Estimating the Reducing Power of Carbon Nanotubes and Granular Activated Carbon Using Various Compounds

: Determining the degree of the reducing power of multi-walled carbon nanotubes (MWC-NTs) and granular activated carbon (GAC) enables their effective application in various ﬁelds. In this study, we estimate the reducing power of carbon nanotubes (CNTs) and GAC by measuring the reduction degree of various compounds with different reduction potentials. MWCNTs and GAC materials can reduce Cr(VI), Fe(III) and PMo 12 O 40 3 − , where the reduction potentials range from +1.33 V to +0.65 V. However, no reduced forms of PW 12 O 403 − and SiW 12 O 404 − compounds were detected, indicating that the reducing power of MWCNTs and GAC is insufﬁcient for reduction potentials in the range +0.218 V to +0.054 V. MWCNTs exhibit a short reduction time (5 min), whereas GAC exhibits a gradually increasing reduction degree of all the compounds assessed until the end of the reaction. This indicates a higher reduction degree than that of MWCNTs systems. Acidic initial pH values favor reduction, and the reduction degree increases as the pH becomes lower than 4.0. Moreover, large quantities of MWCNTs and GAC increase the concentrations of the reduced compounds.


Introduction
Carbon-based materials and their derivatives have been widely studied and used in various applications, due to their excellent adsorption ability, good conductivity, thermal and chemical stability and low cost [1,2]. Particularly, the use of carbon-based materials in water purification has received much attention in pollutant removal via adsorption, redox and catalytic reactions [3]. Among numerous carbon-based materials, granular activated carbon (GAC) and carbon nanotubes (CNTs) have been known as promising materials to remove inorganic [4,5] and organic compounds [6,7], due to their faster kinetics and higher reactivity than other common adsorbents. Moreover, GACs and CNTs can be regenerated and reused easily [8]. GAC materials are synthesized from raw materials such as fruit seeds, coconut shell, wood and lignin, and have high adsorption capacity in pollutant removal, owing to their very large inner surface area (800~1200 m 2 /g) induced by pore structures. The performance is determined by the internal pore structure, size and physico-chemical properties [9]. Another efficient material, cylindrical hollow CNTs, are different in that they have tubular morphologies with graphene layers [10]. Owing to this unique structure, they possess superior physical and chemical properties, such as large specific surface area, electrochemical stability and mechanical and electric properties, leading to many applications [7,11]. These CNTs can be divided into single-walled carbon nanotubes (SWCNTs) and multi-walled carbon nanotubes (MWCNTs) according to the number of layers. Removal of pollutants such as copper, silver, lead and nickel using pristine GAC and CNTs materials via adsorption has been studied extensively [12][13][14].
Recently, the use of functionalized GAC and CNTs produced via oxidative/acidic treatment or addition of metals has also received considerable attention as adsorbents, electron donors and electron-transfer mediators [15][16][17][18][19][20]. Modification of GAC and CNTs via acid treatment improves the adsorption of heavy metals, by inducing electrostatic attraction between the metallic ions and newly introduced functional groups such as -OH, -C=O and -COOH on their surfaces [21][22][23]. Such modified CNTs and GAC can also serve as catalysts or reductants that mineralize the target contaminants [24]. Some studies have previously verified their role as electron donors, exhibiting metal reduction [25][26][27] under certain conditions. However, systematic studies on estimating the reducing power are lacking. Estimating the reducing power of GAC and CNTs may provide insights into their applicability in redox reactions. To address this issue, we evaluated the reducing power of GAC and CNTs using various compounds with different reduction potentials (i.e., Cr(VI) (E 0 = +1.33 V), Fe(III) (E 0 = +0.77 V) and PMo 12 O 40 3− (E 0 = +0.65 V), PW 12  ) in this study, and estimated their reduction potentials ( Figure 1) [28][29][30]. Chromium commonly requires the removal and/or reduction of Cr(VI) to Cr(III) in water treatment due to its toxicity and Fe(III) was selected owing to its easy-accessibility [31,32]. For other polyoxometalates, given that they have unique chemical structure, polynuclear metal-oxo clusters with discrete and chemically modifiable structures, the intention of choosing these compounds was to investigate if CNTs and GAC are able to reduce them regardless of their unique chemical structures [33]. In addition, the effect of several experimental factors such as the initial pH and dosage of GAC and CNTs on the reducing power were investigated.
