Perchlorate
Perchlorate can form naturally in the atmosphere, with trace levels of perchlorate in precipitation (ATSDR 2008). Five perchlorates can be formed in large amounts, i.e. magnesium, potassium, ammonium, sodium, and lithium perchlorate. Perchlorate finds its application in explosives, batteries, adhesives, and bleach. Perchlorate primarily affects the thyroid organ and interferes in the uptake of iodine (ATSDR 2008; Colinas et al. 2016). An overview of the experimental parameters and optimized conditions from batch adsorption experiments for perchlorate is presented in Table 6.3.
Effect of pH
The effect of pH with adsorption capacity of perchlorate varies with the adsorbent. The adsorption was maximum at pH 2 with biochar (Fang et al. 2013) and pH 4 with Mg/(A1 + Fe) hydrotalcite (Yang, Gao, Chu, et al. 2012), independent of pH in the range of 4-10 with the calcined product of Mg-AlCO, layered double hydroxide (Lin et al. 2014) and near isoelectric point with oxidized carbon nanotubes (Fang and Chen 2012).
The perchlorate removal with the nano iron oxide-doped activated carbon was more favorable in acidic medium (max at pH 6; however, increases slightly on rise of pH from 4 to 6) (Xu et al. 2015). This was attributed to the electrostatic attraction. At low pH, the surface of the adsorbent carries more positive charge. On rise of pH (after pH 6), the reduction in adsorption capacity was attributed to competition with hydroxide ions and electrostatic repulsion due to negative charge on the surface of the adsorbent.
Similarly, the perchlorate removal efficiency increased with the calcined product of Mg/(A1 + Fe) hydrotalcite-like compound on increasing pH from 2 to 4 (Yang, Gao, Chu, et al. 2012). The low percentage removal at lower pH was due to the destruction of the adsorbent structure. In the pH range of 4-10, the adsorption efficiency pla-teaued. On increase of pH after 10, the adsorption efficiency declined significantly.
There was little impact of pH on the Fe pillared bentonite, though in the case of Al and Fe-Al pillared bentonite, there was a significant decrease in abatement of the perchlorate uptake above pH 8 (Yang, Xiao, et al. 2013). The increase of the electrostatic repulsion was held as one of the significant reasons for the decline in the percentage removal.
The maximum perchlorate adsorption with biochar (negatively charged at most pH values) occurred at pH near the isoelectric point (Fang et al. 2013). At pH higher
TABLE 6.3
Summary of Parameters and Optimized Conditions for Batch Adsorption of Perchlorate
|
Adsorbent |
Adsorbate |
Surface Area (m2/g), Pore Volume (cm’/g), Pore Size (nm) pHzpc |
Experimental Conditions |
Adsorption Capacity (mg/g) |
Thermodynamic Parameters |
Kinetic Model and Curve Fitting |
Isotherm Model and Curve Fitting |
Maximum Adsorption Conditions |
References |
|
Calcined LDH |
cio4- |
Surface area = 124.8 Pore volume = 0.954 |
pH = 6 Agitation speed = 120rpm Concentration = 1-450 mg/l (isotherm study) Contact time =24h Temperature =25°C |
280 |
Langmuir Nonlinear (Freundlich under high and low concentration range Linear) |
pH =no effect in pH range of 4-9.5 |
Lin et al. (2014) |
||
|
I lexadecy 1 py ridi ni um-modiiied montmorillonite |
cio4- |
Concentration =0.l-0.6mM (isotherm study) Contact time =24h Temperature =298-318 K |
0.081 mmol/g |
AH” = -49.65kJ/mol AS0 = -115. J/mol/K AG° = negative |
Pseudosecond order Linear |
Freundlich and D-R isotherm |
Temperature = 298 K |
Luo ct a). (2016) |
|
|
Calcined product of Mg(Al-Fe) hydrotalcite |
cio4- |
pH = 2-12 Dose =up to 0.13-0.33 g/1 Agitation speed =200rpm Concentration = 200-10,000 pg/1 Contact lime = 30-2880 min Temperature =25°C |
3605-5001 pg/g |
Pseudosecond order Linear |
