Principle of iron removal by contact oxidation using iron-based active filter membranes

Thread Content

Principle of iron removal by iron-based active filter membrane contact oxidation I. Introduction In China’s groundwater iron removal technologies, the aeration contact oxidation method is widely used for iron removal. The aeration-contact oxygen-based iron removal method involves allowing ferrous ions in aerated groundwater to enter the contact filter layer together with dissolved oxygen, where oxidation of these ferrous ions takes place under the catalytic action of the filter layer, along with their retention. Natural manganese sand iron removal is a contact oxidation method for iron removal that has been widely used in China ； The methods of removing iron using artificial rust sand and naturally formed rust sand are another type of contact oxidation-based iron removal technique that was successfully tested in our country in the 1970s. 　　In the past, the author conducted systematic experiments and research on the iron removal method using natural manganese sand. In recent years, both domestically and internationally, research has been conducted on the iron removal processes of artificial rust sand using quartz sand as a carrier, as well as naturally formed rust sand. These research findings have advanced the contact oxidation iron removal process, improved its efficiency, and promoted its widespread use. 　　There has been a developmental process in people's understanding of the mechanism of iron removal through contact oxidation. Since the 1930s, when pyrolusite sand was begun to be used as a contact oxidation iron-removal media for groundwater, manganese dioxide has been regarded as a catalyst – this is known as the classical theory. As early as the early 1960s, while studying the use of natural manganese sand for iron removal, the author discovered the contact catalytic effect of the \"active filter membrane.\" This was confirmed through numerous model tests and practical experiments, and in 1974 the principle of iron removal via contact oxidation using an active filter membrane was formally proposed, further deepening our understanding of this phenomenon. 　　In recent years, the author has conducted further research on the basic characteristics of iron removal by contact oxidation using iron-based active filter membranes. Experiments show that the new filter media possess a certain iron-removal capacity at the beginning, but this capacity does not last; after some time, it begins to decline. The iron concentration in the water after filtration increases accordingly ； As operating time increases, the iron removal capacity of the filter media gradually improves, the quality of the water after filtration gets better, and ultimately the filter media acquire a stable iron removal capacity. It ultimately possesses stable iron-removal capability. The filter media that ultimately possesses a stable iron-removal capacity is referred to as \"mature\" filter media ； The process of transformation from new filter media to \"mature\" filter media is known as the \"maturation\" process of the filter media. In fact, the maturation process of the filter media is precisely the process of the formation and accumulation of an active iron-based film on its surface. This paper will explore issues such as the iron removal capability of the new filter media, the formation and accumulation process of the active filter membrane, as well as the iron removal characteristics of the active filter membrane in the mature filter layer. 　　II. Iron removal by the new filter media When anaerobic groundwater containing iron, which had not been aerationed, was filtered through the new filter media layer, it was found that the layer initially possessed a certain capacity to remove ferrous ions. Figure 1 shows the removal of ferrous ions from water using new natural manganese sand. The removal of ferrous ions by new quartz sand or anthracite is similar to that of natural manganese sand. The new filter media can remove iron under anaerobic conditions, indicating that it has an adsorption effect on ferrous ions in water. 　　The adsorption capacity of new filter media for ferrous ions in water depends on the type of filter media. Table 1 shows the dynamic adsorption capacity of several new filter media for ferrous ions in water under anaerobic conditions. As can be seen from Table 1, Masan manganese sand has the highest adsorption capacity, while quartz ore sand has the lowest. 　　　　　　　　　　　　　 Dynamic adsorption capacity of new filter media for ferrous ions Table 1 Names of filter media, Particle size of filter media in mm, Iron concentration in water in mg/l, pH of water, Water temperature (°C), Adsorption capacity in mg/l: Mashan manganese sand: 1.0–1.25, 14–18, 6.1, 6, 5000; Jinxi manganese sand: 1.0–1.25, 14–18, 6.1, 6, 1000; Yangquan anthracite: 1.0–1.25, 14–18, 6.1, 6, 250; Heilongjiang bituminous coal: 1.0–1.25, 14–18, 6.1, 6, 250; Songhua River sand: 1.0–1.25, 14–18, 6.1, 6, 250; Beidaihe quartz ore sand: 1.0–1.25, 14–18, 6.1, 6, 24. Experiments show that ferrous ions adsorbed on the surface of the new filter media can be oxidized to ferric ions in the presence of dissolved oxygen. However, the ferric hydroxide formed on the surface of the new filter media differs significantly in properties from the iron-based filter membrane with strong catalytic activity formed on the surface of the mature filter media. First, the ferric hydroxide formed on the surface of the new filter media has a very dense structure. Comparative tests between the new filter layer and the mature filter layer showed that when both layers retained the same amount of iron (2 g in one test), the hydraulic impedance of the mature filter layer was 40 times higher than that of the new filter layer. Therefore, the