Water 2021, 13,1959 2 of 10 nickel using pristine GAC and CNTs materials via adsorption has been studied extensively [12][13][14]. Recently, the use of functionalized GAC and CNTs produced via oxidative/acidic treatment or addition of metals has also received considerable attention as adsorbents, electron donors and electron-transfer mediators [15][16][17][18][19][20]. Modification of GAC and CNTs via acid treatment improves the adsorption of heavy metals, by inducing electrostatic attraction between the metallic ions and newly introduced functional groups such as -OH, -C=O and -COOH on their surfaces [21][22][23]. Such modified CNTs and GAC can also serve as catalysts or reductants that mineralize the target contaminants [24]. Some studies have previously verified their role as electron donors, exhibiting metal reduction [25][26][27] under certain conditions. However, systematic studies on estimating the reducing power are lacking. Estimating the reducing power of GAC and CNTs may provide insights into their applicability in redox reactions. To address this issue, we evaluated the reducing power of GAC and CNTs using various compounds with different reduction potentials (i.e., Cr(VI) (E 0 = +1.33 V), Fe(III) (E 0 = +0.77 V) and PMo12O40 3− (E 0 = +0.65 V), PW12O40 3− (E 0 = +0.218 V) and SiW12O40 4− (E 0 = +0.054 V)) in this study, and estimated their reduction potentials ( Figure 1) [28][29][30]. Chromium commonly requires the removal and/or reduction of Cr(VI) to Cr(III) in water treatment due to its toxicity and Fe(III) was selected owing to its easy-accessibility [31,32]. For other polyoxometalates, given that they have unique chemical structure, polynuclear metal-oxo clusters with discrete and chemically modifiable structures, the intention of choosing these compounds was to investigate if CNTs and GAC are able to reduce them regardless of their unique chemical structures [33]. In addition, the effect of several experimental factors such as the initial pH and dosage of GAC and CNTs on the reducing power were investigated.

Experimental Setup and Procedure
All experiments on redox reactions of carbonaceous materials were conducted in 100 mL-Erlenmeyer flasks under continuous N 2 purging to maintain an anaerobic condition. Each flask contained a 100 mL reaction solution prepared from an aliquot of a predetermined amount of stock solution in desired concentrations (0.1-1.0 mM). The initial pH was adjusted using 1 M HClO 4 or 1 M NaOH solutions. To initiate the redox reactions, a designated amount of carbonaceous materials was added to the solution to achieve the concentration of 0.1−2.0 g/L, and the suspension underwent vigorous magnet-mixing. In addition, 1 mL of suspension was withdrawn at each sampling and immediately filtered using a 0.45 µm PTFE syringe. The suspension was then analyzed to measure the concentrations of dissolved Fe(II), Cr(III) and Cr(VI), and reduced polyoxometalates (POM) anions.