Freundlich model Linear |
pH =4-10 Dose = 1.3 g/1 Contact time = 1440 min Temperature = 25 °C |
Yang, Gao, Chu, et al. (2012) |
230 Batch Adsorption Process of Metals and Anions
TABLE 6.3 (Continued)
Summary of Parameters and Optimized Conditions for Batch Adsorption of Perchlorate
|
Surface Area |
Kinetic |
||||||||
|
(m2/g), Pore |
Adsorption |
Model and |
Isotherm |
Maximum |
|||||
|
Volume (cm3/g), |
Experimental |
Capacity |
Thermodynamic |
Curve |
Model and |
Adsorption |
|||
|
Adsorbent |
Adsorbate |
Pore Size (nm) pHzpc |
Conditions |
(mg/g) |
Parameters |
Fitting |
Curve Fitting |
Conditions |
References |
|
Oxidized double |
cio.- |
pH = 1-10 |
2.62 and 3.55 |
pH = neutral pH |
Fang and |
||||
|
walled CNTs |
Agitation speed |
(for contact |
and pH not |
Chen (2012) |
|||||
|
= 30rpm |
period of 2h |
adjusted in |
|||||||
|
Contact time |
and 8h |
experiments |
|||||||
|
= less than 1 h |
oxidation) |
||||||||
|
Temperature = 25 °C |
|||||||||
|
Fe-bent |
cio4- |
Surface area zzzzz |
pH =4-10 |
0.102. 0.107 |
Fe Bent |
Pseudo |
Langmuir |
pH=4-7 |
Yang. Xiao. |
|
Fe-Al-bent |
= (Fe-bent = 190 |
Dose = 1 g/1 |
and 0.117 |
(AH’=15.684 |
second |
model linear |
4-10 in case of Fe |
et al. (2013) |
|
|
Al-bent |
Fe-Al-bent = 168 |
Agitation speed |
mmol/g for |
kJ/mol. AS“ |
order |
- bent |
|||
|
Al-bent = 152) |
= 200rpm |
Al-bent. |
= 61.478J/K mol) |
model |
Contact time |
||||
|
Pore |
Concentration |
Fe-Al-bcnt. |
For Fe-Al-bent |
linear |
= 100 min |
||||
|
volume = (Fe- |
= 0.05-0.7 mmol/1 |
Fe-bent |
(AH° |
||||||
|
bent =0.223 |
Contact time |
= 14.048 kJ/mol. |
|||||||
|
Fe-Al-bent = 0.191 |
= 1-600 min |
AS" |
|||||||
|
Al-bent =0.151) |
Temperature |
= 54.248J/mol) |
|||||||
|
Pore size |
= 25 °C |
Al-bent (AH“ |
|||||||
|
(Diameter) |
15.470 kJ/mol. |
||||||||
|
= (Fe-bent = 5.62 |
AS“ 58.060J/mol) |
||||||||
|
Fe-Al-bent = 4.52) |
AG“ negative for |
||||||||
|
Al-Bent = 4.01 |
all |
||||||||
|
Iron hydroxide- |
cio.- |
Surface area |
pH =4-10 |
0.113 mmol/g |
Pseudo |
pH = 6 |
Xu et al. |
||
|
doped granular |
= 488-632 |
Dose = 1 g/l |
second |
Time |
(2015) |
||||
|
activated carbon |
Pore volume |
Agitation speed |
order |
= 10 h but most |
|||||
|
= 0.251-315 |
= 200rpm |
model |
of experiments |
||||||
|
Pore size |
Concentration |
only |
conducted till |
||||||
|
= 2.13-2.01 |
= 0.1 mmol/1 |
Linear |
24h |
||||||
|
Contact time = 36h |
Temperature = 25 °C
Remediation of Anions
TABLE 6.3 (Continued)
Summary of Parameters and Optimized Conditions for Batch Adsorption of Perchlorate
|
Adsorbent |
Adsorbate |
Surface Area (m2/g), Pore Volume (cm’/g), Pore Size (nm) |
pHzpc |
Experimental Conditions |
Adsorption Capacity (mg/g) |
Thermodynamic Parameters |
Kinetic Model and Curve Fitting |
Isotherm Model and Curve Fitting |
Maximum Adsorption Conditions |
References |
|
Calcined product of Mg-FeCO, LDH |
cio4- |
pH = 2-12 Dose = 0.13-3.33 g/1 Agitation speed =200rpm Concentration = 2000 pg/l Contact time = 24 h Temperature =25°C Mg to Fe ratio = 3 |
2568.4 pg/g |
Pseudosecond order model Linear |
Freundlich model Linear |
pH =4-10 Ek> sc = 1.33 g/1 |
Yang. Gao. Deng, et al. (2012) |
|||
|
Nanoiron hydroxide-doped granular activated carbon |
cio4- |
pH,er = 8.4-8.8 |
pH =2-10 Dosc= 1 g/1 Agitation speed =200rpm Concentration =0.0816-0.4897 mmol/1 (isotherm study) |
0.088-0.169 mmol/g |
Pseudosecond order model Linear |
Langmuir model |
pH =2-3 |
Xu et al. (2013) |
Contact time=36 h Temperature=25 °С
232 Batch Adsorption Process of Metals and Anionsthan the isoelectric point, the adsorbent showed a low level of perchlorate adsorption. On the basis of these phenomena, it was suggested that electrostatic attraction was not the major cause of the adsorption.
