Arsenic and Fluoride in Groundwater, Prevalence and Alternative Removal Approach

: Contamination of drinking water by arsenic and ﬂuoride is a global problem, as more than 300 million people in more than 100 countries have been affected by their presence. These elements are considered the most serious contaminants in drinking water and their removal is a worldwide concern. Therefore, the evaluation of three alternative approaches—electrocoagulation, adsorption by biomaterials, and adsorption by metal oxide magnetic nanoparticles (MNPs)—was performed for arsenic and ﬂuoride removal from groundwater. Arsenic removal from synthetic and groundwater (well water) was accomplished with the three processes; meanwhile, ﬂuoride removal from groundwater was only reported by two methods. The results indicate that an electrocoagulation process is a good option for As (>97%) and F (>90%) removal in co-occurrence; however, the operational conditions for the removal of both pollutants must be driven by those used for ﬂuoride removal. As (80–83%) and F (>90%) removal with the biomaterials was also successful, even when the application objective was ﬂuoride removal. Finally, MNPs (Co and Mn) were designed and applied only for arsenic removal and reached >95%. Factors such as the pH, the presence of interfering ions, and the initial concentration of the contaminants are decisive in the treatment process’s efﬁciency.


Introduction
Due to the subsoil geological composition (natural causes), arsenic (As) and fluoride (F − ) are widely distributed throughout the Earth's crust, as well as at different concentrations in all-natural waters [1][2][3][4]. Recently, numerous studies have reported that the co-occurrence of these two pollutants, mainly in semi-arid and arid areas of the planet, is more frequent than previously assumed, representing a severe health risk [5][6][7][8]. The co-occurrence of both pollutants can be due to geogenic and anthropogenic causes [9]. Globally, it is estimated that more than 300 million people ingest water contaminated with As and F − [10,11], in levels above the maximum permissible limit (MPL) set by the World Health Organization (WHO), which is 0.010 mg/L for As and 1.5 mg/L for F − . The most affected countries by As and F − in drinking water are in Africa, Asia, and Central and Latin America [8,12].
High As groundwater concentration above the MPL is the primary source of arsenic exposure and can cause several health problems [13]; its effects are associated with skin lesions, some types of cancer, peripheral neuropathy, diabetes, and cardiovascular diseases [2,14,15]. In the case of F − , it is an essential element for humans, and low F − concentrations in water could lead to dental decay, while high F − could produce dental trations in water could lead to dental decay, while high F − could produce dental and skeletal fluorosis, and it has been associated with neuronal damage and fertility problems too [1,16]. Similarly to arsenic, groundwater is also the primary source of fluoride intake [9].
In Mexico, groundwater is the main source of water supply [17]. In the country's arid region (central and northern), these waters exceed the MPL for As and F − , with the states of Durango, San Luis Potosí, Chihuahua, Zacatecas, and Jalisco standing out [8]. In Durango, Mexico, Alarcón-Herrera et al. [18] reported As concentrations in the city's water sources ranging from 0.028 to 0.092 mg/L, a situation that has not been rectified, according to the results obtained in the last sampling carried out in 2019 from the city's water wells (data not shown). Likewise, F − reported concentrations range from 0.3 to 24.1 mg/L [19].
Different technologies can remove these contaminants and prevent damage to human health; however, the treatment method selection will depend on the volume to be treated, the type of water, and the available resources. Figure 1 shows the most commonly used treatment methods to remove As and F − from drinking water [12,[20][21][22][23][24]. Precipitation-coagulation methods can remove high concentrations of pollutants [24,25] and are competitive in terms of treatment costs; however, they generate large amounts of sludge and dissolution of elements in treated effluents [21,23]. Membrane separation processes operate at low temperatures and are not pH-dependent but require high pressures, and their costs are often high [10,20,25]. Ion exchange methods can be selective, work with trace concentrations, and, in some cases, resins can be regenerated; however, they are highly pH-dependent and their costs are high [12,25]. The adsorption advantages include low cost (profitability) and the adsorbent materials' versatility, as well as the possibility of reusing them [20,21]. The disadvantages described are often pH dependence and interference from ionic species [20,22].
