AN ALTERNATIVE TO POWDERED ACTIVATED CARBON FOR THE REMOVAL OF EOSIN DYE FROM AQUEOUS SOLUTIONS

This paper reports the use of powdered activated carbon (PAC) and raw coal fly ash (RFA) in the removal of eosin dye from aqueous solution in batch processes. Operational parameters such as contact time, initial dye concentration, pH and temperature were investigated. Adsorption equilibrium was established in 120 min for the two adsorbents. Langmuir and Freundlich isotherms were used to fit the adsorption data. Langmuir model gave the best fit in both cases. The adsorption capacities of PAC and RFA were found to be 62.28 mg/g and 43.48 mg/g, respectively. The highest percentage of eosin dye removal for both PAC (98%) and RFA (90%) was observed at pH 2. Pseudo first-order and pseudo second-order kinetic models were used to fit the adsorption data. Pseudo second-order kinetic model gave the best description of the adsorption of eosin dye onto the two adsorbents. Thermodynamic parameters, ∆H, ∆S and ∆G confirmed the physical nature, spontaneity and the endothermic nature of the adsorption process. A regeneration technique and a process calculation for evaluating the adsorbent dose required were carried out. This study has shown that RFA is a good alternative adsorbent in the removal of eosin dye from aqueous solution.


INTRODUCTION
Dyes are coloured compounds suitable for colouring textiles, wool, and other materials.Natural dyes like indigo have been in use for over 5000 years.Synthetic dyes have replaced natural dyes because of their low cost and vast range of new colours in spite of their negative environmental effects [1].Dyes of synthetic origin have a complex aromatic structure; their resistance to biodegradation is usually not a concern in the use of natural dyes [2].The textile industry ranks first in use of dyes for colouring fibers.Today, over 9000 types of dyes have been incorporated into the handbook for dyers called the color index [3].Soluble eosin dyes, for instance, have found increasing application in ink production, apart from their textile usage.In fact, all conventional red inks are dilute solutions of eosin [4].The huge demand for dyes in industries has resulted in the introduction of these poorly biodegradable, hazardous dye molecules into our ecosystem; hence decisive and effective remedy becomes imperative.
Activated carbon is the most popular adsorbent, which is capable of adsorbing many dyes with a high adsorption capacity [5].It is very expensive and the cost of regeneration is high because desorption of the dye molecules is not easily achieved [6].Various low-cost adsorbents have been investigated as alternatives to activated carbon [7].Fly ash is a waste material originating in large amounts from high-temperature combustion of coal and wood.It exhibits good sorption capacity, non-toxicity, hydrophilicity, biocompatibility and susceptibility to biodegradation [8].Also, fly ash has been found to be an efficient adsorbent for water treatment in the removal of dissolved organic carbon [9].Its alkalinity makes it useful in water treatment to precipitate metallic ions.Moreover, fly ash is also used in the cement industry due to its cementation property [10].Indeed, it has been shown that fly ash possesses interesting properties for the removal of metal ions and organics; the residual sludge produced is used in cement industries and in landfill sites [11].Application of fly ash as adsorbent in treating wastewater not only solves the problem of dye removal but also eliminates waste disposal.More importantly, fly ash is readily available at little or no cost and hence does not require any complicated regeneration process; unlike activated carbon [12].It has been studied on several types of dyes such as omega chrome red, basic violet 10, congo red, methylene blue and rhodamine blue [13][14][15].To the best of our knowledge, no report has been documented on the removal of eosin dye using raw coal fly ash.We therefore investigate the adsorption of eosin dye from aqueous solution using both powdered activated carbon (PAC) and raw coal fly ash (RFA) as adsorbents.Parameters studied include; initial dye concentration, pH and temperature.The kinetics of the adsorption process and the thermodynamic parameters governing the adsorption process were evaluated and discussed.A regeneration technique is proposed to investigate the reuse of the adsorbent for better economy of the process.Finally, a process calculation is hereby reported to know the amount of adsorbent required for efficient removal of eosin dye.

Dye properties and preparation
Analytical grade eosin dye (99.99% purity) used in this study was supplied by Merck and used without further purification.Chemical structure and properties of eosin dye are shown in Figure 1 and Table 1 respectively.10-100 mg/L concentrations of eosin dye were prepared.Doubly distilled water was used in preparing all solutions.