The change of pH (pHfinal-pHinitial) during adsorption of perchlorate with raw carbon nanotubes was positive (Fang and Chen 2012). However, adsorption with oxidized carbon nanotubes led to change of pH, almost nil at adsorption below pH 5 and negative above pH 5. The increase in pH is suggested to be an indication of the ion exchange phenomenon (Xu et al. 2013).
The pH of solution increased with addition of biochar (Fang et al. 2013). The increase of the pyrolysis temperature for biochar led to increase in the pH of the solution. The increase in the pH of the solution by addition of the adsorbent is attributed to ash content.
The pH dependent zeta potential curve of biochar prepared from fir wood chips depicts an S-shaped curve with two inflection points (Fang et al. 2013). The presence of two inflection points suggests that two different forms of the functional group exist on the surface, influencing the dissociation of the functional groups at two different pH values. The coupling of the zeta potential points with the FT-IR led to the identification of two points at hydroxyl and carboxyl groups.
Effect of Coexisting Ions
The presence of anions in the solution declined the adsorption efficiency of perchlorate (Yang, Gao, Chu, et al. 2012; Yang, Xiao, et al. 2013; Xu et al. 2013). Phosphate, sulfate, and nitrate influenced the removal of the perchlorate on A1-, Fe- and Fe-Al pillared bentonite (Yang, Xiao, et al. 2013). The reduction was due to competitive adsorption, and the decline in the electrostatic effect was due to decline in the charge on the surface of the adsorbent.
The effect of anions is in the order of NO1“
Effect of the Material
The adsorption of perchlorate with uncalcined magnesium aluminum carbonate layered double hydroxide was independent of the Mg/Al ratio (Lin et al. 2014). However, adsorption with calcined magnesium aluminum carbonate layered double hydroxide was dependent on the Mg/Al ratio. Adsorption rises with increase in Mg/Al ratio. This led to the conclusion that adsorption increased with decline in surface charge density. This is in contrary to the general rule. The reason is attributed to the nondependency of adsorption on electrostatic interaction.
The Fe content in Mg/(A1-Fe) hydrotalcite enhanced the adsorption of perchlorate (Yang, Gao, Chu, et al. 2012). The Fe3+ content led to the formation of a brucite-like sheet structure. The Al3+ substitution with Fe3+ was proposed to enhance the positive surface charge. The enhanced positive charge increased the bond strength between layered double hydroxide and anion in the interlayer. However, excessive Fe3+ destroyed the brucite-resembling sheet structure. Hence, an optimum ratio of Mg2+/Fe3+ is required for maximum charge density.
The removal efficiency of perchlorate also improved with calcination temperature up to 550°C with the calcined product Mg/(A1-Fe) hydrotalcite. The calcination temperature after 600°C declined the adsorption efficiency. The lack of recovery into uncalcined adsorbent configuration (memory effect) for the material prepared at high temperature (> 600°C) contributed to decline in adsorption efficiency. The calcination temperature should be adequate for removal of the interlayer constituent, i.e. carbonate. However, at very elevated calcination temperature, creation of MgO occurred, and adsorbent reconstruction to uncalcined structure state stalled.