At this time, the electrocoagulation process for As and F − separate removal has been widely studied on a batch scale [26][27][28][29]. However, due to the co-occurrence of both contaminants in water, recent years' efforts have been to develop such simultaneous removal processes and in continuous flow reactors [30][31][32]. In addition, novel prototypes have also been developed in terms of their configuration, which seek to provide alternatives to separate treated water from the sludge generated more efficiently, thus making the process more applicable.
In the same way, adsorption processes are the most widely applied treatment for pollutant removal, such as As and F − , and technologies that allow the reuse and recycling of all types of materials are preferred. Hence, the study of biomasses (usually agro-industrial waste) modified and/or doped with metal oxides and hydroxides has become an attractive option [33][34][35]. Waste biomasses are abundant in functional groups and rich in lignin, hemicellulose, cellulose, and pectin. These biopolymers exhibit the physicochemical characteristics, selectivity, and reactivity required for anion capture [33,36]. Furthermore, when subjected to impregnation with cations [37][38][39][40], these biomasses improve their adsorption capacities, thus becoming low-cost, environmentally friendly bioadsorbents or biocomposites with competitive removal efficiencies [33,41]. Precipitation-coagulation methods can remove high concentrations of pollutants [24,25] and are competitive in terms of treatment costs; however, they generate large amounts of sludge and dissolution of elements in treated effluents [21,23]. Membrane separation processes operate at low temperatures and are not pH-dependent but require high pressures, and their costs are often high [10,20,25]. Ion exchange methods can be selective, work with trace concentrations, and, in some cases, resins can be regenerated; however, they are highly pH-dependent and their costs are high [12,25]. The adsorption advantages include low cost (profitability) and the adsorbent materials' versatility, as well as the possibility of reusing them [20,21]. The disadvantages described are often pH dependence and interference from ionic species [20,22].
At this time, the electrocoagulation process for As and F − separate removal has been widely studied on a batch scale [26][27][28][29]. However, due to the co-occurrence of both contaminants in water, recent years' efforts have been to develop such simultaneous removal processes and in continuous flow reactors [30][31][32]. In addition, novel prototypes have also been developed in terms of their configuration, which seek to provide alternatives to separate treated water from the sludge generated more efficiently, thus making the process more applicable.
In the same way, adsorption processes are the most widely applied treatment for pollutant removal, such as As and F − , and technologies that allow the reuse and recycling of all types of materials are preferred. Hence, the study of biomasses (usually agro-industrial waste) modified and/or doped with metal oxides and hydroxides has become an attractive option [33][34][35]. Waste biomasses are abundant in functional groups and rich in lignin, hemicellulose, cellulose, and pectin. These biopolymers exhibit the physicochemical characteristics, selectivity, and reactivity required for anion capture [33,36]. Furthermore, when subjected to impregnation with cations [37][38][39][40], these biomasses improve their adsorption capacities, thus becoming low-cost, environmentally friendly bioadsorbents or biocomposites with competitive removal efficiencies [33,41]. Likewise, the development of materials at the nanometer level allows an increase in the active sites for adsorption and the contact surface area, improving the removal capacity of these elements. The adsorbents' characteristics determine their adsorption potential, high surface area, and chemical stability (which participate in adsorption and improve the adsorbent's adsorption capacity). However, most of the reported works have not experimented with groundwater, and their effectiveness cannot be proven since it is known that groundwater contains ions that can compete for the active sites.
Thus, this study aimed to evaluate the efficiency of three different methods with novel approaches for arsenic and/or fluoride removal from groundwater. The effects of different parameters, such as pH, initial As and F − concentration, and contact time, among others, were evaluated on a laboratory scale, with synthetic water and natural groundwater, and the main aspects considered were removal efficiency, operational parameters, management, and ease of process implementation.