ferric hydroxide formed on the surface of the new filter media is much denser than the active filter membrane on the surface of the mature filter media. Secondly, the ferric hydroxide formed on the surface of the new filter media does not possess strong contact catalytic activity. Figure 2 shows the comparative tests on the maturation process of the three new filter media. As can be seen from the graph, due to the certain adsorption capacity of the new filter media, there is a certain iron removal effect during the initial stages of filtration. However, as their adsorption capacity is gradually exhausted, the iron concentration in the water after filtration continues to rise. As the iron removal process by filtration proceeds, an iron-based filter film with contact catalytic activity begins to form on the surface of the filter media. Due to the accumulation of this active filter film material on the surface of the filter media, the latter gradually matures. The iron concentration in the water exiting the filter layer begins to decrease again, thus exhibiting a peak-like pattern. Tests showed that although the adsorption capacities of these three new filter media vary greatly, their maturation periods are essentially the same. If the ferric hydroxide formed on the surface of the new filter media possesses contact catalytic activity. Therefore, new filter media with a high adsorption capacity retain more iron, and should be able to mature more quickly, that is, they should have a shorter maturation period; however, this is not the case in reality. Therefore, the ferric hydroxide formed on the surface of the new filter media does not possess strong contact catalytic activity; it is different in nature from the iron-based filter membrane substances on the surface of mature filter media, which do have strong contact catalytic activity. 　III. Maturation process of the filter media After iron-containing groundwater is aerated to increase its oxygen content, it is filtered through a layer of new filter media. Due to the adsorption capacity of this new filter media, it possesses a certain ability to remove iron. Meanwhile, a catalytically active iron filter film began to form on the surface of the filter media. Therefore, during its maturation process, the new filter media possesses both adsorption-based iron removal and contact oxidation-based iron removal functions. In the initial stage of filtration with the new filter media, the iron removal effect through contact oxidation is minimal; therefore, adsorption is the primary method for removing iron. As the adsorption capacity of the filter media decreases, the iron removal efficiency drops, and the iron concentration in the water flowing out of the filter layer gradually increases. On the other hand, the iron-removal capacity of the active filter membrane formed on the surface of the filter media continues to increase. When the rate of increase in this iron-removal capacity exceeds the rate of decrease in the adsorption-based iron-removal capacity, the iron concentration in the water flowing out of the filter layer begins to decline. Since the contact oxidation iron removal process using active filter membranes is an autocatalytic process, the increase in the iron removal capacity of the filter membranes occurs in an accelerated manner, resulting in a curve showing a slight upward curvature after the peak in the iron concentration of the water emerging from the filter layer, until this concentration drops to the required level. Thereafter, the iron concentration in the water exiting the filter layer stabilized at a very low level, indicating that the filter media had reached maturity. In this way, the maturation process of the filter media can be divided into three stages. In the first stage, adsorption-based iron removal by the new filter media plays a dominant role, and this stage is referred to as the adsorption phase ； The second stage is characterized by the dominant catalytic iron removal effect of the iron-active filter membrane, as well as an accelerated progression, and is referred to as the accelerated catalytic phase ； The third stage is characterized by the stable catalytic iron removal action of the iron-active filter membrane, known as the stable catalytic phase, as shown in Figure 3. The stable catalytic iron removal process continues for a considerable period of time, until the filter media is finally fully matured. The surface of fully matured filter media is covered by an iron-based active filter membrane, causing it to turn yellow; hence it is often referred to as rust sand. 　　Due to the different adsorption capacities of the filter media, their maturation processes also vary ； Filter media with low adsorption capacity have a shorter adsorption phase, and the peak value of the concentration profile of the water exiting the filter layer is also higher ； Filter media with high adsorption capacity have a longer adsorption phase, and the peak value of the effluent is also lower. When the adsorption capacity of the filter media is high and the iron concentration in groundwater is low, the peak value of the effluent concentration may drop below the requirements set by water quality standards; in such cases, the filter can supply water of qualified quality as soon as it is put into operation. 　　Under the conditions shown in Figure 2, we also conducted maturation tests on filter media such as Beidaihe quartz sand, Songhua River river sand, and Heilongjiang bituminous coal; the test results were generally consistent with those in Figure 2. The total length of the adsorption and accelerated catalytic sections for these six types of filter media is approximately 4–5 days; during this time, the iron concentration in the water flowing out of the filter layers can be reduced to below 0.3 mg/l. However, the quality of the water output is not yet stable, but after 7 days, iron removal becomes stable. 　　