Characterization resultsIn the pyrolysis phase, a significant portion of cellulose and lignin undergo conversion into volatile material, impacting the ultimate yield and augmenting the surface area. Consequently, approximately 29% of the carbonaceous skeleton formed from both residues was produced as a result of this process. This finding aligns with other studies that similarly employed ZnCl2 in a 1:1 ratio27,28,29. As shown in Table 1, the pore volume of the charred skin was 4.223 × 10–1 cm3 g−1, the pore diameter was 2.70 nm, and the surface area was 624.728 m2 g−1. Carbonization using zinc chloride also resulted in favorable textural characteristics for an adsorbent from the core (Table 1), with an even greater surface area compared to that of the skin (SBET = 1125.43 m2 g−1). Furthermore, the material showed good pore development (Vp = 3.241 × 10–1 cm3 g−1; Dp = 2.321 nm). The surface area is one of the main factors influencing the adsorption capacity of a material. However, carbonization with ZnCl2 may limit the good performance compared to that of PROP for both adsorbents due to certain factors. Compared with the results obtained with other carbons developed in the literature, we observed that the surface area obtained from carbonization with zinc chloride is significantly greater. For example, in a study carried out by Liu et al.30 using coconut shells as precursors, the surface area was reported to be 803 m2 g−1 after chemical activation with ZnCl2. Another study carried out by Ferrari et al.31 using grape seed waste reported a surface area of 542 m2 g−1 after carbonization with ZnCl2.
Table 1 Details regarding the pore volume and specific surface area of carbons produced utilizing ZnCl2 as the activating agent.Figure S1 enables the identification of the primary functional groups present on the surface of the materials. Initially, the functional groups within the carbonized material derived from the corozo core were analyzed. (Fig. S1A), and all functional groups identified in the second material (Fig. S1B) will not be reported. Therefore, the residues before and after carbonization share similar chemical groups. The 3441 cm−1 O–H bond region occurred in the same way as it remained after heating but with lower intensity32. The binding of CH (2918 cm−1), which was exclusively detected in the precursor material, suggested that the pyrolysis step led to the consumption of this group33. The disappearance of these functional groups corroborates the loss of volatile material that occurs during the pyrolysis process34. The presence of CO groups (1743 cm−1) in both materials is generally attributed to the presence of ketones35. The carboxylate stretching vibrations are associated with the spectral band at approximately 1636 cm−136. Aromatic rings are also found at 1452 cm−1 in the precursor material34. The band present in the 1041 cm−1 region corresponds to the secondary alcohol stretching of CO bonds37. Regarding the carbonization of the core (Fig. S1B), only the bands at 2851 cm−1 were not observed in Fig. S1A. Asymmetric vibrations of CH2 in the precursor were found at 2851 cm−138. Therefore, both materials share similar functional groups corresponding to structures that are formed by cellulose, and when lignin is added, the pyrolysis step reduces the heterogeneity of the surface, resulting in the formation of carbonaceous material with carbon as the final product.X-ray diffraction serves as a structural analysis methodology facilitating the elucidation of the crystalline structure of a material based on the acquired diffraction pattern. Figure S2 shows the X-ray diffraction patterns of the original sample and the charred sample for both residues. In the case of the sample in question, both patterns (Fig. S2A and B) show the presence of amorphous carbon in the form of a long diffraction band between 15° and 30°. Amorphous carbon is a disorganized and irregular form of carbon that does not have a definite crystal structure39. However, carbonization can lead to the formation of more organized and crystalline structures, which may explain the decrease in diffraction bandwidth and increase in intensity (Fig. S2B). The empty spaces present in the amorphous structure can be occupied by adsorbate molecules, which are substances adsorbed on the surface of the material40. This can be useful for the adsorption of substances in purification or separation processes, as in the case of this study41.The micrographs before the carbonization step (Fig. S3A and C) show that the surface of the residual corozo biomass is composed of irregular particles of different sizes. After carbonization with ZnCl2, the particles maintained their irregular shape and varied in size (Fig. S3B and D). However, the surface shape was modified, becoming smoother and more regular with some randomly distributed cavities. Similar morphological modifications have been reported in the literature in several studies of carbonized materials of plant origin for the purpose of removing different compounds42,43,44,45.Propranolol adsorption equilibrium and application of the PSMThe experimental results and the model predictions are shown in Fig. 1. In both cases, the adsorption capacity tends to increase according to the propranolol equilibrium concentration and the temperature. These results indicate that ACC and ACP have strong affinities for propranolol. For the temperature, this indicates that both systems are endothermic, indicating that the temperature causes some modification of the material or that the propranolol molecules present