Water Sampling and Characterization
Evaluations of each treatment system were carried out with groundwater collected from a well, "San Luis" (24.05587778, −104.59307222), which supplies water to the Durango city, México residents, and a local facility. Samples were collected in high-density polypropylene containers, which were preserved at a temperature of 4 • C. The samples were characterized in terms of their physicochemical parameters following the Standard Methods for examining water and wastewater, APHA, AWWA, and WEF, in the Environment Laboratory of the Centro de Investigación en Materiales Avanzados S.C.-Durango.
Historical data provided by water authorities at Durango city (four years of data, 2012-2016) were analyzed to evaluate the persistence of As and F − in groundwater. Moreover, well water samples were collected and analyzed during recent years (2016-2019).

As and F Quantification
For As and F − quantification in water, the APHA standard method was followed, 3114 C. Continuous Hydride Generation/Atomic Absorption Spectrometric Method and 4500-F − C. Ion-Selective Electrode Method, respectively. As was determined by the atomic absorption method (AAS); a GBC XplorAA Dual model with hydride generator was used by pre-reduction of the sample. F − was determined by the ion-selective method (ISE) with an ionic adjuster in Thermo Scientific Orion Versa Star equipment.

Electrocoagulation Method
Electrocoagulation tests for As and F − removal were carried out in a 1.5 L capacity reactor, under the operating conditions described in a previous work in which both contaminants were simultaneously removed [29].
The experiments were carried out by applying a current density of 4.5 mA/cm 2 for 15 min of treatment time. As and F − concentrations in groundwater were 0.042 mg/L and 4.24 mg/L, respectively. Sample initial pH was adjusted to 3 and 5 using HCl (1 M). Additionally, the water conductivity was adjusted with NaCl to a value around 1200 µS/cm.
Three scenarios were evaluated: (a) two aluminum electrodes (pH i = 3, pH i = 5); (b) two iron electrodes (pH i = 5); (c) four electrodes, two aluminum and two iron (pH i = 3, pH i = 5). In all cases, the As and F − removal was monitored to determine the optimal operating conditions. The experiments were carried out in triplicate and the average values were reported.
At the end of each experimental run, duplicate samples were collected. Subsequently, filtration (0.3 µm) was carried out for As and F − determination. Finally, the removal percentage was calculated according to Equation (1): where C i is the initial concentration and C f is the final concentration.
The sludge was collected after filtration and dried at 105 • C for one hour [27,42] to determine the amount of sludge produced.
Operating cost = a × C electrode + b × C energy (2) a coefficient corresponds to the wholesale cost of aluminum, Al (2.003 USD/kg) [43], or iron, Fe (0.169 USD/kg) [44], and the b coefficient relates to the electricity cost (USD/kWh). For example, according to the Federal Electricity Commission [45], the kWh cost for the middle industrial tariff (High Demand in Medium-Voltage Time) for March (2021) at Durango city, Mexico, was MXN 1.36, corresponding to 0.066 USD/kWh, according to the current exchange rate [46]. Each anode used was weighed before and after the experimental run to determine the electrode consumption. Subsequently, the final weight was subtracted from the initial weight. The energy consumption in the process was calculated from Equation (3).

Bioadsorbent Preparation
The selected biomasses were orange (Citrus sinensis) and apple (Malus domestica) peels, obtained, respectively, from a juice shop and a local market in Durango, Dgo, Mexico. The fresh apple (Red Delicious-Canatlan) was washed with soap and water to remove dust or agrochemical residues. The peel was then separated from the pulp with a knife. The orange peels (Naranja Valencia) were donated, cleaned, and pulp-free.
The peels of each fruit were subjected to solar drying in a wind tunnel dryer for 5 h at 40 • C in batches. The dried peels were then passed through a blade mill and sieved (250-500 µm). Next, the dried powders were washed with deionized water to remove water-soluble compounds. They were then dried at room temperature for 24 h. After this, the dry powders were subjected to saponification with calcium hydroxide Ca(OH) 2 , carboxylated with chloroacetic acid ClCH 2 COOH, and impregnated with oxychloride zirconium Cl 2 OZr·8H 2 O. Finally, the biomasses were washed with deionized water between each step and dried at room temperature for 24 h.
To verify that the bioadsorbent preparation process was successful, they were characterized by FTIR-ATR, BET, and SEM-EDX. Additionally, the materials in contact with the fluoridated water were characterized to check the As and F − ions captured by the bioadsorbents.

Adsorption Experiments
Batch adsorption studies were carried out in polypropylene bottles (250 mL) in triplicate with a working solution volume of 100 mL. Sodium fluoride NaF solutions with concentrations of 2, 4, 6, 8, and 10 mg F/L were prepared. The working pH was 3.8-4.2 (optional) at room temperature. The agitation rate was 320 rpm, the adsorbent dose was 1 g/L, and the contact times were 60, 180, 300, 480, 1140, and 2880 min. Langmuir and Freundlich's adsorption isotherm models were compared with the completed results. Additionally, one-factor Anova and Tukey comparisons were performed.

Adsorption for As/F in Solution
To determine whether the materials are able to remove As and F − simultaneously, a solution of 0.047 and 4 mg/L of As and F − , respectively, was prepared. The experimental conditions were kept the same as in the adsorption capacity experiment.

Groundwater As and F − Adsorption
Finally, experiments were carried out to determine the bioadsorbents' capacity to remove As and F − ions in a natural matrix (well water). The experimental conditions were the same as described above. However, the working pH for the experiment's first section was left unadjusted, i.e., the pH at which each bioadsorbent modifies the groundwater pH (8.7) and the pH at which both adsorbents modify is approximately 6.7. In the second part of the experiment, we worked with the same groundwater, but the pH was adjusted to 3.5 with hydrochloric acid HCl.

