1 Introduction

The presence of antibiotics in ground and surface water has become ubiquitous in recent times. These groups of pollutants generally referred to as “emerging pollutants” have been widely reported in aqueous environments. Antibiotics which are largely used in veterinary and human medicines have entered into water bodies through various anthropogenic activities. Unfortunately, receiving water bodies have been reported to contain these antibiotics in their metabolized and un-metabolized forms even at low concentrations. The ecological health suffers a setback by the persistence of these undesirable contaminants which accumulate in water bodies [1, 2]. The trending ability of bacteria to be resistant to antibiotics has become another growing threat to scientists as the presence of these pharmaceutical products is increasing rapidly in the environment [3,4,5,6,7,8,9].

Ciprofloxacin, a fluoroquinolone antibiotic, has been used in therapeutic applications both in veterinary and human medicine and has been reported to have concentrations reaching 31 mg l−1 or above in discharges from pharmaceutical production plants [10,11,12]. Due to its solubility in water, bacteria-inhibiting ability, and high molecular ring structural stability, ciprofloxacin is not easily removed from water systems [13, 14]. Antibiotics removal from waste sources has reportedly been achieved by various methods such as photodegradation, biodegradation, advanced oxidation, and adsorption [15, 16]. Unlike other treatment technologies, the adsorption process has proven better by its simple design, ease of operation, cost-effectiveness, and convenience in antibiotics removal from aqueous environments [17]. In this present study, the adsorption process is employed as an efficient, simple, and economical method for ciprofloxacin removal from contaminated water [18,19,20]. The cost associated with commercially available activated carbon utilized in adsorption has brought about the need to develop inexpensive adsorbent from agricultural waste materials usually of little or no economical values. Examples of such adsorbents that have been used for adsorption include rice and wheat husk [21], ground nutshell [22], waste biomass [23], and banana stalk [24, 25]. Recent advances in the adsorption of CIP have employed various adsorptive techniques such as the use of activated sludge laced with biochar derived from apple tree [26], surfactant-modified sepiolite [27], nanohybrid GO/O-CNTs [28], TiO2 nanotube/graphene oxide hydrogel [29], Fe3O4/graphene oxide nanocomposite [30], and biogenic palladium nanoparticles [31] as sorbent materials. Other adsorbent materials that have been employed for the removal of other pollutants include starch/PVA composite films, ZnO–NR–AC, slag, porous carbon, carbon nanotubes, fullerenes, tire-derived carbons, multi-walled carbon nanotubes, ZnO/CuO nanocomposite, Hg-doped ZnO nanorods, bagasse fly ash, Fe@Au core–shell, ZnO/Ag nanocomposite, waste rubber tire, and rice husk [32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66].

Ortho-phosphoric acid (PA), H3PO4 is a common activating agent whose use has been extensively reported for preparing activated carbons from agricultural products [67,68,69,70]. PA promotes bond cleavage in the biopolymers and dehydration at low temperatures, followed by extensive cross-linking that binds volatile matter into the carbon products and thus increases the carbon yield. Benaddi et al. [71] showed that the mechanism of PA activation of biomass feedstock occurs through various steps: cellulose de-polymerization, biopolymer dehydration, formation of aromatic rings, and elimination of phosphate groups. This produces activated carbon with good yields and high surface areas. Activation conditions thus depend on the nature of the precursor, i.e., on the relative amounts of cellulose, hemicelluloses, lignin, and ashes. In the present study, authors prepared activated carbon (AC) from a banana stalk (BS) by activation with H3PO4; the optimum BS/PA weight/volume ratio allowing complete activation of BS was evidenced when 25.0 ± 0.01 g BS impregnated with 500 cm3 of 0.3 mol/dm3 ortho-phosphoric acid (H3PO4). The best yield was obtained using this weight/volume ratio. Alteration in the ratio of BS/PA did not produce good yield during pyrolysis [72]. Hence, this ratio was used for the acid activation.

