A STUDY OF THE ADSORPTION OF DYES FROM THEIR AQUEOUS SOLUTIONS BY MORUPULE COAL 1

M.S. Nadiye-Tabbiruka,Fortunate Phenyo Sejie, S.H.Coetzee and E.J. Salamula Department of Chemistry, University of Botswana, Private bag 00704, Gaborone Botswana nadiyemst@yahoo.com Department of Chemistry, University of Botswana, Private bag 00704, Gaborone Botswana fsejie@yahoo.com Department of Physics, University of Botswana, Private bag 00704, Gaborone Botswana 1. COETZEES@mopipi.ub.bw 3 Department of Chemistry, College of Math and Science, Makerere University matsciejs@gmail.com ABSTRACT


INTRODUCTION
There are many water pollutants including radioactive isotopes, heavy metals, inorganic fertilisers and organic compounds from dying and printing textile industry and in research laboratories [1][2][3][4]. However, pollution of water by organic compounds is among the most serious health hazards even when present at low concentrations [5][6][7][8]. Water organic compounds' pollutants increase the development of antibiotic resistance of pathogens thus causing danger to the living systems [9,10]. Chemical pollution of surface water presents a threat to the aquatic environment with hazardous effects such as genotoxicity, carcinogenicity, Neurotoxicity and disturbance to the energy transfer by lowering ATP production [11]. Coloured materials and organic dyes constitute the focus of many environmental concerns because of their nonbiodegradable and polluting nature [10,12,13].
A variety of decontamination methods, including Coagulation, Flocculation, Ozonation, Reverse Osmosis, and adsorption, can be used to remove the above dyes as well as other water contaminants in purifying water. Among these, adsorption is one of the cheapest and most effective for removal of large soluble molecules from water. Therefore, it is widely used for the removal of organic compounds from polluted water [23]. Consequently, activated carbon is the most effective and commonly used adsorbent for decolourisation of water.

Preparation of the adsorbent
Coal in form of lumps was collected from Morupule Colliery which is situated along Palapye-serowe road.


Lumps were ground to a fine powder using a mortar and pestle. The powder was steam activated, allowed to cool in a desiccator until use. Fourier Transform Infra-Red (FTIR Perkin-Elmer) was used to identify the vibration frequency of the functional groups of the materials while Scanning Electron Microscope (SEM) was used to study the microstructure and morphology of the material 

Adsorbate
Methyl Orange (Anionic, water soluble azo dye) powder (C14H14N3NaO3S, figure 1) supplied by SAARCHEM, South Africa was used without further purification. A solution of the dye was made by dissolving a known amount into double distilled de-ionised water. Other required dye concentrations were prepared by serial dilution. Several of these concentrations were used to obtain an absorbance vs concentration calibration curve at λ=463nm using a UV-Visible spectrophotometer spectronic 20. This curve was used to obtain dye concentrations of experimental solutions.  4). This experiment was repeated for various MO concentrations, reaction temperature and pH.

Adsorption thermodynamics' studies
In this section, a batch technique was used. MO (1.5g) was added to each of several 250 mL screws capped bottles containing 150 ml of methyl orange of varying initial concentrations. The bottles were kept in a thermostatic water bath shaker at a speed of 220 rpm for 232 min at a constant temperature. After the samples had reached equilibrium, the solutions were filtered using 0.45μm filter paper and their residual concentrations were determined spectrophotometrically as before. The amount adsorbed at equilibrium, qe (mol/g), was calculated by using equation 1 given below:

1
Where Co is the initial concentration of the dye and Ce is the equilibrium concentration (mol/L), V (L) is the volume of methyl orange solution and m (g) is the weight of MC. Equilibrium adsorption Qe values were used to plot an adsorption isotherm.

Characterisation of the adsorbent
Samples of coal were taken for scanning electron microscopy, and FTIR.   The presence of bands at 3689.5cm -1 represents the free OH stretch. The =C-H stretch is found at 2985.8cm -1 .The -C-H stretching is at 2899.5 cm -1 and bending is at 1419.0cm -1 . Coal also shows bands at 1592.7cm-1 which are assigned to the N-H bending vibration. The presence of the S=O bend is confirmed by a peak at 1038.40cm -1 The c=c assymetric strech is found at 902.15cm -1 . The FTIR of the coal after adsorption was not useful as Morupule coal has a complex structure with functional groups that are also characteristic of the dye investigated.