Magnetic Nanoparticles Synthesis
All the precursors used in the experiments were of analytical reactive grade and were used without further purification. The MNPs were prepared by the chemical coprecipitation method [47]. Finally, the products were taken to the oven at 100 • C for 24 h to grind them subsequently.
The CoFe 2 O 4 , MnFe 2 O 4 , and Fe 3 O 4 specific surface areas were calculated using the Brunauer, Emmett, and Teller equation (BET) on nitrogen N 2 adsorption/desorption data collected on a Quantachrome Nova Corporation 1000 series equipment surface analyzer. The X-ray diffraction (XRD) pattern was recorded with an X-ray Diffractometer A PANalytical model Empyrean of Malvern with a K-Alpha Cu anode of 1.54 nm at an amperage of 40 mA and a 45 kV voltage, with a scanning step of 0.02 in 2θ degrees, which was used to determine the crystal structure of the MNPs. The surface morphologies of the MNPs were observed under a scanning electron microscope (SEM), using a FEI Nova NanoSem200 with a low vacuum detector.

MNP Adsorption Experiments
Batch experiments were performed to evaluate the NPs' adsorption capacities for arsenic removal. Kinetic experiments were carried out at different adsorbent doses (0.01 and 0.1 g/L) and contact times (1, 15, 30, 60, and 90 min). Groundwater with 0.045 mg/L of As initial concentration was used with constant agitation speed (480 rpm) at room temperature. At the end of each experiment, the aqueous solutions with MNPs were separated using a high gradient magnetic separation (HGMS) column.

Results
Historical data of As and F − concentration in groundwater provided by the municipal authorities were analyzed in the context of a bi-national project. Moreover, more than 25 Durango city wells and 25 sampling points at Guadiana Valley were sampled on several occasions during three years (2016-2019). Groundwater As concentration ranged from 0.0002 to 0.188 mg/L and fluoride concentration ranged from 0.16 to 10 mg/L. Just three sampled wells reported As and F − levels adequate for human consumption during all the sampling sessions. Additionally, 80% and 88% of the total sampled points in Durango city did not satisfy the maximum permissible limits for human consumption water (NOM-127-SSA1-1994, 2000) for both As and F − , respectively. Table 1 shows the physicochemical characterization of the well water used in the adsorption experiments (biomaterials and nanomaterials). The results reported could be registered as mean values of groundwater quality in this region.

Electrocoagulation Method
The evaluated parameters and the electrocoagulation test results are shown in Table 2. The experiments with the aluminum anode showed F − removal efficiencies of 41.51% and 55.50% at initial pH 3 and 5, respectively. However, it was impossible to achieve a final concentration below 1.5 mg/L that was in line with the maximum permissible limit established for drinking water.
The maximum F − removal was 93.16% when the initial pH was 3, and aluminum and iron anodes were used together in the process. The increase in F − removal can be attributed to the electrodes' arrangement; the aluminum anode increased its active area. Therefore, there was higher Al +3 release into the solution. As a result, it was possible to form a more significant amount of aluminum coagulant species. On the other hand, it was observed Processes 2021, 9, 1191 6 of 15 that when the anode material was iron, the removal of F − was not favored since it was practically null. It is essential to mention that the use of iron anodes has been focused on As removal [42,[48][49][50], while aluminum anode has been used to remove F − [51,52], As [49][50][51][52][53], and both pollutants simultaneously [27]. In the investigations where As removal was carried out with aluminum anodes, the authors mentioned above point out that removal occurs when As is adsorbed on amorphous Al(OH) 3 complexes, as is the case with F − .
−2 (mg/L) 59.76 As * (mg/L) 0.032 * AsT = As +3 + As +5 . As was practically wholly removed in all the scenarios, except when the aluminum electrode was used at an initial pH of 3, with a notable difference with respect to the other processes evaluated, since, in this particular one, there was no As removal at all. In this scenario, the production of aluminum species was not adequate, as the amount of sludge was also minimal, so the low As removal could be attributed to the fact that the pH conditions benefitted F − removal over As removal. The final pH in this assay was 4.57; the F − ion was predominantly present in the solution from pH 3 onwards [54], while at pH below 7, the neutral arsenite species H 3 AsO 3 was predominantly found [27]. After pH 7, the arsenite species H 2 AsO 3 − and arsenate species HAsO 4 −2 began to occur. At the same pH value, the aluminum species Al(OH) 3 was present, which aided the As removal through the use of aluminum anodes when the pH was close to neutral, as occurred in the scenario in which the initial pH was 5, and an aluminum anode was used, finding complete arsenic removal, with the final pH of the solution being 7.47. On the other hand, when the iron anode was involved in the process, the Fe(OH) 3 (s) species, to which As removal was attributed, was present in the solution from pH 3 and remained constant up to pH 12, which allowed As removal over a broader pH range. For As and F − removal, the optimum operating conditions were an initial pH of 3 and the use of aluminum and iron as anodes together. The estimated cost was 0.279 USD/m 3 with a sludge production of 0.667 kg/m 3 . Under these conditions, sludge production was the highest compared to the other scenarios evaluated. Moreover, sludge generation has been considered a drawback due to the disposal cost and environmental impact. According to the price of sludge handling by confinement in Mexico (0.035 USD/kg) [55], the cost of the process would increase to 0.302 USD/m 3 .
Recent research has focused on adding value to these by-products to increase the feasibility of applying this water treatment technology. Thakur et al. [54] evaluated the possibility of reusing sludge, produced in a simultaneous As and F − removal process, as a building material by adding 10% by weight of the sludge to clay to create a composite, finding that such material can be considered inert and applicable in various construction areas. Santana et al. [56] point out in their study that alumina sludge from the electrocoagulation process for F − removal can be used to produce microcapsules that protect aluminum from corrosion.

