Role of phase transformation of barium perborates in the effective removal of boron from aqueous solution via chemical oxo-precipitation

Abstract

This work developed a chemical oxo-precipitation (COP) process for the removal of boron from aqueous solution. Hydrogen peroxide was utilized to convert boric acid to perborate anions, which precipitated directly with barium as barium perborates, and so efficiently reduced the boron level via phase transformation at room temperature. Several experimental parameters were investigated to evaluate the appropriate conditions for the phase transformation, including the pretreatment time and pH (of the mixture of boric acid with hydrogen peroxide), the reaction pH and the molar ratios of hydrogen peroxide and barium to boron. The results revealed that the reaction time of COP was shortened due to the speciation of perborate anions during pretreatment. The reaction pH_r was crucial for COP because the phase transformation of barium perborates only occurred at a pH range of 8.5–10. Under optimal conditions – initial [H₂O₂]/[B] = 2, [Ba]/[B] = 0.8, pH_p = 10.5 ± 0.1, time_p = 20 minutes, and pH_r = 10 ± 0.2 – the boron concentration was reduced from 1000 ppm to less than 3 ppm in 3 hours.

---

1. Introduction

Boron is an essential element for many purposes in various industries; for example, it is used as a moderator in nuclear reactors, in borosilicate glass, in fertilizers, in ceramics and in the manufacture of TFT-LCDs and semiconductors.^1,2 The manufacture of such products generates a variety of wastewaters that contain high concentrations of boron. For instance, wastewater from the glass industry contains boron and fluorine;³ effluent from the manufacturing of TFT-LCD polarizers contains 800 ppm of boron with a pH value around 6;⁴ fired power plants discharge boron-containing wastewater because boron is present as an impurity in coal;^5,6 electroless plating processes using sodium borohydrides and dehydrogenation of borohydrides via hydrolysis reactions generate alkaline boron wastewater.^7,8 Moreover, our recent case studies show that a semiconductor company in Taiwan discharges alkaline wastewater (pH ∼ 12.2) containing 1000 ppm of boron. Streams at pH 9.2 containing 600–1000 ppm of boron are generated when applying aqueous borax solutions as lubricant carriers in wire drawing processes. Considering the potential toxicity of boron and its impact on the environment,⁹ the World Health Organization (WHO) in 1993 established a guideline maximum level of boron of 0.5 ppm for drinking water, which was revised to 2.4 ppm in 2011.¹⁰ National standards vary, being 5 ppm in Canada,¹¹ 4 ppm in Australia,¹² 1 ppm in the European Union and Japan^13,14 and 0.6–0.9 ppm in the U.S.¹⁵

Various types of separation technologies for boron removal have been developed, including precipitation, electrocoagulation, adsorption, ion exchange and membrane processes. Ion exchange methods using boron-selective resins are suitable for removal of scarce boron levels. Nevertheless, the high costs and complicated regeneration processes limit the implementation of resins.¹⁶ Although adsorptions using industrial wastes materials and doubled-layered hydroxides (LDHs) as sorbents are economical, they suffer from low adsorption capacity and poor regenerability.^17,18 Membrane processes have been widely used in desalinations, however, high operation pH is required to achieve high boron rejection that often results in scaling problems owing to the precipitation of metal hydroxides.^19,20 Conventional precipitation methods are capable of reclaiming boron as borate ores from solutions under moderate temperatures.²¹ Using lime alone can achieve 87% of boron removal from solution containing 750 ppm boron at 60 °C in 8 hours.²² With the aid of extensive dosages of co-precipitants such as sulfuric acid and phosphorous acid, more than 90% of boron can be removed.^23,24 Electrocoagulation utilizes sacrificial anodes to generate metallic hydroxide precipitates and remove pollutants by adsorption, surface complexation or co-precipitation.²⁵ Several kinds of anodes have been studied for boron removal, including aluminum, iron, zinc and magnesium.^26–29 However, large amounts of sludge are generated in either conventional precipitation or electrocoagulation when dealing with wastewater with high boron concentration, because the boron is mainly removed by sorption on flocs rather than direct precipitation.

