The biomass adsorbent used in this study was obtained at Damyang (Korea) and cut into particle size in the range of 2 - 3 mm to prepare bamboo chips. In the activation process, the bamboo chips were impregnated by 50 wt% KOH solution for 3 days and then dried at 105˚C for 2 days. The obtained biomaterial was placed in a furnace, followed by heating to the carbonization temperature 450˚C at an increasing rate of 5˚C/min and maintaining at the temperature for 30 min under N₂ atmosphere. Activation temperature 750˚C, and maintaining at the temperature for 60 min. Flaked form of BAC was milled using a ball mill and then sieved into a narrow range of particle sizes (0.42 - 0.50 mm). The obtained BAC was washed with deionized water until the pH of neutral. Finally, the BAC was dried in an oven at 105˚C for 2 days. All of the reagents, including 2,4-DB (Acros Co. USA), sodium hydroxide and acetic acid were reagent grade.

Adsorption column experiments were carried out in a fixed bed, which was made of a glass column of 2.50 cm in diameter and 50 cm in length. A stainless sieve was attached to the bottom of the column, followed by a layer of glass beads. A known quantity of BAC was packed in the column, to obtain the desired bed height of adsorbent of 10 cm. The column was then filled up with 5 mm size glass beads, in order to provide a uniform flow of the solution through the column. 2,4-DB solution of known concentration, 1.06 mol/m³, was pumped upward through the column at the desired flow rate of 10 - 20 mL/min, controlled by a peristaltic pump. The 2,4-DB solutions at the outlet of the column were collected at regular time intervals for analysis, and the concentrations were determined using a UV spectrophotometer (Shimadzu 2401PC, Japan) at λ = 284 nm. The column was lined with a water jacket, and the experiments were performed at 298 K.

The specific surface area, pore-volume and average pore size of the BAC are measured from N₂ adsorption/desorption isotherm at 77 K (Quantachrome, USA). The specific surface area, pore-volume and average pore size obtained from Figure 1 were found to be 1664 m²/g, 0.72 cc/g and 1.74 nm, respectively. As can be seen Figure 1, in the low pressure area, the adsorbed amount increases rapidly, which is due to the micro pores filling. Also, there is no distinct hysteresis loop in the high and medium pressure. It indicates that the proportions of meso pores and macro pores are relatively low in BAC.

The adsorption equilibrium amounts of 2,4-DB on BAC was investigated in terms of initial pHs. 1 mol of HCl and NaOH solutions were used to adjust initial pH (3.2, 7, 10) of the solution. The initial pH of the solution is a major factor influencing the adsorption equilibrium capacity of compounds that can be ionized. Acid or alkali species may change the surface chemistry of the adsorbent by reacting with the surface groups, which may lead to significant pH dependent alterations in the adsorption equilibrium. The adsorption equilibrium amounts of 2,4-DB on BAC was evaluated by measuring the adsorption equilibrium data in terms of the initial pH of 2,4-DB solution. The amounts of 2,4-DB adsorbed at equilibrium were calculated from the following mass balance equation:

q = ( C o − C e ) V W (6)

where, V is the volume of solution (m³), and W is the weight of adsorbent (kg).

Figure 2 shows the adsorption equilibrium isotherm of 2,4-DB on BAC. Adsorption amounts of 2,4-DB on BAC increased with decreasing initial pH of the solution. The similar results were observed in herbicides adsorption on biochars [22] [23] The adsorption equilibrium uptake of 2,4-DB on BAC was 1.61 (pH 3.2), 0.9 (pH 7) and 0.88 mol/kg (pH 10) at 298 K.

The kinetics of 2,4-DB on BAC were analyzed using pseudo-first and pseudo-second order models. The pseudo-first-order rate expression of Lagergren can be expressed as:

d q d t = k 1 ( q eq − q t ) (7)

where, k₁ is the rate constant of first-order adsorption (min⁻¹), and q_eq and q_t are the amounts of adsorbed 2,4-DB on BAC at equilibrium, and at time t, respectively. After integration, and applying boundary conditions, t = 0 to t = t and q_t = 0 to q_t = q_eq, the integrated form of Equation (7) becomes:

log ( q eq − q t ) = log q eq − k 1 2.303 t (8)

A straight line of log(q_eq − q_t) versus t suggests the applicability of this kinetic model (Figure 3). If the plot was found to be linear, with a good correlation coefficient, this would indicate that Lagergren’s equation is appropriate to 2,4-DB adsorption on BAC. The Lagergren’s first-order rate constant (k₁) and q_eq,cal determined from the model are presented in Table 1, along with the corresponding correlation coefficients. It was seen that the pseudo-first-order model better represented the adsorption kinetics, and the calculated q_eq,cal values agreed with the experimental q_eq values (Table 1). This suggests that the adsorption of 2,4-DB follows first-order kinetics. A similar result was reported for the adsorption of 2,4-D from aqueous solution by granular activated carbon [24].

The pseudo-second-order equation is also based on the sorption capacity of the solid phase, and is expressed as [25]:

d t d q = k 2 ( q eq − q t ) 2 (9)

where, k₂ is the rate constant of second-order adsorption (g/mg min). For the same boundary conditions, the integrated form of Equation (9) becomes

t q t = 1 k 2 q eq 2 + 1 q eq t (10)

If second-order kinetics are applicable, the plot of t/q_t against t of Equation (10) should give a linear relationship, from which q_eq,cal and k₂, can be determined from the slope and intercept of the plot (Figure 4). The k₂ and q_eq,cal determined from the model are presented in Table 1, along with the corresponding correlation coefficients. It can be seen from Table 1 that there is no agreement between q_eq experimental and q_eq calculated values for the pseudo-second-order model.

