Preparation, characterization and coagulation behaviour of polyferric magnesium silicate (PFMSi) coagulant

Coagulation process plays an important role in water or wastewater treatment facilities (Moussas & Zouboulis 2008). When coagulant is added to water, the suspended solids and colloidal particles undergo a process of destabilization, collision, aggregation, and sedimentation, and are finally removed (Jiang & Lloyd 2002). More importantly, part of the natural organic matter, even some toxic and hazardous pollutants, can be removed (Cheng & Chi 2002; Cheng et al. 2008; Mikola et al. 2013; Yang et al. 2013).

The coagulants could be classified into inorganic and organic coagulants (Sun et al. 2012), and the former is more used in treatment facilities. Nowadays, the major types of inorganic coagulants are the metal salts. The common coagulants of metal salts are based on aluminium and iron, i.e., AlCl₃, Al₂(SO₄)₃, FeCl₃ and FeSO₄, etc., which can be hydrolyzed in water to form positively charged irons and complexes. The inorganic polymer coagulants such as polyferric sulphate (PFS) and polyaluminum chloride (PAC) are pre-hydrolyzed coagulants, their coagulant performance being improved by the OH bridges and large number of inorganic macromolecular compounds with a high positive charge. However, given the unsatisfactory humic substances removal efficiency and the residual color for conventional coagulants, the composite coagulant has become the research focus in recent years.

The composite coagulant is prepared by combining a metal or inorganic ion, and the aim is to overcome the problems of the individual coagulants, such as small flocs, the neurotoxic effects associated with high residual Al concentration and the high residual color associated with high residual Fe concentration, etc. Previous study also confirmed that flocculating activity and floc properties are significantly improved by introducing various ions in PFS to produce a new type of composite coagulant (Sun et al. 2010; Zhu et al. 2011).

The iron-based composite coagulants possess the advantages of non-toxic, larger flocs and superior coagulation performance (Jiang & Graham 1998). On the other hand, magnesium salts are regarded as an efficient class of coagulant in the removal of color. Additionally, a number of researchers have revealed that enhanced removal of impurities or pollutants has been observed in the presence of magnesium. For instance, a good coagulation performance could be achieved if enough magnesium ions are present in the system of lime treatment (Liao & Randtke 1986; Folkman & Wachs 1973). More importantly, the magnesium can be removed from the precipitated sludge and recycled through the process (Gao et al. 2007), indicating a great reduction in chemical costs.

Moreover inorganic polymeric coagulants, which possess the advantage of higher efficiency than the conventional coagulants, but still much lower efficiency compared to organic polymeric coagulants, have started to be used in treatment facilities (Tang & Shi 2002). As another type of inorganic polymer coagulant, the polysilicate complex coagulant based on polysilicic acid (PS) and metal salts has been developed since the 1990s (Hasegawa et al. 1991). This is a complex of a positively charged metal coagulant and a negatively charged PS flocculant (Fu et al. 2009). The combining of PS with metal ions can increase the molecular weight of the coagulants and thereby, promote the aggregation capacity. On the other hand, the amount of residual metal ions is greatly reduced. (Ohno et al. 2004; Fu et al. 2006). There are two conventional methods to prepare metal-polysilicate complex coagulants, i.e., co-polymerization and composite-polymerization (Li et al. 2013a, 2013b). The former, by hydrolyzing after mixing, is relatively more simple and maneuverable than the latter by mixing after hydroxylation (Xu et al. 2009). Moreover, the coagulant prepared by the first method is more effective. For example, the polyferric silicate sulphate prepared by co-polymerization exhibited better performance in removal of turbidity, UV₂₅₄ and phosphate than the composite-polymerization ones (Moussas & Zouboulis 2008). Therefore, the co-polymerization method was selected to prepare the coagulant. Although there are a number of researches focusing on using metal salts and PS to prepare the coagulants, the combination of magnesium, iron and PS is really rare, there has been no report of the synthesis of composite coagulant polyferric magnesium silicate (PFMSi) by co-polymerization.