Chemical oxo-precipitation (COP) has been demonstrated to remove boron efficiently from solutions that contain high boron concentrations at room temperature.^30,31 The core mechanism of COP is the peroxolysis of boric acid or borate ion by hydrogen peroxide, creating perborate anions that can directly precipitate with alkaline earth metal as perborate salts with rather low solubility. As displayed in Fig. 1, various peroxy–boron species are formed in equilibrium by boric acid, hydrogen peroxide and protons through a series of nucleophilic substitutions. Perborate anions, particularly B(OH)₃OOH⁻ (3) and B(OH)₂(OO)₂B(OH)₂²⁻ (7), dominate at pH 8.5–12.5 (as will be discussed later). Among alkaline earth metals, barium salt is the most effective precipitant for removing boron that can eliminate boron level from 1000 ppm to 3 ppm in 4 hours. The high efficiency of COP has been attributed to the phase transformation of the precipitates from amorphous Ba(B(OH)₃OOH)₂ and BaB(OH)₃OOB(OH)₃ to crystalline BaB(OH)₂(OO)₂B(OH)₂ (eqn (1) and (2)).³¹

BaB(OH)₃OOB(OH)_3(s) + H₂O₂ → BaB(OH)₂(OO)₂B(OH)_2(s) + 2H₂O (1)

Ba(B(OH)₃OOH)_2(s) → BaB(OH)₂(OO)₂B(OH)_2(s) + 2H₂O (2)

Fig. 1 Thermodynamic equilibrium of peroxo compounds and corresponding constants of boric acid, hydrogen peroxide, perborates and proton.^32–35

Crystalline barium perborate has been assumed to be more stable and less soluble than amorphous forms, favoring the elimination of boron. However, the suitable conditions for phase transformation of barium perborates are still obscure due to the lack of information on precipitation behavior of perborate salts. This work aimed to study the influence of the major parameters, including pretreatment and reaction pH and molar ratios of [H₂O₂]/[B] and [Ba]/[B], on the efficacy of COP from the views of phase transformation, perborate speciation and precipitation behaviors. The correlations between reaction pH and phases of precipitates were verified with the aid of characterizations, including XRD, Raman spectra and TEM.

2. Materials and methods

All reagents were analytical grade, and used without further purification. Synthetic wastewater was prepared using boric acid (B(OH)₃, Showa). Hydrogen peroxide (H₂O₂, 35 wt%, Merck) was the agent for the pretreatment of boric acid or borate ions, and barium chloride (BaCl₂·2H₂O, Panreac) was the precipitant. All solutions were prepared with deionized water that had been purified using a laboratory-grade RO-ultrapure water system.

The COP batch experiments were carried out with a jar test at 150 rpm and room temperature. The pretreatment stage, mixing H₂O₂ and boron solution, was implemented before precipitation using barium salts. During the pretreatment, a known dosage of H₂O₂ (based on the initial molar ratio of [H₂O₂]/[B]) was added to 500 mL of solution that contained 2000 ppm boron. The pH of the solution was adjusted to the desired value (pH_p) using NaOH (Panreac), and HCl (Shimakya, 32%). After the mixture of boric acid and H₂O₂ had been stirred for a specific time (time_p), 500 mL of BaCl₂ solution was added into the boron solution to yield various molar ratios of [Ba]/[B] and to initiate the precipitation. The reaction pH of the mixture (pH_r) was monitored and controlled using NaOH (4 M) and HCl (32 wt%) every 15 minutes. The reaction proceeded for 4 h, and 5 mL of the solution was withdrawn at specific intervals and filtered through a 0.20 μm PVDF syringe filter (CHROMAFIL). Dissolved oxygen concentration (DO) was measured using a dissolved oxygen meter (Oxi 3210, WTW). The filtrate was titrated using 0.01 M permanganate (KMnO₄, Showa) in 100 mL of 3% H₂SO₄ solution to determine the concentration of the aqueous active oxygen (peroxo-containing compounds). The boron and barium concentrations were analyzed using an inductively coupled plasma optical emission spectrometer (ICP-OES, ULTIMA 2000, HORIDA) set to detect at wavelengths of 249.672 and 413.225 nm, respectively. The precipitates collected were rinsed several times and then dried at 60 °C for one day. The crystal structure of the collected precipitates were verified by an X-ray diffraction analyzer (Rigaku, RX III) with Cu Kα radiation (40 kV, 30 mA). The morphology of the precipitates were observed by a transmission electron microscope (Hitachi, H-7500) at 80.0 kV. As for the TEM sample preparation, the suspensions of precipitates were prepared by agitating aqueous solution containing small amount of dried precipitates ultrasonically for 20 minutes. After casting 20 drops of the suspensions on the TEM grids, the grids that contained samples were dried at 60 °C for one day.

