Preparation and Characterization of Activated Carbon Produced from Oil Bean (Ugba or Ukpaka) and Snail Shell

Industrial and municipal waste water can contain many impurities, which are divided into different pollutant groups (among others dissolved and undissolved substances, easily degradable organic substances, persistent organic substances, plant nutrients, heavy metals, and salts). The aim of effluent treatment is the elimination of these undesired impurities and the restoration of natural water quality. The contaminations are usually pre-cleaned with other treatment processes such as flocculation, sedimentation adsorption and/or biological treatment. The activated carbon is usually used in a last processing step to remove the most difficult impurities like heavy metals other pharmaceutical micro pollutants. The activated carbon used for cleaning (granular or powder) has to fulfill many different tasks: removal of organic-chemical substances and colorants, reduction of trace substances like chemicals or pharmaceuticals, enormous decrease of residual Chemical Oxygen Demand. Human activities such as mining and associated smelting, burning of fossil fuels, and industrial uses of mercury (Hg) in paints, batteries, medicine, and dentistry have significantly increased the global reservoir of atmospheric metals since the beginning of the industrialized period [1]. High concentrations of heavy metals have been found in sediments and fish in the vicinity of small-scale mining activities using amalgamation as their main technique. The concentration in most fish fillets in these areas exceeds the recommendations of the United States Food and Drug Agency [2] A major concern about the increasing background concentration of heavy metals in the environment is that any inorganic form (less toxic) can be converted into organic form, especially methyl-mercury, one of the most toxic compounds known to humans. Therefore the main objective of the problem is to treat waste water before discharging to water source, thus decreasing the threat and deterioration to the environment and promising better sustainability of the environment. There are many technologies that have been developed for the purification and treatment of wastewater and among them is adsorption. But the hitch is the cost of establishing an adsorption column for this purpose. This research tends to device a cost effective means of producing activated carbon from cheap agricultural wastes.


Introduction
Industrial and municipal waste water can contain many impurities, which are divided into different pollutant groups (among others dissolved and undissolved substances, easily degradable organic substances, persistent organic substances, plant nutrients, heavy metals, and salts). The aim of effluent treatment is the elimination of these undesired impurities and the restoration of natural water quality. The contaminations are usually pre-cleaned with other treatment processes such as flocculation, sedimentation adsorption and/or biological treatment. The activated carbon is usually used in a last processing step to remove the most difficult impurities like heavy metals other pharmaceutical micro pollutants. The activated carbon used for cleaning (granular or powder) has to fulfill many different tasks: removal of organic-chemical substances and colorants, reduction of trace substances like chemicals or pharmaceuticals, enormous decrease of residual Chemical Oxygen Demand. Human activities such as mining and associated smelting, burning of fossil fuels, and industrial uses of mercury (Hg) in paints, batteries, medicine, and dentistry have significantly increased the global reservoir of atmospheric metals since the beginning of the industrialized period [1]. High concentrations of heavy metals have been found in sediments and fish in the vicinity of small-scale mining activities using amalgamation as their main technique. The concentration in most fish fillets in these areas exceeds the recommendations of the United States Food and Drug Agency [2] A major concern about the increasing background concentration of heavy metals in the environment is that any inorganic form (less toxic) can be converted into organic form, especially methyl-mercury, one of the most toxic compounds known to humans. Therefore the main objective of the problem is to treat waste water before discharging to water source, thus decreasing the threat and deterioration to the environment and promising better sustainability of the environment. There are many technologies that have been developed for the purification and treatment of wastewater and among them is adsorption. But the hitch is the cost of establishing an adsorption column for this purpose. This research tends to device a cost effective means of producing activated carbon from cheap agricultural wastes.

