Trace element ions, such as Cr(VI) and F, are of particular interest due to their environmental impact. Both ions exhibit an anionic nature in water that can show similar removal tendencies except for their significant differences in ionic radius, speciation forms, and kosmotropic-chaotropic behaviors. Accordingly, partial freezing was performed to examine the comparative freeze separation efficiencies of Cr(VI) and F from aqueous solutions. Freeze desalination influencing parameters such as initial ion concentration, salt addition, and freeze duration were explored. Under optimal operating conditions, freeze separation efficiencies of 90 ± 0.12 to 95 ± 0.54% and 58 ± 0.23% to 60 ± 0.34% from 5 mg/L of Cr(VI) and F, respectively, were demonstrated. The salt addition into the F-containing solutions revealed more F ion intercalation into the ice, initiating the decrement of freeze separation efficiency. The influences of structuring-destructuring (kosmotropicity-chaotropicity) and the size-exclusion nature of ice crystals were used to explain the plausible mechanism for the difference in freeze separation efficiency between Cr(VI) and F ions.

1. Introduction

Fluoride (F) is a highly reactive ion that seldom exists in a free state. The distribution of elevated fluoride concentrations is reported in several countries in groundwater, including India [1], Ethiopia [2], China [3], USA [4], and in most rift valley areas of Africa [57]. Fluoride contamination of drinking water is common, especially from geogenic sources, mainly in calcium-poor aquifers [8, 9]. Fluoride is vital in preventing dental caries when found at the safe limit. However, it causes detrimental health effects such as tooth molting and dental and skeletal fluorosis when found beyond the safe limit [10]. Accordingly, the World Health Organization (WHO) has set maximum permissible level of 1.5 mg/L F in drinking water [11]. On the other hand, Cr(VI) is a common pollutant discharged into the environment from both anthropogenic and natural sources. Thus, surface and groundwater contamination is a persisting challenge in many countries. However, the problems are being exacerbated in developing countries that lack ternary treatments for their industries [12]. Cr(VI) has been shown to have genotoxicity, mutagenicity, and even cause cancer in humans [13, 14]. To mitigate such effects, promising treatment options for Cr(VI)-contaminated water and wastewater are highly sought after.

Several treatment techniques have been investigated in removing Cr(VI) and F from water, including adsorption [15], ion-exchange [16], electrodialysis [17], electrocoagulation [18], membrane techniques [19], and freeze desalination [20, 21] have been investigated in removing Cr(VI) and F from water. Freezing, the process of cooling to change liquid water into ice (a solid form of water), and follow-up dissolution of ice in a separate system, are alternative physical processes that can be used for the separation of water contaminants. This is based on the difference in freezing points between fresh and contaminated water. When partial freezing is achieved, contaminants are separated from the ice due to the impossibility of permeating into the ice crystal lattice [22]. Freeze separation technologies are believed to have promising advantages in different applications. Among these are the energy savings when compared to conventional thermal processes [23]; the use of renewable energy resources and liquefied natural gas (LNG) [24, 25]; the application of combined desalination processes to reduce energy consumption [23]; and low operating temperature, which minimizes scaling and corrosion [26].

Freeze desalination is commonly employed to remove salts from seawater and lakes. Recently, the technique has also been applied to the removal of several contaminants, including pharmaceuticals [27], ammonia [28], and ions from water [29]. Previously, we reported the separation efficiencies of individual Cr(VI) and F ions in a closed freezer unit from aqueous solutions, simulated water (synthetic solutions that mimic natural waters), and tap water samples [20, 30]. In fact, the reasons for selecting these two ions existing in anionic form are based on the assumption that Cr(VI) is relatively high in hydrated radius, whereas fluoride is minimal in anions. Thus, this study affirmed the hypothesis of size exclusion in freeze desalination and other related parameters. Specifically, the focus of the study was to assess the comparative freeze separation efficiencies of both ions, Cr(VI) and F, simultaneously using home-use refrigerators.