In this paper, a relatively novel coagulant, PFMSi, was prepared by the method of co-polymerization. In order to investigate the coagulation mechanism, the characterization of the products was studied. The structure and morphology analysis were achieved by obtaining X-ray diffraction (XRD), Fourier transform infrared (FT-IR) spectra and scanning electron microscopy (SEM) microphotographs. Moreover, the destabilization ability of the coagulants and the coagulation kinetics were evaluated by the zeta potential values and the ratio R curves respectively. Finally, coagulation experiments were carried out to determine the effects of molar ratios of Fe/Si, Mg/Si, [OH]/[M] and coagulant dosage on the coagulation performance in terms of removal of turbidity (NTU), UV₂₅₄ (cm⁻¹) and total organic carbon (TOC) (mg/L). Meanwhile, the conventional polymerized and non-polymerized coagulants were also used to make a comparison. For the aforementioned indexes, 1 NTU is defined as 1 mg SiO₂ in 1 liter water, UV₂₅₄ represents the absorbance of some organic matter in water irradiated by ultra violet with A wavelength of 254 nm, TOC can be defined as the content of total carbon of resolvable and suspended organic matter in water.

All reagents, i.e., concentrated sulfuric acid (98 wt%), FeSO₄·7H₂O, MgSO₄·7H₂O, NaClO₃, NaHCO₃, NaOH, Fe₂(SO₄)₃, polymeric aluminium (PAC) and Ferron™ were analytically pure chemicals except for the water glass (industrial grade, ω (SiO₂) = 26%, ρ = (1.36–1.41) × 10³ kg/m³, modulus = 3.1–3.4). Deionized water with conductivity lower than 0.5 μS/cm was used to prepare all the solutions.

The raw water used in this study was collected from the Pearl River in Guangzhou. Table 1 summarizes the physico-chemical parameters of the raw water and the water treated by PFMSi.

With rapid stirring using an electric stirring apparatus, water glass diluted with deionized water was introduced into sulfuric acid (H₂SO₄, 20 wt%), then the fresh PS was obtained after 1 h of polymerization. FeSO₄·7H₂O and MgSO₄·7H₂O were firstly dissolved in deionized water to obtain sulfate solution, and in order to keep a strong acid condition, sulfuric acid (H₂SO₄, 98 wt%) was added. Then the sulfate solution was mixed with fresh PS, which was stirred with an electric stirrer, and a small amount of NaClO₃ was added subsequently to ensure that all the Fe²⁺ was oxidized to Fe³⁺. Finally, a certain amount of NaHCO₃ solution was added to increase the degree of polymerization, and the product of PFMSi was obtained after 3 h of stirring.

To obtain a more efficient coagulant, the factors, Fe/Si (the molar ratio of Fe and Si), Mg/Si (the molar ratio of Mg and silicon) and OH/M (the basicity) were studied.

The liquid products were freeze-dried for 48 h by a vacuum freeze drier (Virtis, USA) and then placed into a desiccator to cool for another 2 h. The solid PFMSi was ground using a mortar and pestle for further characterization studies.

The crystal characteristics of solid PFMSi were analyzed by Empyrean X-ray diffractometer (PANalytical B.V., The Netherlands) with Cu K radiation in the range of 5–65 ° 2θ at a scan rate of 1 °/min. The surface morphology of the coagulants was observed and analyzed by SEM micro-photographs using a Merlin scanning microscope (Zeiss Company, Germany). Samples of PFMSi were characterized by FT-IR spectroscopy for the determination of structure characteristics using a CCR-1 Spectrophotometer (Thermo-Nicolet, USA) and the spectra were recorded in the range of 4,000–400 cm⁻¹.

In this study, Fe–Ferron™ complexation timed spectrophotometry was used to measure the Fe(III) species distribution with a UV-5800 spectrophotometer (METTASH, China); the selective wavelength is 600 nm. In general, the Fe(III) species that reacted rapidly with Ferron™ was considered to be Fe(III) monomeric species (Fe_a), over the next 2 h and later, the reacted Fe(III) species were considered to be polymeric species (Fe_b) as well as colloidal and precipitated species (Fe_c) respectively (Song et al. 1997).

The PFMS coagulant used was diluted with deionized water to a concentration of 2 g/L, as Fe.