Abstract
Two agro wastes were selected and carbonized at 600°C for 45 min and 800°C for 30 min and each was divided into three different portions. Each portion was activated with HCl, H 2 SO 4 and H 3 PO 4 as activating agents. The activated carbons were characterized for some important parameters such as pH, ash content, nitrogen, carbon content, sulphur, fat, fibre, protein, moisture content, carbohydrate, oxygen, hydrogen, sodium, potassium and pore volume. Wastewater from battery industry was collected and treated with the activated carbons with a view to determining the extent of the heavy metal adsorption ability. The results of the characterization shows pH range of 6.71 to 6.82, while the pore volume ranged from 3.9 × 10 -5 to 2.4 × 10 -5 m 3 /g for Oil Bean activated carbon and 8.7 × 10 -6 to 6.2 × 10 -6 m 3 /g for Snail shell activated carbon. The percentage yield of activated carbon before activation is 25.79 to 27.27 for Oil Bean and 61.85 to 86.11% for Snail shell activated carbon. FTIR results shows a surface reorganization of the activated carbon with a formation of new functional groups after chemical activation. The adsorption data generated fitted well into the Freundlich isotherm model since most values of the determination coefficient (R 2 ) >0.500 indicating a heterogeneous adsorption of heavy metals from aqueous solution. It was observed also that those carbon activated with H 3 PO 4 were better adsorbents in most of the activated carbons produced irrespective of the heavy metals, followed by those activated with HCl while those activated with H 2 SO 4 were the least. The ANOVA indicates that there exist a positive significance relationship between the reliability factor (R 2 ) and the Langmuir constants in almost all the activated carbon types produced. So also it was for Freundlich isotherm constants except few of them. This study has shown that activated carbon produced from oil Bean shell and Snail Shell can compete favourably with traditional activated carbons in treating industrial waste especially from battery industries using HCL, H 2 SO 4 and H 3 PO 4 as activating agents.
Exactly 2 g of wet sample was weighed into the Platinum Crucible and placed in a Muffle Furnace at 850°C for 3 hours.
The sample was cooled in a desiccator after burning and weighed. Nitrogen content determination [4] Principle: the method is the digestion of sample with hot    concentrated sulphuric acid in the presence of a metallic catalyst (Selenium powder). Organic nitrogen in the sample is reduced to ammonia. This is retained in the solution as ammonium sulphate. The solution is made alkaline, and then distilled to release the ammonia. The ammonia is trapped in dilute acid and then titrated.
Procedure: Exactly 1 g of each sample was weighed into a 30 ml kjehdal flask (this was gently done to prevent the sample from touching the walls of the Flask) and then the flasks were closed (stoppered) and shaken. Then 1 g of the kjeldahl catalyst mixture was added. The mixture was heated in a digestion rack under fire until a clear solution was seen. The clear solution was then allowed to stand for 30 minutes in order for it to cool. After cooling about 100 ml of deionized water was added to avoid caking and then transferred to the kjeldahl digestion apparatus. A 500 ml receiver flask containing 5 ml of boric acid indicator was placed under a condenser of the distillation apparatus so that the tap was about 20 cm inside the solution. The 10 ml of 40% sodium hydroxide was added to the digested sample in the apparatus and distillation commenced immediately until distillation reaches the 35 ml mark of the receiver flask, after which it was titrated to pink color using 0.01N Hydrochloric acid.

Sulphur content determination [5]
Digestion of samples: 10 g of the dried sample was Weighed into a digestion flask and 20 ml of the acid mixture (650 ml of concentrated HNO 3 ; 80 ml Perchloric acid; 20 ml concentrated H 2 SO 4 ) was added. The flask was heated until a clear digest was obtained. The digest was then diluted with distilled water to the 250 ml mark. Appropriate dilutions were then made for each element to be determined. Sulphate was analysed according to APHA standard method [6].
Procedure: A 250 cm 3 of the sample was evaporated to dryness on a dish. The residue was moistened with a few drops of Conc. HCl and 30 cm 3 distilled water. The mixture was boiled and then filtered. The dish was rinsed and the filter paper washed with distilled water severally and both filterate and washings added together. This was heated to boiling and then 10 cm 3 of 10% BaCl 2 solution was added, drop by drop with constant stirring. The mixture was digested for about 30 minutes, filtered and the filter paper washed with warm distilled water. It was then ignited, cooled and weighed in a pre-weighed crucible.

Determination of heavy metal concentration
Heavy metal concentration was carried out using Varian AA240 Atomic Absorption Spectophtometer. The samples were prepared for AAS analysis according to Adrian [5]. 2 g of the dried sample was weigh into a digestion flask and 20 ml of the acid mixture (650 ml conc HNO 3 ; 80 ml Perchloric acid; 20 ml conc H 2 SO 4 ) was added. It was heated until a clear digest was obtained. The digest was diluted to the 100 ml mark with deionized water.

Proximate analysis methods
Moisture content: Procedure: A petri-dish was washed and dried in the oven • Exactly 2 g of the sample was weighed into the petri dish • The weight of the petri dish and sample was recorded before drying in the oven • The petridish and sample were put in the oven for 30 minutes and the weight was noted • The drying procedure was continued until a constant weight was obtained 1 2 W W % moisture content 100 weight of sample(2 g) Where W 1 =weight of petri dish and sample before drying W 2= weight of petri dish and sample after drying
Procedure: Empty Platinum Crucible was washed, dried and the weight was recorded.
Exactly 2 g of the wet sample was weighed into the Platinum Crucible and placed in the Muffle Furnace at 500°C for 3 hours.
The sample was cooled in a Dessicator after burning and it was weighed and weight recorded.