Effect of concentration on kinetics
The effect of varying concentration on methyl orange adsorption kinetics was investigated in the concentration range 1.56x10 -5 M to 1.019x10 -4 M. Figure 4 presents the kinetics data at 299.9K. Clearly, at all concentrations the process is at equilibrium within two hours, with the higher concentrations taking longer and giving a higher adsorption capacity than the lower concentrations. The adsorption capacities of the three lower concentrations are very low and very close whereas the three higher concentrations are noticeably much higher and not as close.
The process is initially rapid but slows down as it approaches equilibrium adsorption for the various initial concentrations.

Figure 4 Effect of Concentration on the adsorption of Methyl Orange onto Morupule coal at 298K
The increase in rate with increasing concentration can be attributed to an increase in the driving force or increased concentration gradient as the initial concentration is increased from 1.56E-5M to 1.019x10 -4 M. The initial concentration provides a driving force to overcome the effect of mass transfer resistance to adsorbate (methyl orange) in the aqueous solution and solid adsorbent phase [25,26]. Figure 4 shows that in most cases, equilibrium was reached within the first 50 minutes but the process was allowed to continue for over three hours. The initial rapid section can be attributed to the initial large numbers of vacant easily accessible adsorption sites at the solid surface. As contact time increases the process slows down as the surface concentration of easily accessible vacant sites decreases; furthermore, there could be a contribution from steric repulsion forces between adsorbed dye molecules as well as with incoming ones. Figure 5 shows an increase in the rate and capacity of adsorption of methyl orange as the temperature is raised from 208K to 308 K indicating a dependence upon temperature indicating that the adsorption process is probably activated. However, increasing temperature may cause some swelling effects within the internal structure of Morupule coal creating more sites for more uptake of methyl orange thus falsifying the temperature effect. From Figure 6, an increase in pH reduces the adsorption of methyl orange. In acidic media there is high population of hydrogen H + ions. It is thought that adsorption of methyl orange is favoured at lower pH values as seen figure7, probably because at these pHs, methyl orange assumes the structure shown below in figure 7.

Figure 7 methyl orange at low pH (acid conditions)
As it can be seen, protonation of the tautomer nitrogen atoms gets them involved in conjugated double bond systems that delocalises the positive charge. This is able to interact with the negatively charged surface of Morupule coal, and hence adsorption is increased. From figure 8 above, it can be seen that methyl orange exist as an anion at higher pHs and also OH ions are abundant in the solution. Consequently there will be repulsion of the dye by the negatively charged surfaces of the adsorbents. Furthermore reduction in adsorption could be due to competitive adsorption between the strong base (OH -) and the dye anion at higher pH values.

Lagergren Pseudo 1 st order Kinetic model
Lagergren pseudo-first-order rate equation (equation 2 below), the earliest known model describing the adsorption rate based on the adsorption capacity [27], was used to analyse experimental data. − = − .

2
Where 1 is the pseudo first order rate constant in g/mg* min is the amount adsorbed at equilibrium, is the amount adsorbed at time t and t is the time.
A plot of log − against t gives a straight line and 1 can be determined from the slope and from the intercept.
The theoretical and the equilibrium amount adsorbed qe can be determined from the plot. showing that multiple pseudo first order kinetics sorption processes are occurring. Each of the two lines represents a pseudo first order kinetic mechanism [28][29][30]. The first section can be associated with the binding of methyl orange to the easily accessible surface of the adsorbents, the second section/stage to adsorption of methyl orange onto the less accessible surface. This can be associated with adsorption of the dye molecules after diffusing through a thin layer of the bound dye molecules to reach the available surface [28][31] or due to a gradual swelling of the coal to avail more surface.
From table 1 below, it is observed that the theoretical amount adsorbed significantly differs from the experimentally determined amount adsorbed . The χ 2 values, which measure the difference between the model data and the experimental data, are very large, indicating that the adsorption process cannot be described using the Lagergren pseudo first order model. This is also supported by the low R 2 values for this model as seen from table 1.

3
A plot of against t gives a straight line from which 2 and the theoretical equilibrium amount adsorbed can be determined. Here 2 is the pseudo first order rate constant in mol/g min.
is the amount adsorbed at equilibrium, is the amount adsorbed at time t, t is the adsorption time,     Similar results were obtained in the study of bio sorption of yellow 4GL on to kaolinite [1], methylene blue and Congo red on to Activated carbon [39], humic and tannic acid on to chitosan [40], and methyl orange onto Calcined lampindo volcanic mud [41].