The adsorption capacity of double-walled carbon nanotubes enhanced on its oxidation (Fang and Chen 2012). The phenomenon occurred without any significant change of surface area. The enhanced removal was proposed due to the introduction of the oxygen-containing functional groups onto the carbon nanotube surface on oxidation. The change in the surface functional groups was supported by XRD, FT-IR, XPS, and Raman spectroscopy. The FT-IR peaks at 1660, 1724, and 3413cm-1 corresponding to unsaturated ketone, carboxylic acid and saturated ketone, and hydroxyl functional groups intensified after oxidation. This shows that oxidation leads to the introduction of additional functional groups having oxygen atoms. The D band (1338cm-1) and G band (1580cm-1) in Raman spectra relate to amorphous carbon and the C-C bond in graphene, respectively. The Raman spectra depict the increase of the D-band to the G-band on oxidation, suggesting the increase in surface defect. The XPS analysis also suggests the increase in carbon with oxygen atoms along with decrease in graphitic carbon. The oxygen-containing sites help in additional H-bonding and electrostatic interaction.
The lower adsorption of perchlorate on sodium bentonite as compared to Al or Fe or Fe-Al pillared bentonite was attributed to the increased number of vacant adsorption sites (Yang, Xiao, et al. 2013).
Mechanism
At different pH values, different mechanisms were working on the adsorption of perchlorate with activated carbon (Xu et al. 2016). At pH 2, the pH which was less than pHl№ the charge on the surface of the adsorbent was positive and the nitrogen-comprising groups were in the form of -N+ and -NH2+. This helped in the effective adsorption of perchlorate. In the pH range of 4-8 (larger than the isoelectric point), the surface of the adsorbent was negatively charged, and the nitrogen-comprising groups lost charge and change their form from -N+ and -NH,+ to -N and -NH, respectively. In addition, -COH and -COOH were the major oxygen-comprising groups, and adsorption is derived by hydrogen bonding. At pH 8-10, a fraction of -COH and -COOH groups were deprotonated to -CO and -COO groups, which led to decrease in adsorption.
The memory effect was proposed for adsorption of perchlorate with calcined magnesium aluminum carbonate layered double hydroxide (Lin et al. 2014). The phenomenon was supported by XRD, SEM, and FT-IR analysis. The peak corresponding to the hydroxide layer structure wiped out after calcination and remerged after perchlorate adsorption. Moreover, the distance between Mg and Al in the parent layered double hydroxide were the same as compared to reconstructed layered double hydroxide. The reconstruction of the hexagonal structure post adsorption was also confirmed by SEM analysis. The FT-IR peak of carbonate at 1062 cm-1 in the parent LDH shifted to 1097cm-1 after calcination and reverted back to 1062cm-1 after adsorption.
However, the phenomenon of the memory effect on Mg/Al layered double hydroxide was not ideal. Excessive broadening of the peaks happened at (018) and (015) reflection, leading to the occurrence of stacking faults (Lin et al. 2014). The reconstructed layered double hydroxide showed a strong peak at 936cm-1 in FT-IR spectrum, which was close to perchlorate in solution, i.e. 935 cm-1 than to perchlorate of solid state at 954cm-1. This portrays that perchlorate in the interlayers is in the free state as compared to that in the perchlorate solution.
The memory effect is also observed with calcined Mg/(A1-Fe) hydrotalcite after adsorption with perchlorate (Yang, Gao, Chu, et al. 2012). The mechanism of adsorption on carbon nanotubes of perchlorate also varies with the pH of the solution (Fang and Chen 2012). At very low pH, the surface is protonated, but the enhanced electrostatic attraction prefers smaller ion, i.e. chloride (volume=0.047 cm3) than perchlorate (volume = 0.082 nm3). This leads to the trace amount of perchlorate adsorption. On increase of pH but lower than that of the isoelectric point, the competition from chloride ions declined along with deprotonation of the surface groups. The deprotonation of the surface groups led to increase in the H-bonding. At near-neutral pH, the carbon nanotube surface was slightly negatively charged (pH was +0.85 more than pHIEP) and electrostatic force of attraction was eliminated. The adsorption of C1O4-happened by H-bonding rather than by electrostatic attraction. On further increase of pH, perchlorate adsorption declined due to increased electrostatic repulsion.