Determination of the Percent fixed carbon
Percentage fixed carbon=100-% ash content Crude fibre determination 2 g of sample was de-fat with petroleum ether (this is done if the fat content was more than 10%). It was boiled under reflux for 30 minutes with 200 ml of a solution containing 1.25 g of H 2 SO 4 per 100 ml of solution. The solution was filtered through linen on a fluted funnel. It was washed with boiling water until the washings were no longer acid. The residue was then transfers to a beaker and boiled for 30 minutes with 200 ml of a solution containing 1.25 g of Carbonate free NaOH per 100 ml.
The final residue was filtered through a thin but close pad of washed and ignited asbestos in a Gooch crucible. The Gooch crucible was dried in an electric oven and weighed. It was incinerated, cooled and then weighed. The loss in weight after incineration × 100 is the percentage of crude fibre.
weight of fibre % crude fibre 100 weight of sample = ×

Determination of crude fat (Soxhlet fat extraction method)
This method is carried out by continuously extracting a sample with non-polar organic solvent such as petroleum ether for about 1 hour or more.

Procedure:
A 250 ml clean boiling flasks was dried in oven at 105-110°C for about 30 minutes. It was transferred into a desiccator and allows to cool. It was labeled and weighed. The boiling flask was filled with about 300 ml of petroleum ether (boiling point 40-60°C). The extraction thimble was plugged lightly with cotton wool. The soxhlet apparatus was assembled and allowed to reflux for about 6 hours. The thimble was removed with care and petroleum ether was collected draining into a container for re-use. The flask was removed and dried when the flask were almost free of petroleum ether at 105°C-110°C for 1 hour. It was transferred from the oven into a dessicator and allowed to cool, then weighed.
weight of Flask oil weight of Flask % fat 100 weight of sample

Determination of crude proteins [4]
Principle: The method is the digestion of sample with hot concentrated sulphuric acid in the presence of a metallic catalyst. Organic nitrogen in the sample is reduced to ammonia. This is retained in the solution as ammonium sulphate. The solution is made alkaline, and then distilled to release the ammonia. The ammonia is trapped in dilute acid and then titrated.
Procedure: Exactly 1 g of sample was weighed into a 30 ml kjehdal flask (gently to prevent the sample from touching the walls of the flask) and then the flasks were stoppered and shaken. Then 1 g of the kjedahl catalyst mixture was added. The mixture was heated cautiously in a digestion rack under fire until a clear solution was seen. The clear solution was then allowed to stand for 30 minutes and allowed to cool. After cooling, about 100 ml of deionized water was added to avoid caking and then 50 ml was transferred to the kjedahl distillation apparatus. A 100 ml receiver flask containing 5 ml of 2% boric acid and indicator mixture containing 5 drops of Bromocresol blue and 1 drop of methylene blue was placed under a condenser of the distillation apparatus so that the tap was about 20 cm inside the solution. 5 ml of 40% sodium hydroxide was added to the digested sample in the apparatus and distillation commenced immediately until 50 drops gets into the receiver flask, after which it was titrated to pink colour using 0.01N Hydrochloric acid.  [3] with slight modification was used to study the adsorption capacity of each activated carbon prepared at different temperature and time from each of the agro waste. Dose of Activated Carbon was 2 g, Volume of Aqueous Solution used 10 ml, Different Initial Heavy Metal Concentration were prepared for different metal ranging from 35 mg/l to 150 mg/l, pH 7, Temperature was about room temperature, Contact Time 24 hours, Agitation Speed 120 rpm.

