1 Introduction

Banana is one of the most consumed fruits in the world, with emphasis on the African continent, which in 2019 reached 21.5 million tons, representing 18.4% of world consumption. Regarding production, Uganda is the largest producer of plantains with approximately 9.6 million tons, followed by Ghana and Rwanda, however, it is estimated that each hectare of cultivated banana generates about 220 tons of waste [1, 2].

This organic waste is mostly discarded in rivers, lakes, and dumps, causing environmental risks, due to the release of greenhouse gases, so their recovery and reuse are necessary [3]. The main waste generated is the banana pseudo-stem, which has promising characteristics for application as an adsorbent for the removal of contaminants, due to its low cost, being a renewable resource, and having surface chemical groups, which are favorable to the adsorption process [4, 5].

In parallel, there is an increase in demand for African textiles and clothing globally, due to its colorful, elegant, and representative pattern, in which more and more stores are integrating African influences into their collection. It is estimated that from 2019 to 2024 there will be a 5% growth of the textile industry in Africa [6]. In this context, dyes are widely used in the production process of the textile industries, they are non-biodegradable and stable compounds, with toxic, mutagenic, and carcinogenic potential, in addition to being resistant to light, heat, and oxidizing agents. It should be noted that 10–15% of the dyes used do not adhere to textile fibers and, therefore, are discarded with the effluent [7,8,9]. Therefore, the release of wastewater containing dyes, even in small concentrations, can lead to ecological problems and human health, such as the increased risk of cancer, respiratory problems, skin irritations, and reduced oxygen levels in the water, threatening fauna and flora [10,11,12].

Regarding dye wastewater treatment techniques, adsorption stands out as a low-cost and high removal efficiency, since wastewater treatment plants do not effectively remove these contaminants, also, it makes it possible to obtain alternative adsorbents from solid waste from industrial and agroindustrial activities, and nature, promoting more sustainable and technological uses for them [13, 14].

Fresh biomass from banana pseudo-stem is an alternative biosorbent to remove different contaminants, such as lead (II) [15], methylene blue [4], and safranin dye [16], with a maximum adsorption capacity of 34.21, 33.78 and 21.74 mg g−1, respectively. Also, biomass from banana pseudo-stem can be applied in the production of activated carbon [17], biochar [18], hydrogels [19], and modifications through chemical treatments [20].

Thus, taking into account the projected growth for African textile industries and the continuous generation of banana pseudo-stem solid wastes. This paper aimed to study the banana pseudo-stem waste as an alternative adsorbent for the removal of textile dye. For this, the characterization of the superficial area, morphology, and chemical structure was carried out, also, to obtain the best operational conditions (pH, temperature, and agitation speed) and kinetic and equilibrium data.

2 Materials and Methods

2.1 Obtainment and Preparation of Biosorbent

Biomass used comes from the banana (Musa sp.) pseudo-stem, collected in the western region of the state of Paraná, Brazil. The dye used was the BF-5G blue reactive dye (Texpal, Brazil). All reagents used were of analytical grade.

The biomass waste was dried by convective drying (CienLab, CE-480, Brazil) at 50 °C for 24 h. The biomass was ground in a hammer mill (SPLabor, SP-33, Brazil) with #2 to #20 mesh size.

2.2 Characterization of Banana Pseudo-steam Waste

Surface area and porosity analyzes of biosorbent samples were determined by Brunauer, Emmett and Teller [21], using Quantachrome autosorb − 1 − CMS−1 apparatus, similar to the conditions described in [22]. Scanning electron microscopy (SEM) was performed using a high-resolution electron microscope (Shimadzu superscan, SS-550, Japan) with magnifications of 2000 and 5000 times.

The chemical structure of the biosorbent before and after dye adsorption was evaluated using Fourier transform attenuated total reflection spectrometry (Agilent, Cary 630 FTIR, USA) in the range 4000–650 cm−1, at 4 cm−1 resolution and 48 scans per spectrum.

2.3 Batch Adsorption Tests

Firstly, the point of zero charge (pHPZC) of the banana pseudo-stem was determined by potentiometric titration, following the methodology proposed by Dav-anche et al. [23] For this purpose, two conical flasks were used, each initially containing 5 g of biosorbent and 100 mL of 0.1 mol L−1 NaNO3 solution. Then, the solution was titrated from one conical flask with 0.1 mol L−1 NaOH and the other with 0.1 mol L−1 HNO3. The pH value was checked before and during the titration, noting the volume of the titrant solution that was added for each pH change until it remained constant.