2. Materials and Methods

Separately, a stock solution of 1000 mg/L of Cr(VI) and F was prepared using deionized water. The different concentrations of each ion have been prepared by diluting the stock solutions using deionized water. The required simultaneous Cr(VI) and F solutions were prepared with subsequent dilutions and used fresh during the experiment. Ion strength adjustment buffer (TISAB) was used during fluoride determination. Different levels of Cl ion, from NaCl, (from 0–40 mg/L) have been prepared using 5 mg/L F solutions as a solvent.

2.1. Methods

Fluoride in meltwater was determined using a fluoride ion selective electrode (ISE, Mettler-Toledo AG) based on a standard procedure [31]. Meltwater to buffer (TISAB) ratio of 1 : 1 mixture was measured. The TISAB buffer was prepared by adjusting the pH to a value between 5 and 5.5. The pH was measured using a multiparameter probe (Hatch, HQ40 d).

Cr(VI) in the meltwater was analyzed using a double beam UV-Vis spectrophotometer (SPECORD 200, Analytik Jena, Germany). Thus, 250 mg 1,5-diphenylcarbazide (BDH, England) was dissolved in 50 mL acetone (HiMedia, India) and stored in an amber bottle. The pH of the sample solution was adjusted to 2.0 ± 0.5 using 0.5 M sulfuric acid. Then, 2 mL of the 1,5-diphenylcarbazide solution was added to each sample (100 mL) and the mixture was allowed to stand for 10 min in the dark to obtain full color development. The absorbance was then measured at 540 nm. The experiments were carried out in triplicate, and the results were accompanied by error bars based on the standard deviation of the triplicate measurements.

2.2. Evaluation of Freeze Desalination

Freeze separation efficiency (E) and freezing ratio (RF) were presented as follows:where E is the freeze separation efficiency, RF is the freezing ratio, Vo is the initial volume of melt water in F or Cr(VI)-containing solution, Vs is the volume of the solid phase (ice) after melting (mL), Co initial concentration of F or Cr(VI) solution (mg/L), and Cs concentration of F or Cr(VI) in meltwater (mg/L).

3. Results

3.1. Effect of Freezing Time

Theoretically, ice formation is assumed to be free of salts. Thus, fresh water can be produced by partial freezing and separating the ice from the concentrated residues. The effect of freeze duration on Cr(VI) and F separation efficiency is illustrated in Figure 1. It was demonstrated that as the freeze duration increased for both ions, the separation efficiency remained constant for the respective ion removal until the ions started to intercalate within the ice crystals. As shown in Figure 1, after 2.5 hours of partial freezing, 58–60% F and 90–95% Cr(VI) separation was observed. When the freeze duration was further increased, the freeze separation efficiency of the ions decreased significantly. Higher concentrations of Cr(VI) and F introduced into the ice could be attributed to salinity buildup at the interface, which causes ion introduction/intercalation [20, 25, 30]. This result is with the exception of separation efficiency trends. The optimal duration that provides a small volume of residual water rejection and high quality meltwater was considered at 2.5 h. Therefore, follow-up experiments were performed at a fixed freeze duration of 2.5 h. Liquid residue rejection by a volume of 10–20% was observed for the reported 95% Cr(VI) and 60% F separation efficiencies (Figure 2). Similar meltwater recovery from saline waters (144 mg/L), meeting the standard of drinking water, was achieved using hybrid desalination process [32]. In the freezing process, the source temperature has also significant impact on the freezing time [33]; however, in this case, the experiment performed at fixed initial temperature.