Calculations
Adsorption Efficiency=C 0 -C/C 0 × 100 C=concentration at equilibrium heavy metal C 0 =initial concentration of the heavy metal

Isotherm studies (Freundlich and Langmuir Isotherms)
The batch technique was selected to obtain equilibrium data because of its simplicity. Batch adsorption was performed at the same temperatures and initial Heavy Metal (HM) concentrations to obtain equilibrium isotherms. For isotherm studies, adsorption experiments was carried out by shaking 1 g of activated carbon samples with 100 ml flasks filled with 10 ml of HM solution at a concentration range 35 mg/l to 150 mg/L at a fixed temperature in a thermostated shaker bath for a known period of time. After equilibrium the suspension was filtered and the metal solution was then analyzed using AAS (Varian AA240). In order to obtain the adsorption capacity, the amount of ions adsorbed per mass unit of activated carbon sample (mg/g) was evaluated using the following expression: Where q e is the amount adsorbed at equilibrium (mg/g), C 0 is the initial metal ions concentration (mg/L), C e is the equilibrium metal ions concentration (mg/L), V is the volume of the aqueous phase (L), and m is the amount of the activated carbon used.
Determination of pore volume 1 g of each activated carbons was immersed in water and boiled for 15 minutes in order to displace air from the pores of the activated carbons. The samples were superficially dried and reweighed. The pore volume was calculated from the difference in weight (dw) divided by the density of water (e) at room temperature.

Results and Discussion
The proximate analysis of the samples was carried out (Table 1), and it was observed that the pH values of the samples ranged from 6.71 for OBAC and 6.82 for Snail shell indicating slightly neutral samples. It was also observed that OBAC had higher sulphur content (37035 mg/g). Also, the percentage ash, fat, fiber, nitrogen and protein contents were relatively low for both samples which are comparable to literature values [7,8]. Oil Bean had higher moisture content, 56.5% but recorded low percentage carbohydrate unlike its counterpart. The elemental analysis of the samples ( Table 2), showed that Snail Shell recorded the highest carbon content (45.33%), indicating that it may be a good adsorbent. From Table 3, percentage yield after carbonization, we observe that Oil Bean had a low percentage yield after carbonization having recorded percentage yield below 50%, but snail shell had a percentage yield of 61.85% and 86.11% for the carbonization temperature of 800°C and 600°C respectively. This is in line with other studies [9]. This is also an evidence of the high carbon content recorded by snail shell in the elemental analysis presented in Table 2. It can also be observed that the percentage yield at a higher temperature is lower than that at lower temperature even when carbonized at a longer period. This is because increase in temperature resulted in more volatile components of the precursor materials being lost, and hence decreases in percentage yield [9]. After chemical activation, the percentage yield (

Surface chemistry study
The type and net charge of functional groups bonded to the carbon surface is important in understanding the mechanism of adsorption of ionic adsorbates on activated carbons [10]. The adsorption capacity of activated carbon is influenced by functional groups on the carbon surface. The functional groups on the surface of activated carbon analyzed by the FTIR demonstrated the existence of carboxyl, hydroxyl, amine groups, amide groups, alkyl, aromatic C=C, Nitrile, Phenol and carboxylic groups ( Table 6). The FTIR spectra of the carbonised and activated carbons showed some differences from each other. This demonstrates that after the activation, shifting occurred both to higher and lower wave numbers. This shifting indicated that there were binding processes taking place on the surfaces of the activated carbons during activation. FTIR spectra were obtained on a JASCO FTIR-3500 spectrometer. The analysis conditions used were 16 scans at a resolution of 4 cm −1 measured between 400 and 4000 cm −1 . The FTIR spectra of both carbonised (unactivated) and activated OBAC revealed complex surfaces as by the presence of several peaks observed ( Table  6). The activated carbons showed more peaks than the carbonized Oil Bean shell. This is an indication that some reorganization of the surface oxides on subjection to chemical activation has taken place ( Table 6). The peak observed around 1177 cm -1 for the carbonised OBAC shifted to around 1780 cm -1 in the activated OBAC. So also it was for the C=C aromatic bonding observed around 1574 cm -1 on the unactivated Oil Bean carbon shifted to 1750 cm -1 . The O-H stretch observed around 3223.80 cm -1 in the unactivated carbon and the C-H observed at 822.92 cm -1 also underwent various shifts as well. These results are similar to the ones reported for sawdust [11]. These functional groups could act as chemical binding agents where carboxyl, hydroxyl and amine groups could dissociate negatively charged active surface. This means that these functional groups could attract the positively charged objects such as heavy metal ions [12]. Looking at the FTIR spectra, it showed that chemical modification was very effective on chemical structure of activated carbon. Functional groups are formed during activation by interaction of free radicals on the carbon surface with atoms such as oxygen and nitrogen, both from within the precursor and from the atmosphere [13]. The functional groups render the surface of activated carbon chemically reactive and influence its adsorptive properties [14]. Surface oxidation is an inherent feature of activated carbon production. It results in hydroxyl (-OH), carbonyl (=CO), and carboxylic (-COOH) groups that impart an amphoteric character to the activated carbon. The FTIR spectra of both unactivated and the activated Snail Shell carbon also revealed complex surface going by the presence of several peaks ( Table 7). The activated snail shell carbon showed more peaks than the unactivated snail shell carbon. This is an indication of the strong surface oxide reorganizations resulting from the interactions of the atoms of the activating agents (HCl, H 2 SO 4 and H 3 PO 4 ) used. There was shifting of observed peaks as well as disappearance of some observed peaks by the unactivated snail shell carbon prepared. The C=O of the aldehyde functional group was not observed in the unactivated carbon but was found in the activated SSAC. So also it was for C=C aromatic stretch. From Table 8, it can be deduced that carbonization at 600°C gave a better percentage adsorption compare to those of 800°C irrespective of the activating agents. In the adsorption of Cd by the OBAC, there was a better adsorption by those carbonized at 600°C since most of the percentages are higher than those of their counter parts excepts for some that were activated with HCl (Table 9).
In Table 10 Pb adsorbed better in OBAC carbonized at 800°C and as well had a high percentage of adsorption. This is shown pictorially by Figures 5-7. Just like Pb adsorption by OBAC, Mn was adsorbed better by those carbonized at 800°C even though the percentage adsorption was not that high (Table 11 and Figure 8). The AC carbonised at 800°C had a better percentage of adsorption except for those activated with H 3 PO 4 , but they all generally had a low percentage of adsorption (Table  12 and Figure 9).      In the adsorption of Pb by SSAC, the AC carbonized at 600°C had a better percentage adsorption irrespective of the activating agent used though they all had low percentage of adsorption (Table 13 and Figure  10). The adsorption of Cd by SSAC carbonized at both temperature values and activated with different reagents/acids was very low in all though those carbonized at 600°C had a better percentage adsorption in most (Table 14 and Figure 11). The percentage adsorptions of Mn by SSAC were generally low though those carbonized at 600°C had a better adsorption especially those activated with H 3 PO 4 and H 2 SO 4 (Table 15 and Figure 12).

