3.1 Physical Properties
Table 4 evidences the physicochemical properties of the three materials were computed through the TGA technique. Results show a significant decrease in the moisture content after the BPF submitted to treatment with solution NaOH. Maia et al. [41] also observed same results when studying the viability of producing activated carbon from banana peel.
Table 4
Parameters
|
BPF
|
BFA
|
BFAC
|
Moisture content (%)
|
6.85
|
3.87
|
6.76
|
Ash content (%)
|
31.44
|
49.95
|
80.67
|
Table 4.
Regarding ash content, the banana peel obtained a value of 31.4%, which refers to the amount of inorganic matter or minerals in the peel. An increased ash content after fiber activation and activated carbon preparation was evidenced, which is associated with high thermal stability achieved by materials [34, 41]. Besides, it was exhibited that the low ash content of the precursor makes it a good starting material for preparing activated carbon. The activated carbon yield ranged from 18.15% for BFAC. A similar result was obtained by Hock et al. [51].
3.2 Morphological analysis
SEM images, shown in Fig. 2., were employed to investigate the surface morphology of the samples. Figure 2a shows that the banana peel fibers have an irregular structure, and such irregularity can be associated with the grinding process [52]. Additionally, the presence of surface pores can be observed, and the dark regions suggest an increase in the contact area, favoring the chemical interaction between the fiber and the active agent (NaOH). Similar behavior was reported by Nipa et al. [53] when they studied papaya bark fiber for the adsorption of methylene blue.
Figure 2
From Fig. 2b, the SEM images indicated the modification provided by the activation treatment with NaOH. The BFA surface showed more irregularity and more development of the dark regions. When evaluating the morphology and functionality of activated hydro char, Sultana et al. [54] observed this trend. Maia et al. [41] also reported similar behavior when studying potential banana peels for activated carbon fabrication. Figure 2c illustrates the observed development of a highly porous and heterogeneous structure in the BFAC. According to Maia et al. [41], the increase in surface area is be attributed to the removal of volatile components, resulting in the formation of fixed carbon mass with the enlargement of the pore networks with different sizes.
Determining pore size is essential when proposing activated carbon as an adsorbent material. Pore size measurement was performed using the ImageJ software. Based on the IUPAC classification, pores can be classified according to their diameter into micropores (< 0.2 µm), mesopores (between 0.2 and 0.5 µm), and macropores (> 0.5 µm). The average pore diameter for BFAC was 2.127 ± 0.95 µm, and the existence of macropores and mesopores was reported by Khoung et al. [55] when investigating K2CO3 activation of bamboo fiber. The authors reported that pore volume with a radius of 1 to 2.5 nm was significantly enhanced by K2CO3 activation. Apart from stimulating the development of small mesopores, the chemical treatment also favored pore enlargement (2.5–15 nm radius). Figure 2d represents the porous structure obtained by BFAC, highlighting that the prepared material has a relatively large surface area, one of the prerequisites for excellent adsorption material for MB adsorption.
3.3 ATR-FTIR analysis
Figure 3 shows the ATR-FTIR analysis to identify the functional groups of the BPF, BFA, and BFAC samples.The stretching vibration of the O-H group is represented by a characteristic absorption band in the range of 3500 − 3200 cm− 1, as observed in the FTIR spectrum of the fiber obtained from banana peels (BPF) [41, 56].
The absorption bands at 2920 cm− 1 and 1620 cm− 1 are associated to the aliphatic (-CH) and aromatic (C = C) groups that are also characteristic of cellulose [41, 57].
In addition, the presence of absorption peaks at 1160 cm− 1 and 1020 cm− 1 can be ascribed to the elasticity vibrations of -CH bonds commonly found in organic compounds like alcohols, esters, ethers, and phenols [57].
After the chemical activation process, the activated fiber spectrum (BFA) decreased in the absorption band range of 3500 − 3200 cm− 1. It also observed a reduction in the absorption peak at 2920 cm− 1 and a disappearance of the peak around 1800 cm− 1. The aforementioned modifications are associated with the disruption of functional chains within the lignocellulosic structure [58].
