Abstract

Water pollution is characterized by the presence of a wide variety of substances found in significant quantities in this environmental compartment, which requires the installation of high-performance treatments. Among the multiple technics of treatment, one can mention the adsorption on activated carbons which enable the retention of organic molecules. This led to the development of high-cost granular activated carbons to specifically reduce some micro-pollutants or macro-molecules. This work aims at designing a low-cost activated carbon adsorbent material by giving it properties to adsorb a broad spectrum of the pollutant load of effluents rich in organic molecules such as landfill leachate. The first phase of this study involved the synthesis of activated carbon from avocado core with optimization of manufacturing conditions, followed by physical and chemical characterization. Activated carbon with the highest iodine index value (1506 mg.g^-1) and the highest specific surface area (862 m².g^-1) was obtained. FTIR results confirms the presence of carboxylic functional groups, lactones and phenols. Phenol removal tests were carried out in batch mode to evaluate the adsorbent performance. Results showed that equilibrium was reached during a contact time of 60 min for an adsorption capacity of 48 mg.g^-1. The study of adsorption kinetics revealed the adequacy of the pseudo-second order kinetic model. These results show that avocado kernels would be good adsorbent for organic compounds in solution.

1. Introduction

Pollutants are constantly being discharged into the environment, generating a lot of organic and mineral residues. Over the past few years, research which has focused on previously neglected wastes, have shown that they can have added value through the production of second-generation materials. This approach has aroused interest in the development of new recovery processes, particularly those used for the production of adsorbents, which have experienced real boom ¹. Activated carbon is a material known for its high adsorption capacity and is one of the most commonly used materials in many liquid or gas phase applications, for its interesting structural properties (porosity, specific surface area, etc.). It is currently widely used in the industrial, healthcare and para pharmaceutical sector ², as well as for water treatment. The structural properties and efficiency of activated carbon are highly related on the raw materials in their manufacture. However, it should be noticed that this efficiency is relative to the use for which it is intended, and consequently determines the choice of basic or raw materials ³. The manufacturing of activated carbon, the expected performances in related to various uses go necessarily through the identification and choice of raw materials. They are mainly derived from carbon-containing materials of plant origin, such as cocoa pod shells ⁴ coconut shells, date pits ⁵, tea waste ⁶ wheat and rice ⁷, eucalyptus ⁸, animal waste (animal bones, blood and hair) or minerals waste (coal, coke, peat). The manufacturing process can be carried out by both physical and chemical activation. In general, the activation reaction is a process that uses an oxidizing agent at high temperature to produce a highly porous product. The physical activation follows two phases. The first phase, known as carbonization, is generally carried out at relatively moderate temperatures (400 to 500°C). The aim is to remove organic compounds and form a carbon. The material then undergoes extensive activation by treatment with steam, carbon dioxide or a mixture of these gases at 800°C ^{9, 10}. The second phase leads to the opening of pores and the production of a very large specific surface area (between 500 and 3000 m².g^-1), whereas the original carbonisate generally has only a low specific surface area (from 50 to 250 m².g ^-1) ¹¹.

In chemical activation, raw carbonaceous materials such as sawdust, peat, molasses and vegetable waste are mixed with chemicals such as sulfuric acid, calcium chloride, zinc chloride, phosphoric acid, potassium hydroxide, etc. After carbonization (400 to 500°C), the carbon is separated from the inorganic activating substance by dissolving the latter in a suitable washing. Zinc chloride [12-14] ¹² and phosphoric acid [15-17] ¹⁵ are the activating agents most commonly used.

A number of parameters can be used to characterize activated carbon and determine conditions of use. These include pore volume and pore size, specific surface area, grain size, hardness, density, iodine number, ash content ³. The main characteristics of activated carbons from different materials are compared in Table 1.

Recently, studies have focused on the conversion of agricultural and forestry biomass into value-added products, in particular adsorbents for the mitigation of environmental pollutants. For example, research has revealed the effectiveness of avocado kernel based activated carbon in the adsorption of phenolic compounds ¹², cadmium ion ¹⁸, ammonium ¹⁹, methylene blue ²⁰.

As part of our research into the valorization of Togolese biomass, we have turned our attention to avocado kernels for the production of activated carbon by chemical activation with orthohosphoric acid, with a view to manufacturing a new material composite for the treatment of waters with a high organic and mineral load. The choice was made not to produce an expensive material whose manufacturing process requires high-temperature pyrolysis, but to limit activation to the use of a chemical agent and a temperature rise around 500°C.

