The coconuts were collected in February of 2016, in semi-open field at the Jack Norment Camp in Caacupé, Paraguay. The pulp was processed by the Botany and Physical Chemistry Departments at Universidad Nacional de Asunción. First, the epicarp was removed and mesocarp cuts were made. The slightly insoluble fibers were then dried at 40 °C for two days (González et al., 2018; Yubero et al., 2016a; b). The size of the dry fiber was sieved for size selection, to be between 297 and 350 µm. To conserve the fiber, 0.5 g of sodium azide was added and homogenized in about 4 g of the size-selected fiber.

Fourier transform infrared spectroscopy (FTIR) analysis from 500–4000 cm^–1 was conducted on the native pulp and the pulp in contact with a 25 ppm Pb(II) solution for 5 min, 30 min, and 5 h, using the Shimadzu Europa GmbH IR Prestige-21. Thermogravimetric analysis (TG-DTA) was carried out using TA Instruments ST 2960 Simultaneous DTA-DTG on a sample of m = 5.6 mg of native pulp/sodium azide mix. The sample was heated at β = 20 °C min^–1 under air atmosphere (100 mL min^–1) in an open aluminum crucible (70 mL) during thermogravimetric analysis (TG-DAT analysis. Scanning electron microscopy (SEM) images of the pulp azide mix after contact with Pb(II) were taken using FEG-SEM JEOL Model 7500F.

The distilled water was tested with a PHS-38W microprocessor from Bante Instruments. The pH was found to have a pH of 5.84 at 25.4 °C. As lead has a solubility between a pH of 2 to 7, distilled water was used in the following solutions (Wang et al., 2017). Lead nitrate, Pb(NO₃)₂, were used in the preparation of all solutions.

A 1 L solution of 50 ppm Pb(NO₃)₂ was prepared using 50 mg of anhydrous Pb(NO₃)₂. Solutions of 30, 40, and 50 ppm Pb(NO₃)₂ were prepared by adding 30, 40, and 50 mL of the 50 ppm Pb(NO₃)₂ solution to 50 mL volumetric flasks and filling them with distilled water.

To prepare the pulp for batch experiments, 200 mg samples of the pulp were washed three times with 25 mL of distilled water. The batch experiment was carried out by adding a sample of 200 mg of pulp to a 13 mL solution with an initial concentration of either 30, 40, or 50 ppm Pb(II) to conical centrifuge tubes. Each sample was rotated for a required time of either 5 min, 30 min, and 5 h at 8 rpm. Next, the pulp and solution were centrifuged for 5 min at 6000 rpm. The supernatant was extracted and the final concentration of Pb(II) was tested using AA-6300 Shimadzu atomic absorption spectrometer. It must be noted that this experiment was performed in duplicate (i.e. two samples of pulp in contact with 30 ppm Pb(II) for 5 min, two samples in contact with 30 ppm Pb (II) for 30 min).

The Langmuir (Eq. 1) and Freundlich (Eq. 2) adsorption isotherms were applied to the batch data using the following equations.

(1)

(2)

where KL is the Langmuir constant (L mg^–1), q_m is the maximum adsorbed capacity (when the adsorbate monolayer has formed, mg g–1), K_Fis the Freundlich constant for adsorption capacity (L mg^–1), n is Freundlich constant for adsorption intensity, and ci is the initial concentration adsorbate in solution (mg L^–1). qe is the adsorption capacity at equilibrium (mg g^–1), solved for using Eq. 3.

(3)

where cf is the final solution (mg L^–1), m is the mass of adsorbent (g), and V is the volume of solution (L).

Removal efficiency (Eq. 4) was also calculated for batch data using the following equation.

(4)

A 5 mL syringe with needle was used as the column. A sheet of packing plastic was placed at the bottom of the syringe and 200 mg of pulp were loaded into the column. The syringe needle was connected to a tube that fed through a peristaltic pump that regulated a flow rate of 4 mL min^–1, as seen in Fig. 1.

Twenty-five mL of distilled water were passed through the column to stabilize the column bed. After the pulp was stabilized, a 35 ppm Pb(II) solution was continuously passed through the column. Over a 3-time period, 15 mL samples of effluent solution were collected every 15 min in conical centrifuge tubes. In between sampling times, the effluent was collected in a waste beaker. The levels of Pb(II) in the conical centrifuge tube samples were tested using atomic absorption spectroscopy.