Quantitative description of sorption isotherm
Traditionally, the sorption of ions by adsorbents has been quantitatively described by parameters obtained either directly from isotherms or by least square analysis with Freundlich and Langmuir isotherms. OBAC=Oil Bean Activated Carbon      Table 9: Percentage Adsorption of Cd by OBAC.    The data from this study was subjected to Freundlich and Langmuir models to determine sorption parameters and identify the model which best fit or describes the adsorption by these prepared activated carbons.
The Freundlich model is a case for heterogeneous surface energies and it gives an exponential distribution of active sites present in the activated carbon (AC). This form of the equation was used to relate the amount of heavy metal ions sorbed from the metal solutions and the linear form of the model is: logq e =logK f +1/nlogCe     Where q e is the amount of sorbed metal ions in mg/g, Ce is the equilibrium concentration of the metal ions. n and K f are the Freundlich constants which respectively indicates the adsorption intensity and the adsorption capacity of the AC [15]. They are calculated from the slope and intercept of the plot of logqe versus log Ce (Table 16).
The Freundlich adsorption intensity (n) for the OBAC carbonised at 600°C ranged from 1.14 to 15.6 ( Table 64). The higher the value of n, the stronger bond between the adsorbent and the adsorbate, a desired parameter in waste water treatment [16] n is >1 for Ni adsorption on each OBAC (Table 64). A similar result was obtained for coconut shell carbon adsorption of Methylene Blue [17]. Figure 3 shows the result of the adsorption isotherm for Ni on OBAC carbonised at 600°C and activated with HCl. From the figure, the straight line plot indicates the occurrence of Ni metal adsorption from the sample. The graph shows the adsorption isotherm is of relative good Ni removal. This can be proved by looking at the coefficient of determination, R 2 =0.603. The Goodness of fit of an experimental data is measured by the determination coefficient (R 2 ) [18]. The R 2 for all the isotherms are presented in Table  64. The slope of the linear plot is also good and suitable for testing the adsorption effectiveness. The value of n obtained from the slope of the linear plot ranged from 1.91 to 15.9 which indicate the strength of the OBAC as an adsorbent. When n>1, the adsorption coefficient increases with increasing concentration of the solution which lead to an increase in hydrophobic surface characteristic after monolapisan and when n<1, K f decreases with concentration [19]. Appendix 4 shows the Atomic Absorption Spectrophotometer (AAS) results of the heavy metal analysis (Figure 13). The Langmuir isotherm model suggests that sorption occurs on homogeneous surface by monolayer sorption  Where Ce is the equilibrium concentration (mg/l), q e is the amount of metal ions adsorbed at equilibrium (mg/g) and the Q o and b are the Langmuir isotherm constants which gives the adsorption capacity and the energy of adsorption respectively [14]. The linear plot of 1/q e versus 1/Ce indicates the Langmuir model plot and the values of Q o and b were calculated from the slopes and intercepts of the Langmuir plots. In Figure  3, the determination coefficient did not show a good fit. The Langmuir adsorption capacity (Q o ) of the three type of OBAC for Ni ranged from 0.035 to 1.4144 l/kg while the Langmuir energy of adsorption (b) ranged from 0.32 to 5.895 which reflect the retention intensity and the number of sites available for a sorbate (Figure 14 and Table 17). Figure 3 indicates a good determination coefficient (0.693). Also, the adsorption intensity was relatively high (12.9) as well as a good adsorption capacity of 1.8 (Figures 15, 16 and Tables 18, 19). For the adsorption of Ni by OBAC carbonised at 800°C, the K f ranged from 2.5 to 2.9 which is an indication of good adsorptive capacity. Also, the adsorption intensity ranged from 0.001 to 0.739. The determination coefficient of the Freundlich plot indicates poor fit except for that activated with HCl. The adsorption data for OBAC carbonised at 600°C had a better fit for the Freundlich model. The Freundlich intensity of adsorption (n) of Ni on OBAC was relatively high, but the reverse was the case for the Freundlich adsorption capacity of the (K f ) values OBAC. The Langmuir isotherm for Ni adsorption onto OBAC carbonised at 800°C showed a poor fit considering the determination coefficient (R 2 ) for the plot of the experimental data obtained for adsorption of Ni on OBAC. Their adsorption capacities ranged from 0.0035 to 0.012 while the energy of adsorption ranged from 0.8 to 17.7 (Tables 20-22 (Tables 29, 30). The experimental data for the adsorption of Pb on OBAC had a good fit on the Langmuir isotherm model having had the determination coefficient ranging from 0.004 to 0.9950. This is in exception of that activated with H 2 SO 4 at both 600°C and 800°C carbonization temperature and that activated with H 3 PO 4 at 600°C. It also had low Q o values as well as low b values indicating poor Langmuir adsorption capacity and adsorption energy (Tables 31-34). The determination coefficients for the adsorption of Mn by OBAC where generally low and so the Freundlich isotherm plot did not fit well except for that carbonized at 800°C and activated with HCl (R 