Banana stalk, a lignocellulosic agricultural waste (which constitutes a pollutant in itself if not well disposed of), has been used in several studies as precursors in producing activated carbon [73, 74]. BSAC has been used as an efficient adsorbent in removing various pollutants such as dyes, and heavy metals in previously reported works, but the adsorption of emerging contaminants such as CIP using BSAC has not been explored. The use of carbon materials prepared from banana stalk for the sorption of ciprofloxacin antibiotics from aqueous solution, to the best of our knowledge, has not been reported. The novelty of this work is in the remarkable adsorption capacity demonstrated by the prepared adsorbent from banana stalk on ciprofloxacin (CIP) adsorption when compared with other biomass. Similarly, banana stalk can cause serious environmental threat if not properly managed; it can produce greenhouse gas if dumped in wet conditions. Usually, farmers threw the banana stalk waste in rivers and ponds where it degraded slowly and formed methane, and other gases, spreading foul smell thereby affecting the nearby ecosystem and the natural balance of atmospheric gases. It is abundantly available and has significant adsorption capacity against water-soluble pollutants (CIP inclusive). Therefore, the selection of banana stalk as an effective adsorbent material is a smart choice for sustainable future. This work, therefore, investigates the use of banana stalk as a cost-effective precursor material for preparing functionalized adsorbent for the uptake of ciprofloxacin antibiotic from aqueous solutions.

2 Materials and methods

2.1 Chemicals

Orthophosphoric acid (H3PO4), ciprofloxacin, NaOH, NaHCO3, Na2CO3, and HCl are the analytical grade chemicals used for this investigation and were not needing purification before use. The chemicals used were obtained from Aldrich Chemicals, Germany, except for ciprofloxacin (96%) which was obtained from Bond Chemicals, Aawe, Oyo State, Nigeria (Fig. 1). All chemicals were used as purchased.

Fig. 1
figure 1

Chemical structure of ciprofloxacin

Full size image

2.2 Sample collection and pretreatment

Banana stalk agricultural waste was obtained from a local farm in Osun State, Oyan (a south-west town in Nigeria). The stalks were washed thoroughly with water, and impurities were further removed by rinsing with distilled water. To increase sample surface area, it was ground with mortar and pestle after the sample was sun-dried to constant weight. Uniform particle size was achieved by sieving with a 106-μm mesh size sieve.

2.3 BSAC preparation

The modification was done by reacting to 1000 ml of 0.3 M orthophosphoric acid with 50 g of banana stalk. The reaction was carried out on an electric hotplate until a slur was formed. After the slurry formation, it was carbonized at 250 °C for 3 h until a char was formed. The char formed was then washed to neutral pH with distilled water after which it is oven-dried to fixed weight at 105 °C to eliminate the moisture content and other adhering volatile materials [75]. An airtight container was used to store the banana stalk–activated carbon (BSAC) obtained for subsequent investigation.

2.4 Adsorbate preparation

Adsorbate solution stocks were prepared by the dissolution of 1.0 g CIP crystalline powder into 1000 ml of distilled water from which desired working solutions were prepared using serial dilution. Some physicochemical properties of CIP are reported in Table 1.

Table 1 Physicochemical properties of ciprofloxacin
Full size table

2.5 BSAC characterization

2.5.1 Fourier transform infrared spectroscopy

The (FTIR) spectrometer (Perkin-Elmer FTIR-2000) was employed to record the spectra of raw banana stalk. The measurement of the spectra was obtained using a KBr disc between 4000 and 400 cm−1 wavelength. This analytical method was used to elucidate the characteristic functional group(s) available on the raw banana stalk sample surface.

2.5.2 Scanning electron microscopy

The morphological characteristics of the sample surface were observed with the scanning electron microscope (SEM). This electronic microscope produces enlarged images of samples [76]. Interactions of the sample’s atomic constituents with the shaft of electron result into the production of different signals. Details of the sample’s constituents, scenery, and the surface property were obtained from the signals generated by a scanning electron microscope [77].

2.5.3 Boehm titration

Functional groups that contain oxygen on the adsorbent surface (BSAC) were investigated using Boehm titration [78]. The workings of the Boehm titration are principled on the fact that surface oxygen groups on activated carbon have varying acidities and can be brought to neutrality using bases of different strengths [79].

2.5.4 pH point of zero charge (pHpzc)

The pH at which all charges on the adsorbent surface is zero is the pHpzc [80]. It measures the adsorption efficiency of adsorbents in removing species from solution. The pHpzc is reached when no changes occur with pH on interacting with the adsorbent(s) [81, 82].

