Environment
pH Dependance of Scorodite Formation in As(V) Solution Using Magnetite as the Solid Iron Source
2022 Volume 63 Issue 9 Pages 1287-1293
Details
2022 Volume 63 Issue 9 Pages 1287-1293
Arsenic-containing wastewater, which is mainly generated in nonferrous metal-smelting plants, threatens the natural environment and human health owing to its high toxicity and mobility. A promising arsenic immobilization method to solve this problem is to immobilize arsenic as scorodite. In this study, the method of scorodite synthesis using magnetite as the iron source was investigated, which can be performed using low-cost materials under atmospheric pressure without using an autoclave. The generation and crystallization behaviors of the intermediate of the gel-like precursor of scorodite were investigated by monitoring the concentration of each iron ion in solution and characterizing the precipitate. By clarifying the effect of the solution pH for each reaction step, the mechanism and suitable conditions of this process are discussed.
Arsenic (As) is very harmful element, and it is widely distributed on Earth. Nonferrous metal-smelting plants are one of the main sources of high As-containing substances. Among As-containing substances, including slag, sludge and dust, wastewater especially threatens the natural environment and human health owing to its high toxicity and mobility. Therefore, development of methods to immobilize As in wastewater in a stable form that is suitable for final disposal has been strenuously carried out. Currently, the iron co-precipitation method is widely used for As immobilization. Although this method is favorable because As can be removed to a sufficiently low concentration to meet effluent standards, a very large volume of As-containing sludge is generated, which has a large environmental impact.1) In addition, because As is immobilized in an unstable form of fixed iron compounds, As elution and secondary pollution of the environment are of concern.2) A promising arsenic immobilization method to solve this problem immobilization of As as scorodite (FeAsO4·2H2O). Scorodite is stable in a wide pH range and has a solubility product of less than 10−20 mol−2L2, and it has the lowest bioaccessibility of common As-bearing mine-waste minerals at pH > 4.3,4) Because scorodite has larger removal capacity of As and lower demand for iron than iron co-precipitation, the amount of As-containing sludge can be greatly reduced by using this As-immobilization method. For these reasons, introduction of the scorodite synthesis process is expected, especially at sites where high concentrations of As-bearing wastewater are generated.
Conventionally, synthesis of crystalline scorodite has been investigated by the hydrothermal method with an autoclave.5–9) Although coarse scorodite particles with good crystallinity can be obtained by the hydrothermal method, from an economic point of view, synthesis of scorodite under ambient pressure and at low temperature is more attractive. Demopoulos and co-workers10–14) investigated scorodite synthesis under atmospheric pressure below 100°C by adding seed crystals of scorodite. Fujita et al.15–19) reported synthesis of scorodite with excellent crystallinity at ambient pressure and 95°C as a scorodite synthesis method under low-temperature and atmospheric-pressure conditions. In this method, ferrous sulfate is added to the As-containing solution as the Fe-ion source, and scorodite with good crystallinity is formed by in situ oxidation of Fe2+ with air blowing. In addition to temperature, other conditions, such as the pH, Fe/As ratio and reaction time, have been reported to affect scorodite crystallization. Furthermore, the reaction mechanism has been reported, where crystalline scorodite synthesis proceeds via an amorphous intermediate, which is called the gel-like precursor in this paper, and it contains more Fe(II) than Fe(III).20)
In recent years, scorodite synthesis methods by adding solid iron oxide, such as hematite (Fe2O3),21–24) ferrihydrite (Fe2O3·0.5H2O),25) limonite (FeO(OH)·nH2O),26) and magnetite (Fe3O4),27–30) to solutions containing As ions have attracted much attention. These methods are favorable because they can be performed under atmospheric pressure using low-cost Fe compounds. Among the scorodite synthesis methods using solid iron oxides, the magnetite method is promising because of its multiple advantages, including scorodite synthesis using inexpensive by-products such as steel slag without using additional materials such as aqueous solutions of Fe salt reagents, the good absorbability of As ions, and the possibility of separation from the precipitate by magnetic separation.29) However, the details of the reaction, especially generation and crystallization of the amorphous intermediate of scorodite,14,31,32) have not been elucidated. To clarify the reaction process and the effects of the synthesis conditions, especially for the gel-like precursor that is difficult to locally analyze owing to its small size and unfixed composition, such as the Fe(II)/Fe(III) ratio, it is necessary to monitor the change in the concentration of each species of aqueous Fe ions using solutions prepared with reagents instead of solutions obtained from industrial processes containing As(III) and As(V). In this study, the effect of the pH on the reaction process, including formation and crystallization of the gel-like precursor, were investigated for synthesis of scorodite using magnetite.
