Column Study for Adsorption of Copper and Cadmium Using Activated Carbon Derived from Sewage Sludge

Among the water-polluting substances, heavy metals stand out due to their carcinogenic and toxic effects on the creatures and environment. +is study aimed to scrutinize the effectiveness of sewage sludge-based activated carbon in the removal of copper and cadmium from aqueous solutions in column study. Detection of breakthrough curves and related parameters was conducted by varying bed depths (3, 6, and 9 cm). +e solution with an initial metal concentration (IMC) of 100 ppm was pumped to the column at a flow rate of 2mL/min. In the process of copper removal, the breakthrough points for depths 3 cm, 6 cm, and 9 cmwere achieved at 10min, 15min, and 60min, respectively, whereas breakthrough points of similar depths in cadmium removal process were achieved at 5min, 10min, and 30min, respectively. Adsorption kinetics were analyzed using the Adams–Bohart, Yoon–Nelson, and+omas kinetics models.+e Adams–Bohart model described only the initial part of breakthrough curves.+e +omas model represented the adsorption process with coefficients of determination (R) ranging between 0.90–0.95 for cadmium removal and 0.89–0.96 for copper removal, while the coefficients of determination of Yoon–Nelson ranged between 0.89–0.94 for cadmium and 0.95–0.97 for copper. Yoon–Nelson was fitted well with copper removal data, while removal of cadmium data was best described by the+omas model.+is study demonstrated that using sewage sludge-based activated carbon to remove heavy metals is an alternative, more cost-effective option to reach the objectives of sustainable development.


Introduction
Pollution caused by industrial and anthropogenic activities is one of the major challenges of this century [1]. e environmental pollution concept is referred to the deficit in the environmental process of the ecosystem caused by contamination of its components, either atmosphere, soil, or aquatic system [2]. However, heavy metals are one of the most common environmental polluting substances, particularly copper and cadmium, which at high concentrations are toxic and detrimental to the environment due to their carcinogenic and nonbiodegradable features [3,4]. Yet, although copper at low concentration is an essential metal for the growth of plants [5], excessive copper is associated with many health and environmental problems [6]. Nowadays, copper is the most abundant and wildly prevalent heavy metal used in many industrial activities such as metal finishing, electroplating, and plastics [7].
ere is a continuous increase in the amount of heavy metals present in wastewater; therefore, many techniques are used to remove them, such as electrocoagulation (EC), adsorption using synthetic and natural adsorbents, magnetic field implementation, advanced oxidation processes, and membranes. However, choosing the most applicable and efficient techniques is very critical and depends on many factors [8]. Adsorption is an efficient and polishing technique to remove heavy metals, but the manufacturing of granular and powdered activated carbon is an expensive process [9]; therefore, using low-cost adsorbents is a promising technique in removing a wide range of pollutants [10]. However, different types of waste materials, such as nutshell [11], bagasse [12], wood, sawdust, and tea leaf, have been used as precursors to produce activated carbon as adsorbent [13].
e key features which evaluate the efficiency of any adsorbent are its adsorption capacity, structural, and physical properties. Among a wide variety of starting waste materials, sewage sludge has been reported to produce an effective lowcost adsorbent due to its carbonaceous property [14,15]. e adsorption process is usually carried out in batch studies, fixed-bed column studies, fluidized bed continuous studies, or moving beds. However, among continuous studies, a fixed-bed column is preferred due to its feasibility, efficiency, and low-cost process [16]. e process in column studies is described by a breakthrough curve [17]. Breakthrough curves are modeled using the data from the adsorption experiments. Additionally, many models were generated to investigate the breakthrough profiles and explain the dynamic adsorption behaviour of heavy metals, such as the Adams-Bohart, Yoon-Nelson, and omas models [18]. e purpose of this study is to assess the efficiency of sewage sludge-based activated carbon for the removal of copper and cadmium ions from aqueous solutions under different operation runs in column studies.