2.6 CIP adsorption studies

2.6.1 Batch adsorption studies

The effectiveness of BSAC in removing CIP from aqueous solution was investigated at 303 K, 313 K, and 323 K using the batch adsorption method. The effect of solution temperature, contact time, and initial concentration of CIP on the adsorption process was investigated. Concentrations of 10–50 mg/l were studied. A total of 0.1 g of BSAC was carefully weighed into a 200-ml flask containing various CIP solutions (100 ml in each flask). The samples were agitated for 7 h in a water bath oscillator at a specific temperature until equilibrium was reached. Aliquots were drawn and residual concentration measurements were taken using a spectrophotometer (UV-visible, Model 6715: JENWAY) at a preset wavelength of 273 nm. Equation (1) was used to evaluate the amount of CIP adsorbed onto BSAC.

$$ {q}_e=\frac{\left({C}_o-{C}_e\right)V}{W} $$
(1)

where the concentrations of ciprofloxacin at equilibrium and initially are Ce and Co (mg·l−1) respectively at the liquid-phase. The volume of solution and mass of BSAC used is given as V (l) and W (g) respectively.

2.6.2 Adsorption kinetics

Aliquots were drawn at specific intervals for this study and CIP measurements were taken. Eq. (2) was used to evaluate the residual concentration of CIP removed at preset times.

$$ {q}_t=\frac{\left({C}_o-{C}_e\right)V}{W} $$
(2)

where the concentrations at liquid phase of ciprofloxacin initially and at preset times are Co and Ct (mg·l−1) respectively. The volume of solution and mass of BSAC used is given as V (l) and W (g) respectively. Error bars were incorporated into different plots in this study to show the range of uncertainty or the level of accuracy of the data. The error bar also shows the range of data on the plot. The error bars in this study were evaluated for the dependent variable using the standard error obtained from the data.

3 Results and discussion

3.1 Effect of CIP initial concentration, solution temperature, and agitation time

To investigate the influence of agitation time on the adsorption process as well as the initial CIP concentration absorbed, 100 ml of initial concentrations ranging from 10 to 50 ml of the CIP solution were prepared into Erlenmeyer flasks (200 ml). An even quantity of 0.1 g of BSAC was carefully weighed into each flask, tightly closed with glass stoppers and orderly put into an isothermal water bath shaker (Model NE5.28D). The shaker was preset at 303 K and at 120 rpm rotating speed until equilibrium was attained. This process was repeated at higher temperatures (313 K and 323 K respectively) to determine the effect of solution temperature on the uptake of CIP. The amount of CIP removed increased as agitation time increased at all initial CIP concentrations. Also, as initial CIP concentration increased, the adsorbate uptake increased. The first 50 min showed rapid sorption followed by a slower sorption rate until equilibrium was reached (figure not shown).

The speedy sorption rate at the initial phase can be credited to the availability of unoccupied sites on the BSAC surface. This resulted in forces strong enough to drive the CIP molecules to overcome transfer resistance between the solid and the aqueous phase. The consequent decline in the adsorption rate must have been due to repulsion or electrostatic hindrance that exists between the adsorbent surface and CIP molecules which makes the remaining site difficult or unavailable for adsorption [83, 84]. This phase must have necessitated the adsorption process to reach equilibrium and also indicated the maximum adsorption capacity for each concentration [85].

It was observed that the different concentrations reached equilibrium at different times with the lowest initial concentration reaching equilibrium first. Experimental data shows that time to equilibrate increased with increasing initial CIP concentration. This affirms that the time to reach equilibrium for the sorption process is largely influenced by the initial antibiotic’s concentration. For the removal of CIP by BSAC, the maximum adsorption removal was 9.81 mg/l at the initial CIP concentration of 10 mg/l and 45.88 mg/l at 50 mg/l initial CIP concentration. This shows also that at higher initial CIP concentration, appreciably high amount of CIP molecules was removed at equilibrium. The higher concentration gradient in the higher CIP concentrations produced a better driving force for the process [86, 87]. Cyril Dube et al. (2018) reported the same observation for CIP removal from aqueous solutions [88].

3.2 Activation of adsorbent

The dehydration of agricultural waste materials at low temperatures through the use of orthophosphoric acid in the activation of adsorbent enhances bond cleavage in the materials [67,68,69,70]. This process of activation involves the dehydration of biopolymers, re-polymerization of cellulose with the removal of phosphate groups, and the formation of aromatic rings which yield high carbon content and good surface area activated carbon [71]. A carbon material yield generally is dependent on the nature of precursor material. Activated carbon was prepared by activating 50.0 g of BS with 1000 cm3 of orthophosphoric acid (0.3 mol/ dm3) in this study. This weight to volume ratio has been reported to produce the best yield [72].