The activation process proved to be effective in modifying the fiber surface. Following the carbonization process, a notable change in the intensity of the bands associated with both the in natura and activated fibers was observed in the spectra. In the activated carbon spectrum (BFAC), it was remarkable the disappearance of the -OH vibration band (3500–3200 cm− 1). According to Da Silva et al. [59], this shows that the pyrolysis process removed the oxygen, and its aromatic structures were broken down to become a carbon-rich solid material. Furthermore, a reduction in the intensity of peaks associated with other functional groups was observed, indicating the breakdown of the complex lignocellulosic matrix into simpler compounds.
The results obtained guarantee that the synthesis process suggested by this work results in a potential adsorbent for removing dyes since it presents similar characteristics to recently published studies that used lignocellulosic waste as a precursor [60–62].
Figure 3
3.4 TGA analysis
Figure 4 and Table 5 show the data obtained from the TGA and DTG curves materials (BPF, BFA, and BFAC). Figure 4a evidences BPF lost 8.16 dry wt% of its initial weight before reaching 130°C, corresponding to the water loss and moisture from the BPF [63]. At temperatures exceeding 130°C, the degradation of lignocellulosic compounds began [64]. The second peak is located in the range of temperature between 130 and 220°C, and in it, the degradation of hemicellulose [63] is 16.56 wt% of the total. The third stage corresponds to the decomposition and degradation of cellulose (cleavage of cellulose glycosidic bonds), located in the temperature range between 220 and 340°C [65]. The fourth peak observed in the analysis corresponds to the degradation process of cellulose and lignin, occurring within the temperature range of 340 to 600°C. Lignin is the component with the most complex structure, and its decomposition range is the widest, occurring from 200°C to 900°C. The degradation of lignin is more significant near 400°C, where a small peak, overlapping with the end of the cellulose degradation, can be observed in this region
Table 5
Degradation stages and main mass loss for banana peel fiber (BPF), banana peel activated fiber (BFA), and banana peel activated carbon (BFAC). Data obtained for the heating rate at 10°C.min− 1
Sample
|
Tonset (ºC)
|
Thermal event (ºC)
|
Mass loss (%)
|
Tpeak (ºC)
|
Residue650ºC (%)
|
BPF
|
162
|
30–130
|
8.16
|
55
|
31.44
|
130–220
|
11.50
|
190
|
220–340
|
27.77
|
300
|
340–600
|
19.41
|
378
|
BFA
|
211
|
130–170
|
2.33
|
143
|
49.95
|
325–410
|
11.77
|
370
|
BFAC
|
-
|
25–180
|
9.38
|
65
|
80.67
|
For BFA, there are present two main thermal events (Fig. 4a): the first one between 130 − 170°C is associated with depolymerization of hemicellulose and pectin. The second was found over the temperature range 325 − 410°C, caused by the thermal degradation of the cellulose chains and involves lignin degradation [66].
For the BFAC, the peaks corresponding to hemicellulose, cellulose, and lignin degradation disappear, as observed in the DTG curve, indicating that there is no organic matter in the sample because it starts its decomposition/degradation easily, with an active decomposition zone between 180 and 600°C. It is known that the temperature pyrolysis region (180–400°C) associates with the decomposition of easily-degradable compounds like carbohydrates, while the high-temperature region (400–600°C) corresponds to the decomposition of compounds with high molecular weight like aromatic compounds and polyphenols [59].
According to the results (Table 5), the Tonset for BFA was superior to BPF, showing that after the activated fiber has the increase in thermal stability. For BFAC, the second and third degradation stages range (180–400°C) disappear, showing a high residue at 650°C (Table 5), accounting for a significant percentage (80.67%) associated with the complete carbonization in the nitrogen atmosphere. This result suggests the degradation of external functional groups from the activated carbon. In previous work, Maia et al. [41] studied a method similar to produce activated carbon from banana peel. In this work, the same behavior was found, reducing the stability thermal, with modifications in the methodology to prepare the activated carbon by NaOH activation followed by pyrolysis at 400°C for 1.5 h.
Figure 4
Similar results were reported by Serna-Jiménez et al. [67] by employing a simple and low-cost method based on the chemical activation/carbonization of banana peel waste with sulfuric acid and potassium hydroxide solution. Other authors prepared activated carbon using waste using different methods and evidence of lower thermal stability than obtained in this work [68, 69].
Table 5.