2. Materials and Methods

The avocado is the fruit of the avocado tree (Persea americana) of which the seed, commonly called the stone, is used for a variety of purposes, not only to renew plantations, but for its health and cosmetic benefits. Despite these applications, these avocado kernel end up in the Togolese environment as urban waste. Hence the advantage of recycling them, giving them another added value. The kernels were collected from fruit sellers in the town of Atakpame in the Plateaux region where they are abundant. In fact, these sellers would constitute the main intermediaries in the acquisition and recovery of avocado kernels.

Parameters such as moisture content, volatile matter and ash content complying with the AFNOR XP CEN/TS 14774-3, XP CEN/TS 15148, and AFNOR XP CEN/TS 14775 standards, have been determined in order to have immediate knowledge of the raw material.

Part of the entire proceedings, not an independent document. Please do not revise any of the current designations.

Avocado kernels were washed with distilled water and then oven-dried at 110°C for 24 hours. They were then crushed in a blender and sieved until the fraction smaller than 1 mm was obtained; the retained grains are impregnated in a mass of orthophosphoric acid at 85% in aqueous solution, at an impregnation ratio of 2, for a period of 24 hours and pyrolysis was carried out in a nitrogen atmosphere. Figure 1 shows the various stages in the activated carbon preparation process.

In order to select the best activated carbon to meet the needs of this study, three parameters were studied during the synthesis process: carbonization temperature (T), carbonization time (Δt) and activating agent concentration (x). The eleven (11) activated carbons manufactured were coded according to the AC_T -Δt-x model where AC designates activated carbon, T represents the carbonization temperature, Δt is the carbonization duration and x is the activating agent concentration. The different carbons obtained according to the preparation conditions are shown in Table 2.

The choice of experimental conditions is justified by literature data, according to which several authors have shown that the advantage of chemical activation is low pyrolysis temperatures and lower energy cost. Good-quality activated carbons with a very large porous structure, high specific surface areas and a broad spectrum of adsorbed molecules were prepared from plant biomasses using orthophosphoric acid as the chemical activating agent [21-23] ²¹ which was also the choice made in this study.

The iodine index was tested on all the prepared carbons in order to evaluate the influence of the preparation conditions of the activated carbons on this performance indicator and to identify optimal synthesis conditions. The purpose of the iodine index value test is to evaluate the porosity of an activated carbon; it’s a precise and easy-to-implement test that approximates the specific surface area and provides information on the micro porosity of the carbons. This indicator was obtained using the AWWA B 600-78 standard method ^{16, 24, 25}.

In addition, activated carbons preparation mass yields were calculated in order to assess the influence of mass loss on carbons performance.

The authors use a number of parameters to characterize activated carbons. The pH at zero charge point (pH _PCN) determined by the Lopez-Ramon method ²⁶, provides information on the pH of the aqueous solution in which the solid exists under zero electrical potential. Infrared spectrometry was used for qualitative determination of acidic or basic surface groups, while Boehm’s method ²⁷ was used for the quantitative determination. Chemical composition was determined by CHNS/O elemental analysis and activated carbons surfaces were observed by scanning electron microscopy (SEM). Finally, the crystalline phases are identified using X-ray diffraction. The specific surface area, which provides information on the total surface area per unit mass of adsorbent accessible to the molecules ²⁸ was determined by the BET method. Equipment and operating conditions are summarized in Table 3 for some techniques.

In order to access the adsorption capacity of the activated carbon with the best performance indices, phenol adsorption was tested. In fact, due to its simple dosage, several authors [29-32] ²⁹ have been interested in this molecule to find out more about its adsorption capacities; the data in the literature therefore enable us to compare the effectiveness of the carbons synthesized in this study. Phenol (C₆H₆O, molar mass of 94 g.mol^-1) and its derivatives are among the European Union’s priority pollutants and the United States Environmental Protection Agency ³³.

Adsorption was followed until equilibrium time. The activated carbon/phenol solution mixture is stirred at defined time intervals in the range varying from 0 to 80 min. The tests were carried out in batch of 50 ml (20 mg of activated carbon for 2 mg of phenol in 50 ml of distilled water). The residual concentrations Ci, were determined on the filtrates by UV spectrophotometry at the wavelength λ _max = 270 nm and adsorbed quantity (mg.g^-1) is calculated using equation (1). Organic product concentration is given by Beer Lambert's law, from a calibration curve.

(1)

C₀ and C_i are respectively the initial and residual concentrations (mg.l^-1) of phenol; v the volume (l) of the solution; m the mass of the adsorbent (g).

To study the adsorption kinetics of phenol on carbon, pseudo-first order, pseudo-second order and intraparticle diffusion models were applied.