3.2. Effect of Initial Concentration

Figure 3 shows the effect of concentration variation on freeze desalting. The freeze separation efficiency was approximately 60% for 1–5 mg/L F solutions and 48% for 100 mg/L F solutions. Similarly, for concentrations of 1–10 mg/L Cr(VI), the separation efficiency was about 95%. As the initial concentrations of Cr(VI) and F ions increased, the efficiency of freeze separation decreased for relatively higher concentrations. The decrease in freeze separation efficiency may be caused by changes in interface phenomena and the influence of intermolecular effects in the multi-ion system, as described in the literature [21, 34]. In a separate experiment under similar conditions, the separation efficiency of Cr(VI) reached 97%, and F in the range of 62–75%, which was slightly higher than that of the bi-ionic system where Cr(VI) and F were treated together in a solution [20, 21, 30]. Another factor in heterogeneous systems could be the faster rate of ice formation after the first stage of ice nucleation that favor the intercalation of ions into the ice phase. The nature of charge and size of F, as well as kosmotropicity, which contributes to the stability and structuring of water-water molecules [35, 36], and the chaotropic nature (e.g., Cr(VI)) [37] could all contribute to differences in freeze separation efficiencies of Cr(VI) and F ions. In addition to Cr(VI)’s chaotropicity, its larger size and lower tendency to form H-bonding compared to F ions can have an effect on freeze desalting efficiency (Table 1). Kosmotrope is the term explaining ion hydration which contributes to the stability and structure of water-water interactions, leading to the favorable interaction of ions with water molecules. Such phenomena could enhance the incorporation of F into the ice phase during partial freezing. On the other hand, chaotrope stands for expressing weakly held hydrated ions with low charge density and weaker interactions with water molecules [29].

3.3. The Influence of Salt Addition

The use of freeze desalination technology for practical purposes is most common in desalting saline water, which includes seawater, brackish water, and so on. On top of that, to investigate the effect of impurities/multi-ions on freezing technology with the ultimate goal of desalting saline water sources, salt addition (NaCl) was investigated. As shown in Figure 4, the added salt concentration significantly decreased the separation efficiency of F. The tendency of poor crystal structure and morphology modification of ice upon salt addition could be the cause of F intercalation into ice. In addition, the change in the interface phenomena and catalyzed rate of ice formation after the initial ice nucleation process could be the reason for the higher F inclusion into the ice phase [44]. Freeze desalination processes for several ions in different initial concentrations have been demonstrated in previous studies [31, 45]. The enhancement of ice crystal growth kinetics in the presence of salt ions has been demonstrated by Adeniyi et al. [46]. The microscopic models showed the effect of polarizability in driving the weakly hydrated Cl ion into water [45]. Halde [47] described the inclusion of impurities during progressive freezing, demonstrating that variables such as freezing rate and added chemicals have a strong influence on impurity migration [47]. Similar observation has been made by Melak et al. [20]. In parallel to this fact, the study by Shen et al. [48] demonstrated that supersolid formation is related to the effect of ionic polarization that shortens and stiffens hydrogen bonding, whereas NaCl addition lengthens and weakens the nonbonding interactions of hydration cells [48].

4. Discussion

4.1. Kosmotropicity-Chaotropicity

The freeze separation efficiency of ions is connected to water-ion interactions in the interface, involving several forces [37]. The Hofmeister series arranges ions as a monotonic function of their surface charge density. The dominant forces on ions in water are short-range chemical forces, as evidenced by neutron diffraction, X-ray spectroscopy, and ab initio molecular orbital studies; only the water molecules in close proximity to the ions are affected (e.g., for small ions, i.e., ∼5 Å) [45, 49, 50]. The interaction of water with ions is related to hydrogen bonds in a linear fashion and hydration-dehydration processes [45, 51, 52]. F ion has five bound water molecules, where four of these are tightly bound with water molecules in the first hydration layer, leaving one for the second layer [5355].