Analytical Methods
The Brunauer-Emmett-Teller surface area, pore volume, and average pore size of carbonaceous materials were determined by adsorbing N 2 gas molecules at 77 K using an ASAP 2420 system (Micromeritics Instrument). The concentration of dissolved Fe(II) (Supplementary Materials) was determined using the 1,10-phenanthroline method [34]. In addition, Cr(VI) was analyzed by the 1,5-diphenylcarbazide method [35]. The concentration of total chromium (Cr(III) plus Cr(VI)) in the solution was measured when necessary, according to the method for the Cr(VI) measurement described above. However, in this case, the remaining Cr(III) was intentionally oxidized to Cr(VI) by adding several drops of 10 mM KMnO 4 oxidant to the sample. The concentration of Cr(III) was determined from the difference between the concentrations of total chromium (Cr(III), Cr(VI)) and Cr(VI). The details of Fe(II) and Cr(VI) measurements are described in Section 1 of the (supporting information, SI). The reduced POM (PW 12 O 40 4− , SiW 12 O 40 5− and PMo 12 O 40 4− ) were monitored by measuring the UV/visual absorption spectra at 730 nm using a UV−vis spectrophotometer (S-3100, Sinco). All experiments were conducted in duplicate or triplicate, and the means with standard deviations were taken as the results.

Estimating the Reducing Power of CNTs and GAC Using Various Compounds
To compare the reducing power of CNTs, two MWCNTs (Hanwha and Comocat) and an SWCNT (nano-lab) were applied in the reduction of Fe(III) to Fe(II) (Figure 2). By comparing the two differently manufactured MWCNTs (Hanwha vs. Comocat), no significant difference in the Fe(II) concentration was observed in 20 min of reaction. At a longer reaction time (60 min), MWCNT-Hanwha showed gradually increasing Fe(II) concentration up to 0.1 mM and no significant change was observed in the Fe(II) concentration for MWCNT-Comocat. In the comparison of the reducing capacity of MWCNTs with that of SWCNTs, a small discrepancy was observed in the reduction degree of Fe(III) until 20 min, whereas SWCNTs measured the lessen and more Fe(II) concentrations at 60 min than the systems of MWCNT-Hanwha and MWCNT-Comocat, respectively ( Figure 2). Considering that MWCNTs are less toxic and more cost-effective than SWCNTs [36,37], we used the MWCNT-Comocat as a representative CNT for the reduction of other compounds. The reducing power of MWCNTs and GAC was estimated and compared by measuring the concentration of the reduced forms of all the assessed compounds. In addition, the reduction degree (%) of the compounds was quantified using Equation (1) at reaction times of 5 min and 60 min, which are representative of early and longer reaction times, respectively. For the PMo12O40 3− compound, the reduction degree was estimated by comparing the degree of absorbance measured at the initial and sampling times.
MWCNTs and GAC materials were able to reduce compounds having reduction potentials E o = +1.33 V (Cr(VI → III)) to E o = +0.65 V (PMo12O40 3− → PMo12O40 4− ). However, the reduced forms of PW12O40 3− and SiW12O40 4− were not detected in both systems, indicating that MWCNTs and GAC could not reduce compounds with reduction potentials less than +0.218 V (Figures S1 and S2 in SI). By comparing the reducing power of MWCNTs and GAC at the early stage (5 min reaction time), MWCNTs showed a higher reduction degree (%) than that of GAC for Fe(III) and PMo12O40 3− . However, no difference in the reduction degree (%) for Cr(VI) was observed between MWCNTs and GAC systems ( Figure 3a). However, at the long reaction time (60 min), GAC showed higher reducing capability than that of MWCNTs, showing 70%, 40% and 30% reduction of Cr(VI), Fe(III) and PMo12O40 3− , respectively ( Figure 3b). Thus, it can be said that while MWCNTs exhibit strong reducing power in short periods, GAC requires longer reaction times to exhibit a continuous increase in the reducing power. The reducing power of GAC is mostly higher than that of MWCNTs at the final reaction time (Figures 1, S1 and S2). Even though GAC possesses a higher specific surface area than MWCNTs (Table 1), it contains numerous inaccessible micropores in aggregated form, which requires the diffusion of target compounds into the internal porous structure [38][39][40]. In contrast with GAC materials, MWCNTs exhibit a high available surface area for the target compounds on both their external surfaces and loose bundles/aggregate forms [40,41]. Therefore, MWCNTs may be advantageous for fast redox reactions, whereas GAC requires diffusion and contact time for the adsorption and reduction of the compounds due to the unique porous structure. This hypothesis is also supported by the results of chromium adsorption (Crtotal = Cr(VI) + Cr(III)) ( Figure 4). MWCNTs showed the fastest adsorption of Crtotal in 5 min, and no significant change of the Crtotal adsorption degree until 60 min was observed. GAC exhibited steadily increasing Crtotal adsorption degree during the whole reaction time (Figure 4). Interestingly, even though GAC showed a higher reduction degree of Cr(VI) than that of MWCNTs, a lesser The reducing power of MWCNTs and GAC was estimated and compared by measuring the concentration of the reduced forms of all the assessed compounds. In addition, the reduction degree (%) of the compounds was quantified using Equation (1) at reaction times of 5 min and 60 min, which are representative of early and longer reaction times, respectively. For the PMo 12 O 40 3− compound, the reduction degree was estimated by comparing the degree of absorbance measured at the initial and sampling times. concentration/absorbance o f the reduced f orms measured initial concentration/absorbance o f the target compound × 100% = reduction degree(%) MWCNTs and GAC materials were able to reduce compounds having reduction potentials E o = +1.33 V (Cr(VI → III)) to E o = +0.65 V (PMo 12 O 40 3− → PMo 12 O 40 4− ). However, the reduced forms of PW 12 O 40 3− and SiW 12 O 40 4− were not detected in both systems, indicating that MWCNTs and GAC could not reduce compounds with reduction potentials less than +0.218 V (Figures S1 and S2 in SI). By comparing the reducing power of MWCNTs and GAC at the early stage (5 min reaction time), MWCNTs showed a higher reduction degree (%) than that of GAC for Fe(III) and PMo 12 O 40 3− . However, no difference in the reduction degree (%) for Cr(VI) was observed between MWCNTs and GAC systems ( Figure 3a). However, at the long reaction time (60 min), GAC showed higher reducing capability than that of MWCNTs, showing 70%, 40% and 30% reduction of Cr(VI), Fe(III) and PMo 12 O 40 3− , respectively (Figure 3b). Thus, it can be said that while MWCNTs exhibit strong reducing power in short periods, GAC requires longer reaction times to exhibit a continuous increase in the reducing power. The reducing power of GAC is mostly higher than that of MWCNTs at the final reaction time (Figure 1, Figures S1 and  S2). Even though GAC possesses a higher specific surface area than MWCNTs (Table 1), it contains numerous inaccessible micropores in aggregated form, which requires the diffusion of target compounds into the internal porous structure [38][39][40]. In contrast with GAC materials, MWCNTs exhibit a high available surface area for the target compounds on both their external surfaces and loose bundles/aggregate forms [40,41]. Therefore, MWCNTs may be advantageous for fast redox reactions, whereas GAC requires diffusion and contact time for the adsorption and reduction of the compounds due to the unique porous structure. This hypothesis is also supported by the results of chromium adsorption (Cr total = Cr(VI) + Cr(III)) ( Figure 4). MWCNTs showed the fastest adsorption of Cr total in 5 min, and no significant change of the Cr total adsorption degree until 60 min was observed. GAC exhibited steadily increasing Cr total adsorption degree during the whole reaction time (Figure 4). Interestingly, even though GAC showed a higher reduction degree of Cr(VI) than that of MWCNTs, a lesser adsorption degree of Cr total was observed in the GAC system than the MWCNT systems, even at a long reaction time (60 min). Likely, not all of the adsorbed Cr(VI) onto the MWCNTs is reduced, and thus, a portion of the adsorbed Cr(VI) may remain as an adsorption phase without reduction. However, this requires further verification. 