Adsorption studies
Batch experiments were carried out to measure the adsorption characteristics of eosin dye on both PAC and RFA, respectively.2 g of each adsorbent was added to 100 mL of eosin dye solution in separate 250 mL conical flasks of varying concentrations (10-100 mg/L).These mixtures were stirred in a water bath shaker (120 rpm) over varying time periods to determine the equilibrium time at different temperatures (30-60 0 C).After stirring, the suspensions were filtered using filter paper.The residual concentration of eosin dye solution in each case was determined using a calibration curve prepared at a maximum wavelength of 517 nm using a UV-visible spectrophotometer (model 6850 JENWAY).The effect of solution pH on eosin dye removal was investigated similarly as described above by changing the initial pH (2-12); adjusting by 0.1 M HCl or 0.1 M NaOH.For thermodynamic studies, the adsorption procedure was carried out at different temperatures (30-60 0 C).Lagergren first-order and Ho McKay pseudo-second-order kinetic models were used to examine the controlling mechanism.Langmuir and Freundlich isotherms were applied to evaluate the adsorption capacities.All adsorption data reported in this paper were the average values of three runs.

Characterization of fly ash adsorbent
The microstructure of RFA played an important role on its adsorption capacity. Figure 2 shows the SEM images of the RFA samples.As illustrated in Figure 2a, RFA consists mainly of compact or hollowed spheres of different sizes and some unshaped fragments ascribed to unburnt carbon.It was also seen in the micrograph that some of the smaller size particles adhered on bigger size particles.Figure 2b illustrates a number of pores on the surface of the RFA revealing the potential adsorption power.The specific surface area and pore volume of the fly ash obtained from the N 2 adsorption isotherms were found to be 9.62 m 2 /g and 0.026 cm 3 /g, respectively.

FTIR measurements
The FTIR spectroscopic characteristics are shown in Table 3.This was done using NICOLET 6700 FT-IR Spectrometer Thermo Scientific.The FTIR spectroscopic analysis indicated a broad band at 3420 cm −1 , representing bonded -OH groups.The band observed at about 2915-2848 cm −1 was assigned to the aliphatic C-H group.The peak around 1631 cm −1 corresponds to the C=O stretch.The peaks observed at 1546 and 1511 cm −1 correspond to CO 3 2-group.Symmetric bending of CH 3 is observed to shift to 1448, 1435 and 1345 cm −1 .The peaks observed at 1232, 1141 and 634 cm −1 were assigned to Si-O stretching and Si-O bending, respectively.As seen in Table 3, the spectral analysis before and after dye adsorption indicated that mostly the bonded -OH groups, C=O stretching, secondary amine group and symmetric bending of CH 3 were involved in eosin dye adsorption.There were clear band shifts and intensity decrease of the band at 3420, 1631, 1546 and 1448 cm −1 .These findings suggest that there is attachment of dyes on the RFA [16][17][18].and b).With increase in agitation time, the external mass transfer coefficient increases, resulting in quicker adsorption of the eosin dye molecules by the adsorbents.However, the adsorption capacity gradually increases until equilibrium is reached at 120 min.At equilibrium, the amount of eosin dye desorbed from the adsorbent is equal to those adsorbed onto it, i.e. they are in a state of dynamic equilibrium; hence an indication of the maximum adsorption capacity of the adsorbent at the experimental conditions [19].A similar trend has been reported in literature [20].

Effect of pH
The percentage of eosin dye adsorbed by the adsorbents at different pH is shown in Figure 4 for an initial dye concentration of 10 mg/L.pH plays an important role in the adsorption capacity by influencing the chemistry of both the eosin dye molecule and the adsorbent (PAC or RFA) in aqueous solutions.Eosin is a dipolar molecule at low pH, as shown in Figure 1.RFA contains oxygen donor sites on its surface, e.g.hydroxyl groups and carbonyl groups (Table 3).These groups are nucleophilic in nature.With a decrease in the pH of the eosin dye solution, more dye (a) (b) molecules are protonated and get adsorbed on the surface of RFA.It was observed from Figure 4 that at pH 2, eosin dye adsorbed onto PAC and RFA were 98% and 90% respectively for an initial dye concentration of 10 mg/L. Figure 4 also shows that the percentage adsorption decreased with an increase in pH.The percentage removal of PAC and RFA were 90 %, 80 % and 85 %, 65 % at pH 7 and 12, respectively.This is due to the fact that the eosin dye molecules become nucleophilic at higher pH (basic pH) resulting in less adsorption on the nucleophilic sites of the fly ash.Similar reports have been documented in literature [21][22].