3. Results and discussion

3.1. Effects of pretreatment conditions on COP performance

The phase transformation of amorphous Ba(B(OH)₃OOH)₂ or BaB(OH)₃OOB(OH)₃ to crystalline BaB(OH)₂(OO)₂B(OH)₂ was a critical reaction that considerably reduced the level of dissolved boron. The transformation was driven by the degree of supersaturation, so a phase with the lowest solubility dominated the precipitate at equilibrium.³⁶ To shorten the reaction time of COP, the pretreatment stage had to cause proper speciation of perborate anions to create the least soluble barium perborate phase (BaB(OH)₂(OO)₂B(OH)₂) as much as possible at the beginning of precipitation stage. Therefore, the mixing time of boric acid and H₂O₂ was varied to determine the effect of time_p on boron removal, as shown in Fig. 2(a), while both pH_p and pH_r were maintained at 10.5. The time required for the boron concentration to approach its minimum value decreased as time_p increased. COP with pretreatment for 0 min reduced 1000 ppm boron to less than 5 ppm in 4 h. In contrast, COP with pretreatment for 20 and 30 min removed more than 99.5% of the boron in 3 h. The variable time_p presumably influenced the speciation of the perborate anions. Restated, B(OH)₂(OO)₂B(OH)₂²⁻ required time to reach equilibrium with boric acid and H₂O₂ based on Fig. 1 before it could effectively precipitate with barium salt. When B(OH)₂(OO)₂B(OH)₂²⁻ was the form of most of the soluble boron, the duration of the subsequent phase transformation was minimized. The sequence of the reagents addition is a decisive factor influencing the efficacy of COP. In the previous studies, the hydrogen peroxide was added after the mixing of barium salt and boron solution (without pretreatment step), and therefore the complexation and speciation of perborate anions and precipitation with barium ions occurred simultaneously. However, Fig. 3 shows that the reduction ratio of aqueous peroxo to boron (Δ[Peroxo]/Δ[B]), an index to represent the peroxo content (the extent of boron complexing with hydrogen peroxide) in precipitates defined in eqn (3), was only 0.7 at the beginning in our earlier study.³¹ By contrast, the initial Δ[Peroxo]/Δ[B] of COP with pretreatment was 0.9, indicating the peroxo contents in precipitates were higher than that without pretreatment step. Furthermore, Δ[Peroxo]/Δ[B] reached unity with pretreatment more rapidly than without pretreatment. It meant that the time required for phase transformation of barium perborates to BaB(OH)₂(OO)₂B(OH)₂ ([Peroxo]/[B] = 1) was shortened. In other words, the stage of pretreatment accelerated the phase transformation of barium perborates as more stable precipitates with high peroxo content formed at the beginning of COP.

Δ[Peroxo]/Δ[B] = ([Peroxo]₀ − [Peroxo])/([B]₀ − [B]) (3)

Fig. 2 Effect of pretreatment time (● 0 min, △ 10 min, ◆ 20 min, ▽ 30 min) on (a) boron and (b) dissolved oxygen concentrations. (Initial boron concentration = 1000 ppm, initial [H₂O₂]/[B] = 2, [Ba]/[B] = 0.8, pH_p = 10.5 ± 0.1, pH_r = 10.5 ± 0.2).

Fig. 3 Comparison of Δ[Peroxo]/Δ[B] from COP with and without pretreatment (earlier study in ref. 31). (Initial boron concentration = 1000 ppm, initial [H₂O₂]/[B] = 2, [Ba]/[B] = 0.8, pH_p = 10.5 ± 0.1, pH_r = 10.5 ± 0.2).

Dissolved oxygen (DO), as presented in Fig. 2(b), can be used to detect the loss of peroxo content in an aqueous phase. At the beginning of precipitation herein, the DO was significantly elevated by the pretreatment because the pH_p of the mixture that contained perborate and excess H₂O₂ was maintained at 10.5, at which H₂O₂ easily decayed. Furthermore, the deprotonated HO₂⁻ may have promoted the self-decomposition of H₂O₂.³⁷

H₂O₂ + HO₂⁻ → H₂O + O₂ + OH⁻ (4)

The variation of DO seems to be related to the boron level. When the boron concentration sharply decreased to 30 ppm and was then maintained at approximately 20 ppm, the DO without pretreatment was around 10 mg-O₂ L⁻¹, and with a time_p of 10–30 min, it decreased from 20 mg-O₂ L⁻¹ to 11 mg-O₂ L⁻¹. However, as the residual boron concentration declined to less than 10 ppm, DO raised remarkably from 10 mg-O₂ L⁻¹ to higher than 20 mg-O₂ L⁻¹. Since the boron level (∼30 ppm or 2.78 mM) was much lower than the peroxo concentration (including perborates and H₂O₂, which had a total concentration of 90 mM), H₂O₂, rather than perborate anions, was the dominant peroxo species before the boron level declined to below 10 ppm. Hence, the decomposition of H₂O₂ was hindered in the early stage of precipitation, and DO was simultaneously suppressed. This effect was attributable to the consumption of H₂O₂ by the phase transformation of BaB(OH)₃OOB(OH)₃ to BaB(OH)₂(OO)₂B(OH)₂, which may compete with the self-decomposition of H₂O₂.³¹ Therefore, DO can be regarded as an index of the termination of the precipitation process.

