A new insight into the restriction of Cr(VI) removal performance of activated carbon under neutral pH condition

Chromium abounds in nature and is highly toxic in the form of and through bioaccumulation (Li et al. 2018). The chromium pollution of water, land, and environment has attracted the interest of experts as electroplating plants, stainless steel manufacturing plants, leather manufacturing plants, and refractory plants have progressively appeared (Wang et al. 2019; Aparicio et al. 2021; Samuel et al. 2021). Cr(III) presents less mobility, toxicity, and solubility than Cr(VI) and generates sparingly soluble chromium hydroxides (Dognani et al. 2019). The reduction of Cr(VI) to Cr(III) is rapid at acidic conditions; meanwhile, the readily available electron is required for the reduction process (Deng et al. 2020; Fang et al. 2021b). It is well known that removing Cr(VI) from water by reduction and precipitation is a viable option (Barrera-Diaz et al. 2012; Qian et al. 2019; Wu et al. 2019; Zeng et al. 2019). Conventional reducing agents are sulfur and iron salts (Zhang et al. 2021a, 2021b); post-treated effluent that contained sulfate and iron salts would contaminate water and soil. Furthermore, the mandatory wastewater discharge would result in high costs.

The bulk of published studies on Cr(VI) removal by low-cost and readily accessible activated carbon (AC) demonstrated that removal capacity was greater in acidic conditions than alkaline conditions. This suggests that the Cr(VI) elimination is strongly pH dependent (Al-Othman et al. 2012; Zhou et al. 2016; Valentín-Reyes et al. 2019). The removal efficiency of Cr(VI) by activated carbon prepared from teakwood sawdust was 100% at pH 2, while it was below 20% at pH 10 (Ramirez et al. 2020). Table 1 shows the capacity of several ACs to remove Cr(VI) in acidic and alkaline environments. At low pH, the removal capacity was favored because the positively charged surface of Cr(VI) advanced the adsorption of anion Cr(VI) (Norouzi et al. 2018). It is worth noting that the pH-speciation of Cr(VI) reveals that dominated below pH 6, whereas dominates above pH 7 (Rakhunde et al. 2012). Furthermore, prior studies have found that positively charged AC produced by protonation at a low pH value tends to attract chromate anions, which is thought to be the main mechanism for Cr(VI) adsorption (Mohan et al. 2005). It was discovered that as pH dropped, the reduction and adsorption process enhanced simultaneously.

To our understanding, there have been few investigations on systematic chromium adsorption and reduction study when pH rises over 7, to shed light on a substantial drop in AC performance. Most studies have only focused on the unfavorable effect of electrostatic repulsion between chromate anions and negatively charged AC surface at high pH values (Niazi et al. 2018; Ma et al. 2019; Liu et al. 2020). Besides, an earlier study failed to elucidate the removal paths at pH higher than 6 because the surface negatively charged bamboo bark-based AC that unfavored the adsorption of Cr(VI) anions, hence the reduction process of Cr(VI) to Cr(III) at high pH was omitted (Zhang et al. 2015). Cr(III) speciation as a function of pH was depicted clearly by Lopez-Valdivieso's study showing that Cr(OH)₃(s) predominates at pH over 6.4 (Fang et al. 2021a). The effect on the removal of Cr(VI) under alkaline conditions of the AC surface loaded Cr(III) precipitate was especially neglected. Early reported studies on the synthesis of eskolaite (α-Cr₂O₃) nanoparticles through AC following adsorption of Cr(VI) have shown that Cr₂O₃ was the reduced species of Cr(VI) on AC (Cruz-Espinoza et al. 2012; Ibarra-Galván et al. 2014). Recently study showed that the Cr₂O₃ reduced the adsorption rate of Cr(VI) significantly (Wang et al. 2020).