To determine the best operational conditions for the adsorption of the dye by the biomass of the banana pseudo-stem waste, the pH of the dye solutions was modified (1–11) by the addition of 0.1 mol L−1 HCl or NaOH and the effect of temperature (30, 40, and 50 °C) and agitation speed (30, 60, and 90 rpm) were also evaluated. The tests were carried out using 300 mg of banana pseudo-stem waste, added in 50 mL of the BF-5G blue reactive dye solution at 100 mg L−1, kept in a mechanical shaker (SPLabor, SP-222, Brazil) under constant agitation for 2 h.

Afterward, the samples were centrifuged and read in a spectrophotometer (Varian, Cary 50 Scan, United States) was performed at 610 nm, that is the wavelength of the maximum absorption of BF-5G blue dye (Supplementary material). An external calibration curve (y = 48.220x) with R2 = 0.999 was used to quantification.

From the concentration data obtained, the amount of dye adsorbed was calculated according to Eq. 1.

$$\mathbf{q}=\frac{\mathbf{V}({\mathbf{C}}_{0}-\mathbf{C})}{\mathbf{m}}$$
(1)

where C0 and C (mg L−1) are the adsorbate concentrations in the initial and final aqueous solution, respectively, V (L) is the volume of solution and m (g) is biosorbent weight.

The kinetic and equilibrium tests were carried out in batch, with samples in triplicate, according to the best operational conditions defined in the preliminary tests (30 °C, 60 rpm, and pH 1). For kinetics, 50 mL of the 100 mg L−1 dye solution and 0.3 g of biosorbent were added in 125 mL conical flasks.

The isotherm assays were performed using 50 mL of dye at 200 mg L−1 at 30 °C, 60 rpm, and pH 1, with the weight of the banana pseudo-stem waste ranging from 0.02 to 0.60 g. We consider that the equilibrium was reached after the concentration of the dye did not vary in 3 consecutive measurements at 1 h intervals.

In modeling mathematical, kinetic models of Pseudo-first order (PFO) [24], Pseudo-second order (PSO) [25], and Elovich [26] were fitted to the kinetic experimental data, (Eqs. 2–4). For isotherms, Langmuir [27], Freundlich [28], and Redlich–Peterson [29] models were fitted to the experimental equilibrium data, (Eqs. 5–7, respectively). The kinetic and equilibrium equation models are shown in Table 1.

Table 1 Kinetic and equilibrium models equations
Full size table

The modeling mathematical was performed using the software Origin 8.0 (Originlab Corporation, USA).

3 Results and Discussion

3.1 Characteristics of Banana Pseudo-stem

Table 2 presents the isotherm of nitrogen adsorption–desorption of the banana pseudo-stem waste.

Table 2 Nitrogen adsorption–desorption isotherm by BET/BJH analysis of the banana pseudo-steam
Full size table

In Table 2, the values of surface area (10.08 × 10–2 m2 g−1) and average pore diameter (1.639 × 101 Å) for the banana pseudo-stem, indicate a material with a predominance of micropores, once it is considered a microporous material if it has a lower diameter at 2 nm [30,31,32]. In this sense, biosorbent has a low-surface area due to a very small number of pores, allowing the occurrence of multilayer adsorption [33].

The microporous surface area of plants and wastes in raw form studied as biosorbents present values close to those obtained in the present work, such as pistachio hull powder and orange peel, varying from 0.77 and 1.46 nm [34, 35].

Also, according to Maeda et al. [36] the BF-5G blue reactive dye molecule has a diameter of 26.16 Å, therefore, the biomass of the banana pseudo-stem has a smaller diameter than the dye, suggesting that adsorption occurs on the material surface and does not on its inner surface. A similar behavior was reported by Msaadi et al. [37] who observed the molecules adsorbed on the surface by means of XPS measurements, and Sassi et al. [38], that confirmed through thermodynamic and isothermal analysis the physisorption of adsorbate particles. Figure 1 shows the morphological analysis of the banana pseudo-stem by SEM.

Fig. 1
figure 1

Characterization of the banana pseudo-stem by SEM with magnifications of: a 2000 and b 5000 times

Full size image

Scanning electron microscopy (SEM) images (Fig. 1) show that the particles of the banana pseudo-stem waste, present aligned and organized fibers, which leads to a smaller surface area, similar behavior is reported by Casqueira and Lima [39], using banana pseudo-stem in the adsorption of Cr(III).

Figure 2 shows the Fourier-Transform Infrared (FTIR) analysis of the banana pseudo-stem before and after the adsorption of the dye BF-5G blue reactive dye.