During the pretreatment stage, the pH of the mixture of boric acid and H₂O₂ fundamentally determined the speciation of perborate anions according to the works of Davies and Deary, which revealed that the presence of perborate anions under alkaline conditions was dependent on the concentration of boron and peroxide.^33,34 As indicated in Fig. 4, pH_p critically influenced the reaction time required to reduce the boron concentration to below 10 ppm via phase transformation. At pH_p in the range of 10 to 11, perborate anions, mostly in the form of B(OH)₂(OO)₂B(OH)₂²⁻, readily precipitated as BaB(OH)₂B(OO)₂B(OH)₂. Restated, the pH_p determined not only the speciation of perborates in pretreatment stage but also the perborates species in the initial precipitates, and therefore shortened the time for barium perborates to transform.

Fig. 4 Effect of pretreatment pH on removal of boron by COP process. (Initial boron concentration = 1000 ppm, initial [H₂O₂]/[B] = 1.5, [Ba]/[B] = 0.6, pH_r = 9.5 ± 0.1, pretreatment time = 20 min).

3.2. Relationships between reaction pH and phases of precipitate

With pH_p fixed at 10.5, Fig. 5 plots boron levels at 15 min and 4 h as functions of pH_r. The boron levels at 15 min were determined by the degree of supersaturation of barium perborates. Therefore, the concentration of dissolved boron followed the solubility of precipitates, and decreased from 1000 ppm to lower levels in 15 min. The results are nearly the same as our previous work in which up to 98.5% of boron removal was reported.³⁰ Notably, at pH_r 8.5–10.5, more boron could be eliminated, yielding a concentration of less than 10 ppm in 4 h, while boron levels at other values of pH_r remained nearly the same for 4 h. The considerable difference between the boron levels at 15 min and 4 h arose from the phase transformation of the barium perborates at a pH_r (8.5–10.5) that could reduce the boron level to less than 10 ppm. Fig. 6 displays X-ray diffractograms and Raman spectra of the precipitates of COP at pH_r 7.5–12.5. When pH_r was fixed at 7.5, the precipitates that were collected at both 15 min and 4 h were amorphous phases and no apparent band was observed in Raman patterns, consistent with the fact that boron levels remained unchanged after 4 h at pH_r values of less than 8.5. In contrast, pH_r values of 9 and 10.5 supported the transformation of barium perborate from an amorphous phase in 15 min to a well-crystallized phase after 4 h (although the crystalline structure cannot yet be characterized using the standard JCPDF database), and minimized the boron concentration in the end solution. Meanwhile, the Raman spectra in Fig. 6(b) reveals that a band at 870 cm⁻¹ is attributable to the symmetric stretching of (O–O) in –OOH bond, and two intensive bands at 710 cm⁻¹ and 932 cm⁻¹ represent the symmetric stretching of O–O and asymmetric stretching of [–B(OO)₂B–] of cyclic perborate anion for the crystalline precipitates (samples at pH 9 and 10.5 after 4 h treatment).^38,39 None of the characteristic bands appears for the amorphous samples. Therefore, the chemical compounds of crystalline precipitates are assumed to be BaB(OH)₂(OO)₂B(OH)₂ based on the characterization of Raman spectra. At pH_r 12, a witherite phase (BaCO₃, PDF#71-2394) predominated the diffraction peaks of the precipitates, suggesting that much of the barium salt was consumed by the dissolved carbonate at higher pH_r (above 11, as anticipated owing to the constant boron level in 15 min – 4 h). Since the jar test was performed in an open system, the carbonate (CO₃²⁻) concentration could rise to 10^3.4 M at pH 12.5 at equilibrium ([H₂CO₃*] = K_HP_CO2 ∼ 10⁻⁵ M at 1 atm, 298 K).⁴⁰ BaCO₃, whose solubility product is rather high (pK_sp = 8.29),⁴¹ strongly competed with barium perborate and thereby inhibited the removal of boron. The phenomenon was confirmed by the strong band at 1057 cm⁻¹ on Raman spectra of precipitates at pH 12.5, which was attributable to the symmetric stretching of carbonate anion.⁴²

Fig. 5 Effect of reaction pH on concentration of boron after 15 min (●) and 4 h (▲) of COP treatment. (Initial boron concentration = 1000 ppm, initial [H₂O₂]/[B] = 2, [Ba]/[B] = 0.75, pH_p = 10.5 ± 0.1, pretreatment time = 20 min).