Compared to the consensus that the electrostatic repulsion led to poor Cr(VI) removal efficiency of AC in alkaline conditions, the effect of AC surface-coated Cr₂O₃ precipitate on removal performance was not fully understood. This work aimed to study the effect of powdered AC (PAC) surface-formed Cr₂O₃ precipitate on Cr(VI) removal, SEM-EDX (scanning electron microscope-energy dispersive X-ray analysis), and XPS (X-ray photoelectron spectroscopy) were used to investigate the surface morphology and the chemical properties. Desorption and regeneration experiments were used to confirm the role of Cr₂O₃. The insight gained from this study would help to expand the longevity of AC and the recovery of Cr via AC.

To scrutinize the composition of loaded-chromium on PAC after Cr(VI) removal, three types of de-passivation agents were examined to desorb the adsorbed/reduced chromium on PAC. The formation process of the chromium layer on PAC at pH 3 and 7 was inspected by carrying out consecutive desorption tests following each Cr(VI) adsorption. SEM-EDX (JSM-6610LV, JEOL, Japan) and XPS (K-Alpha, Thermo Scientific, USA) were employed to characterize the distribution of chromium and the chemical species of Cr and C on PAC at pH 3 and 7, respectively. The difference of removal mechanisms under the two pH values was delineated by XPS and Raman spectroscopy (DXR, Thermo Scientific, USA).

All the chemicals were analytical grade and the aqueous solutions were prepared with deionized water through Barnstead pure II water purification system (Thermo Scientific, USA). Potassium dichromate (K₂Cr₂O₇) was purchased from J.T.Baker and used for preparing 1,000mg/L Cr(VI) as a stock solution. 1.0 M H₂SO₄ and 1.0 M NaOH were used to adjust the pH of the aqueous solutions. H₂SO₄, NaOH, KCl, and 1,5–diphenylcarbazide were provided by J.T.Baker. Commercial available AC was provided by Calgon company, its chemical composition was 97% carbon and 3% inorganic residual. The AC was treated by ball milling and obtained a PAC with an average size of 4 μm, a specific surface area of 929 m²/g, and a pore radius of 15.9Å, PAC used in this study was reported in our previous work (Fang et al. 2021b). Following the grinding step, the PAC was dried and stored in a desiccator.

The adsorption capacity of PAC loaded with chromium on Cr(VI) after the last elution assessment was examined, and all the adsorption experiments were conducted with 1,000mg/L Cr(VI).

A colorimetric approach employing 1,5–diphenylcarbazide and a UV-Visible spectrophotometer (Thermo Scientific, USA) coupled with a 1cm quartz cell was used to determine Cr(VI). 100 μL filtered solution was diluted to 10ml and mixed with 0.1mL 49% H₂SO₄, 0.1mL 42.5% H₃PO₄, and 0.4mL 0.2% 1,5–diphenylcarbazide solution, sequentially. The mixed solution stood for 10min and was then measured by a UV-Visible spectrophotometer under 540nm. The absorbance of deionized water was used as a reference. With prepared 0, 0.02, 0.05, 0.1, 0.2, 0.4, 0.6, 0.8, and 1.0mg/L Cr(VI), a standard curve of concentration versus absorbance was constructed; this standard curve was used to determine the Cr(VI) concentration of the sample. The total concentration of aqueous Cr was analyzed by atomic absorption spectrometry (AAS, Varian Spectra 220FS), a 50 μL filtered solution was diluted to 10mL and then sprayed into the flame of air-acetylene. The chromium ground state atoms formed under a high-temperature flame produce selective absorption of the 357.9nm characteristic spectrum of chromium hollow cathode lamps, and the absorbance value is proportional to the concentration of Cr. The standard curve of total Cr was built in the same way as Cr(VI), and the concentration of total was determined by the standard curve as well. The presence of soluble Cr species in the solution was Cr(III) and Cr(VI), the concentration of aqueous Cr(III) was confirmed by the difference between total Cr and Cr(VI).

The surface morphology of PAC after Cr(VI) adsorption under pH 7 was characterized by SEM-EDX and elements mapping. As seen in Figure 1, a chromium layer adsorbed mostly on the PAC surface. A similar observation was reported recently (Wang et al. 2020).

