Adsorption Removal Characteristics of Hazardous Metalloids (Antimony and Arsenic) According to Their Ionic Properties
3.1. Metalloids Analysis Using ASV
We used ASV to confirm the behavior characteristics of each ion. Previous studies indicated no known existing method for analyzing total Sb using ASV. Thus, before researching the removal characteristics of antimony, it is vital to set up an analysis method using ASV. Considering the form of Sb(V) in water depending on pH, previous research results showed that Sb(V) exists in the form of Sb(III) in a highly concentrated HCl solution (5 M or more) [17]. However, if the sample is preprocessed with HCl, the mercury film coated on the working electrode of the ASV dissolves, preventing measurements. To prevent the mercury film from dissolving, the sample preprocessed with a 5 M HCl solution was re-neutralized into a basic solution. However, as the pH became neutral, Sb(III) was oxidized back to Sb(V). These results demonstrate that Sb(III) is unstable in water and is easily oxidized to Sb(V); when organic acid is added to the solution, Sb(III) is maintained without oxidization [18], L-ascorbic acid, which is used as an electrolyte for ASV, was added for preprocessing. For the final preprocessing, 45 mL of 5 M HCl and 1.4 g of L-ascorbic acid were added to 10 mL of the sample and left for 10 min, and 45 mL of 5 M KOH was added here and cooled to room temperature. After preprocessing, mercury film dissolution of the working electrode did not occur, and a trend line at the level of R2 = 0.9969 was stably obtained Sb(III), As(III), and total As were analyzed using the measurement method distributed by the ASV manufacturer, and the R2 was Sb(III): 0.9693, As(III): 0.9980, and total As: was found to be 0.9925 (Figure 1).
3.2. Metalloid Adsorption Removal Characteristics
To interpret the antimony adsorption processing results, the surface areas of the selected adsorbents, PAC, zeolite, SP825 made of styrene-divinylbenzene copolymer, and MAC were analyzed using a BET analyzer. Consequently, the specific surface area of PAC was 1098 m2/g, zeolite was 678 m2/g, SP825 was 977 m2/g, and MAC was 1904 m2/g.
To find the adsorption characteristics of PAC, MAC, zeolite, and SP825, the maximum adsorption amount was determined through an isotherm study. After writing and comparing the Langmuir adsorption isotherm (1) and the Freundlich equation (2), a more appropriate adsorption isotherm was determined. First, the Langmuir isotherm had a stable linear equation, with a R2 value of the adsorption isotherm greater than 0.96 for all antimony ions. In contrast, the Freundlich isotherm showed linearity with an R2 value of about 0.94 and was not suitable. Accordingly, it was more appropriate to apply the Langmuir isotherm rather than the Freundlich isotherm for adsorption removal of antimony (see Table 2). The qm values for Sb(III) were 0.126 mg/g for PAC, 0.159 mg/g for MAC, 0.056 mg/g for zeolite, and 0.035 mg/g for SP825. The qm value for Sb(V) was 0.062 mg/g for PAC, 0.090 mg/g for MAC, 0.038 mg/g for zeolite, and 0.118 mg/g for SP825. They are listed in order of size as follows:
- –
-
Maximum adsorption amount for Sb(III) (qm): MAC >PAC > Zeolite > SP825
- –
-
Maximum adsorption amount for Sb(V) (qm): SP825 > MAC > PAC > Zeolite
Table 2.
Langmuir and Freundlich isotherm parameters for Sb.
Table 2.