21, 13,1959 5 of 10 adsorption degree of Crtotal was observed in the GAC system than the MWCNT systems, even at a long reaction time (60 min). Likely, not all of the adsorbed Cr(VI) onto the MWCNTs is reduced, and thus, a portion of the adsorbed Cr(VI) may remain as an adsorption phase without reduction. However, this requires further verification.  In Figure 3a, we consider that Cr(VI) tends to be relatively readily reduced among the compounds assessed due to its high reduction potential (E o = +1.33 V). GAC can reduce Cr(VI) to Cr(III) in a short time (5 min), equivalent to that of MWCNTs and showing no difference in the Cr(VI) reduction degree between MWCNTs and GAC. In addition, at long reaction time (60 min), GAC showed about 2.5 and 1.5 times higher reduction degree (%) for Cr(VI) and PMo12O40 3− than MWCNTs, respectively. The MWCNTs exhibited approximately 1.1 times higher reduction degree of Fe(III) than that of GAC showed ( Figure  3b). It is hypothesized that the dissolved Fe(III) ions may be beneficial for adsorption to on the reactive surface area and/or the pore structure of MWCNTs, due to its smaller molar mass (molar mass: 55.845 g/mol) than other compounds (Cr2O7 2-, molar mass: 386.92  adsorption degree of Crtotal was observed in the GAC system than the MWCNT systems, even at a long reaction time (60 min). Likely, not all of the adsorbed Cr(VI) onto the MWCNTs is reduced, and thus, a portion of the adsorbed Cr(VI) may remain as an adsorption phase without reduction. However, this requires further verification.  In Figure 3a, we consider that Cr(VI) tends to be relatively readily reduced among the compounds assessed due to its high reduction potential (E o = +1.33 V). GAC can reduce Cr(VI) to Cr(III) in a short time (5 min), equivalent to that of MWCNTs and showing no difference in the Cr(VI) reduction degree between MWCNTs and GAC. In addition, at long reaction time (60 min), GAC showed about 2.5 and 1.5 times higher reduction degree (%) for Cr(VI) and PMo12O40 3− than MWCNTs, respectively. The MWCNTs exhibited ap- In Figure 3a, we consider that Cr(VI) tends to be relatively readily reduced among the compounds assessed due to its high reduction potential (E o = +1.33 V). GAC can reduce Cr(VI) to Cr(III) in a short time (5 min), equivalent to that of MWCNTs and showing no difference in the Cr(VI) reduction degree between MWCNTs and GAC. In addition, at long reaction time (60 min), GAC showed about 2.5 and 1.5 times higher reduction degree (%) for Cr(VI) and PMo 12 O 40 3− than MWCNTs, respectively. The MWCNTs exhibited approximately 1.1 times higher reduction degree of Fe(III) than that of GAC showed (Figure 3b). It is hypothesized that the dissolved Fe(III) ions may be beneficial for adsorption to on the reactive surface area and/or the pore structure of MWCNTs, due to its smaller molar mass (molar mass: 55.845 g/mol) than other compounds (Cr 2 O 7 2− , molar mass: 386.92 g/mol and PMo 12 O 40 3− , molar mass: 1822.470 g/mol). This leads to the increasing reduction degree, which is almost equivalent to that of GAC [41,42].

Effect of Initial pH on the Reducing Power of MWCNTs and GAC
The generated concentrations of the reduced forms such as Cr(III), Fe(II) and PMo 12 O 40 4− were measured at various pH conditions from pH 2 to pH 7 ( Figure 5). In the range of pH values 5-7, the reduced forms of the compounds were not detected in both the MWCNTs and GAC systems. For the MWCNT systems, significantly increasing concentrations of the reduced compounds were observed as the pH decreased from pH 4 to 2. However, GAC did not show the reduction of Cr(VI), Fe(III) and PMo 12 O 40 3− in this pH range ( Figure 5). This indicates that the reducing power of MWCNTs was more sensitively influenced by the pH condition than that of GAC systems. 6 of 10 g/mol and PMo12O40 3− , molar mass: 1822.470 g/mol). This leads to the increasing reduction degree, which is almost equivalent to that of GAC [41,42].