• Pseudo-first order kinetics

This kinetic model was proposed by Lagergren ³⁴ and assumes that the adsorption at right time t is proportional to the difference between the quantity adsorbed at equilibrium and the quantity adsorbed at time t.

(2)

and are the quantities (mg.g^-1) of adsorbate adsorbed at equilibrium and at time t respectively and (ml.min^-1) is the adsorption rate constant.

After integration, equation (2) becomes:

(3)

• Pseudo-second order kinetics

This model assumes rapid adsorption of the solute at high-energy sites and relatively slow adsorption at low-energy sites.

The pseudo-second order kinetic model is expressed according to equation (4).

(4)

and are the quantities (mg.g^-1) of adsorbate adsorbed at equilibrium and at time t respectively and (g.mg^-1.min^-1) is the adsorption rate constant.

Integrating equation (4), we obtain equation (5):

(5)

• Intra-particle diffusion

The intra-particle diffusion model or Weber and Morris model assumes that the diffusion process is the only limiting step controlling adsorption.

(6)

is the intra-particle diffusion rate constant (mg.g^-1.min^-0.5) and C a constant (Boundary layer thickness).

For the adsorption capacity of activated carbon, the Langmuir and Freundlich models were applied.

• Langmuir model

The solid is assumed to have a limited adsorption capacity Q_max linked to a limited number of adsorption sites (monomolecular layer).

The adsorption isotherms on activated carbon are described according to the Langmuir model corresponding to the equation (7).

(7)

is the quantity (mg.g^-1) of phenol adsorbed at equilibrium; is the equilibrium concentration (mg.l^-1); : the monolayer adsorption capacity (mg.g^-1 ); : Langmuir's constant (l.mg ^-1) related to the free energy of adsorption.

The linear Langmuir transform used is the relation.

(8)

• Freundlich model

This theory describes a multilayer adsorption of the adsorbate on the adsorbent surface. This isotherm is represented by the equation (9).

(9)

is the quantity (mg.g^-1 ) of phenol adsorbed by one gram of adsorbent; is the equilibrium concentration (mg.l^-1); are the Freundlich for adsorption capacity constants (l/g) and adsorption affinity respectively.

The logarithmic form gives the linear equation (10).

(10)

3. Results and Discussion

With a low moisture content (5.3 ± 1.2%), the avocado kernel used could have a high calorific value. The low ash content (3.5 ± 1.6%) implies that the biomass consists essentially of organic matter, i.e. the element carbon. The high volatile matter content (96.5 ± 1.6%) offers the advantages of high calorific value and a large number of functional groups ³⁷ for the production of activated carbons. These results suggest that the avocado kernel is a good precursor for activated carbon production.

Mass yield values decrease with increasing furnace residence time and carbonization temperature (Table 4). This is a classic thermochemical phenomenon in thermochemistry; an increase in heat or in the duration of its exposure, more of the macromolecules making up the biomass are lost. A lower yield implies an increase in the porous network due to the significant elimination of the organic part of the coal. Generally speaking, the mass yield of synthesis is inversely proportional to the loss of volatile matter.

The activated carbons prepared had iodine values ranging from 680 ± 74 mg.g^-1 to 1300 ± 74 mg.g^-1. Seven (7) of the prepared activated carbons are of good quality because according to the SILEX INTERNATIONAL technical description on the quality criterion of an activated carbon, the iodine index must be higher than 950 mg.g^{-1 25}. Comparing these values with those in the literature (Table 5), we deduce that the mass yield and the iodine value are strongly related to the biomass and activating agent.

Principal component analysis was used to study the effects of different experimental factors (carbonization temperature, furnace residence time and activating agent concentration) on the quantity and quality (synthesis mass yield and iodine value) of the activated carbons obtained. The results in Tables 6 and 7 (after varimax rotation) shows that the mass yield, the iodine value and the carbonization temperature are strongly correlated to the factors F1 and D1; similarly, acid concentration is strongly correlated with the factors F2 and D2.

Carbonization temperature, activating agent concentration, mass yield and iodine number are strongly correlated. Increasing temperature and concentration are proportional to the iodine value. A negative effect on mass yield was observed. This is because increasing the concentration of activating agent can modify the structural characteristics and surface functional groups of the adsorbent material. On the other hand, weak correlations are observed with carbonization time. Similar correlations are found in the literature ^{14, 23, 40}.

Activated carbon with the highest specific surface area AC₅₀₀-2-65 was used to test phenol adsorption capacity.

Figure 3 shows the evolution of the removal rate and the residual phenol concentration (C_e) of the solution as a function of time. Rapid adsorption is observed for the first 20 minutes, followed by slow adsorption before reaching equilibrium. The rapid adsorption observed at the start of contact could be explained by the presence of a large number of available sites on the adsorbent. But it could also be linked to the physicochemical characteristics of the material and especially to the nature of the material’s porosity ⁴¹.