3.3 Adsorbent characterization

3.3.1 Surface chemistry of BSAC

The surface properties of the prepared banana stalk were studied using FTIR. This essentially showed spectroscopic characteristics of the sample including major peaks assigned to functional groups that influence the adsorption process and suggest a possible mechanism of CIP interactions with the BSAC surface. Notable bands associated with specific functional groups that may have aided the efficient uptake of CIP were observed at 3429.55 cm−1 (bonded O–H group), 1616.40 cm−1 (carbonyl group), and 3012.91–3257.88 cm−1 (the secondary amine group) from the FTIR spectra of BSAC (Fig. 2). Previous studies have also reported the influence of these functional groups on adsorption processes [89,90,91]. Other observed bands on the BSAC which may have participated in the interactions include 1402 cm−1 (symmetric bending of CH3), 1114.89 cm−1 (−C=O=C stretching of ether), 1242.20 cm−1 (–SO3 stretching), and 648.10 cm−1 (–CN stretching). Ogunleye et al. (2014) reported a similar characterization result of BSAC [74]. The FTIR spectra for BSAC before and after adsorption clearly shows that the BSAC surface interacted with the CIP molecules as corresponding disappearance and appearance of new bands were observed. A slight shift in adsorption bands was also observed, all indicating the changes in spectroscopic properties of the BSAC surface having effectively adsorbed the CIP molecules. Notable changes in spectra bands after adsorption include the shift from 3429.55 to 3425.69 cm−1 (bonded O–H group), 3012.91–2924.18 cm−1 (secondary amine group).

Fig. 2
figure 2

a FTIR spectrum of acid-modified banana stalk before adsorption. b FTIR spectrum of acid modified banana stalk after adsorption of CIP

Full size image

3.3.2 pH and pHpzc determinations

The pHpzc of BSAC was found at 4.5 (as shown in Fig. 3). This shows that the BSAC surface is positively charged in the solution up to pH 4.5, and above this becomes negatively charged. The pHpzc influenced the uptake of CIP molecules in the adsorption experiment. The pHpzc of BSAC shows that cationic adsorbent uptake will be favored at pH values above the pHpzc [92]. In aqueous solution, CIP molecules undergo both protonation and deprotonation reactions which give them the ability to produce three possible species of ions, which includes cationic when the pH of the solution is less than 5.9, zwitterionic when the solution pH is within the range 5.9–8.9, and anionic when solution pH is greater than 8.9 [93]. When the pH of the solution is less than 5.9, more cationic species exist which reduces CIP uptake as a result of electrostatic repulsion between the BSAC surface and the molecules of CIP. Consequently, when the value of pH is above 8.9, anions of CIP exist in solution which is favorable for the adsorption of CIP due to electrostatic exchange between the BSAC surface and CIP molecules. Nevertheless, at this pH region, there was a strong competition between the OH− group and the CIP anionic specie for the BSAC surface which led to decreased CIP adsorption.

Fig. 3
figure 3

Zeta potential vs. pH curve of BSAC

Full size image

However, the zwitterionic species of CIP exists when the pH value is between 5.9 and 8.9 which play the most influential role in the solution as there is no repulsion or competition between the BSAC surface and the CIP molecules. Therefore, the positive BSAC surface and the zwitterionic CIP specie electrostatic interactions contributed to the high adsorption efficiency of the process. Cheng et al. (2018) and Aborode (2020) reported a similar trend on the adsorption of CIP antibiotics using halloysite nanotubes and activated kaoline respectively [94, 95]. Result from the pH studies taken in triplicates is depicted in Fig. 4 which shows that the maximum adsorption took place between pH values between 5.9 and 8.9 and at an optimum value of 8.

Fig. 4
figure 4

Influence of pH on the adsorption of CIP by BSAC at 303 K

Full size image

3.3.3 Proximate and elemental analysis

Proximate and elemental analysis was performed using a thermogravimetric analyzer (Perkin-Elmer TGA7, USA) and elemental analyzer (Perkin-Elmer series 11, 2400, USA) respectively. Proximate analysis results revealed that the percentages of moisture, fixed carbon, volatile, and ash content for BS were 14.36, 5.83, 73.91, and 5.90 respectively. On the other hand, BSAC showed low moisture and volatile content with a correspondingly high amount of fixed carbon and lower ash content with percentages of 2.91, 18.64, 73.92, and 4.53 respectively. The higher fixed carbon content value suggests that the adsorbent is favorable for the uptake of CIP antibiotics. The elemental analysis for the raw banana stalk (RBS) showed low carbon (9.32%), hydrogen (4.38%), sulfur (0.10%) content, and high nitrogen-oxygen content combined (86.20%). The BSAC showed a very high amount of carbon (75.28), with low hydrogen (1.44%), sulfur (0.05%), and nitrogen-oxygen content combined (23.23%). This further revealed the suitability of BSAC for the uptake of CIP molecules from aqueous environments.