higher energy or change the solubility. The shape of the isotherm may be related to the L2 shape according to the Giles classification, without the presence of a plateau. This indicates that the adsorption capacity will steadily decrease with increasing concentration at equilibrium due to the decrease in the number of vacant sites46. The choice of the best PSM was based on statistical indicators (Table S.1) and the stereographic parameters of the mode. In this case, it was found that all the employed models presented good statistical indicators (no major significance was found) regarding the correlation factor (R2 > 0.98), average relative error (ARE < 7.84%), and minimum squared error (MSE). Thus, the Bayesian information criterion (BIC) was used as an indicator for selecting the model; the lower the BIC value is, the more suitable the model. Therefore, from the comparison of the BIC values, the MLO model was selected, and the DLO and TLO models were discarded. MLO indicates that the propranolol molecules are adsorbed, forming a monolayer on the surface of the ACC and/or ACP, and it is also assumed that the adsorption energy is the same for all the adsorbed molecules21,47.Figure 1Adsorption isotherms for the propranolol onto the ACC (A) and ACP (B) systems for the temperatures of 298.15 to 328.15 K.Figure 2A illustrates the correlation between the anticipated adsorption capacity and the system temperature. Upon scrutinizing the parameters, it was observed that the adsorption capacity exhibited an upward trend in both systems, directly mirroring the inherent characteristics of the adsorption system. For ACC/propranolol, the adsorption capacity increased from 105.82 to 168.70 mg g−1 as the experimental value increased from 100.24 to 112.94 mg g−1, which corresponds to an ARE of 14.28%. For ACP/propranolol, the predicted adsorption capacity increased from 124.49 to 160.0 mg g−1, with experimental values corresponding to 117.2 to 145.45 mg g−1, corresponding to an ARE of 7.89%. Overall, this indicates that the adsorption capacity predicted by the MLO is in agreement with the experimental data. The evolution of the receptor site (Nm) density according to temperature is shown in Fig. 2B. Similar behaviors emerge for the Nm parameters, which indicates that propranolol tends to have similar behavior for both systems. For the ACC/propranolol system, the receptor density started at 77.76 and reached 168. 70 mg g−1 at the highest temperature of 328.15 K. In the ACP/propranolol system, Nm also increased linearly with temperature, starting at 120.0 to 194.56 mg g−1 according to the temperature. The increase in Nm with temperature can be attributed to two different effects: (i) a decrease in the number of molecules adsorbed per site or (ii) the appearance of new receptor sites due to the temperature effect. Therefore, it is possible that ACP presented a greater density of molecules per site than ACC22,48.Figure 2Stereographic parameters evolution according to the temperature for the ACC/propranolol and ACP/propranolol systems; (A) anticipated adsorption capacity in correlation with the system temperature, (B) densities of the receptor sites (Nm) according to the temperature, (C) number of molecules versus temperature, (D) adsorption energy versus temperature, (E) main adsorption mechanism of propranolol for ACC and ACP.The number of molecules per site (n), also called the stoichiometric coefficient, can be considered the most important parameter that explains the adsorption of propranolol from different points of view. From a chemical point of view, this parameter describes the degree of aggregation of propranolol for both22. In this study, a similar trend was found for both adsorbents, where the number of molecules tended to decrease with temperature (Fig. 2C). For the ACC, the values decreased from 1.36 to 0.8176, and for the ACP, the values ranged from 1.03 to 0.82. The magnitude and change of these values indicate different behaviors. The first is the magnitude, which indicates the orientation of the adsorbed propranolol molecules, which can be separated into two possibilities: when n > 1, the propranolol is adsorbed horizontally, and when n < 1, the propranolol molecules are adsorbed in parallel. For the ACC, the initial value is 1.36 at 298.15 K, followed by 1.06 at 308.15 K, meaning that at the early initial temperature, the propranolol molecules will be adsorbed in a horizontal manner. As the temperature of the system increases, the manner in which the propranolol molecules are adsorbed in this system shifts in a parallel way, since the values for n reach 0.933 and 0.817 for 318.15 and 328.15 K, respectively. Similarly, for the propranolol/ACP system, the initial values for n were found to be 1.03 and 1.06 for 298.15 and 308.15 K, respectively. However, as the temperature starts to increase from 308.15 K, it is possible to quickly shift the position of the adsorption since the values are under one unit. The second observation is the overall change in the n values, which decreases almost linearly according to the system temperature evolution, revealing the reverse behavior of the receptor density. This further corroborates the theory that temperature causes the generation of new receptors that were previously hidden due to thermal expansion.The application of MLO also provides information regarding the concentration of half-saturation, which is employed for the determination of the adsorption energy, as shown in Fig. 2D. The adsorption energy is a useful parameter since it indicates the nature