The mechanism of adsorption of perchlorate on nanoiron hydroxide-doped granular activated carbon follows electrostatic attraction, ion exchange, and surface complexation (Xu et al. 2013). The analogous trend of the zeta potential and pH in the perchlorate adsorption suggests the adsorption follow-up by electrostatic phenomenon. The increase of pH of the solution after adsorption suggests that the ion exchange phenomenon also happens during adsorption. In addition, sulfate ions are also present in the solution, which surged with decline of perchlorate anion in the solution. The relative proportion of the mechanism is estimated by desorption of the adsorbent with an alkaline pH of 11. The perchlorate desorption efficiency of 76% suggests that electrostatic interaction and ion exchange (outer sphere complexation) were dominant mechanisms for perchlorate adsorption, and inner sphere complexation accounts for nearly 24% of the perchlorate removal.
The maximum perchlorate removal with carbon nanotubes happened near the isoelectric point (+0.85) rather than at pH < pHIEP (Fang and Chen 2012). This observation led to the conclusion that electrostatic phenomenon cannot be the principal governing factor at neutral pH, and hydrogen bonding is considered to be the principal governing factor near neutral pH.
Desorption
The regeneration of silver 4,4'-bipyridine perchlorate (SBP) is achieved by increasing the ratio of nitrate to perchlorate and the temperature (Colinas et al. 2016). This led to the shift of the equilibrium and conversion of silver 4,4'-bipyridine perchlorate to individual components. A total of 96% of the material is regenerated and returned to its shape. The regeneration of nanoiron oxide embedded activated carbon can be achieved under alkaline pH (Xu et al. 2015).
Sulfate
Sulfate occurs naturally and is also produced commercially (WHO 2017). Sulfate is nontoxic, and WHO has also not set up a guideline value for sulfate, due to its level found in water having no health concern. However, high levels of sulfate (600 mg/1) can impact the gastrointestinal system (Silva et al. 2012).
Effect of pH
The removal of sulfate decreases (Gu et al. 2016; Fukushi et al. 2013; Moret and Rubio 2003) with rise in pH. In some cases, the decline in the percentage removal with pH was not linear. The percentage removal increased up to a maximum with rise of pH and declined thereafter, e.g. maximum at pH 4 (Dong et al. 2011) and rise up to pH 3 and remained constant up to pH 9 and declined thereafter (Namasivayam and Sangeetha 2008).
The higher removal at lower pH is attributed to the protonation of the surface, and declination of the removal at higher pH is attributed to the competition from hydroxyl groups (Dong et al. 2011; Moret and Rubio 2003). However, in the case of adsorption with activated carbon, the low removal at low pH was attributed to the competition from chloride ions (Namasivayam and Sangeetha 2008).
The pH affects the formation of inner and outer sphere complex (Gu et al. 2016). The pre-edge intensity in Extended X-ray absorption fine structure (EXAFS) is considered proportional to inner sphere complex formation, and it becomes weaker with increase of pH from 3 to 7 during the adsorption of sulfate on ferrihydrite. Hence, the fraction of the inner sphere complex of adsorbed sulfate on ferrihydrite decreases with rise in pH, while maintaining the ionic strength constant. The increase in pH led to increase in the dominance of > FeOH and > FeO~, and these groups disfavored the inner sphere adsorption. The reason is attributed to stronger and harder Fe-O bonds as compared to FeOH2+ and decreased electrostatic interaction via decrease in the positive surface charge.
Effect of Coexisting Ions and Surface Modification
The presence of ions in the solution resulted in the decline of adsorption capacity (Howarth et al. 2016; Namasivayam and Sangeetha 2008). The order of influence of anions on the adsorption of sulfate with activated carbon is as follows: molybdate > chlorate > nitrate > chloride > phosphate (Namasivayam and Sangeetha 2008).