3.3.4 Surface morphology

Images from the SEM of BS and BSAC are respectively shown in Figs. 5a and 5b. The pores of the activated sample show that they were well-developed when compared with that of the raw sample. The improved numerous pores observed can be attributed to the influence of the acid activation in the adsorbent preparation. The better performance of BSAC in removing CIP from aqueous solution can also be attributed to the prominent and well-developed pores which allow the trapping of the CIP molecules to the surface of the BSAC.

Fig. 5
figure 5

SEM image of (a) raw BS magnification: × 1000, (b) BSAC magnification: × 1000

Full size image

3.3.5 Boehm’s titration

Oxygen-containing functional groups usually characterize activated carbons as they influence the properties of the surfaces of carbons and their quality as adsorbents [96]. Surface acidity and basicity of adsorbents are evaluated by the Boehm’s titration which assumes that NaOH, Na2CO3, and NaHCO3 neutralize acidic groups while HCl neutralizes groups that are basic. The acid and basic groups are lactonic (0.2418 mmol/g), carboxylic (0.1832 mmol/g), and phenolic (0.2749 mmol/g). The total acidic groups were 0.699 mmol/g and that of the basic groups was 0.1582 mmol/g. These results clearly show that the BSAC surface contains more acidic groups than basic groups. The surface is therefore predominantly acidic as suggested by the pHpzc (4.5) value obtained [73]. This complements that cationic adsorbate will be favored in the adsorption process.

3.3.6 Adsorption kinetic studies

Four kinetic models were used in this study to investigate the kinetics of adsorption of CIP antibiotics unto BSAC. They are the pseudo-first-order (PFO), pseudo-second-order (PSO), Elovich, and the intraparticle diffusion models. Table 2 shows the adsorption kinetic data for the four models used. The PSO model described adsorption data best with a good correlation between qe(exp) and qe(cal) (Table 2), unlike the PFO model. A similar report was given by Li et al. (2018) on CIP removal using biochar obtained from used tea leaves [97]. Comparing the values of R2 for the kinetic models used, the sorption of CIP onto BSAC was in the sequence: pseudo-second-order > pseudo-second-order > Elovich. The poor agreement of the qe(exp) and qe(cal) in the PFO model and the resulting low values of R2 for the Elovich model makes them unsuitable to satisfactorily explain the uptake of CIP from aqueous media by adsorption.

Table 2 Kinetic model parameters of CIP adsorption onto BSAC at 323 K
Full size table

The rate constant (kdiff) for the intraparticle diffusion model increased with increasing initial CIP concentration as shown in Table 2. The low R2 values obtained showed that the model poorly fitted the adsorption data in addition to the value for the boundary layer thickness, C (mg/g), obtained in Table 2. These values showed a deviation of the linear plot from the origin. A multistep adsorption process is proposed with a gradual adsorption phase preceded by an instantaneous adsorption phase. The intraparticle diffusion is suggested to be the rate-determining step in the sorption process [98, 99]. These findings agree with studies on adsorption of CIP by rice straw biochar [100].

3.3.7 Adsorption isotherms

Studies on adsorption isotherm were investigated by fitting experimental data to five different models namely Langmuir, Temkin, Dubinin-Radushkevich, Freundlich, (two-parameter isotherm models), and Sips isotherm (three-isotherm model) respectively. The R2 values of the different isotherms were compared to ascertain the isotherm that best explained the adsorption process. Results from the adsorption models used in this study are presented in Table 3. Figure 6 a, b, c, and d also show various plots from isotherm models. The Langmuir isotherm assumes the surface is covered as monolayer by an adsorbate when maximum monolayer adsorption has occurred. The best adsorption capacity that correlates with monolayer coverage on the BSAC can be evaluated by this adsorption isotherm. Qm for the Langmuir isotherm was 49.75 mg/g at 323 K with an R2 value of 0.9824. This indicates that the experimental data fitted well into the isotherm model. The separation factor (RL) value ranged between zero and unity which indicated the favorability of the process and the good fitting of the data to the Langmuir isotherm [101].