2 =0.826). The metal adsorption intensity as well as the Freundlich adsorption capacities were all relatively high, though those carbonized at 800°C where better than their counterparts, except for that activated with HCl at 600°C, which was very high in its adsorption capacity (K f =4.9 × 10 5 ) (Tables 35, 36). For Mn adsorption on OBAC, the experimental data generated did not have a good fit on Langmuir isotherm model in all the types of activated carbon produced except for that activated with HCl and carbonized at 800°C (R 2 =0.892). The R 2 ranged from 0.008 to 0.892. The Langmuir adsorption energy (b) was relatively high while the Langmuir adsorption capacity constant (Q o ) was low (Table 37-40). The determination coefficient from Ni adsorption on SSAC carbonized at 600°C and 800°C were all low and indicates a poor fit of the adsorption data on Freundlich isotherm model. R 2 ranged from 0.57 to 0.927. It was only that carbonized at 600°C and activated with H 2 SO 4 and that carbonized at 800°C and activated with HCl that had a good fit (R 2 =0.710 and 0.927 respectively). The Freundlich adsorption capacity constants were very high indicating a good adsorption capacity of the SSAC. So also it was for the Freundlich adsorption intensities which ranged from 0.23 to 1.07 (Tables 41-43). The adsorption data for Ni on SSAC had a poor fit on the Langmuir isotherm model as the coefficient of determination ranges from 0.029 to 0.800. The Q o were very low but the reverse was the case for the Langmuir adsorption energy constant (Tables 44-46).
The Freundlich model did not fit well with the adsorption data of the SSAC for Pb, taking a look at the determination coefficient values recorded, which ranges from 0.031 to 0.555, but the n and K f values were relatively high indicating good adsorption intensities and adsorption capacities (Tables 47-49). The Langmuir plot for adsorption of Pb on SSAC followed the same pattern just like the adsorption of Ni by SSAC. The R 2 values were low and the range is from 0.029 to 0.451. Also, the Q o which is the Langmuir adsorption capacity constant were low. It ranged from 0.00013 to 0.061. The Langmuir adsorption energy constant (b) ranged from 5.1 to 120.9 indicating good adsorption energy for the adsorption of Pb on SSAC (Tables 50-52). The adsorption data of Cd on SSAC were also very poorly fitted on the Freundlich model as the determination coefficient (R 2 ) ranged from 0.001 to 0.890. The adsorption intensity parameter was also low (0.11 to 1.33). Unlike the adsorption intensity term, the adsorption capacity term were very high and so indicates the poor fit nature of the isotherm plot (Tables 53-58). The adsorption of Mn on SSAC data were also very poorly fitted on the Freundlich model since the determination coefficient (R 2 ) ranged from 0.099 to 0.563. The n and the K f values are relatively high indicating a high adsorption capacity as well. They ranged from 0.82 to 19.61 and 1.1 to 586.1 respectively (Tables 59, 60). The adsorption of Cd and Mn by SSAC experimental data fitted poorly on the Langmuir isotherm model plot. Their respective R 2 ranges from 0.002 to 0.852 and 0.152 to 0.813. the adsorption energy constant were relatively high for both, but the Q o values were very low indicating low adsorption capacity of SSAC prepared from snail shell (Tables 61-63). Freundlich isotherm model as earlier mentioned is a case for heterogeneous surface energies and it gives an exponential distribution of active sites. The Freundlich constants n and K f which respectively indicates the adsorption intensity and adsorption capacity were calculated from the slope and intercept of the plot of logq e versus logC e which have been presented in the previous pages and the parameters summarized in Table 64. The Freundlich binding capacities for Activated carbon produced from Oil Bean ranged from 1.0 to 7.6 × 10 5 (Table 64). The increasing value indicates greater adsorption capacity. The value of n is a function of the strength of the adsorbent materials used and it gives an indication of the favourability of adsorption. As n>1, favourable adsorption condition is observed. When the value of n is high, it showed the adsorption bond is weak and when n>1, the absorption coefficient increases with increasing concentration of the solution which lead to an increase in the surface characteristic after monolayer adsorption. When the value of n<1 K f decreased with concentration in most of the activated carbon types. This is in line with previous studies [19]. The goodness of fit of an experimental data is measured by the determination coefficients or data realiability (R 2 ) Zaid and Mohammed [18]. The R 2 for the isotherm studied as summarized in Table 64                                           model suggests that sorption occurs on homogeneous surface by monolayer sorption without interaction between sorbed ions. Table 66 presents the determination coefficients, the Langmuir constants after experimental data were fitted. The values of b and Q o were evaluated from the intercepts and slopes respectively for the 24 types of OBAC. It was deduced from the table that the coefficient of determinations were all poorly fitted into the Langmuir model equation except for few of them. The range of the R 2 values was 0.00 to 0.993 with most falling below 0.500.