Preparation of the Adsorbent.
Sewage sludge-based activated carbon, prepared in a previous study [1], with a surface area of 377.7 m 2 /g, was used as an adsorbent in this study. Stock solutions (1000 ppm) of cadmium Cd 2+ and copper Cu 2+ were prepared using cadmium nitrate Cd (NO 3 ) 2 and copper nitrate Cu (NO 3 ) 2 [2]. Afterward, different concentrations of Cd 2+ and Cu 2+ were prepared by diluting the stock solution to the desired concentrations [3]. All the chemicals used were of analytical grade. e specific surface areas were measured by the BET equation based on the Brunauer-Emmett-Teller method using N 2 adsorption/ desorption isotherm. Table 1 illustrates the adsorption data to determine BET.

Column Studies.
Column study was conducted in downflow fixed-bed column (1.5 cm diameter), packed with activated carbon at 3 cm, 6 cm, and 9 cm. Glass wool was used at the bottom of the column to prevent the leaching of activated carbon and clogging of the drainage area. It was also placed on the top of the adsorbent bed to gently distribute the solution onto the adsorbent surface and to maintain a consistent flow. e initial concentrations of copper and cadmium solutions were fixed at 100 mg/L. pH was set at 5.0 using sodium hydroxide (NaOH) and hydrochloride acid (HCL). e optimum values of batch adsorption conditions were adopted in the column study. e solution was pumped into the column at a constant flow rate of 2 ml/min in a downflow direction by a peristaltic pump (Eyela Poller Pump, RP-1000). Effluents were collected every 5 min for the first hour and at intervals of half-hour for the remaining period until the concentration of metal ions in the effluents reached the exhaust point. Measurement of residual concentrations was conducted using a spectrophotometer (DR3900). e column study was conducted at room temperature, and all column experiments were performed in triplicate.

Calculation of Column
Parameters. Important parameters, such as the total weight of metal adsorbed q total (mg), volume of effluent treated V t (ml), total mass of metal entering the column M (mg), concentration of metal adsorbed C a d and q e (mg/g), and the total experimental uptake capacity, are determined using the plot of time (t) against C t /C 0 according to equations (1) to (5), where C t denotes outlet concentration, C 0 is the feed concentration, m is the mass of adsorbent in column (g), and Q is the flow rate (mL/ min) [4].

Modeling of Column Studies
Many models are used to analyze the column adsorption performance depending on regression coefficient (R 2 ) and the best fit of the straight line [5]. In this study, three models, Adams-Bohart, Yoon-Nelson, and omas, were applied to investigate the column adsorption process and performance.

Adams-Bohart
Model. e ability of this empirical model is limited to the first part of the breakthrough curve [6]. Equation (6) represents the formula of the Adams-Bohart model. is model was initially applied to gas-charcoal adsorption, and thereafter, it applies to check the dynamic behavior of the column [7].
where k AB (L/mg.min) stands for kinetic constant, whereas the depth is denoted by Z (cm) and N 0 (mg/L) symbolizes the maximum adsorption capacity. U 0 (cm/min) is designated for the linear velocity of the solution. Parameters of correlation coefficients (R 2 ) k AB and N 0 can be determined using the graph of ln (C t /C 0 ) versus t.

Yoon-Nelson
Model. One of the common and simple models which are wildly used to study the breakthrough curve behavior during the adsorption process is the Yoon-Nelson model. Parameters of k YN and τ are calculated using where k YN (L/min) stands for the rate of constant and the time required for 50% of the breakthrough curve is denoted by τ (min). A plot of ln (C t /(C 0 − C t )) versus time τ is used to calculate the model parameters of k YN and τ from the slope and intercept, respectively. e calculated τ from the experiment was compared with the value obtained from the linear model [8].

omas Model.
omas model is commonly used to study the process of the adsorption and predict the breakthrough curve in fixed bed column. It is based theoretically on Langmuir kinetic and the assumption that the driving force is subjected to second-order reversible rate kinetics [9,10]. e linear form of the omas model is presented by where C 0 , C t , k TH , q 0 , m, and Q are the initial metal concentration, the final concentration or outlet concentration at time t, omas constant rate (mL/min.mg), the maximum uptake capacity (mg/g), adsorbent mass in the column (g), and the flow rate mL/min, respectively. e values of parameters k TH and q 0 are found from the intercept and slope of the linear equation generated from the plot between t and ln (C 0 /C t − 1).