To further inspect the chemical species of Cr on the surface of PAC, XPS analysis was employed. Figure 2(a) and 2(b) show the XPS spectra, which were fitted and deconvoluted into multiple peaks by CasaXPS (version 2.3.23). The peak referenced as C 1 s at 284.8 eV, the Shirley type was designated as the background subtraction. As presented in Figure 2(a), the Cr 2p peak due to Cr(VI), denoted that the Cr(VI) was adsorbed onto PAC. The XPS spectrum of PAC after being treated with Cr(VI) at pH 3 (PAC-pH 3) and 7 (PAC-pH 7) was built as presented in Figure 2(b). The Cr 2p region of the photoelectron spectrum was both detected for PAC-pH 3 and PAC-pH 7, which was consistent with the EDX spectrum shown in Figure 1. Cr 2p involves two energy levels, 2p 1/2 and 2p 3/2. The XPS spectrum of PAC-pH 3 can be divided into the Cr1, Cr2, and Cr3 peaks, where the binding energies (BE) value of Cr1 peak of PAC-pH 3 was 587.5 eV, which was very close to that of Cr₂O₃ (587.4 eV ± 0.2) (Grohmann et al. 1995). The BE for Cr2 and Cr3 of PAC-pH 3 were 579.2 and 577.8 eV, respectively, which can be attributed to Cr(VI) (Murphy et al. 2009; Ren et al. 2016; Zhang et al. 2019). Two contributions of Cr1 and Cr2 for the Cr 2p region of PAC-pH 7 were 587.7 and 578.0 eV, matching well with the binding energy for Cr(VI) and Cr₂O₃ (Biesinger et al. 2004; Ren et al. 2016; Chen et al. 2019). Due to XPS detection depth being no more than 4 nm from the sample surface, it can be said that the chromium layers on the surface of PAC-pH 3 were mainly constituted by Cr(VI) and PAC-pH 7 was mainly constituted by Cr₂O₃(s) (Chowdhury et al. 2012). Consistent with the present results, previous studies have demonstrated that the reduction and adsorption participated principally in the Cr(VI) removal on biomass (Wu et al. 2010; Cui et al. 2011). Moreover, the peak area ratio (Cr1 versus total peaks) as determined by CasaXPS was 69.93% and 39.91% Cr₂O₃(s) on the surfaces of PAC-pH 7 and PAC-pH 3, respectively. This higher content of Cr₂O₃ on PAC-pH 7 clearly evidenced that more Cr₂O₃ formed on the PAC at pH 7 than at pH 3, impeding the diffusion of Cr(VI) into the PAC, leading to a lower level of Cr(VI) removal. Besides, the ratio of O/C on PAC, PAC-pH 3, and PAC-pH 7 were 0.075, 0.113, and 0.157, respectively (Table 2). This indicates more O on the PAC after adsorption at pH 7 than at pH 3, due to more Cr₂O₃ precipitate. As noted in Table 2, the ratio of Cr/C on PAC-pH 3 was higher than that on PAC-pH 7, which further substantiated that Cr(VI) removal efficiency under pH 3 was superior to that under pH 7 and indicated that not only was there Cr₂O₃ on the PAC surface but also Cr(VI).

The surface functional groups of PAC before and after Cr(VI) adsorption were investigated using high resolution C1 s spectra. The deconvolution of C 1 s produced four peaks, as shown in Figure 3. For PAC, there were four components: C = C (284.8 eV), C-O-C/C-OH (285.5 eV), C-O (286.7 eV), and COOR (286.7 eV) (290.0 eV) (Jieying et al. 2014). Similarly, the four peaks of PAC-pH 3 were assigned to the C = C (284.8 eV), C-OH (285.8 eV), C-O (286.5 eV), and COOR (288.7 eV) (Jia & Wang 2015; Ma et al. 2018). Meanwhile, the four peaks for PAC-pH 7 were allocated to C = C (284.8 eV), C-OH (285.8 eV), C-O (286.5 eV), and COOR (289.5 eV) (Chen et al. 2020). The relative percentages of C-OH for PAC, PAC-pH 3, and PAC-pH 7 were 23.62%, 6.77%, and 6.53%, respectively, which suggested that the group of C-OH contributed to the removal of Cr(VI). The oxidation of C-OH to C-O by Cr(VI) caused the increase of the C-O group (Su et al. 2019). Nonetheless, following Cr(VI) adsorption at pH 3, the relative percentage of COOR on PAC rose from 15.22% to 20.56% and dropped to 10.41% after Cr(VI) adsorption at pH 7. This inconsistency may be due to Cr(VI) oxidizing the surface of PAC at pH 3 and introducing more COOR groups (Yin et al. 2019), while the Cr(VI) exhibited weak oxidative capacity at higher pH (Gangadharan & Nambi 2014), and the removal of Cr(VI) under pH 7 consumed the COOR groups through the complex.