Fig. 6 (a) XRD patterns and (b) Raman spectra of barium perborate precipitates at various pH_r after 15 min and 4 h of COP treatment. (Initial boron concentration = 1000 ppm, initial [H₂O₂]/[B] = 2, [Ba]/[B] = 0.75, pH_p = 10.5 ± 0.1, pretreatment time = 20 min).

The TEM images as shown in Fig. 7 reveal that the morphology of barium perborate precipitates differed with the pH and reaction time. Fig. 7(a), (c) and (e) show that the precipitates gathered at 15 min under pH_r at 7.5, 9 and 10.5 were all in a form of agglomerates of nearly spherical particles whose size varied from 30 nm to 150 nm. After 4 h, the morphology of precipitates insignificantly changed at pH_r 7.5 (Fig. 7(b)), whereas the precipitates at both pH_r 9 and 10.5 were composed of flakes of few micrometers in length (Fig. 7(d) and (f)). Notably, according to the results of X-ray diffractometry, the nano-sized barium perborate were primarily amorphous, while the well-crystallized barium perborate with flaky or tabular habit grew up to micrometer scale. The micromorphology clearly evidenced that the COP significantly reduced the soluble boron via the recrystallization and grain growth of barium perborate precipitates.

Fig. 7 TEM images of COP precipitates at pH_r 7.5 (in (a) 15 min, (b) 4 h), pH_r 9 (in (c) 15 min, (d) 4 h), and pH_r 10.5 (in (e) 15 min, (f) 4 h) (initial boron concentration = 1000 ppm, initial [H₂O₂]/[B] = 2, [Ba]/[B] = 0.75, pH_p = 10.5 ± 0.1, pretreatment time = 20 min).

3.3. Effects of initial [H₂O₂]/[B] and [Ba]/[B]

The H₂O₂ dosage is one of the most important variables that affects not only pretreatment but also precipitation. According to the equilibrium constants of boric acid and hydrogen peroxide in Fig. 1, the mass balance of the total boron [B_T] and peroxo-containing compounds [Peroxo_T] can be written as the following equations, where K′ (=10^−11.6) is the equilibrium constant of H₂O₂/HO₂⁻.³³

As shown in Fig. 8, the distribution of species as a function of pH with 2000 ppm (185.2 mM) of boron ([B]) varied with the initial H₂O₂ concentration ([Peroxo]). As the initial molar ratio [H₂O₂]/[B] increased, the proportions of boric acid in the form of B(OH)₃ and B(OH)₄⁻ declined whereas those of perborates, especially B(OH)₂(OOH)₂⁻, increased. Accordingly, the quantity and speciation of perborate ions can be controlled by the initial dosage of H₂O₂ in the pretreatment step, leading to a variation in the speciation of soluble boron when barium perborates precipitated. As presented in Fig. 9, the boron levels in the first 10 min of precipitation were lowered by increasing the molar ratio of initial [H₂O₂]/[B].

Fig. 8 Speciation of boric acid and perborate ions. ([B_T] = 2000 ppm, initial [H₂O₂]/[B] = (a) 0.5, (b) 1, (c) 2 and (d) 3).

Fig. 9 Boron and peroxo concentrations as functions of initial molar ratio [H₂O₂]/[B] after 10 min (● B) and 4 h (▲ B, △ peroxo) of COP reaction. (Initial boron concentration = 1000 ppm, [Ba]/[B] = 0.75, pH_p = 10.5 ± 0.1, pretreatment time = 20 min, pH_r = 10 ± 0.2).

