1 Introduction

Ammonia is part of the natural cycle of nitrogen circulation in nature; however, an artificial source of ammonia in the case of water pollution causes excessive eutrophication and algal bloom, which leads to disturbance of the balance of the aquatic ecosystem. Increased ammonia concentration in water has a particularly negative effect on fish development causing gill damage, hyperplasia, and significant reduction of growth rate (Cui et al. 2016; Conley et al. 2009; Długosz and Banach 2018b).

Ammonia occurs naturally in the environment mainly in the form of ammonium salts. It is released in the volatile NH3 form mainly due to decomposition of nitrogen-containing organic substances and as a result of volcanic activity. Another natural source of ammonia is electrical discharges that arise in the atmosphere that allow the reaction of oxygen with molecular nitrogen. As a result of the reaction, ammonia and its oxidation products are formed, which subsequently react with each other to form ammonium nitrate (V) and ammonium nitrate (III). Part of the nitrates produced according to this mechanism reaches the earth’s surface, from which they are absorbed by plants, because they contain nitrogen in a form that is available to them (Appl 1999). Ammonia is widely used in the production of fertilizers and animal feed as well as in the production of fibers, plastics, explosives, paper, and rubber (Šiljeg et al. 2010). The largest uncontrolled emission of ammonia to the environment comes from livestock breeding, manure management, and fertilizer application. In case of insufficient ventilation of the breeding rooms, respiratory diseases appear in the animals, which results in a decrease in the production of milk, meat, and eggs (Alberdi et al. 2016; Henry and Aherne 2014). Currently used techniques of removing ammonia from water and wastewater are biological method using activated sludge to remove nitrogen compounds in sewage treatment plants, division in the air stream, and ion exchange carried out in columns with zeolite filling, chemical precipitation, ion exchange, membrane separation (Zhang et al. 2009; Jorgensen and Weatherley 2003; Shalini and Joseph 2012; Huang et al. 2015).

An alternative to commonly used techniques is sorption of ammonia on a properly selected sorbent. The sorbent should meet a number of requirements: be easily accessible and durable both mechanically and chemically, have high sorption capacity and high affinity for analyte, and enable recycling by lowering process costs (Veliscek-Carolan et al. 2019; Azimi et al. 2019; Bhardwaj and Bhaskarwar 2018; Hu et al. 2019).

The aim of the research was to check the possibility of using raphia fibers to remove ammonia from aqueous solutions. Raphia fibers belong to organic materials containing lignin and cellulose in their composition (Xu et al. 2018). The ammonia removal process was carried out in the batch system, allowing to obtain information about the maximum sorption capacity of the material in relation to the analyzed analyte. Raphia fibers have also been subjected to chemical modification to improve its sorption capacity. Obtained results of the research allowed for adaptation of the sorption isotherm and determination of sorption kinetics allowing to obtain basic information on the mechanisms of ammonia sorption on raphia fibers. The solute removal rate controlling residence time in the liquid-solid interface is expressed by the sorption kinetics. One of the most important factors in the designing of the sorption system is the sorption rate, because the kinetics determine the reactor dimensions what is connected with the solute residence time. In order to understand the sorption kinetics, various models have been studied.

Pseudo-first-order and pseudo-second-order kinetic models are the most used to study metal sorption kinetics to solids (Veneu et al. 2018; Guo and Wang 2019).

2 Materials and Methods

2.1 Materials

The research involved the use of raphia fiber (Raphia farinifera) purchased in the commercial store. Raphia test was prepared by cutting fibers into pieces with the length of approx. 3–4 mm and a five-fold rinsing in demineralized water. One rinse cycle lasted 10–15 min. The aim of the washing was to purify and unify the research material. The most favorable purification conditions were obtained by mixing approx. 20 g of raphia in approx. 1.5 dm3 of demineralized water. All used solutions were prepared in distilled water. Reagents used in the tests were purchased from Sigma-Aldrich and were characterized by analytical purity.

2.2 Surface Modification

The process of surface modification of the raphia fibers was carried out using aqueous salt solutions according to Table 1. Substitution of silicon by aluminum atoms in the crystal backbone leads to an additional negative charge balanced by surrounding counterions (such as Na+, K+, Ca2+, and Mg2+), and these counterions are easily exchanged by other surrounding cations in contact solution. To 10 g of raphia, 100 cm3 of the modifying solution was added and mixed for a limited time.