The jar test was carried out in 1.0 L Plexiglas® beakers using a six paddle stirrer (ZR4–6, China). The same batch of raw water was dosed with PFMSi coagulants of 2 mg/L, 4 mg/L, 6 mg/L, 8 mg/L, 10 mg/L, 12 mg/L, as Fe, respectively. The experimental conditions were the following: the fast stirring time was set to 2 min at a speed of 200 rpm to allow the particles to be destabilized. At this point, a sample was collected for zeta potential measurement. Then a flocculation period (10 min at 50 rpm) and final sedimentation (30 min) followed. After sedimentation, a water sample was taken 3 cm below the surface for further analysis.

Zeta potential measurements were performed on a Nano-zs90 analyzer (Malvern, UK). Turbidity was measured with a 2100Q turbidimeter (HACH Company, USA). UV₂₅₄ was measured using a standard method by a UV-5800 spectrophotometer (METTASH, China) at a wavelength of 254 nm with a 1 cm quartz cell. The water samples were first filtered through a 0.45 μm membrane filter to remove turbidity to avoid interfering with the measurement, and the TOC measurement was carried out using a Liqui TOC analyzer (Elementar, Germany). Besides, the pH value was measured with a PHS-3C meter (Rex Company, China).

An intelligent photometric dispersion analyzer (i-PDA100, EcoNovel Company, South Korea) was used to monitor the kinetics of coagulation and the extent of aggregation. The i-PDA instrument measures the ratio R, also called flocculation index (FI), defined by the dimensionless value of ⁠, where n is the particle number concentration, L is the optical path length of the fluorescent cell, A is the effective cross area of the light beam and C is the scattering cross section of the particles, which is directly related to the mean concentration and size of flocs. In that case, the R values can be effectively utilized to describe the aggregation process of the particles. In addition, the coagulation process consisted of 2 minutes of fast stirring (200 rpm) and 10 minutes of slow stirring (50 rpm) for floc formation.

In order to investigate the effects of PFMSi dosage and Fe/Si molar ratio on the coagulant performance, a series of coagulation experiments were conducted at the pre-determined factor level. The parameters, residual turbidity and removal of humic acid (expressed as UV₂₅₄ absorbance), are measured to evaluate the coagulation efficiency of PFMSi in treating the raw water.

It is known that the charge neutralization is generally considered to be the coagulation mechanism of hydrolyzing metal salts; however, Figure 6(c) shows that the optimal values of turbidity removal efficiencies are not always close to the isoelectric point, indicating a function of adsorption-bridging (Sun et al. 2010). Additionally, the zeta potential exhibits a rising trend with increasing Fe/Si molar ratio and coagulant dosage. This can be attributed to the fact that the higher Fe/Si molar ratio causes the result of stronger ionic strength, which enhances the function of charge-neutralization, leading to higher zeta potential values.

However, differently from PFMSi with various Fe/Si molar ratios except for 2 mg/L PFMSi, the removal efficiency of turbidity exhibits no significant difference with the increasing ratio of Mg/Si, while the removal of UV₂₅₄ presents a similar trend to that of the Fe/Si molar ratio. The reason might be that the magnesium ions and hydrolysis products of PFMSi possess the strong ability to adsorb suspended solids and colloidal particles, which reflects the high removal rate of turbidity at low dosage.

It is worth mentioning that in spite of the higher zeta potential values due to the higher molar ratio of Fe/Si or Mg/Si, the removal efficiencies of turbidity, UV₂₅₄ decrease slightly with the increasing Mg/Si molar ratio, further confirming that the introduction of Fe or Mg could improve the process of charge-neutralization; nevertheless, adsorption-bridging still dominates the whole coagulation process.

The study of the zeta potential verifies that from another perspective, as it decreases gradually with the increasing of r at the same coagulant dosage. The higher r values increase the process of adsorption-bridging, while weakening the charge-neutralization process, thereby the zeta potential decreases, as is shown in Figure 8(c).