Table 3 Adsorption isotherm parameters for CIP removal using BSAC
Full size table
Fig. 6
figure 6

Plot of Langmuir (a), Freundlich (b), Temkin (c), and D-R (d) isotherm at 313 K

Full size image

The value of n > 1 in the Freundlich model shows that the sorption process was simple, favorable, and physical [102,103,104]. Table 3 shows that the kf values increased with increasing temperature which suggested a higher affinity of the BSAC for CIP molecules at higher temperatures. The constant BT of the Temkin model which relates to adsorption heat has a positive value of 295.47 (R2 = 0.6669) suggesting an endothermic process for the adsorption. The mean free energy of the D-R model was within the range of 1.29–3.54 kJ/mol which is lesser than 8 kJ/mol suggesting that the adsorption mechanism is physical [105, 106]. The Qm of the D-R model follows closely the Langmuir Qm and with a value of 47.89 mg/g (R2 = 0.8137). The three-parameter isotherm model, Sips, gave the least Qmax of 43.08 mg/g with an R2 value of 0.9043.

Judging from the R2 values vis-à-vis their Qmax, the adsorption of CIP unto BSAC followed this order of fitness of isotherm model to experimental data: Langmuir (0.9824) > Sips (0.9043) > D-R (0.8137) > Temkin (0.6669) > Freundlich (0.5864). This order implies that the Langmuir isotherm best explains the experimental data. Table 4 shows the CIP adsorption capacities of different adsorbent materials. BSAC gave the highest Qmax value compared with other adsorbents.

Table 4 Maximum monolayer adsorption capacities of CIP adsorption on different adsorbents
Full size table

3.3.8 Adsorption thermodynamic studies

Thermodynamic parameters at three different temperatures were investigated in this study using the equations below.

$$ \ln {K}_L=\frac{{\Delta S}^0}{R}-\frac{{\Delta H}^0}{RT} $$
(3)
$$ {\Delta G}^0=- RT\ \ln \left(55.5\ {K}_L\right) $$
(4)
$$ {E}_a={\Delta H}^0+ RT $$
(5)

Since KL is expressed in l.mol−1, then KL can be recalculated as dimensionless by multiplying it by 55.5 (number of moles of water per liter of solution). Accordingly, the correct ΔG0 value can be obtained from Eq. (4). Here, ΔG0 is the standard Gibbs free energy change (kJ mol−1), R is the universal gas constant (8.314 J mol−1 K−1), T is the absolute temperature (K), and Ea (KJ/mol) is the Arrhenius energy. ΔH0 and ΔS0 values of the adsorption process were calculated from the Van’t Hoff Equation (Eq. 4). The values of ΔH° and ΔS° values were calculated from the slope and intercept of the plot of ln (KL) versus 1/T, they are listed in Table 5. The results from standard free energy (ΔG0), entropy (ΔS0), and enthalpy (ΔH0) changes were employed to report the feasibility, spontaneity, and nature of the sorption process as given by the Eqs. (3), (4), and (5). Table 5 shows the thermodynamic values obtained. The negative ΔG0 values suggest that the uptake of CIP by BSAC was spontaneous and thermodynamically favored. ΔG0 values increased with an increase in temperature which suggests that the adsorption was more spontaneous at higher temperatures. Positive ΔH0 value indicated the process is endothermic and the increased randomness at the solid-liquid interface in the course of the adsorption of CIP unto BSAC was established by the positive value of ΔS0. Sharifpour et al. reported a similar observation for removal of CIP using activated carbon with multi-walled carbon nanotubes [103, 119].