Fig. 2
figure 2

FTIR-ATR of the banana pseudo-stem before and after dye adsorption

Full size image

The Fourier transform infrared analysis by Attenuated Total Reflection (FTIR-ATR) (Fig. 2) indicates that there are no significant differences between the transmittance signals before and after the dye biosorption by the banana pseudo-stem waste, suggesting that there is no chemical change in the biosorbent, which is detectable by this analytical technique.

A signal at the wavelength of 3400 cm−1 is attributed to the vibrations of the O–H group, which are found in cellulose and pectin molecules [33]. A signal at 2900 cm−1 is due to the C–H groups, which are present in polysaccharides [15]. Around 1700 cm−1, the signal related to the C = O elongation is observed, due to hemicellulose and in 1600 cm−1 the signal is related to the O–H bond found in the adsorbed water [40]. Signals between 1250 and 1400 can be related to C–O, C–H, and C–O–H bonds, present in carboxylic acids. In the range of 1100–1150 cm−1, it is associated with C–O and C–O–C bonds, found in alcohols and carboxylic acids, present in cellulose and polysaccharides [33].

3.2 BF-5G Blue Reactive Dye Biosorption by Banana Pseudo-stem Waste

The influence of the variables point of zero charge (pHPZC), pH, temperature, and agitation speed, as well as the results of the kinetic and equilibrium essays of the capture of the BF-5G blue reactive dye by the banana pseudo-stem, are shown in Fig. 3.

Fig. 3
figure 3

BF-5G blue reactive dye biosorption by banana pseudo-stem: influence of: a pHPZC, b pH, c temperature, and d agitation speed, e kinetics at 100 mg L−1, and f isotherms at 200 mg L−1

Full size image

In the analysis of the point of zero charge (Fig. 3a), it is observed that the total surface charge decreases with increasing pH and that the pHPZC is approximately 6.5. The evaluation of the point of zero charge is essential because, from the pH condition, the charge of the adsorbent surface can be defined, therefore when pH < pHPZC the surface is positively charged, favoring anion adsorption, and when pH > pHPZC is negatively charged, favoring the adsorption of cations [41].

In the pH tests (Fig. 3b), the largest removals (95 and 93%) occurred at pH 1 and 2, respectively. From pH 3 to 11, there was a reduction in the removal percentage, which varied from 35 to 69%. BF-5G blue reactive dye, in the acidic range (pH 1–3), is deprotonated in an aqueous medium, generating a high negative charge density, and favors adsorption with the banana pseudo-steam biomass (which is protonated: pH < pHPZC), by electrostatic interactions [33, 42]. In the low acid range (pH 4–6), the greatest reductions in removal percentages occur, due to the greater proximity to the point of zero charge of the banana pseudo-steam (pHPZC = 6.5), resulting in the neutralization of the surface charge of the biosorbent [33]. However, it is noteworthy that in neutral and basic medium (pH 7–11), there is a small increase in the percentage of dye adsorption, indicating that electrostatic interactions may not be the only adsorption mechanism, since, at pH > pHPZC, the surface charges of the biosorbent are deprotonated, that is, there is electrostatic repulsion with the anionic groups of the dye, therefore, the formation of hydrogen bonds is possible.

Temperature is an important parameter because it influences the adsorption equilibrium, due to interference in the agitation of the molecules and thus in the dye-adsorbent electrostatic forces [43].

The results obtained in the temperature test (Fig. 3c), indicate that the increase in this parameter did not increase the removal of dye in the studied range (30, 40, and 50 °C), with removals close to 95% for both. Therefore, the temperature used was 30 °C, due to the lower energy cost, an analogous result was obtained by Rigueto et al. [22] in the adsorption of BF-4B red reactive dye by water hyacinth roots.

In the agitation speed test (Fig. 3d), a higher percentage of removal (95%) was observed at speeds of 60 and 90 rpm. With increased agitation, there is an increase in the removal yield, due to the greater transfer of convective mass of the solute present in the aqueous solution [44, 45]. Therefore, for subsequent tests, preliminary tests determined the best conditions for pH 1, temperature of 30 °C, and agitation speed of 60 rpm.

To verify the effect of time and the adsorption capacity of the biosorbent, equilibrium and kinetic tests were performed. In the kinetic curves (Fig. 3e), it is evident that in the first minutes the adsorption density increases from 0 to 40 mg g−1 and after 100 min, it continues to increase. The equilibrium was verified after 240 min, with a removal rate of 96% of the textile dye. A similar equilibrium time was verified by Freitas et al. [46], in the adsorption BF-5G blue reactive dye by orange peel.