Raman spectroscopy investigation was carried out to evaluate the degree of structural order in carbonaceous PACs, as well as to investigate the difference in Cr(VI) removal mechanisms at pH 3 and pH 7. As depicted in Figure 4, the two sharp and strong bands are associated with the D-band (1,319 cm⁻¹. defect with sp³ bonding) and G-band (1,563 cm⁻¹, graphitization with sp² bonding) (Cuesta et al. 1994). The intensity ratio of D-band (I_D) versus G-band (I_G) is often used to assess the degree of disorder in graphite structure in carbon materials (Lespade et al. 1984). The value of I_D/I_G declined from 1.12 to 1.10 after treatment at pH 3 but increased from 1.12 to 1.14 after treatment at pH 7. As a result, the existence of defects in PAC was strengthened under pH 7, while at pH 3, a well-organized carbon structure formed. This divergence could be explained by the different Cr(VI) adsorption mechanisms at pH 3 and 7. In general, the reduction of surface oxygen-containing functional groups resulted in the increase of amorphous carbon at pH 7 (Cong et al. 2012; Yang et al. 2013; Li et al. 2016; Wang et al. 2017), while the oxidation of hybridized carbon atoms caused structural order to grow at pH 3 (Osswald et al. 2006).

Figure 5 shows a comparison of adsorption performance at pH 3, 7, and 9 for 1,000 mg/L initial concentration of Cr(VI). Adsorption at those three pH values both reached pseudo-equilibrium immediately after 10 mins. Removal capacities for PAC at pH 7 and 3 were 16.8 mg/g and 20 mg/g, respectively, equivalent to 83.86% and nearly 100% removal efficiency. And adsorption capacity under pH 9 was 7.8 mg/g, which was much lower than that at pH 3 and 7. It indicated that the removal performance of PAC on Cr(VI) was higher at low pH, PAC is probably protonated and attracted more anionic Cr(VI).

The desorption of chromium from the Cr-loaded PAC was evaluated using various chemical agents including H₂SO₄, KCl, and NaOH. Abundant chromium-loaded PAC was prepared at pH 7. As shown in Figure 6, the Cr(VI) elimination experiment using PAC was repeated four times. The duration time for every removal cycle was 24 hours. The next removal cycle began once the previous cycle was completed. After four consecutive repetitive adsorption cycles, the content of chromium on PAC reached 48.6 mg/g. The results of the desorption analysis using the three reagents are shown in Figure 7. It is noted that only Cr(III) was dissolved in H₂SO₄ aqueous solutions, whereas Cr(VI) was only desorbed in NaOH aqueous solutions. Negligible Cr(III) or Cr(VI) were detected in the KCl solution when compared to acidic and alkaline aqueous solutions. These findings agree well with Ouki & Neufeld's (1997) findings that 3 g/L Cr(III) and 8.4 g/L Cr(VI) were recovered when exhausted carbon was regenerated under acidic and alkaline conditions, respectively (Ouki & Neufeld 1997). Due to the great stability of the adsorbed chromium on the PAC, desorption of Cr(VI) and Cr(III) from the Cr-loaded PAC with deionized water was minimal. Our results are also in line with those of Jing et al. (2011), who found that the desorption rate of Cr-loaded AC was low with distilled water (Jing et al. 2011).

These findings could be useful in the development of a selective recovery method of Cr(III) and Cr(VI) using acid and alkali aqueous solutions.