Langmuir and Freundlich isotherm parameters for Sb.
| Category | Langmuir | Freundlich | |||||
|---|---|---|---|---|---|---|---|
| qm | R2 | qm | R2 | ||||
| Sb(III) | PAC | 0.126 | 0.1121 | 0.9843 | 0.124 | 0.0482 | 0.9513 |
| MAC | 0.159 | 0.0895 | 0.9673 | 0.155 | 0.0298 | 0.9147 | |
| Zeolite | 0.056 | 0.1562 | 0.9981 | 0.058 | 0.0781 | 0.9468 | |
| SP825 | 0.035 | 0.1701 | 0.9998 | 0.031 | 0.0750 | 0.9678 | |
| Sb(V) | PAC | 0.062 | 0.1593 | 0.9975 | 0.058 | 0.0548 | 0.9384 |
| MAC | 0.090 | 0.1393 | 0.9930 | 0.096 | 0.0487 | 0.9288 | |
| Zeolite | 0.038 | 0.1710 | 0.9996 | 0.033 | 0.0742 | 0.9455 | |
| SP825 | 0.118 | 0.1180 | 0.9863 | 0.129 | 0.0318 | 0.9417 | |
Except for SP825, the adsorption performance of the adsorbent per unit mass in an aqueous solution in which Sb(III) is dissolved showed that the removal efficiency of Sb(III) in aqueous solution varied depending on the surface area rather than the surface properties. Sb(III) is electrically neutral as it exists as Sb(OH)3 in water. Therefore, in adsorption removal, removal occurs only by physical adsorption, so MAC with a relatively high surface area showed a high removal rate, and zeolite with the lowest surface area had the lowest removal rate. In contrast, SP825, which has a lower surface area than PAC or MAC, resulted in a higher removal rate for Sb(V). As Sb(V) exists in the form of Sb(OH)6− in water, factors other than the surface area of the adsorbent play a crucial role in removing Sb(V).
Next, to interpret the arsenic adsorption processing results, the surface areas of the selected adsorbents (CAC, PAC, AA, and zeolite) were analyzed using a BET analyzer. The device showed specific surface area of CAC as 1149.8 m2/g, PAC as 1098 m2/g, AA as 342 m2/g, and zeolite as 678 m2/g.
Table 3 lists the arsenic ion adsorption experiment results for all adsorbents. The adsorption performance of each adsorbent differed and depended on factors other than simple surface area. In particular, in AA, the adsorption efficiency for arsenic is high despite its very small surface area (342 m2/g) compared to other adsorbents. This result is attributed to the effect of zeta potential. Since it has been reported that the zeta potential of AA is positive at pH 9 or lower, arsenic, which is mostly in anionic form in water, is easily adsorbed under the influence of zeta potential [16,19].
In addition, a unique adsorption trend was observed in PAC at a concentration ratio of 1:1 because the adsorption ratio of As(V) temporarily showed a negative value. A previous study reported that the oxidation of As(III) to As(V) is accelerated by iron and manganese oxides present in activated carbon. It is presumed that As(III) is oxidized to As(V) during the adsorption process [20,21,22]. Also, the removal efficiency of As(III):As(V) = 1:1 was higher than that of As(III):As(V) = 2:0 or As(III):As(V) = 0:2 when changing the maximum adsorption amount of total arsenic due to the arsenic ion ratio in CAC. Considering the previous results showing that Fe(0) is favorable for adsorbing As(V) when As(III) is oxidized to As(V) during the adsorption reaction, iron ions present in activated carbon are reduced to form Fe(0), which is expected to increase the efficiency of arsenic removal [23]. Moreover, As(III):As(V) = 2:0 is expected to be less affected by the increase in Fe(0) as As(V) does not exist initially. For As(III):As(V) = 0:2, As(III) does not exist, so it is expected that iron ions were not reduced to Fe(0). Then, an isothermal adsorption experiment was conducted for each arsenic ion ratio targeting CAC to identify the removal characteristics of each arsenic ion, which was the most efficient at a 1:1 ratio, and the adsorption isotherm equation was prepared (see Table 4). Compared with Langmuir adsorption isotherm (1) and Freundlich adsorption isotherm, a more appropriate adsorption isotherm was determined. In the Langmuir isotherm graph, the R2 value of the adsorption isotherm for all arsenic ion ratios was greater than 0.98, indicating a stable linear trend. In contrast, the Freundlich isotherm obtained a trend line equation with high linearity with a R2 value of 0.99 or more for some parts but was not suitable when the As(III) ratio was low.