According to the coagulation performance and zeta potential values, the introduction of silicon has a great effect on the efficiency of the coagulant, i.e. increasing the proportion of polymeric species, and relatively low M/Si (M represents the metal) leads to a higher degree of polymerization. This can be explained by the fact that PS contains a series of silica species, such as dimeric (Si₂O₃(OH)₄⁴⁻), trimeric (Si₃O₅(OH)₅³⁻), tetrameric (Si₄O₈(OH)₄⁴⁻) and the polymeric species (Dietzel 2000; Wang & Tang 2001). For example, a composite coagulant poly-zinc-silicate-sulfate (PZSS) was prepared by introducing PS to zinc salt, and the PZSS showed better coagulation performance than that of conventional ferric and aluminum coagulants in oily wastewater (Zeng & Park 2009). The poly-magnesium-silicate-chloride prepared by combining magnesium chloride and PS presented more efficient coagulation performance in dyeing wastewater treatment than conventional magnesium chloride (Wei et al. 2015).

Therefore, the introduction of silicon increases the polymeric species and molecular weight of the coagulant, and the major mechanism is adsorption-bridging. In that case, the flocs produced are bigger and take less time to precipitate. In contrast, conventional coagulants are predominant in the mechanism of charge-neutralization, while the other two abilities are relatively weaker. As seen from Figure 9(a), when the coagulant dosage is 6 mg/L as Fe, the residual turbidity decreases to about 0.5 NTU using PFMSi, slightly higher than the standard of tap water. This means that it can reduce the load of filtration to a great extent when used in practical projects. Moreover, Figure 9(b) shows that PFMSi exhibits absolute superiority in the respect of UV₂₅₄ removal. Meanwhile, the PFMSi coagulant exhibited a better performance than others in the removal of TOC, which can be observed in Figure 9(c).

As shown in Figure 10, all the coagulants exhibit a similar trend. With the addition of coagulants, the FI R increases rapidly, indicating that the growth and formation of flocs, and then reaches the steady phase. It is evident that the PFMSi shows quicker formation of flocs, as the linear growth phase is shorter than FMS and PAC in duration. For the conventional coagulants FMS and PAC, the formation of flocs is relatively retarded and the produced flocs are smaller, as the R values are lower.

In general, for conventional metal salt coagulants with low molecular weight, the predominant mechanism of particle destabilization and floc formation is charge-neutralization. When introducing the silicon to metal salts under appropriate conditions, hydrated irons and polynuclear hydroxyl complexes are formed, which bond with PS to subsequently form new structures of chains or networks. In that case, the molecular weight of the coagulant is greatly increased, meaning that the presence of PS facilitates the formation of flocs through the adsorption-bridging mechanism, and the coagulation process is enhanced simultaneously.

XRD and FT-IR results suggest that PFMSi coagulant consists of new chemical species, containing the metals Fe and Mg and the nonmetal Si. According to the morphology, the solid PFMSi presents an amorphous phase. The Fe(III) species of PFMSi coagulant is analyzed by Fe–Ferron™ complexation timed spectrophotometry, and the results show that the higher basicity is conducive to the formation of Fe_b and Fe_c. Fe_b plays the most important role in the coagulation process, the optimal r is 0.4. The factors of Fe/Si, Mg/Si and [OH]/[M] greatly influence the coagulation performance, which is reflected in the removal of turbidity and UV₂₅₄. The results indicate that a higher percentage of Si and appropriate basicities can improve the polymerization degree and increase molecular weight, and Mg²⁺ and hydrolysis products that contain magnesium possess the strong ability to adsorb suspended solids and colloidal particles. Coagulation experiments show that the PFMSi coagulant works better than conventional coagulants of PAC and FMS, regardless of the removal of turbidity or the removal of organic matter. The zeta potential values suggest that the adsorption-bridging function of PFMSi coagulant is more prominent than in conventional coagulants. I-PDA analysis reveals that the introduction of silicon leads to the formation of large polymers with high molecular weight, thereby the flocs are bigger in size than the flocs produced by conventional coagulants, which reflects on the R curves, indicating the existence of adsorption-bridging process.

The research was financially supported by grants from National Nature Science Foundation of China (21677052), Major Science and Technology Program for the Industry-Academia-Research Collaborative Innovation (201605122301117, 201604010043), Guangdong Province Science and Technology Project (2016B090918104, 2013B090200016, 2015B020215007, 2015B020235009, 2016B020240005), State Key Laboratory of Pulp and Paper Engineering (2016C03), Joint fund of Guangdong Province (U1401235), and Zhanjiang of Guangdong Energy Co. (ZY-KJ-YX-2016X085F).