For a packed bed adsorber, the main parameters for mass transfer are the axial dispersion coefficient, and the external film mass transfer coefficient. Axial dispersion contributes to the broadening of the adsorption front axially, due to flow in the interparticle void spaces. Usually it comes from the contribution of molecular diffusion, and the dispersion caused by fluid flow. In this study, the axial dispersion coefficient, D_L, for the fixed bed adsorber was estimated by Wakao’s correlation.

D L 2 v R p = 20 ξ ( D m 2 v R p ) + 1 2 = 20 R e S c + 1 2 (11)

where, ξ is void fraction. External film mass transfer is by diffusion of the adsorbate molecules from the bulk fluid phase, through a stagnant boundary layer surrounding each adsorbent particle, to the external surface of the solid. The external film mass transfer coefficient, k_f, in a fixed bed adsorber can be estimated by the Wakao and Funazkri equation.

k f = D m S h 2 R p ( 2.0 + 0.6 R e 0.5 S c 0.33 ) (12)

where, R_e is the Reynolds number defined as R e = 2 R v ρ f / μ , S_c is the Schmidt number, S c = μ / D m ρ f and S_h is the Sherwood number, S h = 2 k f R / D m , and the molecular diffusion coefficients, D_m, of 2,4-DB can be calculated by the Wilke-Chang equation. The rate of adsorption by porous adsorbents is generally controlled by transport within the particle, rather than by the intrinsic kinetics of sorption at the surface. There are various methods in the literature for determining the diffusion coefficient, here, we determined it by comparing the experimental concentration history in a batch adsorber, with the predicted one based on the pore diffusion model. The estimated values of axial dispersion coefficient, external film mass transfer coefficient, and molecular diffusion are listed in Table 2. The breakthrough curves of all species, in general, depend on adsorption equilibrium, intraparticle mass transfer, and the hydrodynamic conditions in the column. Therefore, it is reasonable to consider adsorption equilibrium and mass transport simultaneously, in simulating the adsorption behavior in a fixed bed adsorber. On the other hand, operational factors, such as the initial pH of the solution and flow rate, are important in column design and optimization. In this work, breakthrough curves were obtained under the various experimental conditions mentioned above. In order to demonstrate the effect of the initial pH on the breakthrough curves of 2,4-DB, breakthrough curves under different initial pH values with a constant BAC bed height of 0.075 m, and 2,4-DB initial concentration of 1.06 mol/m³, and flow rate of 6.95 × 10⁻³ m/s are shown in Figure 5. The breakthrough time decreased with increase of initial pH value, since the hydrogen ion concentration (pH) has a major effect on the degree of ionization of the sorbate and the surface properties of the adsorbents. These in turn lead to shift in the sorption capacity of the equilibrium sorption process. Since the flow rate is a very important factor in fixed bed design, the effect of flow rate on the adsorption of 2,4-DB on BAC was investigated, by varying the flow rate at 1.02 × 10⁻², 6.95 × 10⁻³ and 3.56 × 10⁻³ m/s, with a constant BAC bed height of 0.075 m, and 2,4-DB initial concentration of 1.06 mol/m³, and initial pH of 3.2. The effect of flow rate on the breakthrough time curves is depicted in Figure 6, and shows that the column performed better at a lower flow rate, which resulted in longer breakthrough time and exhaustion times. The breakthrough time decreased with increasing flow rate, and the breakthrough curves are steeper for higher flow rates. This result can be explained by the concept of the moving velocity of the mass transfer zone (MTZ), V_mtz, which is defined as follows:

V m t z = ∂ z ∂ t = V 1 + ρ p 1 − ε b ε b ( ∂ q ∂ c ) (13)

Equation (13) suggests that MTZ is a function of interstitial velocity, particle density, bed porosity, and ∂q/∂c. Intraparticle diffusivity is generally believed to be independent of flow rate. From Figure 6, however, the breakthrough curves were found to be steeper at a higher flow rate. This phenomenon is attributed to external film mass transfer resistance. This resistance is smaller when the flow rate is higher, so the length of MTZ is reduced, and a sharp breakthrough curve is generated. In general, the breakthrough curves become steeper, with increasing flow rate and decreasing bed height. Since the intraparticle diffusivity is usually independent of flow rate, this behavior is due to the external film mass transfer resistance. This resistance is weakened when the flow rate is higher, so that the length of the mass transfer zone is reduced, and a sharper breakthrough curve is generated.

For the successful application of an adsorption system, an efficient regeneration of the used adsorbent is very important from the economic point of view. In general, there are many regeneration techniques such as thermal, steam, acid or base and solvent regenerations. The choice of a certain regeneration method should depend upon the physical and chemical characteristics of both the adsorbate and the adsorbent. In this study, distilled water was used as desorbate for the 2,4-DB. As shown in Figure 7, desorption of the 2,4-DB in the BAC was

about 95% using distilled water. The effluent pH increased in the initial stage of adsorption, and decreased to the pH of the initial solution as adsorption proceeded and then increased as desorption proceeded. As discussed previously, the rapid increase of effluent pH in the earlier adsorption stage also implies that large amounts of the 2,4-DB was removed by BAC.