Intra particle diffusion model
This model holds for adsorbents with porous surfaces and in these, the intra particle diffusion across the double layer will be the 'rate determining' step. This model is characterised by 3 consecutive steps.
1. Transport of sorbate from the bulk solution to outer surface of the sorbent by molecular diffusion, known as film diffusion.
2. The transport of sorbate from the particle surface to the interior surfaces, known as internal diffusion. Where is the amount, is the intra particle diffusion constant in mg·gmin0.5 under initial rates, C is a constant A plot of against 0.5 should give a graph of two parts, the linear, second part which is due to intra particle diffusion and the straight portion is due to boundary layer diffusion effects. Figure 11, shows that the intra particle diffusion model fitted from 6 minutes of the adsorption process indicating that the surface of coal was easily available for the first 6 minutes of the process. This is expected, since there is no double layer 0.007308E-6 J u l y 2 4 , 2 0 1 5 as MO is a salt. After this, the oncoming methyl orange molecules had to diffuse through the formed layer of adsorbed molecules to reach the surface of the adsorbent resulting in the presence of surface diffusion, which became more noticeable at higher temperatures (Ki increases with temperature as shown in table 3) due to increment in thermal energy [15,42].

Thermodynamics' analysis and adsorption isotherms
Interaction of adsorbate with adsorbents was investigated by plotting adsorption isotherms. In the current case equilibrium adsorption is plotted against equilibrium concentration at constant temperature. The data obtained was then fitted on various models to obtain a comprehensive understanding of the nature of the interactions involved.

Temkin model
Temkin and Pyzhev's adsorption model based on the effect of indirect adsorbent -dye interactions on adsorption leads to, the heat of adsorption of all the molecules in a layer decreasing linearly with surface coverage.
The quantity adsorbed at equilibrium qe is related to equilibrium concentration ce by equation 5 below, which linearizes to equation 6 below. A plot of qe against lnCe gives the heat of adsorption from the slope and the equilibrium-binding constant from the intercept. A small value of the parameter b model indicates a physisorption process.   Table 4 shows that the heat of adsorption constant b is very small and decreases with increasing temperature indicating that 'physisorption' was taking place [ [38,43,44].]   Table 5 was obtained from figure 13 above which clearly shows that the experimental data fits the Freundlich model reasonably well since the correlation coefficients R 2 >0.96. The low value of Kf (0.1540) obtained indicates that Morupule coal has a low adsoption capacity [45,46].The value of 1/n of 1.1372 obtained indicates that the coal has a heterogeneous surface and favours the adsorption process [42].

Thermodynamic parametres
Where is the surface concentration of dye on adsorbent (mol g -1 ), is the concentration of dye in the equilibrium solution (mol L -1 ), is the activity coefficient of the adsorbed dye, is the activity coefficient of the dye in solution. As the concentration of the dye in the solution approaches zero, the activity coefficient approaches unity, and equation 8 above reduces to 0 = hence a plot of ln versus qe extrapolating qe to zero [47,48]gives an intercept equal to lnKo.    figure 15 above, low values of enthalpy and entropy changes accompanying adsorption were obtained. ∆Hº was found to be 1.2352 kJ/mol indicating a low affinity of Morupule coal for methyl orange molecules, while ∆Sº was found to be 8.06x10 -3 kJ/mol indicating minimal change in molecular structure and motion [50], confirming physisorption.

CONCLUSIONS
In this study, it was found that the sorption tends to attain equilibrium in 50 minutes. The adsorption process is best described by the pseudo second-order rate kinetics model probably originating from two simultaneous mechanisms of the same process. Adsorption of MO on coal increases with increasing temperature, with decreasing pH and with increasing concentration, in the range studied. The increase of adsorption with increasing temperature was initially thought to be due to activation, in which case chemisorption would be occurring. However, desorption studies showed reversibility and hence ruled out chemisorption. The remaining option is the welknown swelling of the coal associated with increasing temperature which exposes more area for the physisorption process. The latter is supported by the very low enthalpy and entropy changes accompanying adsorption as shown in table 6 Section 3.5. The increase of adsorption with decreasing pH probably originate from the increased concentration of cationic species resulting from the protonation of the nitrogen atoms in the MO molecule in acid conditions. The protonation leads to double bond conjugation that delocalises the positive charge which interacts with the negatively charged coal surface( figure 8 section 3.2.3). Morupule coal can be used to remove dyes from water however its area needs to be increased to improve its adsorption capacity.