Water-ion interactions are weak in chaotropes where morphological changes of the ice produced and freezing point depression in the first ice nucleation steps are significantly observed. It is also assumed that the removal tendencies of ions is related to the hydration free energy and hydrated radius of ions [56]. Salt addition can cause supersolid formation and is related to the effect of ionic polarization that shortens and stiffens hydrogen bonding, whereas it lengthens and weakens the nonbonding interactions of hydration cells [48]. Vibrational spectra of F(H2O)3-5 in the O–H stretching coupled with high-level ab initio calculations have been used to indicate the structure and stability of these species in water [57]. The Hofmeister series classified each ion as strongly hydrated (small size and high surface charge density)-kosmotrope and weakly hydrated (large size and low surface charge density)-chaotropic ions relative to the strength of water-water interactions Collins et al. (2007). In relation to this, the ionic radius and Jones–Dole viscosity coefficient (B), which have a high influence on freeze separation efficiency, are well discussed in Table 1 and Figure 5. Assaf and Nau [58] noted that chaotropic ions (for example, I, SCN, and ClO4) possess positive Gibbs free energy of hydration and have weak to unfavorable interactions with water [58]. Fluoride ion is relatively kosmotropic, and it could be mainly due to the direct and polarized arrangement of the surrounding water molecules, making it less likely to be removed from pure water (ice form of water). Whereas Cr(VI) is a chaotrope with characteristic features (large ions of low surface charge density which are weakly hydrated). Kosmotropes remain hydrated near the interface while the chaotropes lose their hydration sheath [37].

4.2. Effect of Ionic Size on Freeze Separation

The effect of ionic radius on freeze desalination has been explained in Table 1 and Figure 6. The freeze separation efficiency is strongly related to the ionic radius of ions, as demonstrated in the literature [29, 61] and in our experimental work. A linear relationship has been achieved between freeze separation efficiency and ionic radius (Figure 7).

The theoretical predicted values of Cr(VI) species are presented in Table 2. The prediction was made under the conditions of 5 mg/L Cr(VI), neutral pH, temperature of 25°C, and oxidation-reduction potential(ORP) using visual MINTEQ 3.1 software. Using the effect of Cr(VI) ionic radius, CrO4−2 species were predicted to be the dominant species (Table 2).

4.3. Energy Feasibility Insights

Freeze desalination (FD) has several benefits as compared to thermal and membrane desalination methods. The FD process requires about one-seventh of the latent heat of vaporization. Furthermore, the use of low temperatures in FD applications reduces corrosion and scaling challenges so common in membrane technologies [62]. The utilization of energy released during the regasification of liquefied natural gas and the attempts on the use of hydromechanical methods in the freezer instead of solvents make the method promising [63]. The energy consumption comparison of different desalination methods with respect to freeze desalination was presented in our previous study [20] to indicate the promising nature of freeze desalination. Whereas the required energy to produce 1 kg of potable water is 0.11 kWh according to EPA [63]. The energy consumption of stand-alone progressive freeze concentration (PFC) and suspension freeze concentration (SFC)-based desalination with latent heat recovery was reported close to the reverse osmosis (RO) method [62]. Thus, potable water was produced from saline waters with FD upon post-treatment of obtained ice, making FD a commercially feasible technology. In fact, it may require more than one stage of freezing to bring the saline water to palatability [64].

5. Conclusion

This study presented the comparative desalination of Cr(VI) and F ions from aqueous solutions. The main reason why the differences in separation efficiencies of Cr(VI) and F during partial freezing were explored. It was found that the coexistence of Cr(VI) and F showed a lesser impact on their separation efficiency. The addition of salts/impurities/further reduced or enhanced the intercalation of more fluoride ions into the ice phase. When Cr(VI) and F were studied in a separate set of experiments under similar conditions, a significant difference in their separation efficiencies was implicated. The plausible reasons for separation efficiency differences of Cr(VI) and F were mainly due to the differences in ionic size and kosmotropic-chaotropic associated factors of ions in water.

Data Availability

All the data and materials as well as software applications or custom code support the paper, complying with field standards. The raw data will be available if necessary during publication (data transparency).

Conflicts of Interest

There are no direct or indirect known conflicts of interest to declare related to the work submitted for publication within the last 3 years (beginning from the research to preparing the work for submission).