Results and Discussion
Besides the bed depth, many factors have a great impact on the adsorption process, such as initial concentration, flow rate, and pH. However, the optimum values obtained from batch studies were applied, and bed depth was varied and investigated in this study.

Effect of Bed Depth on Breakthrough Curves of Copper
Removal. Breakthrough curves were plotted by changing the three-bed depths (Z), 3, 6, and 9 cm, and keeping the flow rate and initial metal concentration (IMC) fixed at 2 mL/min and 100 mg/L, respectively. Table 2 shows the performance of the column under each depth. Results showed that the smaller the bed depth, the faster to achieve the breakthrough point. Depth of 3 cm, 6 cm, and 9 cm achieved the breakthrough points at 10, 15, and 60 min, respectively. is trend can be interpreted as follows: at the lowest bed depth, there are few numbers of adsorbent active sites; therefore, it requires a shorter time to achieve a breakthrough point [11]. Figure 1 shows the measured breakthrough curves for adsorption of copper onto activated carbon. It can be observed that the breakthrough time has different trends with different depths. Breakthrough at a depth of 9 cm took the longest time and higher adsorption capacity. is could be attributed to the availability of a large surface area and sufficient contact time [12]. Increasing the bed depth leads to the dominance of diffusion mass transfer over the axial dispersion; therefore, the breakthrough time increases [13]. Figure 2 shows the exhausting points for the three different depths. It is obvious that the higher the depth is, the longer time it takes to exhaust. Bed depth of 3 cm exhausted at time 510 min, bed depth of 6 cm exhausted at 600 min, and bed depth of 9 cm exhausted at 870 min. is can be attributed also to the higher number of adsorbent sites and larger surface area.
A similar observation was reported in the study of copper removal using coimmobilized activated carbon and Bacillus subtilis in fixed-bed studies [14]. In addition, a similar trend was observed in cadmium removal from wastewater using sunflower carbon calcium-alginate beads in column studies [15]. Sugar beet shreds were investigated in the removal of copper ions from aqueous solutions using the Box-Behnken design on three levels and three parameters. e shape of the breakthrough was asymmetrical S-shaped [16]. Kenaf fibres were also used to remove copper ions from an aqueous solution in a fixed-bed column. It was reported that breakthrough time and exhaustion time increased with the increase in bed depths [17].

Effect of Bed Depth on Breakthrough Curves of Cadmium
Removal. Cadmium removal under three different depths (Z) was studied (refer to Table 3). It was found that, by increasing the bed depth, the breakthrough and exhausting times increased. Figure 3 shows the trend observed in cadmium adsorption. A similar observation was reported in the adsorption of cadmium into biogenic aragonite shellderived adsorbents [18].
From Figure 4, it was found that depths of 3 cm, 6 cm, and 9 cm achieved exhausting points at 450 min, 540 min, and 840 min, respectively. is might be assigned to that in higher bed depth; there is more adsorbent which provides more active sites. erefore, more time was required to achieve saturation.
Many researchers investigated the removal of cadmium using various adsorbents. Sunflower waste carbon calciumalginate beads were investigated in adsorption and desorption of cadmium ion in fixed-bed studies. It was reported that breakthrough curves, adsorption capacity, and exhaustion time increased with the increase in bed depths. is could be due to the availability of more binding sites at higher bed depths for sorption resulting in a higher mass transfer zone. Similar observations have been reported by Qu et al. [19].

Kinetics Models for Column Adsorption Studies.
e required time to achieve a breakthrough point and the trend of the breakthrough curve are the dominant parameters in the investigation of fixed-bed column behavior. To decide which model is well fitted, correlation coefficients are used, where the higher coefficient indicates better fitting [20].