Table 5 Thermodynamic parameters for the adsorption of CIP unto BSAC
Full size table

3.3.9 Adsorption mechanism

Consequent to results obtained from the pH studies, a likely mechanism of adsorption was inferred. The basic functional groups on both the surface of the BSAC and the CIP molecules largely influences the mechanism of the adsorption as the pH of solution was varied between 3 and 11. Maximum adsorption was observed within pH 6.5 and 9.0. At this pH range, electrostatic interaction between BSAC surface functional groups and the CIP charged molecules occurred, i.e., between the negatively charged carbon-oxygen groups (carbonyl and hydroxyl groups) present on the BSAC surface and the positively charged amine group in the zwitterionic specie of the CIP (mechanism shown in Fig. 7 a and b). However, the slight decrease in adsorption at pH between 9.0 and 11.0 was attributed to repulsion between the negatively charged adsorbent surface and the carboxylate group on the CIP that is negatively charged. The electrostatic repulsion is suspected to have dominated the adsorption in this phase. At pH above 9.0, both the CIP and BSAC remains negatively charged in solution. This led to a resultant decrease in the uptake of CIP by the adsorbent. Summarily, a fairly good uptake is observed at pH < 4.5 and pH > 8.5. Hence, a hydrogen bonding mechanism is proposed to have taken place particularly at pH > 8.5 where the adsorbent surface containing carbonyl groups could form hydrogen bonding (and possibly a dimmer) with the carbonyl groups on the CIP molecules (mechanism shown in Fig. 7c). Therefore, hydrogen bonding and electrostatic interactions are proposed mechanisms for the adsorption of CIP onto BSAC surface. Wang et al. (2010), Punyapalakul and Sitthisom (2010), and El-shafey et al. (2012) also reported a similar mechanism in the adsorption of CIP unto montmorillonite, modified silicates, and activated carbon prepared from date palm leaflets respectively [120,109,122].

Fig. 7
figure 7

Adsorption mechanism showing (a and b) electrostatic attraction between the negatively charged carbon-oxygen groups on BSAC surface and the positively charged amine group on the CIP molecule, (c) hydrogen bonding between carbonyl groups on BSAC surface and carbonyl groups on the CIP molecule

Full size image

3.3.10 Regeneration and reusability studies

Adsorbent reusability is an important parameter in determining their economic feasibility [123, 124]. For this to be accomplished, the adsorbent was regenerated first using different solvents and re-used over several cycles to check whether there is appreciable performance decline. Three different solvents were used for regeneration; they are water (H2O), hydrochloric acid (HCl), and sodium hydroxide (NaOH). The potential of the eluents used in regeneration follows the order H2O (95.12%) > HCl (81.62%) > NaOH (55.47%). Since water gave the highest desorption, it was used to desorb CIP molecules from BSAC over several cycles. It was observed that the percentage adsorption of CIP remained fairly constant without significant change for 4 cycles (Fig. 8). For accuracy purposes, 2% decrease in the values of percentage of CIP adsorbed was observed between the 1st and 4th cycles (Fig. 8). Apparently, the results obtained proved that the BSAC desorbed with water after each adsorption step has the ability to work with high efficiency for CIP removal from aqueous solution. The results obtained from this study proved that BSAC has attractive features (e.g., recycling, cheap, flexibility, fast adsorption kinetics, handling, and large adsorption capacity) that encourage their practical applications in wastewater management.

Fig. 8
figure 8

The reusability of BSAC for the removal of CIP for 4 cycles (each cycle stood for 50 min)

Full size image

3.4 Cost analysis

Table 6 shows the cost analysis of BSAC in comparison with commercial activated carbon (CAC). Activated carbon prepared in this study is approximately five and a half times cheaper than the CAC saving cost up to USD 258 per kilogram. The cost implication and the summary of all essential expenses made in the preparation of BSAC are shown in Table 6. The cost of orthophosphoric acid and distilled water accounts for the greater percentage of the total cost in the preparation.

Table 6 Comparison of price between BSAC and CAC
Full size table

4 Conclusion

This study shows the effectiveness of agricultural waste biomass, banana stalk, when functionalized with orthophosphoric acid in the adsorption of ciprofloxacin antibiotic from aqueous media. Well-developed pores from the SEM, high carbon content from the proximate and elemental analysis, and the observed functional groups from the FTIR demonstrated the suitability of the adsorbent in removing CIP antibiotics from aqueous solution. The Langmuir isotherm model best describes the adsorption data, thus establishing that the uptake of CIP was of a monolayer approach on the BSAC surface. Experimental data were best explained by the pseudo-second-order kinetic model and adsorption mechanism was mainly controlled by the intraparticle diffusion model. Thermodynamic studies revealed that the free energy change was negative, while the enthalpy and entropy values were positive, indicating that the sorption process is endothermic and spontaneous. This study reveals that banana stalk, an inexpensive activated carbon precursor, can efficiently remove CIP from aqueous solutions.