Bioadsorbent Synthesis (General Characteristics)
Two bioadsorbent materials were obtained, essentially for F − removal. Both bioadsorbents showed the presence of functional groups such as OH and COOH in FTIR analysis. The BET study revealed mesoporous materials (sizes of 6.0-5.0 nm). Finally, SEM analysis confirmed rough materials, with sizes in the range of 80-339 µm. The EDX confirmed the presence of Zr in the prepared bioadsorbents and F − in the bioadsorbents that had contact with the F − in the aqueous medium.

Bioadsorbents' Efficiency
Both bioadsorbents for F − ions performed chemisorption in monolayers (Langmuir) determined by adsorption isotherms. The orange and apple peel bioadsorbents' q max values were 4.9 and 5.6 mg F − /g, respectively.
Removal efficiencies of 82-84% and 86-94% for As and F − in synthetic water solutions, respectively, were obtained, with 1440 min of contact time (Figure 2). On the other hand, with natural groundwater, the results showed that As removal, with both bioadsorbents, was independent of the reaction medium pH and achieved 81-83% removal efficiency. However, F − removal was pH-dependent, so adjusting the final pH to 3 improved the F − anions' removal with both bioadsorbents. As a result, the F − removal efficiencies increased competitively to a range of 69-85% at a contact time of 60 min and 93-95% at contact times of 1440 min.
However, under the experimental conditions described when both bioadsorbents were assayed with natural groundwater, both materials could reduce the As and F − concentrations below WHO's MPLs. Thus, the materials have a good affinity for As and F − anions; moreover, the presence of other ionic species, typical of a real matrix, did not significantly reduce the removal efficiencies achieved in laboratory solutions.
Overall, the F − anion removal rates achieved by both biomasses are competitive with those reached in similar investigations [37,[57][58][59][60][61]. Nevertheless, just a few studies were conducted with groundwater [58,62]. In addition, the capacities of some materials are reduced by the presence of co-existing ions [63]. Bioadsorbents from orange and apple peel, besides showing simultaneous As and F − anion removal capacity, did not significantly reduce their removal percentages and capacities when tested in groundwater. Because of this, they are considered promising materials to be studied in a continuous removal system on a household scale. Finally, among the advantages of composite materials are the costs, as they are estimated to be <0.3 USD/kg F − ; compared to adsorbents such as activated carbon and alumina, whose costs are around 2 and 1.5 USD/kg F − , respectively, biocomposite adsorbents are not only technically but also economically viable [64].  However, under the experimental conditions described when both bioadsorbents were assayed with natural groundwater, both materials could reduce the As and F − concentrations below WHO's MPLs. Thus, the materials have a good affinity for As and F − anions; moreover, the presence of other ionic species, typical of a real matrix, did not significantly reduce the removal efficiencies achieved in laboratory solutions.
Overall, the F − anion removal rates achieved by both biomasses are competitive with those reached in similar investigations [37,[57][58][59][60][61]. Nevertheless, just a few studies were conducted with groundwater [58,62]. In addition, the capacities of some materials are reduced by the presence of co-existing ions [63]. Bioadsorbents from orange and apple peel, besides showing simultaneous As and F − anion removal capacity, did not significantly reduce their removal percentages and capacities when tested in groundwater. Because of this, they are considered promising materials to be studied in a continuous removal system on a household scale. Finally, among the advantages of composite materials are the costs, as they are estimated to be <0.3 USD/kg F − ; compared to adsorbents such as activated carbon and alumina, whose costs are around 2 and 1.5 USD/kg F − , respectively, biocomposite adsorbents are not only technically but also economically viable [64].