3.5 XRD analysis
Figure 5 evidences the diffractograms for BPF, BFA, and BFAC. The raw fiber (BPF) diffractogram showed the crystalline peaks characteristic of the lignocellulosic materials with diffraction patterns around 2θ = 17° and 2θ = 21°, which correspond to the overlapping peaks referring to the crystallographic planes (1ī0) and (002), respectively. Jawad et al. [70] also similar evidenced peaks, attributed to crystalline carbon, with an expanded lattice due to impurities.
Figure 5
After activation, the activated fiber (BFA) diffractogram presented similar peaks. On the other hand, after the carbonization process, the peaks' disappearance was observed, indicating the amorphous nature of the adsorbent, except for the peak at 21°, which increased in structural disorder compared to the BPF and BFAC. This occurrence can be attributed to the potential introduction of heteroatoms into the carbon network. [71]. Da Silva et al. [59] evidenced similar results when developing activated carbon from pineapple crown wastes. Van et al. [72], also observed a distinctive peak weel indexed, indicating the presence of an amorphous structure with a high degree of graphite crystallization when utilizing activated carbon derived from coconut shells loaded by silver nanoparticles for methylene blue sorption.
Jawad et al. [70] also evidenced the peaks' disappearance after the carbonization process when synthesized activated carbon by the waste banana (Musa sapientum) peels by chemical activation using H2SO4 for methylene blue sorption. In general, similar peaks were observed by Maia et al. [41], when developing activated carbon from banana peel waste obtained via NaOH (at 400 oC for 3h) and pyrolyze. Therefore, this peak can be attributed to the presence of a disordered carbon structure.
3.6 Determination of the zero-charge point (ZPC)
pH is recognized as one of the key factors influencing the adsorption process. [59]. According to Maia et al. [34], the pH of the solution affects the speciation of the adsorbent and the functional groups present. Furthermore, it can interfere with the affinities and competition between the protonated or deprotonated forms of the selected molecule bound to the adsorbent sites. Recognizing such importance, a determination of the point of zero charge was realized, which is responsible for explaining the electrical charge on the surface of the adsorbent material [36, 73]. The pHzpc graph was plotted as “∆pH versus initial pH” and the pHzpc value of the adsorbent was obtained at the intersection point.
Figure 6 shows the behavior of activated carbon in the pH variation and its point of zero charge (pHzcp). The surface charge measurement through pHzpc was recorded as 8.0 for the BFAC sample. For the analysis, it is necessary to understand the material's behavior. When the pH < pHzpc, the adsorbent surface will be positively charged, and when the pH > pHzpc solution, the adsorbent surface will be negatively charged [73]. The alkaline pHzpc of the activated BFAC was contributed by a large amount of alkaline ash elements because of the activating agent or the composition of the BPF.
Figure 6
Maia et al. [34] absorbed methylene blue using activated charcoal from palm trees. The authors found the pHzcp value to be 6.0, so dye adsorption is favored under alkaline conditions.
3.7 Methylene blue sorption: Effect of contact time
The contact time is one of the essential parameters to exert a considerable influence on the removal percentage of the contaminants [74]. The effect of contact time was studied in different contact times ranging from 10 to 70 min at a constant temperature. The experimental adsorption rates as a function of the contact time are shown in Fig. 7.
Figure 7
A variability in the adsorption values was observed, and BAC has a better adsorption performance in a shorter time, corresponding to an MB removal capacity and efficiency of 103 mg g− 1 and 62%, respectively, in only 10 min. After 70 min, these values decreased for 95 mg g− 1 and 57%, respectively.
When developing a sludge-derived biochar for removing of lead (II) and methylene blue, Shahib and coworkers reported that the rapid adsorption rate can be associated with the electrostatic attractions among oppositely charged contaminants and the surface of activated carbon [75]. In other words, this confirms the good accessibility to the binding sites for the dyes on the adsorbent; this factor is relevant in the real world because it reduces residence time [33]. Furthermore, the initial high adsorption rate can be recognized to the active sites on the BFAC surface.
As time increases, a less adsorption rate might be observed and explained due to the lack of available sites; they become unavailable when the adsorption process attains equilibrium [74, 76]. Based on the zero-charge point study, the great adsorption efficiency of MB dye was favored by interaction between cationic dye molecules and anionic functional groups on the adsorbent’s surface. Therefore, activated carbon obtained from banana peel is configured as an adsorbent material rapidly responding to MB’ removal in the environment.