The dosage of H₂O₂ influenced the phase transformation of barium perborates and the removal of aqueous boron by precipitation. Fig. 9 indicates that COPs with initial molar ratios [H₂O₂]/[B] in the range 1.25–2.5 further reduced the boron concentration to less than 5 ppm in 4 h. Based on the stoichiometric ratio associated with the crystalline barium perborate, BaB(OH)₂(OO)₂B(OH)₂, the precipitate contains equal amounts of peroxo and boron. The transformation of BaB(OH)₃OOB(OH)₃ to BaB(OH)₂(OO)₂B(OH)₂ involves the dissolution of BaB(OH)₃OOB(OH)₃ and the re-precipitation of BaB(OH)₂(OO)₂B(OH)₂ in equilibrium with H₂O₂/peroxo-compounds.³¹

BaB(OH)₃OOB(OH)_3(s) → Ba²⁺ + B(OH)₃OOB(OH)₃²⁻, K_sp1 (7)

B(OH)₃OOB(OH)₃²⁻ + H₂O₂ → B(OH)₂(OO)₂B(OH)₂²⁻ + 2H₂O, K₇ (8)

BaB(OH)₂(OO)₂B(OH)_2(s) → Ba²⁺ + B(OH)₂(OO)₂B(OH)₂²⁻, K_sp2 (9)

Thus, with an insufficient dosage of H₂O₂, the efficiency of COP would be relatively low. The residual peroxo levels were less than 20 mM for initial molar ratios [H₂O₂]/[B] of 0.5 and 1, and could not easily drive the complete peroxolysis of B(OH)₃OOB(OH)₃²⁻ to B(OH)₂(OO)₂B(OH)₂²⁻ (eqn (8)) (boron removal <95%). In contrast, barium perborate in equilibrium with 50–150 mM of peroxo-compounds, achieved by increasing [H₂O₂]/[B] to 1.25–2.5, substantially reduced the boron concentration (<3 ppm-B, boron removal >99.7%). However, the COP with an initial [H₂O₂]/[B] of 3 was ineffective in removing boron perhaps because the aqueous perborates, B(OH)₃OOH⁻ and B(OH)₂(OOH)₂⁻, predominated the peroxo compounds (Fig. 8).

Based on eqn (10), the concentration of barium ions determines the supersaturation (β) in the precipitation of BaB(OH)₂(OO)₂B(OH)₂. β is defined as the ratio of Q, which is the product of ionic activities of B(OH)₂(OO)₂B(OH)₂²⁻ and Ba²⁺, to K_sp2.

The Gibbs free energy, which is related to β as ΔG = −RTβ, drives the precipitation and dissolution reactions.⁴³ The dissolved boron and amorphous phase are continuously converted into the crystalline phase of BaB(OH)₂(OO)₂B(OH)₂ until Q equals K_sp2 (ΔG = 0 at equilibrium). As shown in Fig. 10, the conversion of the solute to the solid was fast, reducing the concentration of boron from 1000 ppm to 30 ppm in 10 min. With an initial molar ratio [Ba]/[B] greater than 0.6, the residual barium ions promoted the phase transformation of precipitates, further reducing the boron level in 4 h. Although the residual barium concentration at 4 h was increased sharply by elevating the dosage of barium salts, the removal of boron varied insignificantly with [Ba]/[B] above 0.8. Therefore, the optimal molar ratio [Ba]/[B] was in the range 0.6–0.8, which yielded a final boron level of less than 3 ppm in the end solution.

Fig. 10 Boron and barium concentrations as functions of initial molar ratio [Ba]/[B] after 10 min (● B) and 4 h (▲ B, △ Ba) of COP reaction. (Initial boron level = 1000 ppm, initial [H₂O₂]/[B] = 2, pH_p = 10.5 ± 0.1, pretreatment time = 20 min, pH_r = 10 ± 0.2).

Table 1 summarizes the results of present study and other published papers dealing with high boron concentration (>500 ppm-B) via conventional precipitation (CP) and chemical oxo-precipitation (COP). Generally, CP applied hydrothermal processes (130–150 °C) or heating processes at high pH (60–90 °C, pH 11.2–13) to recover boron as parasibirskite (Ca_​2B₂O₅·H₂O). Even though the addition of sulfuric acid and phosphoric acid could enhance the boron removal by co-precipitation of calcium sulfate and hydroxyapatite (HAp, Ca₁₀(PO₄)₆(OH)₂), extensive dosages of precipitants and co-precipitants are required, resulting in low cost-efficiency and creation of large amount of sludge. By contrast, COP is capable of removing boron efficiently at room temperature with the assistance of hydrogen peroxide. Especially when the phase transformation of barium perborates was accomplished, the boron levels could be eliminated to less than 3 ppm. As compared with existing technologies, particularly CP, COP has been found to be more feasible to treat boron wastewater based on some workable conditions, such as room temperature, lower pH and dosage of precipitants.