Scorodite synthesis was performed by adding magnetite (Fe3O4, <325 mesh, Alfa Aesar, Haverhill, MA, USA) to arsenic As(V) solution prepared with arsenic acid (60 mass% H3AsO4, FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) and ultrapure water in a flask container. A schematic of the reactor for scorodite synthesis is shown in Fig. 1. Ultrapure water with electrical resistivity of more than 18.4 MΩ·cm prepared with a Q-POD Element unit (Merck, KGaA Darmstadt, Germany) was used in the experiments. The volume of the reaction solution was 700 mL, the initial H3AsO4 concentration was 0.67 mol L−1 (i.e., the As(V) concentration was 50 g L−1) and the initial pulp concentration of Fe3O4 was 0.35 mol L−1. Note that the molar ratio of As(V)/Fe(III) contained in Fe3O4 was slightly smaller than 1. During the reaction, the reaction solution was maintained at 95°C with Ar bubbling at a flow rate of 700 mL min−1 to prevent oxidation of the solution by air and agitated at 200 rpm with an impeller to prevent settling at the bottom and ensure uniform dispersion. The pH value of the solution was monitored by a pH meter (GST-5741C, DKK-TOA Corporation, Tokyo, Japan) and adjusted by adding sulfuric acid (H2SO4(aq), 2 mol L−1, FUJIFILM Wako Pure Chemical Corporation). The initial pH value of the solution was adjusted manually, and the pH value during the reaction was adjusted using an automatic titrator (AUT-701, DKK-TOA Corporation, Tokyo, Japan).
Schematic illustration of the reactor for scorodite synthesis.
When a certain time had passed after magnetite addition, the port for the thermocouple was opened and 20 mL of the slurry was sampled with a pipette several times and separated by vacuum filtration into the precipitate and the solution. The precipitate was vacuum freeze-dried using a vacuum freeze-dryer (FD-1000, EYELA Tokyo Physical and Chemical Instruments Co., Ltd., Tokyo, Japan) overnight, and it was then characterized by X-ray diffraction (XRD, D2PHASER, Bruker Corp., Billerica, MA, USA) with Cu Kα radiation (1.542 Å) and observed by scanning electron microscopy (SEM, SU-6600, Hitachi, Tokyo, Japan). The As-ion and total Fe-ion concentrations of the sampled solution were measured by inductively coupled plasma atomic emission spectroscopy (ICP-AES, Spectro Arcos, Spectro Analytical Instruments, Kleve, Germany). The Fe2+ concentration of the sampled solution was determined by the potassium permanganate (KMnO4) titration method with 0.02 mol L−1 KMnO4 solution (FUJIFILM Wako Pure Chemical Corporation), and the Fe3+ concentration was calculated by the difference between the Fe2+ concentration and total Fe concentration.
The overall reaction for scorodite synthesis using magnetite is
| \begin{equation} \text{Fe$_{3}$O$_{4}$} + \text{2H$_{3}$AsO$_{4}$} + \text{2H$^{+}$} \rightarrow \text{2FeAsO$_{4}{\cdot}$2H$_{2}$O} + \text{Fe$^{2+}$} \end{equation} | (1) |
To investigate the effect of the solution pH, synthesis tests were performed with H3AsO4 solutions with the initial pH values adjusted to 0.5 and 1.5 by addition of H2SO4. With and without adding protons (H+) as H2SO4 at the stoichiometry ratio to H3AsO4, the initial solution pH was 0.5 and 1.5, respectively. The change of the solution pH and the concentrations of As5+, Fe2+ and Fe3+ during the reaction from t = 0 to 420 min are shown in Fig. 2. After a certain time, the solution pH sharply increased. The pH increased to 1.5 after t = 75 min and to 3.9 after t = 33 min for initial pH values of 0.5 and 1.5, respectively. The stagnant pH change for a certain period after magnetite addition was not observed in the blank tests of magnetite dissolution, which were performed in H2SO4 solution without H3AsO4 with initial pH values of 0.5 and 1.5 under the same conditions. The dissolution of magnetite started, and the pH increased soon after magnetite addition, as shown in Fig. A1. Note that in this blank test, the unreacted magnetite powder remained within this timescale for both pH conditions. From these results, the time when the pH increase occurs reflects the reaction rate of Fe3O4 and H3AsO4 for each condition. The longer stagnant pH change for lower pH condition, which seemingly contradicts the result of magnetite dissolution that proceeded with higher rate at lower pH condition, is speculated to be because of formation of some type of passivation film by the reaction with H3AsO4 on the magnetite particles. After the pH increase, the As5+ concentrations decreased to 2.0 × 10−3 mol L−1 (i.e., 149 ppm) and 2.1 × 10−4 mol L−1 (i.e., 16 ppm) at t = 420 min for initial pH values of 0.5 and 1.5, respectively. The rate of As5+ removal for an initial pH of 1.5 was likely the result of scorodite synthesis using hematite.22) For an initial pH of 0.5, the Fe2+ concentration monotonically increased. At t = 240 min, the amount of Fe2+ in solution was equal to 70% of Fe(II) included in the added magnetite. In contrast, for an initial pH of 1.5, the Fe2+ concentration hardly increased. The Fe3+ concentration slightly increased at the beginning, but it began to decrease after the pH increase and reached almost zero at t = 240 min.