Table 6 shows a single-factor ANOVA statistical analysis comparing the removal efficiency for each contact time, and the results indicated a nonhomogeneity of the variance due to a P value of less than 5%, confirming the reliability of the analysis.
Table 6
Results of analysis of the Variance (ANOVA), showing the Sum of squares, mean square, and P value using a single factor considering the efficiency of BFAC.
Parameters
|
Sum of squares
|
df
|
Mean square
|
F value
|
P value Prob > F
|
Removal efficiency (%)
|
708
|
6
|
118
|
10
|
2.13E-4
|
Significant at < 0.05 level. Not significant at ≥ 0.05 level |
Table 6.
3.8 Methylene blue sorption: Effect of initial concentration and isotherms
Another crucial parameter possible to evaluate is the effect of dye concentration on activated carbon (BFAC). Adsorption capacity and efficiency were carried out in a range of 25 to 500 mg L− 1 with a contact time of 10 min. The experimental adsorption rates as a function of the initial MB concentration are shown in Fig. 8.
The absorption of the dye exhibited a sharp increase with the escalating concentration of the dye solution. At the initial dye concentration of 25 mg L− 1, the MB removed was 8.7 mg g− 1, which further increased to 46.8 mg g− 1 for an initial concentration of 500 mg L− 1 MB solution. Besides, it was observed that MB adsorption efficiency by BAC decreased with an increase in dye concentration. The highest efficiency in the sorption carried out with the banana peel’s activated carbon is 86.7% at 25 mg L− 1, while the lowest sorption efficiency is 23.4% at 500 mg L− 1. According to previous studies, this behavior is associated with the saturation of adsorbent [41, 75, 76]. At low concentrations, the relationship between active sites of the surface and adsorbate molecules is significant, guaranteeing all the dye molecules will be retained in the adsorbent surface. In contrast, above optimal MB concentration, the active sites required for the dye’s adsorption will lack, which can retard the MB adsorption by activated carbon [34, 41, 76, 77].
Figure 8
For dye concentration, Table 7 shows a single-factor ANOVA statistical analysis comparing the adsoption capacity and efficiency, and it was confirmed the reliability of the analysis because the P value is less than 5%.
Table 7
Analysis of variance (ANOVA) of single factor.
Parameters
|
Sum of squares
|
Df
|
Mean square
|
F value
|
P value Prob > F
|
Adsorbed MB (mg g− 1)
|
3493.1
|
4
|
873.9
|
399.0
|
5.59E-11
|
Removal efficiency (%)
|
8773.3
|
4
|
2193.3
|
116.7
|
2.39E-8
|
Significant at < 0.05 level. Not significant at ≥ 0.05 level |
Table 7.
The interactions between the dye solutions with the adsorbent materials show a specific pathway. In this context, adsorption isotherms are used to describe the sorption mechanism from adsorbate to the adsorbent’s surface [78]. The sorption behavior of the dye on BAC was investigated using Langmuir, Freundlich, and Temkin isotherm models, as depicted in Fig. 9 and summarized in Table 8.
Table 8
Estimated MB adsorption parameters to the BAC using the Langmuir, and Freundlich isotherm models.
Model
|
Parameters
|
Results
|
Langmuir Isotherm
|
qmáx (mg g− 1)
|
417
|
KL (L mg− 1)
|
0.05
|
R2
|
0.996
|
Freundlich Isotherm
|
KF (mg g− 1)
|
57.4
|
n
|
2.85
|
R2
|
0.937
|
Temkin Isotherm
|
KT (L mg− 1)
|
1.2
|
BT (J mol− 1)
|
35
|
R2
|
0.953
|
Figure 9
Table 8.
Based on the data obtained, the Langmuir isotherm model fits the data well MB removal, with R2 = 0.996, indicating that the adsorption of MB on the surface of the adsorbent is uniform, and the coverage of the dye monolayer is dominant [33]. The maximum adsorption capacity of MB is 417 mg g− 1. The separation factor (RL) was also computed from the Langmuir isotherm and utilized to assess whether the adsorption was favorable [77]. The obtained RL ranged from 0.05–0.86 indicating good dye adsorption to the adsorbent (favorable adsorption process).