Effect of temperature
The adsorption of eosin dye on the two adsorbents was carried out at 10 mg/L initial eosin dye concentration at different temperatures.The percentage adsorption increased from 77.71 to 93.52% and 70.33 to 76.50% for PAC and RFA respectively with a rise in temperature from 30 to 60 °C (Figure 5).This is mainly due to the increased surface activity at high temperature suggesting that adsorption between eosin dye onto PAC and RFA obeys an endothermic process.
Figure 5.Effect of temperature on the adsorption of eosin dye unto PAC and RFA.Initial eosin dye concentration: 10 mg/L; temperature: 30-60 0 C.

Adsorption kinetics
Adsorption kinetics was investigated for better understanding of the dynamics of adsorption of eosin dye onto the two adsorbents in order to obtain predictive models that allow estimations of the amount of eosin dye adsorbed with respect to time.

Pseudo first order kinetic model
The pseudo first order equation is generally expressed as follows [23]: ( where q e and q t are the adsorption capacities at equilibrium and at time t respectively (mg/g).k 1 is the rate constant for pseudo-first order adsorption (/min).A plot of ln (q e -q t ) against t at various concentrations resulted in graphs with negative slopes.k 1 and q e are calculated from the slopes and intercepts respectively (Figure 6a).Although, the correlation coefficients were high, comparison of the q e calc. to the q e exp., the values do not agree.Therefore the adsorption of eosin dye using PAC and RFA does not follow the pseudo-first order kinetics (Table 4).Pseudo second order kinetic model The pseudo second order equation is expressed as [24].
the initial adsorption rate, h (mg/g/min) is given by e q k h 2 2

=
(3) plots of t/q t versus t gave linear graphs from which q e and k 2 were estimated from the slopes and intercepts of the plot (Figure 6b).The correlation coefficients were as high as 0.99 and there were good agreement between q e, cal.and q e, exp data obtained.The good agreement shows that the pseudo-second order kinetic equation fits the adsorption data well.The initial sorption rate, h increases with increasing initial dye concentrations (Table 4).where C e is the dye concentration in the solution at equilibrium(mg/L), q e is the eosin dye concentration on the adsorbent at equilibrium (mg/g), q o is the monolayer adsorption capacity of biosorbent (mg/g) and b is the Langmuir biosorption constant (L/mg).A plot of C e /q e versus C e gave a straight line with a slope 1/q o and an intercept of 1/q o b (Figures 7a and b).The R 2 values of Langmuir isotherm were compared with those of Freundlich isotherm.The values indicated that the adsorption of the eosin dye onto PAC and RFA fit the Langmuir isotherm better with maximum monolayer coverage of 62.28 and 43.48 mg/g, respectively.Values of q o and b are calculated and reported in Table 5.The values of q o obtained were compared with those from other adsorbents; a comparative evaluation of the adsorbent capacities of various types of adsorbents for the adsorption of reactive dyes is listed in Table 6.The adsorption capacity of RFA used in this study gave a relatively high uptake capacity of the eosin dye than most adsorbents but lower uptake capacity when compared with PAC, thereby making it a suitable alternative adsorbent for eosin dye removal from aqueous solution.To confirm the favorability of the process, the dimensionless equilibrium parameter (R L ) defined by equation (v) was used (5) where C o is the highest initial dye concentration in solution, is used to confirm the favorability of the adsorption process, that is (0 < R L < 1) favorable, R L = 1 linear, R L = 0 irreversible or R L > 1 unfavorable [25].The values of R L reported in Table 5 obtained were less than one, indicating that the adsorption of eosin dye unto both PAC and RFA is favorable.