C o (mg/l) C e (mg/l) q e (mg/g) E (%) logq
The essential characteristics of Langmuir isotherm can be expressed in terms of a dimensionless equilibrium parameter (R L ) called Separation Factor [19], which is defined as: Where b=Langmuir constant and Co is the highest initial metal concentration (mg/l).            The value of R L indicates the type of the isotherm to be either unfavourable (R L >1), linear (R L =1), favourable (0<R L <1) or irreversible (R L =0). The Langmuir isotherm model for the adsorption of Ni, Pb, Cd and Mn on OBAC generally had low determination coefficients (Table 66). It shows that the Langmuir isotherm plot did not have a good fit in most of the plots. Though the determination coefficient was low, the essential characteristics of Langmuir isotherm (separation factor) R L , showed that the isotherm were all favourable except for that of Pb/ OBAC/800°C/H 2 SO 4 which had a zero separation factor and this means the isotherm is irreversible (R L =0) [19].
The Langmuir adsorption constants were generally high indicating  strong adsorption intensity and as well as good adsorption capacity of the activated carbons. From observation, the two Langmuir isotherm constants seem to have an inverse relationship in most of the OBAC types except for few of them. In the Table 67 which is the summarized Langmuir isotherm model constants for snail shell activated carbon, it shows that they generally had a poor fit as most of the R 2 are below 0.500 unlike that of the OBAC. Though the adsorption capacity constant Q o was very low, the Langmuir adsorption energy was really high. This further confirms the inverse relationship observed earlier in the Langmuir plot of the data from the snail shell activated carbon adsorption experiment. Despite the poor fit of the Langmuir plot, the adsorption was favourable considering the separation factor which is an essential characteristics of the Langmuir plot (0<R L <1).