Table 1 Comparison of the boron removal using conventional precipitation and chemical oxo-precipitation

[B]0 (ppm)	Efficiency and products	Conditions	References
500	99% removal in 14 h, Ca2B2O5·H2O and HAp (Ca10(PO4)6(OH)2)	Hydrothermal process, 130 °C, [Ca(OH)2]/[B] = 17.5, [H3PO4]/[B] = 6.5	44
500	99.2% removal in 2 h, Ca2B2O5·H2O and CaF2	Hydrothermal process, BF4^− as boron pollutant, 150 °C, [Ca(OH)2]/[B] = 10	3
500	99% removal in 10 min, Ca2B2O5·H2O and calcium phosphate species	Hydrothermal process, 130 °C by microwave, pH = 13, [Ca(OH)2]/[B] = 19, [H3PO4]/[B] = 11	24
700	93% removal in 2 h, Ca2B2O5·H2O and CaSO4·H2O	Conventional precipitation, 90 °C, pH = 11.2, [Ca(OH)2]/[B] = 10, [H2SO4]/[B] = 8	23
750	87% removal in 8 h, Ca2B2O5·H2O on unreacted Ca(OH)2	Conventional precipitation, 60 °C, pH = 12.4, [Ca(OH)2]/[B] = 2	22
1000	98.5% removal in 1.5 h, amorphous barium perborate	Chemical oxo-precipitation, room temperature, pH = 10, [H2O2]/[B] = 2, [BaCl2]/[B] = 1	30
1000	99.7% removal in 4 h, crystalline BaB(OH)2(OO)2B(OH)2	Chemical oxo-precipitation without pretreatment, room temperature, pH = 10.5, [H2O2]/[B] = 2, [Ba(OH)2]/[B] = 0.75	31
1000	99.7% removal in 3 h, crystalline BaB(OH)2(OO)2B(OH)2	Chemical oxo-precipitation with pretreatment, room temperature, pH = 10, [H2O2]/[B] = 2, [BaCl2]/[B] = 0.8	This study

---

4. Conclusions

The chemical oxo-precipitation (COP) process is a promising method for remediating wastewaters that contain high boron concentrations at room temperature. The phase transformation of precipitates could enhance the boron removal of COP owing to the low solubility of crystalline barium perborate. The speciation of perborate anions during the pretreatment of boric acid with hydrogen peroxide considerably impacted the reaction time. An increase in the peroxo content of precipitates at the beginning of COP shortened the time to reduce boron to a minimal level. Both pretreatment time and pH_p influenced the efficiency of COP. The characterization of precipitates indicated the phase transformation of barium perborates. The crystallization of precipitates which took place at the reaction pH_r range of 8.5–10.5 significantly affected the performance of COP. The boron level was reduced from 1000 ppm to 3 ppm in 3 h at optimal conditions: initial [H₂O₂]/[B] = 2, [Ba]/[B] = 0.8, pH_p = 10.5 ± 0.1, time_p = 20 minutes, pH_r = 10 ± 0.2.

Acknowledgements

The authors would like to thank the Ministry of Science and Technology, Taiwan for financially supporting this research under Contract No. MOST 105-2622-E-006-013-CC2.