Magnetic Nanoparticles' General Characteristics
The MNPs synthesized had a mesoporous structure with an average pore diameter of 5 nm and average crystal size of 30 nm, in XRD diffraction peaks located at 18, 30, 36, 43, 57, Figure 3 shows the As removal efficiency (%) at two different adsorbent doses to determine the equilibrium between adsorbent and As removal efficiency because the adsorbent dose can limit the quantity of As that could be removed.

Nanoadsorbents' Efficiency
The results show the adsorption capacity as a function of the adsorbent dose, as we expected. For example, an adsorbent dose of 0.1 g/L As removal reaches 97% with CoFe2O4, 94% with MnFe2O4, and 82% with Fe3O4. Instead, with the dose of 0.01 g/L, the removal efficiency was 85%, 78%, and 50%, with CoFe2O4, MnFe2O4, and Fe3O4, respectively. Thus, the adsorption process is strongly linked to the surface area available for the adsorption.  Figure 3 shows the As removal efficiency (%) at two different adsorbent doses to determine the equilibrium between adsorbent and As removal efficiency because the adsorbent dose can limit the quantity of As that could be removed.

Nanoadsorbents' Efficiency
The results show the adsorption capacity as a function of the adsorbent dose, as we expected. For example, an adsorbent dose of 0.1 g/L As removal reaches 97% with CoFe 2 O 4 , 94% with MnFe 2 O 4 , and 82% with Fe 3 O 4 . Instead, with the dose of 0.01 g/L, the removal efficiency was 85%, 78%, and 50%, with CoFe 2 O 4 , MnFe 2 O 4 , and Fe 3 O 4 , respectively. Thus, the adsorption process is strongly linked to the surface area available for the adsorption.
Likewise, Podder and Majumder [65] reports 86% of As +3 removal with MnFe 2 O 4 with a contact time of 80 min and an adsorbent dose of 2 g/L. Additionally, Iconaru [66] reported a maximum removal efficiency of 70% after 24 h of contact time with Fe 3 O 4 (2 g/L) for As +3 . On the other hand, Zhang [67], with a lower adsorbent (CoFe 2 O 4 ) quantity (0.2 g/L), reports 90% (As) maximum removal efficiency in twelve hours. Moreover, after 30 min of contact time for this study, the optimum adsorbent dose was 0.1 g/L with the three MNPs to obtain a final As concentration below the limit of the current Mexican regulation (NOM-127-SSA1-1994).
Effect of the Contact Time on As Removal Figure 4 shows the impact of the contact time on the As removal percentage. The equilibrium was reached after 30 min, but the experiment continued until 380 min; however, no significant increase in As removal was observed. Therefore, subsequent removal studies were carried out up to 60 min. Moreover, in all cases, it was observed that the contact time and the MNPs' saturation were not related to the concentration of the adsorbate in the solution; this could be explained because, at the beginning of the experiment, all the Processes 2021, 9, 1191 9 of 15 adsorbent sites were readily available, and the adsorbate concentration was very high. Therefore, the adsorbate removal was performed quickly in the early stages; then, the removal rate gradually fell until equilibrium was reached in each case. Therefore, the reduction curves were simple, smooth, and continuous, which led to equilibrium and the possibility of multilayer coverage on the adsorbent surface [68].
Likewise, Podder and Majumder [65] reports 86% of As +3 removal with MnFe2O4 with a contact time of 80 min and an adsorbent dose of 2 g/L. Additionally, Iconaru [66] reported a maximum removal efficiency of 70% after 24 h of contact time with Fe3O4 (2 g/L) for As +3 . On the other hand, Zhang [67], with a lower adsorbent (CoFe2O4) quantity (0.2 g/L), reports 90% (As) maximum removal efficiency in twelve hours. Moreover, after 30 min of contact time for this study, the optimum adsorbent dose was 0.1 g/L with the three MNPs to obtain a final As concentration below the limit of the current Mexican regulation (NOM-127-SSA1-1994).  Effect of the Contact Time on As Removal Figure 4 shows the impact of the contact time on the As removal percentage. The equilibrium was reached after 30 min, but the experiment continued until 380 min; however, no significant increase in As removal was observed. Therefore, subsequent removal studies were carried out up to 60 min. Moreover, in all cases, it was observed that the contact time and the MNPs' saturation were not related to the concentration of the adsorbate in the solution; this could be explained because, at the beginning of the experiment, all the adsorbent sites were readily available, and the adsorbate concentration was very high. Therefore, the adsorbate removal was performed quickly in the early stages; then, the removal rate gradually fell until equilibrium was reached in each case. Therefore, the reduction curves were simple, smooth, and continuous, which led to equilibrium and the possibility of multilayer coverage on the adsorbent surface [68].
The maximum As removal with the three MNPs at all initial concentrations assayed was achieved after 30 min. As Figure 4 shows, the As removal (%) is correlated with the initial As concentration in working samples. The As maximum ion removal with CoFe2O4, MnFe2O4, and Fe3O4 (98%, 95%, and 93%, respectively) was achieved in working samples with the lowest As (0.045 mg/L) initial concentration. Podder and Majumder [65] reported that the maximum As ion adsorption was achieved after 80 min with SD/MnFe2O4 composites; in this study, less contact time was enough to achieve the highest removal efficiency (98%). For CoFe2O4 and MnFe2O4, the maximum adsorption capacity was calculated from the Langmuir isotherm and was 250 and 238 mg/g, respectively. The results exceed the value obtained with Fe3O4, 161.3 mg/g. The synthesized nanomaterials of CoFe2O4 and MnFe2O4, which possess a high ion adsorption capacity of As in groundwater ions, were The maximum As removal with the three MNPs at all initial concentrations assayed was achieved after 30 min. As Figure 4 shows, the As removal (%) is correlated with the initial As concentration in working samples. The As maximum ion removal with CoFe 2 O 4 , MnFe 2 O 4 , and Fe 3 O 4 (98%, 95%, and 93%, respectively) was achieved in working samples with the lowest As (0.045 mg/L) initial concentration. Podder and Majumder [65] reported that the maximum As ion adsorption was achieved after 80 min with SD/MnFe 2 O 4 composites; in this study, less contact time was enough to achieve the highest removal efficiency (98%).
For CoFe 2 O 4 and MnFe 2 O 4 , the maximum adsorption capacity was calculated from the Langmuir isotherm and was 250 and 238 mg/g, respectively. The results exceed the value obtained with Fe 3 O 4 , 161.3 mg/g. The synthesized nanomaterials of CoFe 2 O 4 and MnFe 2 O 4 , which possess a high ion adsorption capacity of As in groundwater ions, were convenient for the As adsorption, likewise with easy recovery by magnetic separation.
The adsorption process was best described by the second-order kinetic model for each system. The thermodynamic parameters' values with all the MNPs indicate an exothermic multilayer fission type that could significantly aid the desorption process, making these MNPs a less expensive option.
The results show the As removal efficiencies in groundwater efficiencies for the three nanomaterials. At the end of the contact time, with the three synthesized materials, even at the first 10 min of the assay, the As final concentration was below the limit proposed by the World Health Organization (WHO, 0.010 mg/L).
Treating 1 m 3 of groundwater (C i 0.050 mg /L of As) with CoFe 2 O 4 would cost ≈25 US$/m 3 ; even when the adsorption process has higher operational costs than conventional treatment processes, it may be a suitable alternative for the As removal from aquatic environments such as groundwater.