(a) pH and (b) As5+, Fe2+ and Fe3+ concentrations in the reaction solutions with initial pH values of 0.5 and 1.5 during scorodite synthesis. The reported As5+ removal behavior by the hematite method22) is also shown in (b).
SEM images of the raw magnetite powder and both SEM images and XRD patterns of the precipitates are shown in Fig. 3(a)–(c), respectively. The raw magnetite particles were composed of small particles of sub-micrometer size (diameter of ≤1 µm). For an initial pH of 0.5 at t = 60 min, unreacted fine magnetite particles were observed in the SEM image and the precipitate was identified to be a mixture of magnetite and scorodite from the XRD pattern. At t = 420 min, cauliflower-like faceted crystals were present and the XRD pattern of scorodite was obtained, that is, the precipitate did not contain magnetite. The final Fe2+ concentration was somewhat lower than the expected concentration based on the amount of magnetite addition, as mentioned above. Although this loss of Fe(II) was partially due to extraction of the sampled slurry, it could also be due to contamination of the Fe(II) compound, such as the amorphous hydroxide, in precipitation; however, this was not confirmed in the SEM images.
(a) SEM images of the precipitates in the reaction solutions with initial pH values of 0.5 and 1.5 obtained after reaction for 60 and 420 min. (b) XRD pattern of the each precipitate along with the reference patterns of magnetite (ICDD 01-087-0245) and scorodite (ICDD 00-037-0468).
For an initial pH of 1.5, the precipitates obtained at t = 60 and 420 min were aggregated particles with diameter on the order of 10 µm covered with a fibrous membrane, and the XRD patterns were identified to be mainly scorodite. For scorodite synthesis using hematite as the iron source, synthesis proceeds via an intermediate composed of large particles, which are scorodite crystals covered with a fibrous membrane, like a cocoon.22) The precipitate in this study is thought to be similar to this intermediate product, that is, aggregated scorodite particles covered by a membrane-type precursor containing magnetite particles.
For an initial pH of 1.5, Fe2+ was thought to be incorporated in the precipitate because the Fe2+ concentration did not significantly increase after the As5+ concentration decreased. As mentioned above, in other scorodite synthesis processes, such as the Fe2+ oxidation method19) and hematite method,22) crystal scorodite forms via a gel-like precursor of scorodite, which contains not only Fe(III), but also Fe(II) from X-ray absorption near edge structure analysis. Therefore, the change of the Fe2+ concentration in solution is related to the process of scorodite formation and determined by the balance between incorporation in the generated gel-like precursor and release into the solution accompanied by crystallization. The precipitate obtained for an initial pH of 1.5 was the gel-like precursor containing Fe2+ and remained as aggregated particles without forming faceted scorodite crystals even at t = 420 min. In contrast, for an initial pH of 0.5, the gel-like scorodite precursor was not observed and only faceted crystallized scorodite was obtained. These results suggest that the solution pH affects both the rates of formation and crystallization of the gel-like precursor.
3.2 Synthesis with adjustment of the solution pH value during the reactionTo investigate the effect of the solution pH more clearly, scorodite synthesis using magnetite was performed under continuous pH control. By adding H2SO4(aq), the increase of the pH during the reaction was suppressed and the solution pH was kept at a constant value. The change of the solution pH during synthesis is shown in Fig. 4(a). For the conditions of adjusted pH values of 0.5, 1.0 and 1.5, the solution pH value was adjusted before and after magnetite addition. For the conditions of adjusted pH values of 2.0 and 3.0, to avoid precipitation of jarosite-type hydroxysulfate (MFe3(SO4)2(OH)6, M = H3O+, Na+, K+, Ag+, NH4+, etc.) by base addition, which affects the fixation behavior of As ions,33) the initial pH value was not adjusted and pH adjustment by acid addition was started after the pH increase during the reaction. SEM images of the precipitates at t = 60 and 420 min for each adjusted pH value (0.5, 1, 1.5, 2.0 and 3.0) are shown in Fig. 4(b).
(a) pH of the reaction solutions adjusted to 0.5, 1.0, 1.5, 2.0 and 3.0. (b) SEM images of the precipitates obtained after reaction for 60 and 420 min for each pH condition. (c) As5+, Fe2+ and Fe3+ concentrations in the reaction solutions.