Anova for Langmuir adsorption isotherm constants
Analysis of variance (ANOVA) was conducted on the reliability factor (R 2 ), the Langmuir adsorption capacity Q o and the Langmuir adsorption intensity for all the type of activated carbons produced. For OBAC, the correlation of R 2 and Q o showed no statistically significant relation between them and so also it was for R 2 and b (sig=0.446, 0.755 respectively, P>0.05). For Q o and b, the correlation was negative and insignificant as well. For SSAC, the ANOVA for Q o and R 2 showed no significant relation but the paired samples T-test for b and Q o showed a negative correlation indicating that as b increases, Q o decreases. The T-test between b and R 2 showed a positive significant correlation.

Anova for Freundlich adsorption isotherm constants
For OBAC, the ANOVA showed that there was no significant relationship between R 2 and n, but a significant relationship between R 2 and K f . The paired samples T-test showed a positive significant relationship between R 2 and n but between R 2 and K f and n and K f , there was no significant relationship. The correlation of R 2 and n showed a significant negative correlation (r=-0.545, P<0.05), indicating that as the value of R 2 increases, that of n increases. Between K f and R 2 , there was no significant correlation. For SSAC, the anova and correlation for R 2 and n, K f and R 2 and K f and n, there was no significant relation, t for the paired sample T-test, there was a significant relationship between R 2 and n. Figure 17 shows the SEM images of char samples produced by pyrolysis of Oil Bean as activated carbon precursor at different temperatures. Comparing SEM images of the char samples produced under the two temperatures (600°C and 800°C), it can be seen that the pore development on the surfaces of the carbon were similar in their pores and the pores were not evenly distributed as well. That carbonized at 600°C developed more pores than its counterpart but that at 800°C had larger pore size than the 600°C carbon. Looking at the Figure 18, it can be seen that activation of the carbons with acid developed more micro pores as well expanded the initial pores created during carbonization. After the activation the pores were evenly distributed especially for those carbonized at 600°C but with a rough surface ( Figure 19). The SEM photograph of the snail shell activated carbon showed no visible pores. This may be as a result of presence of impurities such as tar which may have clogged the pores developed by the shell which inhibited the pores structure development ( Figure  20). Table 68 shows the pore volumes of the carbons produced. The pore volumes of the activated carbons were larger than the unactivated carbons in all the type of carbon produced. The activated carbons had good pore volumes which is an indication of good adsorptive capacity except for Snail Shell activated carbons.    (Table 68).

Conclusions
The result of these studies shows that it is possible to prepare activated carbon from oil bean shell, peanut seed, palm kernel shell and snail shell and they can be used effectively in the treatment of wastewater from industries. The precursors are common in these parts of the world and they are cheap to obtain since they are waste generated from their consumption. The proximate analysis of the different precursors reveals that they are effective in the treatment of wastewater since they can easily be homogenized into fine granules and powder as well giving it a large surface area for adsorption. The precursors had good percentage yield after activation at both experimental temperature used. The activated carbon produced showed good adsorption capacity for the industrial waste. The FTIR analysis indicates that there was a reorganization of the chemical bonds present in the surfaces of the activated carbons. This may be attributed to the presence of oxygen atoms on the surface of the activated carbons. For the percentage adsorption of heavy metals by the activated carbon, Oil Bean carbonized at 600°C was better adsorbent for the adsorption of Ni and Cd but for Pb and Mn, it was those carbonized at 800°C that were better off. A similar result was observed for snail shell activated carbon were Pb, Cd and Mn adsorption was better with those carbonized at 600°C unlike Ni adsorption by Snail Shell activated carbon. It was observed that those carbon activated with H 3 PO 4 were better adsorbents in most of the activated carbons produced irrespective of the heavy metals,  followed by those activated with HCl while those activated with H 2 SO 4 were the least. The experimental adsorption data fitted well on the Freundlich isotherm model much better than the Langmuir isotherm model indicating heterogeneous surface energies which gives an exponential distribution of active sites. In most of the activated carbon types, the adsorption intensities were greater than one indicating also a favourable adsorption. So also it was for the adsorption capacities. This fact is supported by the values of the separation factors R L . The ANOVA indicates that there exist a positive significance relationship between the reliability factor (R 2 ) and the Langmuir constants in almost all the activated carbon types produced. So also it was for Freundlich isotherm constants except few of them. There was a great effect of the chemical activation process on the pore development in virtually all the activated carbon types produced as shown by the pore volumes of the AC indicated by both before and after activation.