Discussion
Traditionally, groundwater has been considered a safe (high-quality) source of drinking water, although high concentrations of some heavy metals and other elements, such as As and F − , have been found. Actually, more than 2.5 billion people are exposed to health problems due to the consumption of contaminated water; hence, more research has to be done to accomplish one of the most important challenges of the developed world, providing to the global inhabitants a high quality of drinking water. Mexico has several problems with water, including aquifer depletion, saline intrusion, and As and F − high concentrations in most of the aquifers of its territory. At Durango city, groundwater is the primary source of water, and according to the last sampling, the As concentration ranges from 0.002 to 0.188 mg/L and F − from 0.16 to 10.00 mg/L (CIMAV data). The problem has been known for many years; however, no action has been taken to correct the situation [18].
Until a few years ago, most of the residents were unaware of this problem and the health consequences associated with water consumption with these concentrations of As and F − . In the last ten years, the consumption of bottled water has skyrocketed, increasing the number of shops that sell low-cost bottled drinkable water (30% of the cost) compared to a bottle from a "large" water treatment plant. In other research (data not shown), more than 70 bottled water outlets were sampled; the selling price per liter varied from 0.40 to 0.60 USD, and the most common processes used in these outlets included filtration (with different substrates) and reverse osmosis, as well as ozonation, UV, and/or chlorine for disinfection. However, in more than 50% of the samples taken, the As and F final concentrations were above the MPL established by NOM-127- SSA-1994SSA- (2000, and the problem is worse considering that the bottled water sale regulation is severe, 0.010 mg/L for As and 0.7 mg/L for F (NOM-201-SSA1-2015).
Worldwide growing concern about the presence and toxicity of As and F in drinking water and their health effects have increased interest in looking for alternative and low-environmental-impact methods for their removal at all levels, including the systems operated by the corresponding authority and the other facilities. In Table 3, we present a comparison of the three alternative methods for As and/or F − removal efficiency and consider if it is possible to apply these to solve the problem. In this context, As and F − removal by the electrocoagulation process is emerging as an increasingly viable alternative to solve the presence of both pollutants in drinking water.
Additionally, the bioadsorbent materials' ability to remove F − in aqueous media is attributed to both the biopolymers (functional groups) present in the biomass and the cation supported in it [64,[69][70][71][72]. The capacity that composite materials achieve depends on multiple factors. Among them, the affinity of the cation for the F − ion stands out. The bioadsorbents from orange and apple peels showed a higher capacity (mg F − /g adsorbent) for F − anion removal than other bioadsorbents that did not receive chemical treatments or were impregnated with Ca [7,37,39,40,59,73].
This capacity is comparable to that previously reported by Jha et al. [58] in Zr-loaded biomasses. Their research studied the F − ion adsorption by carboxylated Zr +4 -loaded orange peel, finding a maximum capacity of 5.6 mg/g at neutral pH and with 50 min of contact time. However, the capacities of the bioadsorbents under study are lower than those reported with Al-loaded biomasses [57,74]. Of the three cations mentioned (Ca, Zr, and Al), Zr has multiple advantages. The main one is that Zr +4 has a high selectivity for F − ions, forming stable compounds [34,75,76]. Another relevant advantage is its pH stability and the fact that no adverse effects are reported on human health [69,72,77]. In contrast, even though Al has a high affinity to capture F − ions, it has a limited operating pH and exhibits species solubility in treated water [78,79].
The use of magnetic nanomaterials has improved problems derived from contamination in water, including solutions with these materials removing bacteria, viruses, and pesticides [80]. Several advantages of MNP application for pollutant removal have been reported; one of the most important is their regeneration capacity, allowing their use for numerous cycles. MNPs assayed in this study could be used as an alternative technology for As removal from groundwater. Their higher adsorption capacity could be attributed to the MNP surface area and the existence of iron (hydr)oxides; also, the physicochemical properties of the solution and As speciation must be taken into consideration [81].