Table 1 Modification of the surface of raphia fibers
Full size table

According to the fundamental of ion exchange between solid and liquid phases, the ion exchange process between zeolite frame and aqueous ammonium solution can be expressed by the following equation.

$$ \mathrm{Rf}\hbox{--} {\mathrm{Me}}^{\mathrm{n}+}+{{\mathrm{n}\mathrm{NH}}_4}^{+}=\mathrm{Rf}\hbox{--} {{\mathrm{n}\mathrm{NH}}_4}^{+}+{\mathrm{Me}}^{\mathrm{n}+} $$

where Ze and M represent the zeolite and the loosely held cations in zeolite, respectively, and n is the number of electric charge (Lin et al. 2013).

2.3 Sorption Process

The ammonia sorption process was carried out in a batch dynamic mixing system. The variable parameters of the sorption process were the concentration of ammonia and the time of running the process. The sorption process was carried out in a 60-cm3 PP beaker; 0.5 g of previously prepared raphia fibers and 40 cm3 of ammonia solution were used each time. Five concentrations of model solutions were determined for which further tests were conducted: 1, 3, 5, 7, and 9 mmol/dm3. The sorption time was 0.5, 1, 2, 4, 6, 10, and 15 min. After the sorption process, the solutions were filtered, and then, the ammonia content was determined by titration.

2.4 Physicochemical Characteristics

The analysis of the elemental composition of raphia fibers before and after the sorption process was carried out using the compact energy-dispersive spectrometer X-ray PW4025/00 MiniPal by PANalytical B.V. A specific surface area test was carried out using a ASAP 2010 deaerator station. Before measurement, samples were dried in a helium atmosphere at 110 °C for 8 h, then under vacuum at 100 °C and 0.001 Tor for 8 h. Surface analysis was performed with an LEO 1430 VP scanning electron microscope. SEM photomicrographs were taken on Hitachi TM-3000 equipped with an X-ray micro analyzer EDS. CHN analysis was performed using the Perkin Elmer Type 2400 CHN analyzer. Fourier transform infrared spectroscopy was performed on a Scimitar Series FTS 2000 from Digilab.

2.5 Equilibrium Studies

The equilibrium parameters of ammonia sorption on raphia fibers were modeled on the basis of the isotherms: Langmuir, Freundlich, Temkin, and Dubinin-Radushkevich isotherms (Table 2).

Table 2 Isotherm model equations
Full size table

2.6 Kinetic Models

Investigating the influence of the sorption time allows to determine the mechanisms of the sorption process. In order to determine the effect of contact time of raphia fibers (before and after modification), sorption capacity of raphia fiber in relation to ammonia at different times at various initial concentrations was investigated. Table 3 presents the kinetic models used.

Table 3 Kinetic model equations
Full size table

3 Result and Discussion

3.1 Materials Characterization

The raphia fibers prior to the sorption process have a heterogeneous structure with a fibrous structure (Fig. 1a), whereas the structure of raphia fibers after the sorption process (unmodified—UM) was characterized by a structure with greater homogeneity (Fig. 1b). BET analysis showed that the specific surface area of raphia fibers had a specific surface area of 1.079 m2/g; moreover, the fiber surface is characterized by a mesoporous structure (Fig. 2).

Fig. 1
figure 1

SEM micrograph of raphia fiber: a before sorption, b after sorption

Full size image
Fig. 2
figure 2

Porosity graph of coconut fiber

Full size image

Analysis of the XRF spectrum of raphia without surface modification and after its modification showed the presence of nine elements. The material was marked with the highest content of calcium, silicon, phosphorus, and potassium. In addition, based on elemental analysis of CHN, it was found that the content was equal to C: 46.04%, H: 6.18%, and N: 0.77%. These values are similar to those obtained by other researchers during analyses of organic materials (Bispo et al. 2018; Chwastowski et al. 2017).

The results of the semi-quantitative (non-patterned) analysis are shown in Table 4. The XRF technique allows the determination of elements heavier than sodium (MNa = 22.99 g/mol). Raphia as an organic substance consists mainly of carbon (MC = 12.01 g/mol), which is why the presented percentage composition of elements present in the structure of raphia is inflated and is illustrative (Hunt and Speakman 2015).

Table 4 Elemental composition of raphia without modification and modification
Full size table

Raphia without surface modification from the elements detected by the XRF technique contains the most calcium (44%). Silicon, phosphorus, and potassium are present in a smaller amount with a percentage of 17, 15, and 10%, respectively. The analysis showed the presence of sulfur, titanium, manganese, iron, and nickel.

NaNO3 modification caused a decrease in the content of calcium by 45% and potassium by 100%, which resulted in an increase in the percentage of silicon, phosphorus, and sulfur.

Percentages of calcium, silicon, phosphorus, and potassium are similar and amount to 24, 26, 27, and 15%, respectively.

Modification of K2SO4 caused an increase in the potassium content by 420%, resulting in a 66% decrease in calcium content. The element with the highest percentage share in this sample is potassium—52%.