Modeling of Copper Removal by Adams-Bohart Model.
In this model, some parameters are determined, such as adsorption capacity N o and kinetics constant K ab as listed in Table 4. It was found that the range of R 2 lay between 0.819 and 0.973, indicating that the model does not perfectly match the experimental data and confirming that the Adams-Bohart model is well suited to use only in the first part of the breakthrough curve (10-50% of saturation point) [12]. With increasing bed depth, adsorption capacity N o continues to increase further. is is because longer bed depth contains more active sites and ensures longer contact time [21]. Figure 5 showcases the results from the carried experiments for the first 10% part of the breakthrough curve after fitting to the Adams-Bohart model equation. It can be clearly seen herein that the data for depths of 3 cm and 6 cm are located closely adjacent to trend lines in good agreement. However, for a depth of 9 cm, the points are not the best but relatively in reasonable agreement with the trend line.

Adams-Bohart Model for Cadmium Removal Experiments.
e linear trends of the 10% of the initial part of breakthrough curve experimental data were fitted with the Adams-Bohart model equation as shown in Figure 6. e data for depth 3 cm, 6 cm, and 9 cm are located around the trend lines in good agreement. erefore, the linear relationship can be justified. e obtained parameters for cadmium removal from the Adams-Bohart model are listed in Table 5. e value of k AB increases by an increase in bed height. e highest saturation concentration N 0 of 1074.89 mg/L was obtained at depth 3 cm, and then it was decreased at depths of 6 cm and 9 cm to 516.071 mg/L and 956.43 mg/L, respectively. is is because of the availability of more active sites in higher depths [22]. However, R 2 ranged between 0.9 and 0.96, indicating a good quality of fitting.

Yoon-Nelson Model for Copper Removal.
Yoon-Nelson model was also applied to investigate the behavior of copper adsorption in column studies. Figure 7 illustrates the results from this column study. e linear plots showcase a higher R 2 value, which suggests that the linear relationship is to be a good fit.
Parameters of this Yoon-Nelson model are listed in Table 6. K YN and τ are calculated using the slope and intercept of the trend line resulting from the plot of ln (C t /C 0 − C t ) against time [23] as shown in Figure 7. e longest 50% breakthrough time τ of 462.39 min was obtained at a bed depth of 9 cm, followed by 247.62 min, which was obtained at a bed depth of 6 cm. en the shortest      Advances in Civil Engineering 50% breakthrough time τ of 184.06 min was obtained at 3 cm. It is obvious that, with an increase in bed depth, the 50% breakthrough time τ increases, whereas the rate constant K YN decreases. e lowest K YN of 0.0036 was obtained at depth 9 cm, followed by 0.006 obtained at depth 6 cm, and then the highest K YN was obtained at a depth of 3 cm. Correlation coefficients R 2 with an ideal range of 0.95 to 0.97 confirm that the experimental data were well fitted with the Yoon-Nelson Model.

Yoon-Nelson Model for Cadmium Removal
Experiments. e adsorption behavior of cadmium in column studies was also tested using the Yoon-Nelson model. K YN and τ are calculated using the slope and intercept of the trend line resulting from the plot of ln(C t /(C 0 − C t )) against time as shown in Figure 8.
Parameters of this model are tabulated in Table 7. e longest 50% breakthrough time τ of 389.39 min was obtained at a bed depth of 9 cm, followed by 311.98 min, which was obtained at a bed depth of 6 cm. en the shortest 50% breakthrough time τ of 157.27 min was obtained at 3 cm. It was obvious that, with an increase in bed depth, the 50% breakthrough time τ increases. Similar trend was reported in the modeling of adsorption of hexavalent chromium using Phanera vahlii fruit biomass-based activated carbon. e results showed that the Yoon-Nelson rate constant K YN was decreased from 0.0108 L/min to 0.0086 L/min with an increase of a bed depth from 3 cm to 6 cm. e lowest constant rate of 0.0074 was obtained at the highest bed depth of 9 cm. e correlation coefficients R 2 ranged between 0.87 and 0.94, confirming that the experimental data were fitted with the Yoon-Nelson model, and it can be used to explain overall kinetics in column studies.