The adsorption isotherm of the BF-5G blue reactive dye by banana pseudo-stem is shown in Fig. 3f, and following the Giles’ classification [47], has an L1 profile. In the L isotherm type, it is suggested that adsorption occurs on the material surface, by weak forces such as van der Waals, and as the active sites are occupied, the greater the competition between the solute molecules to fill the empty sites, subclass 1, it is reported that the active sites were not filled [47, 48].

The possible adsorption mechanism between the dye and the biosorbent occurs through electrostatic interactions, due to the effects of pH and pHPZC. Surfaces from biomass have amine and carboxylic groups in their structure, which are protonated in an acidic environment, promoting a positive charge density on the material surface, which causes a strong electrostatic attraction of the dye, which is negatively charged, favoring the adsorption [33].

3.3 Kinetic and Equilibrium Essays

Table 3 shows the parameters of the PFO, PSO, and Elovich models (Eqs. 2–4) fitted to the kinetic experimental data. The highest values of the determination coefficient (R2) were observed in the PSO and Elovich models. The PSO model suggests that the rate-limiting step is the surface adsorption that involves chemisorption, where the removal from a solution is due to physicochemical interactions between the two phases. Also, Elovich indicates that adsorption occurs by internal and external mass transfer, and reinforces that chemosorption on the surface of the biosorbent [49,50,51].

Table 3 Fit of the parameter values of the kinetic models for biosorption of BF-5G blue reactive dye by banana pseudo-stem waste (30 °C, 60 rpm, and pH 1)
Full size table

Table 4 presents the fits of the Langmuir, Freundlich, and Redlich–Peterson isothermal models to the experimental equilibrium data. The determination coefficients (R2) obtained were equal to 0.99 for all models.

Table 4 Fit of parameter values of isotherm models for biosorption of BF-5G blue reactive dye by banana pseudo-stem waste (30 °C, 60 rpm, and pH 1)
Full size table

However, the Redlich–Peterson hybrid model combines elements of the Langmuir and Freundlich equations, and when the parameter β is equal to or close to 1, it tends to follow the fit of the Langmuir model, as in the case of this study (β = 1.68) [52]. The Langmuir model suggests that adsorption occurs in monolayer, with independent active sites, and with equivalent energy, and that each site contains only one dye molecule [48]. As for the Freundlich model, the parameter nF > 1, indicates favorable adsorption, and 1/nF < 1, indicates a normal Langmuir isotherm (1/nF = 0.71) [53, 54].

Therefore, as both models showed good fits, it is suggested that at concentrations lower 200 mg L−1, adsorption is represented by the Langmuir model and limited to the monolayer, however due to the good fit of the Freundlich model and the low area surface and insignificant pore volume obtained, it is possible that at concentrations greater than 200 mg L−1, the possibility of multilayer formation occurs in the adsorption of the dye, and a similar effect was reported by Módenes et al. [33].

In the present work, the maximum biosorption capacity of BF-5G blue reactive dye by banana pseudo-stem in monolayer estimated by Langmuir was 72.21 mg g−1. Other studies have been reported biosorbents with similar capacities in the adsorption of this textile dye, such as soybean hull and malt bagasse, with 72.43 and 42.58 mg g−1, respectively [55, 56]. Thus, the banana pseudo-stem was effective in removing the BF-5G blue reactive dye, with characteristics favorable to adsorption, showing promise for removing dyes from aqueous solutions.

4 Conclusions

In the present work, the use of banana pseudo-stem as an alternative biosorbent for the removal of the BF-5G blue reactive dye from an aqueous medium was studied. The banana pseudo-stem showed a specific surface area of 10.08 × 10–2 m2 g−1. The ideal conditions obtained for the biosorption tests were at pH 1, 30 °C, and 60 rpm.

Regarding kinetics, the Elovich and pseudo-second order models fit the experimental data. Equilibrium was achieved after 240 min with 96% dye removal. In isotherms, the Langmuir model can describe the adsorption equilibrium, with a maximum monolayer adsorption capacity of 72.21 mg g−1. In general, the banana pseudo-stem proved to be effective in removing the textile dye.

In general, it was found that the banana pseudo-stem has promising characteristics for use as a textile dye adsorbent, and therefore, further studies are needed to promote the development of materials based on this waste widely generated in the African continent, enabling regeneration and facilitated reuse after adsorption.