To clarify the influence of pH on the development process of the chromium layer on PAC, H₂SO_4, and NaOH desorption agents were used to determine the content of Cr(III) and Cr(VI) adsorbed on PAC-pH 3 and PAC-pH 7. Figure 8 compares the results obtained from elution tests of PAC-pH 3 and PAC-pH 7 after three consecutive adsorption cycles. It is apparent from the figure that the increment of adsorbed Cr(VI) is higher than reduced Cr(III) at each cycle for both PAC-pH 3 and PAC-pH 7. Hence, it is conceivable to suggest that the adsorption process prevailed for Cr(VI) elimination. This finding was also reported by Daneshvar et al. (2019). Another significant observation was that for PAC-pH 3 and PAC-pH 7, the adsorbed Cr(VI) and reduced Cr(III) decreased as the cycles progressed. PAC-pH 3 showed higher Cr(VI) adsorption and reduction capacity. This may be due to the more generated Cr(III) precipitate accumulating on the PAC surface over time at the neutral condition, sheltering the PAC active sites from Cr(VI adsorption.

The performance of the PAC for Cr adsorption was assessed after chromium was desorbed from the PAC using the H₂SO₄ and NaOH. Re-adsorption experiments were conducted following the third cycle desorption step. As can be seen in Figure 9, 92.43% removal efficiency of Cr(VI) was achieved by PAC-pH 7 after washing with H₂SO₄, whereas only 51.72% removal was attained using NaOH aqueous. Therefore, it can be inferred that the Cr(III) precipitate is mainly responsible for the poor performance of PAC under neutral conditions. The removal performance after acid washing (92.43%) was higher than the preliminary removal efficiency (83.86%); this result indicated that the acid desorption procedure modified PAC properties and introduced surface functional groups. These results are consistent with those of Huang et al. (2009), who improved AC's Cr(VI) removal capacity by modifying it with nitric acid (Huang et al. 2009). As a result, it is proved that PAC's limited removal capability for Cr(VI) at pH 7 was mostly due to Cr(III) precipitate that formed on the surface of PAC. This finding backs the XPS results in that chromium oxide piled up mostly on PAC under neutral conditions. The sulfuric acid proved to be a potential chemical agent for the regeneration of Cr-loaded PAC after treating water contaminated with Cr(VI).

This study aimed to determine the mechanism of the limited sequestration capability of PAC on Cr(VI) under neutral conditions compared to acidic conditions. SEM-EDX substantiated that a chromium layer was formed on PAC, while XPS spectra corroborated the higher Cr₂O₃ content on PAC under alkaline conditions, resulting in poor Cr(VI) removal performance. Conversely, a lower Cr₂O₃ content on PAC under acid conditions is related to the higher Cr(VI) removal capacity. Desorption tests with H₂SO₄ and NaOH solution revealed that the precipitated Cr₂O₃ and adsorbed Cr(VI) can be selectively desorbed, proving that adsorption and reduction processes contributed significantly to the Cr(VI) removal. Consecutive desorption assays proved that the reduction and adsorption capability at 7 declined with time and were both lower than at pH 3. This is due to Cr(III) precipitate and adsorbed Cr(VI) blocking active sites. The superior performance on Cr(VI) removal of Cr-loaded PAC after desorption by H₂SO₄ further confirmed that the restricted removal performance under neutral conditions was ascribed to the formation of a Cr₂O₃ passivation layer on the surface of PAC particles. The insights gained from this work may be of assistance to the recycling of chromium from exhausted AC and extend the lifespan of AC.

Yi Fang acknowledges CONACYT (National Council of Science and Technology, Mexico) for the Fellowship (No. 900339) to pursue Ph.D. studies in Materials Science and Engineering at the Universidad Autónoma de San Luis Potosí, México. Authors acknowledge the partial financial support by the Guangdong College Students' Innovative Project (pdjh2021b0538, X201910580158), and the Zhaoqing Science and Technology Project (2018N005) to carry out this research.

All relevant data are included in the paper or its Supplementary Information.