of the adsorption and provides information regarding the type of adsorption mechanism involved. In this case, it was found that the adsorption energy tends to increase linearly with the evolution of the system temperature, indicating that the adsorption of propranolol for both systems is endothermic. For the magnitude of the adsorption energy, the maximum value is approximately 10.8 kJ mol−1 for both systems, and it is possible to presume that the adsorption is based on physical interactions8.Possible adsorption mechanismFrom the characterization results, propranolol speciation and modeling results were used to propose an adsorption mechanism. For both materials, classical groups such as O–H, C–H, CO, and CH2 indicate the possible formation of aromatic rings49,50,51. Propranolol has two states according to the pH of the solution (Fig. S4), meaning that at pH 10, 99% of the molecules are in neutral form. For the adsorbent surface, the pH values at zero charge are 6.56 and 6.44 for the stone and peel, respectively (Fig. S5). This means that the surface will be negatively charged at pH 10. Finally, the adsorption energy was found to be lower than 40 kJ mol−1, which strongly indicates that the coupling of the propranolol molecules onto the surface of the adsorbents is due to physical interactions. Thus, it is expected that the propranolol molecules are adsorbed through hydrogen bonding (due to the presence of hydrogen in the adsorbate and adsorbents)52, electrostatic interactions (due to the even being a neutral molecule part of its negative region, Fig. S6)23, and anion-π interactions (due to the surface charge and aromatic rings of the propranolol)8 as shown in Fig. 2E.Thermodynamic property simulation resultsThe evolution of the configuration entropy is shown in Fig. 3. Both systems presented similar trends in terms of concentration equilibrium and system temperature. In both cases, it was found that the entropy tends to increase with the equilibrium concentration, reaching maximum values around the concentration at half-saturation, also called entropic peaks. After this point, the entropy starts to diminish until reaching equilibrium (Ce > 100 mg L−1), which is not shown. The lack of equilibrium in the evolution of thermodynamic parameters is directly related to the experimental data not reaching the isotherm plateau, indicating that further adsorption can still occur. Regarding the temperature, both systems tended to follow the same trend, where the entropy increased with temperature due to the endothermic nature of the systems. However, it should be noted that all the stereographic parameters can influence the thermodynamic properties; in this case, minor overlap occurs for the ACP/propranolol systems, as shown in Fig. 3B. This can be related to the number of molecules per site; in this case, the values are similar, and at 308.15 K, the number of molecules per site increases by 0.03.Figure 3Evolution of the configurational entropy according to the system (A being ACC/propranolol and B being ACP/propranolol), equilibrium concentration, and system temperature.The evolution of Gibb’s free energy is shown in Fig. 4A and B. Regarding the concentration at equilibrium effect, it was found that the Gibbs energy tends to quickly increase at lower concentrations, and this effect is directly related to the number of available sites at the initial stage of adsorption. In addition, it was found that the Gibbs energy tends to behave similarly for both systems, where the energy tends to be more negative according to the system temperature, directly reflecting the endothermic nature of the systems, which is mainly related to the appearance of hidden sites. Additionally, the configuration entropy of the Gibbs energy did not reach equilibrium, which is related to the possibility of further adsorption of propranolol.Figure 4Evolution of Gibbs’s free energy according to the system (A being ACC/propranolol and B being ACP/propranolol), equilibrium concentration, and system temperature.Finally, Fig. S7A and B present the progression of the internal energy with respect to the system, equilibrium concentration, and system temperature. Similar to the other results obtained for thermodynamic proprieties, the equilibrium concentration tends to increase the internal energy without reaching equilibrium, which also indicates that the adsorption capacity could be improved. For the temperature effect, the internal energy presents the same behavior as the Gibbs adsorption-free energy. The magnitude tends to increase with temperature and is negative throughout all the simulated ranges. The negative signal here also indicates that the adsorption is endothermic, with the adsorption naturally increasing at higher concentrations.Regeneration performanceThe regeneration results for the ACC and ACP are shown in Fig. 5. The first aspect to be noted is that both adsorbents present a linear behavior; this type of behavior is related to the affinity of the material for propranolol and the low loss of activated sites per cycle. A minor percentage of removal was 71.1% for ACC and 66.2% for ACP, which was expected since ACC presented a greater adsorption capacity than ACP. Finally, it is possible to estimate that ACC has an average loss of 5.77% per cycle, while ACP has a slightly greater loss of 6.76% per cycle.Figure 5Percentage of propranolol removal according to the adsorption/desorption cycles.
Investigation of propranolol hydrochloride adsorption onto pyrolyzed residues from Bactris guineensis through physics statistics modeling