Figure 1.
Column experiment configuration.

The Thomas (Eq. 5) and Yoon-Nelson (Eq. 6) models were applied to column data using the following equations.

(5)

where Co is the original concentration of the solution (mg L^–1), Ct is the concentration of the effluent solution (mg L^–1), Q is the flow rate (mL min^–1), K_TH is the Thomas model constant (mL mg^–1 min^–1), q_TH is the theoretical adsorption capacity, and t is the time (min).

(6)

where C_o is the original concentration of the solution (mg L^–1), C_t is the concentration of the effluent solution (mg L^–1), flow rate of the influent solution (mL min^–1), t is the time (min), K_YN is the Yoon-Nelson constant (min^–1), and τ is the time to reach 50% saturation of the column.

The data for the infrared (IR) spectrum of the native pulp is found in Tab. 1. The native pulp had a characteristic fingerprint region, defined by bands 1–8 in Tab. 1, with representative bands at 1735, 1374, and 1064 cm^–1 (Yubero et al., 2015). These bands correspond with C=N/C=O, C-O-C, and C=S/S=O bonds (Chang, 1981).

Table 1.
Fourier transform infrared (FTIR) spectrum bands of A. aculeata pulp.

When the pulp was in contact with a 25 ppm Pb(II) solution, the bands in the fingerprint region shrink in intensity as the contact time with the lead solution increased, as seen in Fig. 2. This change in intensity, an indicator of change in dipole movement, indicates that the oxygen and sulfur bonds are most likely involved in capturing the Pb(II). Additionally, a band appears at around 2300 cm^–1 in the Pb(II) contact spectra, highlighted in Fig. 2, at a stronger intensity than in the raw pulp IR spectrum. This band could indicate the presence of Pb(II) nitrate.

Figure 2.
Fourier transform infrared spectrum (FTIR) of A. aculeata pulp in contact with 25 ppm Pb(II) solution for 5 min, 30 min, and 5 h; insert: expanded region of 2340–2380 cm^–1.

The results of the scanning electron microscopy (SEM) imaging are found in Fig. 3. These images reveal the pulp to have a rough, fibrous surface texture. The thermogravimetric analysis (TGA) profile of the pulp azide mixture (Fig. 4) was compared to the TGA of A. aculeata pulp in absence of sodium azide, from Fig. 4. It must be noted that in Corrêa et al. (2019) the pulp was heated at 10 °C min^–1 under synthetic air while the conditions for this experiment was 20 °C min^–1. The raw pulp had four decomposition steps, while the thermogravimetric/derivative thermogravimetric (TG/DTG) curves of the pulp sodium azide mixture reveals that there are three distinct decomposition steps (Fig. 4). In the raw pulp TG/DTG curve, the first step occurred between room temperature and 160 °C and was endothermic due to the presence of water in the fibers (Corrêa et al., 2019). Similarly, the first step in the pulp azide mechanical mixture is endothermic between 25 and 100 °C and can be associated with water molecule evolution.

Figure 3.
Field emission gun scanning electron microscope (FEG-SEM) mixed (secondary electron and backscatter electron) image of A. aculeata pulp with Pb(II) in (a) batch and (b) column.

Figure 4.
Thermogravemetric analysis (TGA) profile of A. aculeata pulp, % weight with (a) TG and DTG and (b) TG-DTA; mass sample of 5.6 mg; β = 20 °C min^–1; air (100 mL min^–1); open aluminum crucible (70 L).

The next decomposition step in the raw pulp TGA was exothermic and the largest; it occurred between approximately 190 and 300 °C and can be attributed to the decomposition of hemicellulose and cellulose. This step was followed by two small exothermic mass losses between approximately 390 and 425 °C that can be attributed to the decomposition of lignin (Corrêa et al., 2019). On the other hand, the pulp azide mixture demonstrated thermal stability until about 200 °C, proving that the sodium azide causes a slight stabilization in the material. The following step for the sodium azide mixture is strongly exothermic and can be associated with the kinetically fast decomposition of cellulose and the formation of the metallic Na species from sodium azide, which both simultaneously occur between 200 and 340 °C. The largest percentage weight loss occurs during this step at T^DTA_peak = 336 °C, which is right around the temperature that sodium azide decomposes according to literature. It can be concluded that the presence of sodium azide accelerates the decomposition of the material in this step. The residues from the decomposition reaction of cellulose and sodium azide to metallic Na then decomposes in the step between 350 and 510 °C. The DTG and DTA curves are in agreement with each other and the literature, revealing two small peaks that are associated with the slow kinetics steps of consecutive reactions (Fujimoto et al., 1990). A total of 94.3% of the pulp sodium azide mixture decomposed.