The modification of palygorskite (magnesium aluminum phyllosilicate) with octodecyltrimethylammonium chloride (OTMAC) increased the adsorption efficiency of sulfate ten times (Dong et al. 2011). The increased adsorption efficiency is attributed to improved electrostatic interaction. The unmodified palygorskite has a negative charge on the surface of the adsorbent, and there is repulsion between the sulfate and the adsorbent. The modification of the adsorbent with the octodecyltrimethylammonium chloride led to the formation of positive charge on the surface of the adsorbent. The surface modification led to the first layer of the OTMAC grafted on palygoskite. The second layer of the OTMAC interacts with the first layer OTMAC through the hydrophobic force, leading to the formation of the double layer.
The effect of pH on the adsorption of sulfate on activated carbon is the same for both simulated and natural groundwater (Namasivayam and Sangeetha 2008). However, the presence of calcium in natural groundwater declined the adsorbent’s dose for treatment of sulfate-contaminated groundwater. Calcium precipitates sulfate and reduces the requirement of the adsorbent’s dose.
Mechanism
There are a number of techniques used to determine the mechanism of sulfate adsorption such as EXAFS, differential atomic pair distribution function analysis (d-PDF), (Zhu et al. 2013) and Attenuated total reflection-Fourier transform infrared spectroscopy (ATR-FT-IR) (Johnston and Chrysochoou 2016). The EXAFS spectrum is used for the mechanism of adsorption of sulfate (pH=4) on ferrihydrite (Zhu et al. 2013). The EXAFS spectra of the sample after adsorption on ferrihydrite showed a weak peak at 8.5 Â"1 similar to that of jarosite; this feature was not shown by liquid sulfate solution. This suggests that fewer than three Fe atoms on average surround the sulfur atom of the surface-bound sulfate ion.
The multicurve resolution analysis of ATR-FT-IR indicates the occurrence of the outer and inner sphere sulfate complex on Al-ferrihydrite (Johnston and Chrysochoou 2016). The sulfate species adsorbed at pH higher than 6 depicts two peaks in “multivariate curve resolution ATR-FT-IR” spectra at 1100 and 980cm"1. The two peaks are attributed to the outer sphere complex. However, on adsorption at lower pH, peaks at 1170, 1120, 1050, and 980cm"1 emerged, which is attributed to the inner sphere adsorption process. The increase in the content of aluminum in ferrihydrite increased the content of the outer-sphere oxyanion. This is attributed to the suppression of the availability of inner-sphere binding sites (Johnston and Chrysochoou 2016).
The EXAFS spectra also decipher the distance between the sulfur and Fe atoms (adsorption with ferrihydrite), which is in the range of 3.18—3.19 Â (Zhu et al. 2013). Differential atomic pair distribution function analysis (d-PDF) along with EXAFS analysis suggests the bond length of sulfur and oxygen as the same. This led to the conclusion that the systematic errors between the two methods were insignificant for S-O bond determination. However, the EXAFS analysis underestimates the Fe-S bond length (3.19Â), and d-PDF shows that it is in the range of 3.25 ±0.02 Â.
The surface complexation model was also used to predict the adsorption of sulfate on ferrihydrite. At lower pH, the surface complexation model suggests the existence of inner and outer sphere complex (Gu et al. 2016). The results stimulated by the surface complexation model were supported by results from adsorption experiments and spectroscopic analysis, e.g. the model suggested the decline of the outer sphere complex, increase of sulfate adsorption at lower pH, and absence of the outer sphere complexation maxima at 0.5 M ionic strength.
Mean free energy of adsorption was also used as the parameter for physisorption or chemisorption estimation. The mean free energy for sulfate adsorption with OTMAC-modified palygoskite was calculated as 0.48 kJ/mol. The calculated value was below 8 kJ/mol indicating that the process occurred via physical interaction phenomenon (Dong et al. 2011). The author suggests that the process is controlled by electrostatic interaction rather than a chemical bond.
Desorption
Hydrochloric acid and alkaline pH were used for desorption and reuse of the adsorbent after adsorption with sulfate (Howarth et al. 2016; Namasivayam and Sangeetha 2008). Alkaline pH causes desorption by ion exchange phenomenon (Namasivayam and Sangeetha 2008), and the species adsorbed by chemical adsorption cannot be recovered by desorption via increasing pH value.