Conclusions
The electrocoagulation process allows the simultaneous removal of fluoride and arsenic present in a groundwater matrix, in 15 min, applying a current density of 4.5 mA/cm 2 , with a fluoride removal efficiency of up to 93.16%, as well as practically complete removal of arsenic. The initial pH and the anode material have a determining effect on the removal efficiencies of both contaminants; however, fluoride removal was found to be the most difficult. The highest efficiencies were achieved when combined aluminum and iron were used as an anode at an initial pH of 3. Under these operating conditions, the process has an estimated cost of 0.279 USD/m 3 , considering only the material loss of the electrodes and the energy consumption.
The bioadsorbents prepared from orange and apple peels, under the described conditions, showed the capacity to adsorb As and F − ions simultaneously, both in laboratory solutions and in a real matrix. The experimentation carried out shows that, under acidic conditions, there is a higher percentage of removal in F − ions and that this decreases with increasing pH. Under acidic and basic pH conditions, both bioadsorbents can remove As ions. Neither of the two materials showed a significant reduction in removal capacity in the presence of anions typical of groundwater, so these bioadsorbents are considered to have a high affinity for the adsorbates studied. Finally, the bioadsorbents reduced As and F − concentrations in the groundwater supplied to the population of Durango, Mexico, to levels below the WHO MPL, in a contact time of 60 min.
The adsorption capacity of CoFe 2 O 4 , MnFe 2 O 4 , and Fe 3 O 4 , synthesized by the chemical coprecipitation method, supports their potential application for the As removal from water, reaching final concentrations below the official Mexican standard (NOM-127- SSA1-1994SSA1- , 2000 for each experiment and, furthermore, below the WHO recommended value (0.010 mg/L). The performance of the CoFe 2 O 4 and MnFe 2 O 4 adsorbents was better than that of Fe 3 O 4 . The kinetics of the process for each of the systems were best represented by the second-order reaction model, common for metal adsorbents, where the numerical value of K indicates that the particles remove the metalloid very quickly. As adsorption by MNPs was an exothermic process of the multilayer physisorption type, which can significantly aid the desorption process. Under these operating conditions, the adsorption process cost has been estimated, 25 USD/m 3 .
Finally, the potential for these alternative removal treatments involves the reactor design, continuous flow work, and decreasing electrode passivation, among other important aspects for the electrocoagulation process. For bioadsorbent materials, it is necessary to study the biocomposite lifetime, specifically when the materials could be applied in a continuous flow system. Finally, the challenges of adsorption with MNP are the costs of the materials as well as their disposal at the end of their useful lifetime.