The NaCl modification reduced the silicon content by 71% and potassium by 100%, resulting in an increase in the percentage of calcium by 36%. The element with the largest percentage share in this sample is calcium—60%.

Figure 3 shows the IR interferogram of the raphia without surface modification and with the modified surface made before and after the sorption process. In the range of 3600–3200 cm−1, there is an area characteristic for the vibration of the hydroxyl group (–OH). The peak at 1050 cm−1 corresponds to the amino group (–NH), peak 1650 cm−1 carbonyl group (–COOH), peak 2900 cm−1 group –CH2– (Staroń et al. 2017). The peak at 2360 cm−1 is the result of the presence of CO2. With each measurement, the intensity of this peak increases as opening the measuring chamber of the spectrometer to place the sample caused the diffusion of CO2 from the air in the laboratory inside the device. On the presented interferogram, there are no changes in the peaks for the modified raphia compared to the unmodified material, what results from the modifying substances used. Salts change the acidic and basic surface groups, which changes only the FTIR intensity not spectrum. In addition, slight shifts in the spectrum after the sorption process can be observed, which is possible with overlapping of the peaks associated with NH4+ groups (Petit et al. 2006).

Fig. 3
figure 3

Raphia IR interferogram

Full size image

One of the most important factors affecting the degree of removal of ammonia from the solution is the contact time of the solution with the sorption bed. In Fig. 4, it is observed that with sorption time prolongation, the sorption capacity qt of raphia increases, up to the limit of 0.11 mmol/g for unmodified material, 0.13 mmol/g for modified with NaCl, 0.10 mmol/g for NaNO3, and 0.1 mmol/g for K2SO4. The highest sorption rate occurs at the beginning of the process until about the second minute. Then, it decreases, until reaching the sixth minute of sorption equilibrium between the content of ammonia in the solution and sorbated by raphia.

Fig. 4
figure 4

Sorption of ammonium on raphia fiber over time at different initial concentrations, modification: a without modification, b NaCl 3 h, c NaCl 54 h, d NaNO3 17 h, e K2SO4 16 h

Full size image

Figure 5 shows the degree of removal (Re) of ammonia at the equilibrium for individual concentrations of initial ammonia in the model wastewater. The lower the initial concentration of ammonia, the higher the ammonia removal in equilibrium state Re. The highest removal rate was obtained using raphia modified with NaCl for 3 h. Increase of sorption with unmodified raphia surface by NaCl modification for 3 h occurs only at initial ammonia concentrations below 3 mmol/dm3. The lower the initial concentration of ammonia, the higher the ammonia removal in equilibrium state Re. The highest removal rate was obtained using raphia modified with NaCl for 3 h. Increase in sorption with modified raphia by NaCl for 3 h occurs only at initial ammonia concentrations below 3 mmol/dm3. The modification of raphia surface with NaCl for 54 h, NaNO3, and K2SO4 increases the sorption of ammonia only for a solution 1 mmol/dm3. In other cases, modification of the sorbent surface did not improve the sorption capacity of raphia in relation to ammonia. The modification of raphia surface with NaCl for 54 h, NaNO3, and K2SO4 increases the sorption of ammonia only for a solution of 1 mmol/dm3. In other cases, modification of the sorbent surface did not improve the sorption capacity of raphia in relation to ammonia.

Fig. 5
figure 5

The degree of removal (Re) of ammonia on the surface of raphia

Full size image

3.2 Equilibrium Studies

In order to select a model of ammonia sorption equilibrium on raphia, the obtained data were analyzed in equilibrium. Isotherm equations and determined isothermal parameters for individual sorbents are presented in Table 5. In the case of raphia without surface modification and modified NaCl for 3 h, the best fit to the measurement data was obtained for the Langmuir model (determination coefficient, respectively: R2 = 0.9653, R2 = 0.9510). The sorption equilibrium of raphia-modified surface for 54 h with NaCl and K2SO4 is best described by the Freundlich model (R2 = 0.9634 and R2 = 0.9972, respectively) and raphia modified with the NaNO3-D-R isotherm (R2 = 0.9892).

Table 5 Equations and parameters of the sorption equilibrium for individual sorbents
Full size table

Based on the Langmuir equation, one can determine the RL constant (\( {R}_L=\frac{1}{1+{K}_F{C}_0} \)), from which it can be concluded whether the conditions of the sorption process are favorable (RL = 0: sorption is reversible, 0 < RL < 1: sorption conditions are preferred, RL = 1: the nature of the sorption is linear, RL > 1: the sorption conditions are unfavorable) (Milonji et al. 2002).

For all initial ammonia concentrations, the KL parameter ranges from 0 to 1. On this basis, it can be assumed that sorption conditions of ammonia on raphia are favorable.