3.3. Continuous Adsorption Tests
Figure 2 illustrates that the amount of antimony removed gradually increased as contact time increased. For removal performance for each adsorbent, MAC removed 100% of antimony under the conditions of LV 1 m/h and SV 2 1/h when the adsorbent was filled to 0.5 m. For PAC, up to 92.1% of total antimony was removed under the conditions of LV 1 m/h and SV 1 1/h with the adsorbent filled to 1 m. For SP825, the total antimony removal rate is not that high, at a maximum of 66%, but the removal rate for Sb(V) showed a high removal efficiency of 100% under the conditions of LV 1 m/h and SV 2 1/h. However, when comparing price and performance, MAC and SP825 are more than eight times more expensive than PAC, making it difficult to apply them to actual industrial wastewater treatment. Accordingly, the applicable adsorbent is PAC, and it would be appropriate to operate at LV 1 m/h and SV 1 1/h.
Figure 3 demonstrates that the overall removal efficiency was favorable for AA, but the efficiency of CAC and PAC continued to increase as the contact time increased, reaching the treatment efficiency target under the conditions of LV 1 m/h and SV 2 1/h with a 30-min contact time. Comparing CAC and PAC, the efficiency of PAC was high in a 5- to 20-min contact time with a filling height of 33 cm, but the efficiency of CAC was high during the 30-min contact time with a filling height of 50 cm. This is due to the formation of a detour in the column caused by the short packing height. The removal efficiency of As(III) was advantageous for CAC and PAC, but the efficiency of AA increased as the contact time increased. The high As(III) removal efficiency of these activated carbons is presumed to be due to the oxidation of As(III) to As(V), which was confirmed in previous batch tests. The removal efficiency of As(V) was favorable compared to that of AA, but the continued increase in efficiency of activated carbon as the contact time increased is presumed to be due to the oxidation of As(III) to As(V).
In summary, AA is considered the most efficient substrate for removing arsenic. However, according to recent research results, AA has a positive surface zeta potential and is effective in removing arsenic and antimony in the form of anions in water but has poor removal efficiency for other harmful heavy metals that exist in the form of positive ions [4,16]. Therefore, the optimal adsorbent was selected as CAC, which easily removes highly toxic As(III) through oxidation. The maximum adsorption amount is the highest in the mixed solution of As(III) and As(V), and the zeta potential is negative, so both cations and anions can be removed. The target concentration for arsenic removal was determined at conditions of LV 1 m/h and SV 2 1/h.
3.4. Identification of As(III) Oxidation during the Adsorption Process
In an arsenic removal experiment, the oxidation of As(III) to As(V) by an adsorbent was observed and investigated. Accordingly, the properties of 1 g of CAC which was selected as the optimal adsorbent were analyzed using inductively coupled plasma optical emission spectroscopy (ICP-OES, Optima 8300, PerkinElmer, Waltham, MA, USA) (see Table 5 and Table 6). Hence, iron and manganese, which were reported to oxidize arsenic, were confirmed [20,22,24].
In terms of the proportion of each substance, the amount of iron was the second largest at 41.4%, the amount of manganese was the 13th largest at 0.1%, and the amount of iron was about 360 times higher than that of manganese. Thus, the effect of iron was the main cause of the oxidation phenomenon of arsenic observed for CAC in previous experiments. Additionally, as there was no change in the amount of Fe after arsenic adsorption, it is assumed that no elution of iron existed during adsorption, and that only the oxidation number of Fe changed according to the oxidation reaction of arsenic.
Therefore, 1 L each of 20 mg/L As(III) solution and As(V) solution were prepared, and the pH at this time was neutral pH of about 6.3. Adsorption batch experiments were conducted to investigate the behavior of Fe ions when oxidation reactions did not occur. At this time, stirring was performed at 220 RPM, and samples were collected by filtering 50 mL every 0, 30, 60, 120, 240, 480, and 600 min. and these samples were analyzed using X-ray photoelectron spectroscopy (XPS, K-Alpha+, Thermo Fisher Scientific, Waltham, MA, USA) (see Figure 4).