In the batch study, the removal efficiency (see Eq. 4), or performance in terms of amount of lead removed was found to reach a maximum of 91.9% when a 13 mL solution of 50 ppm of Pb(II) was placed in contact with 200 mg of pulp for 30 min. The Langmuir and Freundlich adsorption isotherms for 5 min of contact with the lead solution are found in Figs. 5 and 6, respectively. It is important to note that most results from the 5 h and 30 min contact time were found to be under the limit of quantification of the instrument (6.30 ppm Pb(II)), so these isotherms were not included in this paper (Silva, 2016). The results from the application of the Langmuir and Freundlich isotherm models applied to the batch results are summarized in Tab. 2.

Figure 5.
Langmuir isotherm of Pb(II) adsorption for 5 min contact time with A. aculeata.

Figure 6.
Freundlich isotherm of Pb(II) adsorption for 5 min contact time with A. aculeata.

Table 2.
Langmuir and Freundlich isotherms of Pb(II) adsorption in batch for 5 min contact time.

The R² value suggest that the Freundlich model fits the adsorption behavior slightly more than the Langmuir model. This indicates that the distribution of surface energy is heterogenous. The Freundlich model also is in accordance with the SEM images, as Freundlich adsorption commonly occurs on rough surfaces (Fig. 3) (Chang, 1981). The n value, the Freundlich constant for adsorption intensity is greater than 1, which indicates that the adsorption reaction is kinetically favorable. The Freundlich constant for adsorption capacity, KF, however, is low compared to other biomaterials in batch, as seen in Tab. 3. The pulp was found to have a low adsorption capacity, qe (Eq. 3), of 2.3 mg g^–1 in comparison to other biomaterials (Tab. 3). Since the adsorption capacity defines the amount of adsorbate taken up by the adsorbent, the low value indicates that it is not a promising biomaterial for filtration (Tab. 3). Future optimization studies regarding ideal conditions for adsorption or chemical treatment of the material are needed to improve the adsorption capacity.

Table 3.
Comparison of Freundlich adsorption parameters for various biomaterials.

The experimental breakthrough curve from the column experiment is found in Fig. 7 and the breakthrough parameters, along with the Thomas and Yoon-Nelson results, are summarized in Tab. 4. The breakthrough curve resembles the typical shape for column adsorption. The breakthrough time, which is defined as the time at which the concentration of the effluent solution is equal to 5% of the influent concentration, occurs in the first 15 min. This signifies an efficient mass transfer rate. The exhaustion time, the time at which the concentration of the effluent solution reaches 95% of the influent concentration, was found to occur between 2.50 and 2.75 h. The long exhaustion time indicates that the pulp-column system would have a long operating life.

Figure 7.
Experimental breakthrough curve for Pb(II) adsorbed on A. aculeata pulp.

Table 4.
The parameters in fixed-bed experiments for adsorption of Pb(II) on A. aculeata.

The Thomas (Eq. 5) and Yoon-Nelson (Eq. 6) are mathematical models for Langmuir adsorption used to describe the performance of the column. The Thomas model is typically used to calculate qth, the theoretical adsorption capacity of the pulp, which describes the maximum solid phase concentration of Pb(II) on the pulp. This was found to be 11.97 mg g^–1. The Yoon-Nelson model is a model for fixed bed, single component adsorption that assumes the decrease in probability of adsorption for each Pb(II) molecule is proportionate to the probability of breakthrough on the pulp by the Pb(II). When applied to the data from the column study, the Thomas and Yoon-Nelson models had low R² and, therefore, did not fit the adsorption behavior of Pb(II) (Tab. 2). The Rq, which is the ratio of the theoretical adsorption maximum in column compared to batch, was calculated using the results from the Thomas model in column and Langmuir model. The ratio of 1.022 shows that the column and batch systems have almost equal efficiency in adsorption of Pb(II), however, as the adsorption does not follow Langmuir behavior and the Thomas and Yoon-Nelson models have poor R² coefficients, the column results cannot truly compare the efficacies of both systems. It is necessary that more column data be collected so that an accurate performance assessment can be made.

at.novak23@gmail.com