At the level of the slow phase that follows and leads to equilibrium, it would reflect the gradual occupation of adsorption sites on the charcoal making them less and less available. This leads to diffusion towards less accessible sites, slowing the rate of adsorption before equilibrium is reached ⁴² Thus, equilibrium is reached after 60 min for an adsorption capacity of 48 mg.g^-1. These results corroborate with the work of ³⁰ on olive kernel-based carbons with an adsorption capacity of 20 mg.g^-1, ⁴³ on commercial activated carbon for an adsorption capacity of 70 mg.g^-1. In contrast, Kaleta (2006) obtained equilibrium at 30 min for a capacity of 61 mg.g^-1. ³¹ synthesized an avocado kernel-based carbon for phenol adsorption and achieded an equilibrium time of 100 min with over 98% removal.

Graphical representations of the pseudo-first order, pseudo-second order and intra-particle diffusion equations are given in Figures 4a, 4b and 4c respectively.

The various parameters of the graphical representations of the pseudo-first order, pseudo-second order and intra-particle diffusion equations are summarized in Table 8.

Analysis of Table 8 shows that the correlation coefficients (R²) of the linear representations of the pseudo-first order and pseudo-second order equations are nearest to the value one. Furthermore, the maximum adsorption value () both obtained experimentally and theoretically using the pseudo-second order equation are closer. Therefore, the second order kinetic law is well suited to simulating the temporal evolution of the phenol concentration in solution. Similar results have been obtained by other authors such as ^{30, 45, 46} when investigating the adsorption of phenolic micro pollutants on activated carbons.

According to Giles' classification, the isotherm is type L (Langmuir class) in sub group 3 without the saturation plateau ⁴⁷. This form isotherm indicates that the adsorbent material would be essentially micro porous and that the adsorption of these compounds takes place via a monolayer chemisorption process at equivalent adsorption sites.

Figures 6a and 6b shows the isotherms applied to the Langmuir (6a) and Freundlich (6b) models for phenol adsorption.

The adsorption isotherm is perfectly described by the Langmuir model. Phenol adsorption is of the chemisorption type, with the formation of a molecular monolayer. The adsorption is located on sites of the equal energy, with no interactions between the adsorbed molecules. Similar results have been obtained by some authors on adsorbent supports from activated carbon from avocado pits ³¹, olive pomace ³², commercial activated carbons ^{43, 48}.

4. Physical and Chemical Characteristics of Activated Carbons

Surface chemical functions and pH at zero charge point were investigated on activated carbons with the best iodine values. Table 10 shows the results obtained.

The results specify the acidic nature of the 3 activated carbons and the absence of basic groups. Similar results were obtained by ²⁵ for activated carbons derived from the core shell of Balanites aegyptiaca, and by ¹⁶ for activated carbons derived from cocoa pod shells. These authors explain that the total absence of basic function is justified by the fact that the activated carbons produced were not brought in contact with oxygen below 200°C or above 700°C. Nor have they been treated with hydrogen or degassed at room temperature as this is where the basic functions are introduced.

These carbons all have zero charge pH values below neutrality; so their acidic nature is certainly linked to the activation method, which introduces a large number of acidic sites. These results show that increasing the concentration of activating agent leads to an increase in the number of carboxyl groups.

In the literature, different pH_PCN values are reported, which may be linked to the activation technique applied and the washing method after elaboration.

The FTIR results (Figure 7) show that the spectra of the four activated carbons are practically similar. The only relative difference observed between them is linked to the intensity of the peaks observed in comparison with the spectrum of the non-activated avocado kernel. Biomass activation has modified the chemical functions on the surface of activated carbons. The main bands are assigned as follows:

- Bands around the 3500-3000 cm^-1 region can be attributed to elongation vibrations of the O-H hydroxyl groups, belonging to phenolic compounds and carboxylic groups.

- The peaks around 2920 cm^-1 more intense on the non-activated core spectrum than in the charcoal spectra are due to elongation vibrations of symmetrical and asymmetrical C-H and suggests the presence of aliphatic structures in the alkanes.

- Bands around 1720-1650 cm^-1 are usually attributed to C=O stretching vibrations of ketones, aldehydes, lactones or carboxylic groups.

- Charcoal spectra also show a strong band at 1600–1580 cm^-1 due to vibrations of C=C bonds in aromatic rings.

- The region around 1410 cm^-1 gives some idea of the relative abundance of the CH₂ and CH₃ groups.