Freundlich isotherm model
The linearized form of Freundlich model is represented by ( 26). ( where q e is the amount of eosin dye adsorbed at equilibrium (mg/g), C e is the equilibrium concentration of the adsorbate (mg/L); k f and n are constants incorporating the factors affecting (a) (b) the adsorption capacity and the degree of non-linearity between the solute concentration in the solution and the amount adsorbed at equilibrium respectively.Plots of log q e versus log C e gave linear graphs (Figures 8a and b) with low R 2 compared with that obtained for Langmuir, indicating that it does not fit the Freundlich isotherm as it did the Langmuir model.The values of k f and n from the graph are reported in Table 5; the value of n greater than one indicates that the adsorption is favorable.

Thermodynamic studies
Thermodynamic parameters: ∆G 0 , ∆H 0 , and ∆S 0 were determined to investigate the feasibility, spontaneity and the nature of the reaction.This was achieved by using the following equations.
Fly ash: an alternative to powdered activated carbon for the removal of eosin dye Bull.Chem.Soc.Ethiop.2013, 27 201 where k o is the equilibrium constant, q e amount removed at equilibrium, C e concentration at equilibrium, T is the temperature in Kelvin and R is the gas constant.A plot of ln k o against 1/T gave linear plots (Figure 9) from which ∆H ° and ∆S° values are obtained from the slopes and intercepts.Results are presented in Table 7. From Table 7, low values of ∆H o which ranges between 1-93kJ/mol signifies that the mechanism of adsorption of eosin dye onto the adsorbents follows physisorption.The positive value of ∆H° affirms the endothermic nature of adsorption.These are in line with the findings of Arivoli and Thenkuzhali [27].The positive values of ∆S° show increased disorder and randomness at the solid/solution interface of eosin dye on the adsorbents.The negative values of ∆G o show that the adsorption is highly favorable and spontaneous.7. Thermodynamic parameters for adsorption of eosin dye onto fly ash at different temperatures.

Repeated adsorption-desorption studies
Desorption of eosin dye was carried out using a batch process at 30 0 C.After equilibrium was reached, the solutions were filtered to remove PAC and RFA respectively.They were then dried in sunlight for 24 h.Each of the dried adsorbents were added to 0.2 M HCl and stirred for 24 h at 120 rpm, separately.These solutions were filtered and dye concentrations in the filtrate were measured using UV-Visible spectrophotometer.It was found that 94.93% and 90.78% eosin dye were desorbed from PAC and RFA respectively.To check the efficiency of the regenerated adsorbents, PAC and RFA were washed thoroughly in distilled water again and sun-dried for 24 h.They were then subjected to adsorption-desorption cycles through regeneration step in between for three additional cycles.About 3% decrease in desorption efficiency was observed for the two adsorbents after the fourth cycle (Table 8).

Process calculation
Adsorption process goes through different stages such as external mass diffusion of solute from the bulk of the system to the adsorbent surface, then the solute diffuses into the pore of adsorbent and finally sorption of solute occurs onto the surface of the pore.It is very difficult to design properly the adsorber size and performance until and unless all experimental data and rate controlling step are available.Therefore, empirical design procedures based on sorption equilibrium conditions are the most common method for predicting the size of adsorber and performance [28].The design objective is to reduce the dye solution of volume V liter from an initial concentration of C o to C 1 (mg/L).The amount of adsorbent is M in g, and the solute loading changes from q o (mg/g) to q 1 (mg/g).At time t = 0, q o = 0 and as time proceeds the mass balance equates the dye removed from the liquid to that adsorbed by the solid.The mass balance equation was written as: under equilibrium conditions, C 1 →Ce and q 1 →q e .Since the sorption equilibrium studies confirmed that the equilibrium data for adsorption of eosin dye onto PAC and RFA follows Langmuir isotherm well, Eq. (x) can be rearranged as, Except for 100% removal conditions, Eq. ( 11) can be used to calculate the amount of adsorbent dose required for the removal of eosin dye from aqueous solution for any initial dye concentration.Figure 10 shows the amount of adsorbent required for removal of eosin dye for different solution volumes at initial dye concentration of 10 mg/L corresponding to a maximum removal of 7 litres of eosin dye at 7.8 and 8.8 g of PAC and RFA, respectively.As expected, the amount of adsorbent required increases with increase in volume of solution to be treated.Again, for a particular volume of eosin dye solution, the required adsorbent dose was less in case of PAC followed by RFA.