Freundlich isotherm model also shows a good fit (R2 = 0.937). The magnitude of the term (1/n or n) indicates the favorability of the relationship between sorbent/adsorbate [41]. The obtained n value was 2.85 and this result confirms favorable adsorption condition. Lastly, the obtained Temkin parameters (R2 = 0.953) indicate that the MB adsorption on the activated carbon of the banana peel occurred by physisorption [79], suggesting that the attractive intermolecular forces between the adsorbent and dye molecules are greater than those between the dye molecules themselves. According to Erkey [80], the binding energy value must be < 8 kJ mol− 1 and the proposed BAC showed a value of 0.035 kJ mol− 1 confirming this mechanism.
According to the literature, various agro-industrial residues have been employed as precursors for the production of activated carbon. In this regard, the selection of the activation method plays a critical role in achieving improved adsorption outcomes for the targeted contaminants [81]. In a recently published paper by the present authors, the methodology adopted was chemical activation, and pyrolysis of the activated banana peel occurred at 400°C for 3 h [41]. Compared with the present study, these parameters were modified to 600°C for 1.5 h. The mentioned modifications ensured different adsorption results, reinforcing the idea that the temperature and carbonization time impact the properties of the adsorbent material. Thus, it is extremely important to continue studies to find a balance between these parameters and obtain better efficiency in contaminant removal.
3.9 Life Cycle Impact Assessment (LCIA)
LCIA Contributional Analysis
As described in the literature, biochar utilization strategies effectively promote zero waste practices and contribute to the development of a circular economy. LCA is a powerful tool for fostering and managing biochar production within different contexts. The more the system boundaries can include inventory considerations to be integrated and comprehensively analyzed in the LCA of biochar production, the more accurate and valuable the interpretation of the results will be. In this cradle-to-gate system boundary study, a relevance check via a contributional analysis was used according to the recommendations of the UNEP/SETAC Life Cycle Initiative Flagship project on hotspot analysis [82].
After conducting the stand-alone framework (the Base-case named BFAC), Scenario 1 gave an idea of the energy intensity of the laboratory process compared to an industrial scale of activated carbon production. However, it is crucial to avoid comparing LCA studies directly, considering the differences in the system boundaries, applications, functional units, and assumptions adopted, among other parameters [83].
Even if the raw material resources are quite different, both processes revealed, as the most relevant environmental impact categories the Terrestrial Ecotoxicity (TEcotox), Global Warming Potential (GWP) and Human Non-Carcinogenic toxicity (HNCT), shown in Fig. 10.
Figure 10
Regarding GWP, the most intensive energy requirements and the consequent CO2 eq emissions are associated with the drying (pretreatment and BFAC) steps. Drying processes are wholly dependent on the moisture content of biomass; energy requirements for drying and consequent GWP emissions are directly proportional. As demonstrated in the Sankey diagram (Fig. 11), electricity was the primary source of GWP emissions, accounting for 98.44% of the total CO2 eq emissions within the lifecycle (Table 9). LCA results showed that steps regarding biomass pretreatment (from banana prata cultivation until BFA production) contributed to 93.17% of the total GWP of the chemical activation/pyrolysis process, mainly due to electricity use. Of these, grinding phase isolated represented 3.24%, peel drying phase counts for 84.59%, and BFA drying for 3.78%. The alkaline pretreatment had a contribution to GWP of 0.63%.
Table 9
Steps of the process, from banana cultivation to BFAC, and cumulative contribution to GWP along the life cycle.
Cumulative contribution (%)
|
|
|
|
Process steps (from banana cultivation to BFAC production)
|
100.00
|
|
|
|
|
|
|
|
|
10. BFAC (dried)
|
99.76
|
|
|
|
|
|
|
|
|
9. BFAC (washed)
|
|
99.76
|
|
|
|
|
|
|
|
8. BFAC (pyrolyzed)
|
|
|
93.17
|
|
|
|
|
|
|
7. BFA (dried)
|
|
|
|
89.39
|
|
|
|
|
|
6. Activated BFA
|
|
|
|
|
88.76
|
|
|
|
|
5. BPF (sieved)
|
|
|
|
|
|
88.76
|
|
|
|
4. BPF (grinded)
|
|
|
|
|
|
|
85.52
|
|
|
3. BPF (dried)
|
|
|
|
|
|
|
|
84.59
|
|
Electricity- BR
|
|
|
|
|
|
|
|
0.93
|
|
2. Banana prata peel (BPF)
|
|
|
|
|
|
|
|
|
0.93
|
1. Banana prata (cultivation)
|
Figure 11
Moreover, the pyrolysis process had a 6.59% of direct contribution to GWP. According to Gahane and coworkers, other authors observed a similar trend. They noticed that electricity contributed roughly 99.80% of the total GWP for an acid + alkali pretreatment [48].