The As(III) solution initially contained a lot of Fe(II) at 52.1%. Over time, Fe(II) is oxidized to Fe(III) due to the difference in reduction potential, and the amount of Fe(II) increased from 43.8% to a maximum of 48.5%. However, the Fe(II) decreased sharply compared to the increase in Fe(III), from 52.1% to 31.2%. The significant increase in Fe(0) from the initial 4.2% to 20.3% is associated with the reduced Fe(II) through the oxidation reaction of As(III). Therefore, the amount of Fe(0) rapidly increased in the first 60 min, which was a similar trend to the rapid decrease in the amount of As(III) in the previous experiment. Afterward, the change in Fe(0) gradually decreased and maintained equilibrium for 120 min, which is assumed to be the same as the reduction rate of Fe(II) due to the reduced As(III) and the oxidation rate of Fe(0) due to external factors. After 480 min, the equilibrium of Fe(0) was broken, and its amount decreased from 20.3% to 8.6%. It is believed that as the amount of As(III) decreases, the reaction in which Fe(0) is oxidized to Fe(II) becomes dominant due to external factors such as dissolved oxygen. Iron oxidation in activated carbon favors As(V) over As(III). In the oxidation-based coagulation and precipitation process, operating costs may be reduced by adding spent activated carbon that has reached the break-through point to the preprocessing process instead of the oxidizer [25,26].
In the As(V) solution, the amount of Fe(II) was highest at 52.1%; after 30 min, Fe(II) was present at 42.6% and Fe(III) increased from 43.8% to 47.8% as Fe(II) was oxidized to Fe(III). Afterward, Fe(II) decreased from the initial 52.1% to 44.8%, and Fe(III) increased from the initial 43.8% to a maximum of 51.6%. Fe(0) was in an equilibrium state as it was maintained within a constant amount of 4–5% without significant change.
When As(III) was oxidized to As(V), the adsorption and removal efficiency increased as As(III) was oxidized by Fe(II). Accordingly, to use Fe(II) for arsenic removal, the reaction rate order and reaction rate constant between arsenic and iron were obtained through isolation, and the amount and the rate of oxidation of arsenic were confirmed (Table 7). The average reaction rate order of arsenic was 1.04, and that of iron was 1.06, and each ion followed a first-order reaction rate function. Also, the reaction rate constant was initially fast in oxidation but gradually slowed, with an average value of 7.45 × 10−4.
3.5. Confirmation of Removal Characteristics of Metalloid Ions According to Zeta Potential
To understand the factors affecting the adsorption of metalloids, the surface zeta potential of each adsorbent at pH 7 was measured using a zeta potential analyzer. CAC had a potential of −75 mV at pH 7, 0 mV at pH 2, and −77 mV at pH 13. PAC had a potential of −48 mV at pH 7, 15 mV at pH 2, and −57 mV at pH 13. AA had a potential of 55 mV at pH 7 (pH 3 = 65 mV, pH 13 = −53 mV), zeolite had a potential of −42 mV at pH 7 (pH 2 = −13 mV, pH 13 = −51 mV), while the values for MAC and SP825 were −49 mV and −4 mV, respectively. Antimony and arsenic mainly exist as anions or neutral ions in water, so the removal efficiency was reduced for adsorbents with a negative zeta potential at pH 7. This can be confirmed by the previous experimental results showing that the removal efficiencies of AA and SP825 were particularly high compared to the surface area of the adsorbent. The maximum adsorbed amount for total As at As(III):As(V) = 2:0 was 14.7% higher in efficiency than for CAC, which had the second-best adsorption performance, and As(III):As(V) = 0:2 had an efficiency 89.6% higher than that of CAC. Additionally, zeolite was about 107% more efficient than MAC in processing Sb(V), which mainly exists in the form of Sb(OH)6− in water. Adsorbents with similar surface zeta potentials have different removal efficiencies depending on the surface area of each adsorbent, and for others, a higher removal efficiency was observed for zeta potentials closer to a positive charge. Accordingly, the removal of antimony and arsenic was efficient when the surface zeta potential was close to positive charge, considering the versatility and economic efficiency, it would be advantageous to apply CAC or PAC in the actual treatment process.