For an adjusted pH value of 3.0, at t = 60 min, the aggregated particles covered with a membrane-type substance remained, like the precipitate obtained for an initial pH of 1.5 without pH adjustment during the reaction. At t = 60 min, for adjusted pH values of 2.0 and 1.5 faceted crystal particles were observed and for adjusted pH values of 1.0 and 0.5, unreacted magnetite particles remained. At t = 420 min, all of the precipitates changed to faceted crystallized particles, including the precipitate covered with a gel-like substance at t = 60 min for an adjusted pH value of 3.0. The XRD patterns of the precipitates (Fig. A2) were similar to those prepared without pH adjustment during synthesis (Fig. 2). At t = 60 min, the precipitates were identified to be a mixture of scorodite and unreacted magnetite, whereas at t = 420 min, the precipitates were composed of only scorodite.
The changes of the concentrations of As5+, Fe2+ and Fe3+ during the reaction are shown in Fig. 4(c). Note that these concentrations were somewhat lowered owing to dilution by H2SO4(aq) addition for pH adjustment, which can be seen by the lower final Fe2+ concentration than that for the result without pH adjustment. The decrease in the As5+ concentration, which reflects formation of the gel-like precursor and scorodite, occurred at the highest rate for an adjusted pH value of 3.0. From the increase in the Fe2+ concentration, the rates of crystallization of the gel-like precursor were highest for adjusted pH values of 1.5 and 2.0, where faceted scorodite crystals formed at t = 60 min. For adjusted pH values of 0.5 and 1.0, the rate of acidic dissolution of magnetite was high, which is indicated by the increase in the Fe3+ concentration in the early stage of the synthesis, and the ratio of the concentrations of Fe3+ and Fe2+ was greater than 2, like the blank test of magnetite dissolution. Although the rates of formation of the gel-like precursor and the decreases in the As5+ concentration were low, faceted scorodite crystals, like those obtained for the other conditions, were present at t = 420 min. The low As5+ concentrations represent the scorodite yields of almost 100% for these conditions.
The experiments performed with different pH conditions revealed that scorodite synthesis using magnetite as the iron source proceeds by the following steps: a gel-like precursor containing Fe(II) and Fe(III) first forms and then the precursor changes to faceted scorodite with release of Fe2+ into the solution. The solution pH is thought to strongly affect the rate of each step, as illustrated by Fig. 5. The first step of formation of the gel-like precursor is advantageous at relatively high solution pH, whereas the second step of crystallization of the gel-like precursor is advantageous at relatively low solution pH. Although the overall reaction of this process is H+ consuming, the orders of the rates of scorodite formation and As5+ removal for each condition did not correspond to the order of the acidity of the solution. This is thought to be due to the difference of the pH dependences of the reaction rates for each step. The results of investigation of the precipitate and As5+ concentration of the solution are summarized at Table 1 for each condition. For synthesis without pH adjustment during the reaction, for an initial pH of 0.5, the rate of As5+ decrease was slow, and for an initial pH of 1.5, faceted scorodite crystals were not obtained. Aggregated particles containing a gel-like substance are not favorable as the form of the As-containing precipitate from the viewpoint of the difficulty in filtration34) and low stability.35,36) With pH adjustment to about 3.0 during the reaction, the As5+ concentration at t = 60 min was relatively low, and a low residual As5+ concentration and formation of faceted scorodite crystals were achieved for an initial pH of 1.5. By using such proper pH adjustment, the possibility of rapid As removal in the form of stable scorodite is suggested for industrial application. Further optimization of this process based on the mechanisms revealed in this study is expected in the future.
Schematic illustration of scorodite formation from magnetite with the pH dependence of the reaction rate for each step.
The effect of the solution pH on scorodite synthesis using magnetite as the Fe source under atmospheric pressure has been investigated. This process is composed of two steps: formation and crystallization of the gel-like precursor. By monitoring the concentrations of Fe2+ and Fe3+, the former step favorably proceeded at relatively high pH (e.g., 3.0 or 4.0) and the latter step favorably proceeded at relatively low pH (<2.0). By proper pH adjustment, rapid As removal with a low residual As5+ concentration of less than 50 ppm and As stabilization in the stable form of faceted scorodite crystals can be achieved.
(a) pH and (b) Fe2+ and Fe3+ concentrations in the reaction solutions with initial pH values of 0.5 and 1.5 during the blank test for magnetite dissolution.
XRD patterns of the precipitates in the reaction solutions with the pH adjusted to 0.5, 1.0, 1.5, 2.0 and 3.0 obtained after reaction for 60 and 420 min along with the reference patterns of magnetite (ICDD 01-087-0245) and scorodite (ICDD 00-037-0468).
This work was supported by JSPS KAKENHI (Grant Number JP19H02753). The SEM observations were performed by SEM-EDX (Hitachi/SU6600 instrument) at the Fundamental Technology Center, Research Institute of Electrical Communication, Tohoku University.