- The spectrum of the non-activated nucleus shows a band in the spectral region between 1190 and 900 cm^-1 with a maximum at 1020 cm^-1. This band was attributed to C–O elongation in acid, alcohol, phenol, ether, and ester groups.

Interpretation of the FTIR results confirms the presence of carboxylic functional groups, lactones and phenols.

By comparing the IR spectra of activated carbons carbonized at the same temperature, we can see that the attack with the more concentrated acid results in a much more homogeneous structure compared, compared with the heterogeneity of non-activated avocado core.

Elemental analysis indicates the carbonaceous nature of the material. The results presented (Table 11) show that all these materials are essentially made up of carbon, which is consistent with activated carbons. Compared with non-activated avocado core, activated carbons have higher carbon contents. On the prepared activated carbons, it can be seen that temperature increases the carbon content at the same time as the oxygen content decreases. These materials would contain oxygen and associated mineral elements (determined by difference). The values obtained are comparable to those in the literature.

The results presented in figure 8 show that the biomass and activated carbons contain other elements such as Al, Si, P and K. Phosphorus is predominant, followed by aluminum. The other elements Fe, Mg, Cr, Na and Ti are found in trace amounts.

In addition, as the concentration of activating agent and carbonization temperature increases, so does the amount of phosphorus in activated carbons, which can be explained by the nature of the activating agent used, which is phosphoric acid.

XRD analysis of the non- activated avocado core (Figure 9) revealed crystallized forms consisting mainly of silica and quartz. On the other hand, the diffractogram of activated carbon AC₅₀₀ -2-65 appears the single peak at 2θ = 27° characteristic of HPO₃ linked to activation with phosphoric acid. Similar observations are made on charcoals based on cotton cake and shea cake chemically activated with phosphoric acid ⁵².

The SEM image is intended to illustrate porosity, especially that created by activation ⁵³. Scanning electron microscopy images of the activated carbon surface and the non-activated core (Figure 10) show similarities with the literature for activated carbons from avocado pits ¹⁹. It is observed on the non-activated core, an irregular surface with scattered grains. On the other hand, SEM images of activated carbons appear relatively more homogeneous with the presence of cavities on the surface, due to heat treatment and activation.

The specific surface areas (S_BET) of some carbons was determined in order to assess their adsorption capacity. The results are shown in Table 12. These results show that AC₅₀₀ -2-65 activated carbon has a high exchange surface. By comparing these values with those found in the literature (Table 13), we can say that this activated carbon complies with standards for the specific surface area of activated carbons, and this would justify its good adsorbent capacity. Furthermore, this activated carbon with the best specific surface area also has the highest iodine index (1506 mg.g^-1), confirming its micro porous nature.

From these two tables we can deduce two essential pieces of information: specific surface area increases with carbonization temperature, i.e. when material loss of is higher, and they are linked to the nature of the precursor and the synthesis conditions. Indeed, activation of avocado kernels with zinc chloride reveals higher BET surfaces areas than with hydrochloric acid. In general, activation with zinc chloride has been shown to produce activated carbons with high porosity, but due to environmental problems linked to the toxicity of residual zinc compounds, this process has been abandoned. However, it is recognized that the suitability of an activated carbon for an application depends less on specific surface area than on pore distribution. That’s why this study focuses instead on developing an activated carbon with a high capacity to adsorb a broad spectrum of molecules.

In addition, AC₆₀₀-2-65 and AC₇₀₀-2-65 activated carbons were prepared and characterized using infrared spectroscopy and scanning electron microscopy to compare the effect of carbonization temperature on the surface behavior of activated carbons. The results (Figures 11 and 12) show that, under the same activation conditions, there is an evolution towards a porous network that becomes more structured as the carbonization temperature rises from 400 to 700°C.

5. Conclusion

The results of this study indicate that it is possible to prepare activated carbons with interesting structural characteristics by acting on the activation parameters. The experimental study, which involved monitoring the influence of phosphoric acid concentration, carbonization temperature and carbonization time, yielded activated carbon with the maximum iodine value (1506 mg.g^-1) and BET surface area (862 m².g^-1), essentially micro porous under the conditions of activation with 65% phosphoric acid at a carbonization temperature of 500°C for 2 h. The study of adsorption kinetics showed that the pseudo-second-order model is the most suitable to describe phenol adsorption, implying a chemisorption-type adsorption mechanism. The high iodine values of the activated carbons we prepared indicate an interesting adsorption capacity. It is clear that some of the activated carbons produced in this study are suitable for use in the treatment.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge the Cooperation and Cultural Action Department of France Embassy in Togo for funding this research works.

References