Cost analysis
Adsorbent cost is an important parameter for comparing the industrial application of adsorbent materials; however it is seldom reported in the literature.The overall cost of the adsorbent material is governed by several factors which include its availability (whether it is natural, industrial/agricultural/domestic wastes or by-products or synthesized products), the processing required and reuse.RFA is available in power plants and thermal stations in abundance; however the handling charges for the collection and transportation will be involved.After considering expenses like transport, chemical, electrical energy and processing cost, the cost of the material would be approximately 0.5 USD per kg.The most popular adsorbent of the present time for the removal of eosin dye is powdered activated carbon (PAC).In Nigeria the cheapest variety of the powdered activated carbon costs about 5 USD per kg.Compared to the powdered activated carbon, the total cost of RFA is about ten times cheaper.Thus the use of RFA as an adsorbent is a thoughtful economic attempt for its valuable use in eosin dye contaminated wastewaters.

CONCLUSION
RFA was an efficient adsorbent for the removal of eosin dye from aqueous solution.The adsorption was highly dependent on various operating parameters, like; contact time, pH and temperature.The percentage eosin dye adsorption was optimum for PAC and RFA at acidic pH 2. Analysis of the thermodynamic parameters suggests that adsorption was spontaneous and endothermic in nature.Equilibrium data fitted very well to Langmuir isotherm equation, with maximum monolayer sorption capacities of 62.28 and 43.48 mg/g of eosin dye onto PAC and RFA, respectively.Isotherm constants were determined to help in the design of a single-stage batch adsorber for the removal of eosin dye using PAC and RFA.The regenerated RFA with 0.2 M HCl gave excellent performance after four cycles making the adsorbent recyclable.The preliminary process calculation reported is very helpful in knowing the amount of adsorbent required for the removal of eosin dye from waste waters.The cost effectiveness of RFA showed that it can be effectively used in the removal of eosin dyes in wastewaters and industrial effluents.

Figure 2 .
Figure 2. SEM images of RFA.Effect of contact timeBatch experiment carried out at different contact times for various concentrations of eosin dye showed that the adsorption capacity of PAC and RFA increases rapidly with increasing contact

Figure 3 .
Figure 3. (a) Effect of contact time on adsorption of eosin dye using PAC: Initial eosin dye concentration: 10 mg/L (b) Effect of contact time on adsorption of eosin dye using RFA: Initial eosin dye concentration: 10 mg/L

Figure 4 .
Figure 4. Effect of pH on the percentage adsorption of eosin dye: Initial eosin dye concentration: 10 mg/L.

Figure 6a .
Figure 6a.Pseudo first order plot of the adsorption of eosin dye onto raw fly ash (RFA) at 30 0 C.

Figure 6b .
Figure 6b.Pseudo second order plot of the adsorption of eosin dye onto raw fly ash (RFA) at 30 0 C.

Figure 7 .
Figure 7. (a) Langmuir isotherm plot of the adsorption of eosin dye onto powdered activated carbon (PAC) at 30 0 C. (b) Langmuir isotherm plot of the adsorption of eosin dye onto raw fly ash (RFA) at 30 0 C.

Figure 8 .
Figure 8.(a) Freundlich isotherm plot of the adsorption of eosin dye onto powdered activated carbon (PAC) at 30 0 C. (b) Freundlich isotherm plot of the adsorption of eosin dye onto raw fly ash (RFA) at 30 0 C.

Figure 9 .
Figure 9. Van't Hoff plot of the adsorption of eosin dye onto raw fly ash (RFA) at 30 0 C. Table7.Thermodynamic parameters for adsorption of eosin dye onto fly ash at different temperatures.

Figure 10 .
Figure 10.Variation of adsorbent mass (M) required to treat the volume (V) of eosin dye solution.

Table 2 .
. Properties of eosin dye.Chemical composition of fly ash.

Table 3 .
FTIR spectral characteristics of RFA before and after adsorption.

Table 4 .
Comparison of pseudo first and second order kinetic models rate constants and experimental and calculated qe values obtained at different initial eosin dye concentration.

Table 5 .
Langmuir and Freundlich isotherm model constants of adsorption of eosin dye unto PAC and RFA at 60 0 C.

Table 6 .
Maximum adsorption capacities of various dyes onto fly ash.

Table 8 .
Percentage eosin dye desorbed after repeated adsorption-desorption studies using 0.2 M HCl.