7	10	9	35	54.54
8	4	3	35	60.02
9	4	9	35	35.38
10	7	3	50	92.06
11	7	6	35	50.81
12	10	6	20	35.24
13	4	6	20	5.18
14	7	6	35	48.62
15	7	9	50	52.84
16	7	3	20	54.09
17	10	6	50	82.27

Table 3

Variance analysis for CPF adsorption onto ZnO/BaTiO3

Source	Sum of squares	df	Mean square	F-value	p-value
Model	10171.25	9	1130.14	82.40	< 0.0001	significant
A-pH	3143.45	1	3143.45	229.19	< 0.0001
B-adsorbent dose	2678.75	1	2678.75	195.31	< 0.0001
C-CPF initial concentration	2065.00	1	2065.00	150.56	< 0.0001
AB	225.45	1	225.45	16.44	0.0048
AC	226.65	1	226.65	16.53	0.0048
BC	32.26	1	32.26	2.35	0.1690
A2	10.42	1	10.42	0.7596	0.4123
B2	1446.43	1	1446.43	105.46	< 0.0001
C2	425.74	1	425.74	31.04	0.0008
Residual	96.01	7	13.72
Lack of Fit	78.59	3	26.20	6.01	0.0579	not significant
Pure Error	17.42	4	4.36
Cor Total	10267.26	16
Std. Dev.	3.70		R2		0.9906
Mean	51.08		Adjusted R2		0.9786
C.V. %	7.25		Predicted R2		0.8749
			Adeq Precision		36.1144

The mathematical equation in terms of coded factor was obtained as the following Eq. (4).

The Pareto graph of the independent factors was drawn with the help of Eqs. (3) and (4). Looking at the Pareto chart in Fig. 5, the most influential factor was seen as A: (25.28%). Then, the other impacted parameters could be listed as the squared adsorbent dose (B2: 22.09%), adsorbent dose (B: 21.55), CPF initial concentration (C: 16.62%).

Fig. 5

Pareto graph for the adsorption capacity of CPF onto ZnO/BaTiO3

The suggested model is well-fitted to the adsorption data because the lack of fit (LOF) is not significant (p-value=0.0579) and this value is also related to pure error containing replicates at center points. The higher R2 value (0.9906) demonstrated a higher regression between the proposed data and empirical results. The obtained higher R2- adj (0.9786) showed a reliable correlation. The difference between R2-adj and R2-pred was calculated as 0.1037. This value implied that the suggested model estimated the outcomes properly.

Generating 3-D surface graphs allow researchers to interpret the relationships between dependent and independent parameters more easily. The desired maximum adsorption capacity was found as 125.29 mgg-1 under optimal conditions such as the adsorbent dose: 3.00 mg, pH value of the solution: 9.88, initial concentration CPF:49.63 mgL-1.

Adsorption studies have revealed that pH of the solution plays an effective factor on adsorption mechanism as a result of the surface charge of the adsorbent is varied with pH [31]. While the pH value of the solution was increasing, the adsorption capacity of CPF onto ZnO/BaTiO3 was increased. The adsorption capacity was obtained from 57 to 126 mgg-1 from pH 4 to pH 10 (Fig. 6 (a)–(b)). Similar result was recorded by Sun et al. (2016) [32]. In order to explain the higher adsorption capacity in basic pH, it is necessary to know the pKa values ​​of the CPF antibiotic. CPF has two pKa values, which are pKa1: 5.90 and pKa2: 8.89 owing to the presence of the carboxyl group and amine groups in piperazine ring, respectively [33]. Thus, three different forms such as cationic CPF+ (pKa1), zwitterionic CPF± (between the pKa1 and pKa2) and anionic CPF− (pKa2) occurred [34]. Thus, the electrostatic interaction between the anionic adsorbate and cationic charged adsorbent occurred in basic pH.

Fig. 6

The binary interaction of inputs on qe a) pH-adsorbent dose, b) pH-initial CPF concentration, c) adsorbent dose-initial CPF concentration under optimal conditions (adsorbent dose: 3.00 mg, pH value of solution: 9.88, initial concentration CPF: 49.63 mgL-1)

While varying ZnO/BaTiO3 dose value from 3 mg to 9 mg, the adsorption capacity was decreased (Fig. 6 (a)–(c)). Obviously, the higher adsorption capacity is obtained by using the lesser adsorbent dose. Besides, increasing the amount of adsorbent brings about both the aggregation of green adsorbent and the improvement of electrostatic repulsive force between green nanoparticles [35]. Thus, CPF molecules are difficult to interact with the adsorbent. Similar data were recorded by Yu et al. (2018) [36] and Li et al. (2017) [37].

Numerous CPF concentrations (20-50 mgL-1) were prepared to assess the impact of initial CPF concentration on the adsorption process. The higher adsorption capacity was gained, when the initial CPF concentration was 50 mgL-1(Fig. 6 (b)–(c)). This situation is clearly explained that, a high initial concentration of CPF results in higher concentration gradient owing to the driving force [38]. Thus, CPF molecules were adsorbed faster to the surface of ZnO/BaTiO3. However, the rapid increment of adsorption capacity was gradually decreased after the certain CPF initial concentration, because of the saturation of adsorbent’s active sites by higher concentration of CPF [39]. The repulsion between the CPF and adsorbent happens by the virtue of the occupation of CPF onto available sites of adsorbent [40].

The high qe of adsorbent is a crucial factor for the utilization of the adsorbent from an economic point of view. ZnO/BaTiO3 exhibited a good adsorbent for the adsorption of CPF. The measured qmax for ZnO/BaTiO3 and green adsorbents recorded in the literature were summarized in Table 4. Hence, ZnO/BaTiO3 is a promising adsorbent and can be used for the removal of antibiotics.

Table 4

The maximum adsorption capacity for ZnO/BaTiO3 and green adsorbents recorded in the literature for CPF adsorption

Adsorbent	qmax (mgg-1)	Reference
amine- functionalized bio graphene	172.6	[41]
Fe2O3–Dy2O3 (FD) c-Fe2O3–Dy2O3 (c-FD) (C: Syzygium aromaticum (Clove))	125 328	[42]
BC-2-650 (magnetic biochar of camphor leaf , ZnCl2/biochar mass ratio of 2, Calcination: 650 °C)	449.40	[43]
ZnO/BaTiO3	125.29 mgg-1	This Study

Conclusion

To sum up, this work examined the valorization of Elaeagnus Angustifolia L. leaf as a high potential reducing agent to prepare ZnO/BaTiO3 nanoparticles. The leaves, which are also described as biowaste, were extracted by selecting the innovative extraction method. Thus, the ultrasound-assisted extraction method offers less extraction time, lower energy consumption, and lower solvent consumption. In this context, the extraction method chosen is also an environmentally friendly method. Box-Behnken design (BBD) gives a better outcome for optimization by reducing the number of experiments. The effects of pH, initial concentration of CPF, nanocomposite dose on CPF-nanocomposite interaction were examined as independent factors. The maximum adsorption capacity of CPF onto ZnO/BaTiO3 was found as 125.29 mgg-1 under optimal conditions (adsorbent dose:3.00 mg, pH value of solution: 9.88, initial concentration CPF:49.63 mgL-1). Pareto graph was indicated that the most influential factor was pH (25.28%).