Table 9.
A similar tendency was observed for the other impact categories evaluated.However, it is worth noting that, in the present study, the use of laboratory scale drying and furnace types of equipment was responsible for an energy consumption efficiency per kg BFAC considerably higher than that reported in other studies. According to Yu et al. [47], net GWP biochar produced from pyrolysis may range from − 442 to -1570 CO2eq/dry ton due to the technology diversity employed and how feedstocks, temperature, and heating rate were operated, as well as the LCA model boundaries or assumptions made. Specific scenarios simulated either the energy consumption in a food waste biorefinery or the sequestration due to stable carbon.
Even if Scenario 2, shown in Fig. 12, represented a rough estimation of the impacts avoided due to the non-disposal of biomass residues for treatment in municipal landfills (since the LCI adopted in this scenario refers to the treatment of MSW containing 31% organic waste), it still brought to light some important aspects to the discussion of circularity of secondary raw materials. More accurate data, however, would require a complementary study to achieve more accurate results.
It is worth noting that negative values (FEcotox, HNCT and MEcotox) represent avoided emissions (environmental aspects) that could lead to related environmental impacts, in case the peels were not sent to a landfill but converted into valuable products as BFAC.
Figure 12
In Scenario 3, a carbon credit, due to 50% stable carbon in BFAC, was simulated, leading to a reduction in GWP from the initial 137,76 kg CO2eq.kg− 1 BFAC to 135,76 CO2eq kg.kg− 1 BFAC with carbon credit, as presented in Fig. 13.
Figure 13
It is essential to mention that this credit reflects exclusively the amount of CO2eq to carbon sequestrated by the mass fraction of peels converted into BFAC. It is essential to mention that this credit reflects exclusively the amount of CO2eq to carbon sequestrated by the mass fraction of peels converted into BFAC, representing 0,75% of the initial amount of peels entering the process, without considering the sequestration corresponding to the other co-products.
In Scenario 4, as suggested in the literature [48], the amount of electricity used in the pyrolysis process falls within 0.4–0.6 kWh per kg of a product; thus, conservatively, 0.6 kWh was adopted to this step. Regarding drying and grinding steps, the same LCI used in Scenario 1 relates to a power consumption of 1.6 kWh of electricity and 0.33 m3 of natural gas to heat 12 kg of water per kg activated carbon produced on an industrial scale. Figure 14 presents the most relevant environmental impact categories, when Base-case is submitted for comparison with Scenario 4 (scale- up to industrial conditions).
Figure 14
There are clear benefits related to scaling up the BFAC processes, particularly regarding energy consumption and related materials consumption and emissions. Terrestrial Ecotoxicity and Global Warming Potential were the most impacted environmental categories. In the meta-analysis conducted by Alhashimi and Aktas [84], the authors concluded that biochar, on average, was found to have negative emissions of -0.9 kg CO2eq.kg− 1 due to carbon sequestration, while activated carbon demonstrated, on average, higher GWP of 6.6 kg CO2eq.kg− 1, what seems to be compatible with the values found in Scenario 4.
That led the authors to propose the complementary Scenario 5, in which this simulation of energy optimization (Scenario 4) was combined with the previous Scenario 2 (non-disposal of biomass residues for treatment in municipal landfills), and Scenario 3 (carbon credit due to 50% stable carbon in BFAC), compared to the Base-case (BFAC), as presented in Fig. 15.
Figure 15
In this new scenario, besides the obvious scale-up benefits, emissions avoided due to revaluation were fundamental to the optimized results in all impact categories evaluated. This prospective LCA focused on verifying the main contributions and feasible points for process improvement. Revaluation of agricultural residues such as banana peel into BFAC, besides contributing to promoting the biological cycle in a circular economy perspective [85], also has the potential to reduce CO2eq emissions, with consequent GWP reduction, when carbon sequestrated is maintained in the form of stable carbon.