4.3.5.
omas Model for Copper Removal Experiments. e values of constant rate (k TH ) and adsorption capacity q 0 for the omas model for copper removal were determined   from the slope and intercept of the trend line resulting from the plot of ln (C 0 /C t − 1) against time as shown in Figure 9. e parameters of the omas model for copper removal data are listed in Table 8. It was found that, at a constant flow rate of 2 mL/min, with an increase in the depth, omas constant rate k TH increased, and adsorption capacity decreased.
It can be observed that the constant rate was increased from 0.00032 to 0.00064 and 0.0007 with the increases in depths from 3 to 6 and 9 cm, respectively. In contrast, adsorption capacity decreased from 4.06 to 1.21 and 0.96 mg/g with increases in depths from 3 to 6 and 9 cm, respectively. Furthermore, the coefficient of correlation (R 2 ) was at its highest at a depth of 9 cm with a value of 0.97, and it was 0.94 for the depth of 6 cm. However, at a depth of 3 cm, it was 0.89, which indicated somewhat poor fitting as shown in Figure 9. e points were not in a straight line for the bed depth of 3 cm. Nevertheless, straight lines can be observed for depths of 6 cm and 9 cm.

omas Model for Cadmium Removal Experiments.
e adsorption behavior of cadmium in column studies was also tested using the omas model. e values of constant rate (k TH ) and adsorption capacity q 0 were calculated from the slope and intercept of the trend line resulting from the plot of ln (C 0 /C t − 1) against time as shown in Figure 10. e inclinations were negative in these relationships as they were in Figure 9.
e parameters of the omas model for cadmium removal data are shown in Table 9. It was found that, at a constant flow rate of 2 mL/min and by increasing in the depths, the omas constant rate was increased, but adsorption capacity has been decreased.
is could be attributed to the increase in mass transfer resistance due to the increase in depth [5].
Cadmium was removed using oil palm shell-based activated carbon. e adsorption capacity achieved using the omas model ranged from 1.4 to 7.4 mg/g with an initial concentration of 200 mg/L and a depth ranging from 3 to    5.5 cm. is can be assigned to the high surface area that formed using oil palm shells compared to sewage sludge [24]. As shown in Figure 10, the correlations coefficients were 0.95, 0.92, and 0.90 for the bed depths 3 cm, 6 cm, and 9 cm, respectively. erefore, the correlation coefficients proved that the omas model was the best to describe the breakthrough curve for cadmium compared to the Yoon-Nelson model. However, the Yoon-Nelson was the best to describe the copper removal breakthrough curve [25].

Conclusions
Adsorption experiments at a constant continuous flow of 2 ml/min were carried out to investigate the removal of copper and cadmium from synthetic solutions onto sewage sludge-based activated carbon. Copper and cadmium adsorption were performed in a fixed-bed column under a fixed preoptimized initial metal concentration of 100 ppm and pH of 5. e columns were packed with activated carbon at three different bed depths, 3, 6, and 9 cm, to monitor the behavior of the breakthrough curves. In copper removal, the highest breakthrough point was achieved at 60 min for bed depth of 9 cm, whereas in cadmium removal, the highest breakthrough point was achieved at 30 min for bed depth of 9 cm. ree kinetics models were applied to investigate the breakthrough curves, including Adam-Bohart, Yoon-Nelson, and omas. However, the Adams-Bohart model described only the initial part of breakthrough curves. Among them, the Yoon-Nelson model was found to be the best model to describe copper removal data, whereas the omas model was the best to describe cadmium removal data. is study showed that sewage sludge could be utilized as a starting material for producing activated carbon which can be applied to remove different types of contaminants.

Data Availability
e data used in this study are available from the corresponding author upon request only for research purposes.

Conflicts of Interest
e authors declare that they have no conflicts of interest.