Effect of Initial pH on the Reducing Power of MWCNTs and GAC
The generated concentrations of the reduced forms such as Cr(III), Fe(II) and PMo12O40 4− were measured at various pH conditions from pH 2 to pH 7 ( Figure 5). In the range of pH values 5-7, the reduced forms of the compounds were not detected in both the MWCNTs and GAC systems. For the MWCNT systems, significantly increasing concentrations of the reduced compounds were observed as the pH decreased from pH 4 to 2. However, GAC did not show the reduction of Cr(VI), Fe(III) and PMo12O40 3− in this pH range ( Figure 5). This indicates that the reducing power of MWCNTs was more sensitively influenced by the pH condition than that of GAC systems. There are two possible reasons why the acidic initial pH condition is favorable for the reduction reactions in both the MWCNTs and GAC systems. Considering that most metals exist in soluble forms such as Fe 3+ and Cr2O7 2− , HCrO4 − and Cr 3+ or Cr(H2O)6 3+ at decreasing pH conditions (<pH 4), increasing pH conditions can generate insoluble metalhydroxide precipitates such as Fe(OH)n 3−n and Cr(OH)x. We hypothesized that soluble forms of metals may be more beneficial to adsorption and reduction reactions in the MWCNTs and GAC systems [43,44]. The other possible reason is that the reduction reaction may consume protons according to the corresponding redox reaction equations, such as in the chromium compound (Figure 1) [45]. In case of POM compounds, it is known that a single electron transfer occurs under neutral pH conditions, and two electrons are transferred under highly acidic pH conditions. Thus, acidic conditions can increase the driving force to trigger reduction reactions [46,47].

Effect of MWCNTs and GAC Dosage on the Reducing Power
Regarding the effect of the dosing amount of MWCNTs and GAC on the reducing power, increasing amounts of carbonaceous materials resulted in high concentrations of There are two possible reasons why the acidic initial pH condition is favorable for the reduction reactions in both the MWCNTs and GAC systems. Considering that most metals exist in soluble forms such as Fe 3+  at decreasing pH conditions (<pH 4), increasing pH conditions can generate insoluble metal-hydroxide precipitates such as Fe(OH) n 3−n and Cr(OH) x . We hypothesized that soluble forms of metals may be more beneficial to adsorption and reduction reactions in the MWCNTs and GAC systems [43,44]. The other possible reason is that the reduction reaction may consume protons according to the corresponding redox reaction equations, such as in the chromium compound ( Figure 1) [45]. In case of POM compounds, it is known that a single electron transfer occurs under neutral pH conditions, and two electrons are transferred under highly acidic pH conditions. Thus, acidic conditions can increase the driving force to trigger reduction reactions [46,47].

Effect of MWCNTs and GAC Dosage on the Reducing Power
Regarding the effect of the dosing amount of MWCNTs and GAC on the reducing power, increasing amounts of carbonaceous materials resulted in high concentrations of Water 2021, 13,1959 7 of 10 the reduced forms of all compounds ( Figure 6). However, GAC materials did not induce the reduction of PMo 12 O 40 3− until 1.0 g/L of GAC was applied. Then, increasing reducing power was observed at 2.0 g/L of GAC concentration. Given that it is relatively difficult to reduce PMo 12 O 40 3− due to its low reduction potential, it is likely that the reducing power of GAC at 1.0 g/L concentration might not have been enough to initiate the reduction of PMo 12 O 40 3− . In contrast with PMo 12 O 40 3− , no difference in the Cr(III) concentration measured between MWCNTs and GAC systems was observed up to doses of 0.5 g/L ( Figure 6). This is also related to the reduction tendency of the target compound, which means that the amount of reductants such as MWCNTs and GAC is not a limiting factor to determine the reduction degree in the presence of easily reduced compounds with high reduction potential. 7 of 10 to reduce PMo12O40 3− due to its low reduction potential, it is likely that the reducing power of GAC at 1.0 g/L concentration might not have been enough to initiate the reduction of PMo12O40 3− . In contrast with PMo12O40 3− , no difference in the Cr(III) concentration measured between MWCNTs and GAC systems was observed up to doses of 0.5 g/L ( Figure  6). This is also related to the reduction tendency of the target compound, which means that the amount of reductants such as MWCNTs and GAC is not a limiting factor to determine the reduction degree in the presence of easily reduced compounds with high reduction potential.