Figure 6 presents graphical representations of sorption equilibrium models for all types of sorbent. For individual cases, the measurement data deviate from the linearity: NaCl-modified raphia for 3 h and NaCl for 54 h in the Temkin model, raphia modified with NaCl for 54 h in the Langmuir model. In other cases, the R2 determination coefficients are above 0.9249. On this basis, it can be concluded that all four models of sorption equilibrium predict the equilibrium of sorption of ammonia on raphia in a reliable manner. This may be due to the concentration of ammonia in the modeled waste at the millimole level.

Fig. 6
figure 6

Isotherms of ammonia sorption on raphia

Full size image

3.3 Sorption Kinetics

In order to confirm the sorption mechanism, an analysis of kinetics of ammonia sorption on raphia was performed. Four kinetic models of sorption were calculated: the pseudo-first-order model, the pseudo-second-order model, the Elovich model, and the intramolecular diffusion model. Table 6 presents sorption parameters of individual kinetic models for a given sorbent: unmodified raphia and four surface modification.

Table 6 Kinetic parameters of different sorption models of ammonium
Full size table

Compared with natural raphia fibers, the modified material showed higher ammonium adsorption capacity. For the lowest ammonium concentration, an increase in sorption capacity was observed for all modifications. The highest increase in sorption capacity was observed for 3-h modification with NaCl, which was equal to 28%. For other modifications, it was average 18–19%. At concentrations of 3 and 5 mmol/dm3, an increase in sorption capacity was observed only on material modified with NaCl for 3 h (~ 15% and ~ 2.5%). For the remaining modifications, the sorption capacity was reduced by 17–19% for 3 mmol/dm3 and 12–22% for 5 mmol/dm3, respectively. At a concentration of 7 mmol/dm3, an increase in sorption capacity was observed for three modifications by ~ 27% (NaCl 3 h), ~ 1.5% (NaNO3), and ~ 0.5% (K2SO4). However, for a concentration of 9 mmol/dm3, an increase in sorption capacity was observed for NaCl modification, ~ 6.5% for 3 h and ~ 4.5% for 24 h. Material modified with NaNO3 and K2SO4 decreased in its sorption capacity by ~ 21% and 14.5%. During the study, it was observed that the rate of reaching equilibrium between the bed and ammonium ions changed. For unmodified material, equilibrium was reached after 4 min for concentration equal to 1 and 3 mmol/dm3, and for 5–9 mmol/dm3 after 6 min. In the case of NaCl modification for 3 h, equilibrium time was reached after 4 min for all tested concentrations. Other modifications resulted in equilibrium being reached after 6 min at all concentrations tested.

The most faithful mathematical description of the sorption kinetics of ammonia on raphia provides a pseudo-second-order model. The R2 determination coefficient for all sorbent types and the entire range of concentrations tested is at least 0.9955 for this model. On this basis, it can be assumed that sorption of ammonia on raphia is of chemical nature (Ho 2006).

Figure 7 presents graphical interpretations of the sorption kinetics model for particular types of sorbent. One can notice a high adjustment of the measurement data to the pseudo-second-order model.

Fig. 7
figure 7

Line graphs of pseudo-second order model of sorption kinetics for ammonium: a UM, b NaCl 3 h, c NaCl 54 h, d NaNO3 17 h, e K2SO4 16 h

Full size image

4 Conclusion

The research confirms that raphia sp. fibers can be used as a biosorbent to remove ammonia from water. The sorption capacity of raphia in relation to ammonia ranges from 0.093 mmol/g for NaNO3-modified raphia to 0.121 mmol/g for NaCl-modified raphia for 3 h. The ammonia removal rate for unmodified raphia for concentrations of 5, 7, and 9 mmol/dm3 is equal to 24, 17, and 17%, while for concentrations of 1 and 3 mmol/dm3 equal to 42% and 51%, respectively. The highest increase in the removal of ammonia from the model solution relative to raphia with unmodified surface was observed for modification of the raphia surface using NaCl for 3 h. At the lowest initial concentration, 1 mmol/dm3, the degree of ammonia removal increased by 60%, while for the highest initial concentration, 9 mmol/dm3, increased by 8%. The Freundlich sorption equilibrium model best describes the sorption of ammonia on all types of sorbent used (R2 equal to at least 0.9248), but the other models (Langmuir, Temkin, D-R) also correspond to the experimental data to a satisfactory degree (R2 equal to at least 0.8706). The kinetics of ammonia sorption from the model sewage on unmodified raphia and all types of surface modification is best described by the pseudo-second-order sorption kinetics model (R2 coefficient is at least equal to 0.9964). On this basis, it can be concluded that the adsorption is of chemical nature.