References

1. D. Mohapatra, G. R. Chaudhury and K. H. Park, Miner. Metall. Process., 2008, 25, 175–180.
2. G. Zelmanov and R. Semiat, Desalination, 2014, 333, 107–117.
3. T. Itakura, R. Sasai and H. Itoh, Bull. Chem. Soc. Jpn., 2007, 80, 2014–2018.
4. L. Kentjono, J. Liu, W. Chang and C. Irawan, Desalination, 2010, 262, 280–283.
5. R. Ochoa-González, A. F. Cuesta, P. Córdoba, M. Díaz-Somoano, O. Font, M. A. López-Antón, X. Querol, M. R. Martínez-Tarazona and A. Giménez, J. Environ. Manage., 2011, 92, 2586–2589.
6. L. S. Ruhl, G. S. Dwyer, H. Hsu-Kim, J. C. Hower and A. Vengosh, Environ. Sci. Technol., 2014, 48, 14790–14798.
7. J. Hannauer, U. Demirci, C. Geantet, J. Herrmann and P. Miele, Phys. Chem. Chem. Phys., 2011, 13, 3809–3818.
8. M. Schlesinger and M. Paunovic, Modern Electroplating, John Wiley, New York, 5th edn, 2010.
9. L. Schoderboeck, S. Mühlegger, A. Losert, C. Gausterer and R. Hornek, Chemosphere, 2011, 82, 483–487.
10. WHO, Chemical aspects, in Guidelines for drinking-water quality, WHO, 2011, pp. 155–201, ch. 8.
11. Health Canada, Guidelines for Canadian Drinking Water Quality—Summary Table, 2012.
12. NHMRC and NRMMC, Australian Drinking Water Guidelines 6, 2011.
13. H. Wakayama, Revision of Drinking Water Quality Standards in Japan, 2010.
14. E. Weinthal, Y. Parag, A. Vengosh, A. Muti and W. Kloppmann, Environ. Sci. Eur., 2005, 15, 1–12.
15. U.S. EPA, Drinking Water Health Advisory For Boron, 2008.
16. M. M. Nasef, M. Nallappan and Z. Ujang, React. Funct. Polym., 2014, 85, 54–68.
17. Z. Guan, J. Lv, P. Bai and X. Guo, Desalination, 2016, 383, 29–37.
18. F. L. Theiss, G. A. Ayoko and R. L. Frost, J. Colloid Interface Sci., 2013, 402, 114–121.
19. V. Freger, H. Shemer, A. Sagiv and R. Semiat, in Boron Separation Processes, ed. M. Bryjak and N. Hilal, Elsevier, Amsterdam, 2015, pp. 199–217.
20. P.-K. Park, S. Lee, J.-S. Cho and J.-H. Kim, Water Res., 2012, 46, 3796–3804.
21. M. Tagliabue, A. P. Reverberi and R. Bagatin, J. Cleaner Prod., 2014, 77, 56–64.
22. C. Irawan, Y.-L. Kuo and J. Liu, Desalination, 2011, 280, 146–151.
23. P. Remy, H. Muhr, E. Plasari and I. Ouerdiane, Environ. Prog., 2005, 24, 105–110.
24. H. C. Tsai and S. L. Lo, J. Hazard. Mater., 2011, 186, 1431–1437.
25. S. Vasudevan and M. A. Oturan, Environ. Chem. Lett., 2014, 12, 97–108.
26. M. H. Isa, E. H. Ezechi, Z. Ahmed, S. F. Magram and S. R. M. Kutty, Water Res., 2014, 51, 113–123.
27. S. Vasudevan, J. Lakshmi and G. Sozhan, Desalination, 2013, 310, 122–129.
28. S. Vasudevan, S. M. Sheela, J. Lakshmi and G. Sozhan, J. Chem. Technol. Biotechnol., 2010, 85, 926–933.
29. G. Sayiner, F. Kandemirli and A. Dimoglo, Desalination, 2008, 230, 205–212.
30. Y.-J. Shih, C.-H. Liu, W.-C. Lan and Y.-H. Huang, Chemosphere, 2014, 111, 232–237.
31. J.-Y. Lin, Y.-J. Shih, P.-Y. Chen and Y.-H. Huang, Appl. Energy, 2015, 164, 1052–1058.
32. Y. S. Sadovskii, T. Solomoichenko, T. Prokop’eva, Z. P. Piskunova, N. Razumova, B. Panchenko and A. Popov, Theor. Exp. Chem., 2012, 48, 163–171.
33. M. C. Durrant, D. M. Davies and M. E. Deary, Org. Biomol. Chem., 2011, 9, 7249–7254.
34. M. E. Deary, M. C. Durrant and D. M. Davies, Org. Biomol. Chem., 2013, 11, 309–317.
35. B. N. Chernyshov, A. Y. Vaseva and O. V. Brovkina, Russ. J. Inorg. Chem., 1991, 36, 460–463.
36. K. Onuma and Y. Sugiura, J. Biotechnol. Biomater., 2015, 5, 2.
37. Z. Qiang, J.-H. Chang and C.-P. Huang, Water Res., 2002, 36, 85–94.
38. J. Flanagan, W. P. Griffith, R. D. Powell and A. P. West, Spectrochim. Acta, Part A, 1989, 45, 951–955.
39. J. Flanagan, W. P. Griffith, R. D. Powell and A. P. West, J. Chem. Soc., Dalton Trans., 1989, 1651–1655.
40. W. Stumm and J. J. Morgan, Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters, John Wiley, New York, 3rd edn, 1996.
41. F. J. Millero, P. J. Milne and V. L. Thurmond, Geochim. Cosmochim. Acta, 1984, 48, 1141–1143.
42. P. Pasierb, S. Komornicki, M. Rokita and M. Rȩkas, J. Mol. Struct., 2001, 596, 151–156.
43. Y. Song, Y. Dai, Q. Hu, X. Yu and F. Qian, Chemosphere, 2014, 101, 41–48.
44. T. Itakura, R. Sasai and H. Itoh, Water Res., 2005, 39, 2543–2548.

---