Conclusions
In this study, the reducing capability of CNTs (MWCNTs) and GAC was indirectly determined using the reduction degree of five compounds with different reduction potentials. Our findings show that both MWCNTs and GAC were able to reduce Cr(VI), Fe(III) and PMo12O40 3− , where the reduction potentials range from +1.33 V to +0.65 V. However, they failed to reduce PW12O40 3− and SiW12O40 4− with reduction potentials +0.218 V and +0.054 V, respectively. This is likely that the difference on the reduction potentials between CNTs/GAC and PW12O40 3− /SiW12O40 4− was not enough to induce electron transfer. In addition, while MWCNTs exhibit potent reducing power in a short period (5 min), GAC consistently increased the reduction degree. This observation led us to hypothesize that MWCNTs have easily accessible external and internal reactive sites for target compounds, whereas GAC requires diffusion and contact time for the adsorption and reduction of the compounds, due to their unique pore structure. This hypothesis was also supported by the adsorption study of Crtotal, where fast saturation was observed in the MWCNTs systems (5 min) and gradually increasing adsorption of Crtotal was observed in the GAC systems. Moreover, acidic initial pH conditions were more favorable for the reduction reaction, and large quantities of MWCNTs and GAC induced a high reduction rate as expected.
Carbon is the most common chemical element and the materials made by carbon can be obtained in an eco-friendly and low-cost procedure [48,49]. Even though carbon-based materials such as MWCNTs and GAC may be not the best beneficial materials in terms of cost when they were used as reductants in water treatment, this study estimating the

Conclusions
In this study, the reducing capability of CNTs (MWCNTs) and GAC was indirectly determined using the reduction degree of five compounds with different reduction potentials. Our findings show that both MWCNTs and GAC were able to reduce Cr(VI), Fe(III) and PMo 12 O 40 3− , where the reduction potentials range from +1.33 V to +0.65 V. However, they failed to reduce PW 12 O 40 3− and SiW 12 O 40 4− with reduction potentials +0.218 V and +0.054 V, respectively. This is likely that the difference on the reduction potentials between CNTs/GAC and PW 12 O 40 3− /SiW 12 O 40 4− was not enough to induce electron transfer. In addition, while MWCNTs exhibit potent reducing power in a short period (5 min), GAC consistently increased the reduction degree. This observation led us to hypothesize that MWCNTs have easily accessible external and internal reactive sites for target compounds, whereas GAC requires diffusion and contact time for the adsorption and reduction of the compounds, due to their unique pore structure. This hypothesis was also supported by the adsorption study of Cr total , where fast saturation was observed in the MWCNTs systems (5 min) and gradually increasing adsorption of Cr total was observed in the GAC systems. Moreover, acidic initial pH conditions were more favorable for the reduction reaction, and large quantities of MWCNTs and GAC induced a high reduction rate as expected.
Carbon is the most common chemical element and the materials made by carbon can be obtained in an eco-friendly and low-cost procedure [48,49]. Even though carbon-based materials such as MWCNTs and GAC may be not the best beneficial materials in terms of cost when they were used as reductants in water treatment, this study estimating the range of their reducing powers was able to provide an insight into their many applications not only as adsorbents but also as reductants and catalysts. In addition, it is expected that these easily available carbonaceous materials can function as activators in reducing Water 2021, 13,1959 8 of 10 peroxymonosulfate and peroxydisulfate oxidants by generating various types. This is an issue that needs further investigation. This will be beneficial to water purification via redox reactions using carbon-based materials.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available upon request from the corresponding author.