In summary, different types of filter media only affect the water quality at the initial stage of iron removal; they basically have no impact on the maturation period of the filter media or their ability to remove iron once they are mature. In other words, for mature filter media, different types of filter media serve the same purpose as carriers for the active iron-removing membrane. This provides a theoretical basis for using inexpensive filter materials such as quartz sand, river sand, and anthracite in contact oxidation processes for iron removal, which holds great economic significance. However, filter media with high adsorption capacity, such as natural manganese sand, yield better water quality at the initial stage of iron removal, which is of great practical significance. Filter media with low adsorption capacity, such as quartz sand and anthracite, result in poor water quality at the beginning of operation; measures must be taken to improve the water quality and accelerate the maturation of these filter media, which is one of their drawbacks. 　　Some use the iron adhesion index on the surface of the filter media (the amount of iron in mg adhering to the surface of 100 mg of filter media) as an indicator of the maturity of the filter media. As mentioned earlier, since different filter media have varying adsorption capacities, the oxidized iron adsorbed on the surface of the filter media does not possess catalytic activity. Filter media with high adsorption capacity achieve a considerable value for the attachment index at the initial stage of iron removal, but at this point the filter media do not yet possess the corresponding level of \"maturity\". Therefore, using the attachment index as an indicator of filter media maturity is not universally applicable to filter media with different adsorption capacities. 　　People* are accustomed to using the reduction of iron concentration in the water emerging from the iron-removing filter layer to below the drinking water quality standard (0.3 mg/l) as an indicator that the filter media is mature. Since the filter layers operate under certain conditions, \"maturity\" is related to specific operating conditions and does not have a uniform standard, making it difficult to compare them with one another; hence it is also imperfect. 　　We believe that using the contact oxidation reaction rate constant per unit filter media surface area, or the contact catalytic activity coefficient of the filter layer, as an indicator of filter media maturity is reasonable. IV. Chemical composition of the iron-based active filter membrane and its basic catalytic characteristics During the removal of ferrous ions, an iron-based active filter membrane gradually forms on the surface of the filter media. During a filtration cycle, if the amount of filter membrane adhering to the surface of the filter media is greater than the amount that is removed during backwashing, the iron content on the surface of the filter media increases, which causes the filter media particles to gradually grow larger. In water treatment plants designed to remove iron from groundwater with high iron concentrations, significant thickening of the filter layer and granulation can be observed. In some such plants, after one year of use, the particle size of certain filter media can increase from 0.6–2.0 mm to 5–6 mm, resulting in a several-fold to dozens-of-fold increase in volume, forming what are known as rust balls. When wet, this rust ball is brownish-yellow, with a layer of loose iron hydroxide (filter film) on its surface. After washing the filter membrane, the surface of the rust ball is smooth and possesses a certain degree of strength. When the rust ball is cut open, its interior is brown and black in color, arranged in concentric rings, and is quite dense. Rust balls usually contain a small core made of fine filtering material, but there are also those that have no core and are composed entirely of iron. 　　After roasting the rust balls obtained from the Jiamusi water plant, it was found that they contained 88% Fe2O3 and 8% SiO2; in addition, they also contained various elements such as Ca, Mg, and Mn. The chemical composition of the loose iron filtration film on the outside of the rust ball is the same as that of the rust ball itself. Based on the process of rust ball formation, it can be concluded that the dense material inside was gradually formed over time by the long-term accumulation of this loose iron film on the surface of the filter media. 　　We also conducted differential thermal and thermogravimetric analyses on the fresh filter membranes and the contents inside the rust balls, and determined their chemical compositions as shown in Table 2. The sample of the fresh filter membrane was the iron sludge precipitated from the backwash water of the production filter (stored for one day before testing). The data from the literature are also listed in the table. As can be seen from Table 2, although the iron filter membrane and the substances inside the rust balls have the same chemical composition, there are significant differences in their chemical makeup. By comparison, it can be seen that the process by which substances within the rust balls formed by the accumulation of iron films on the surface of the filter media is one in which crystal water gradually dissociates, and visually, the structure changes from loose to compact. 　　To understand the difference in catalytic activity between the filter membrane and the substances inside the rust balls. The following comparative tests were conducted. One filter tube was filled with a rust ball equipped with a fresh filter membrane as the filtering material, while another filter tube was filled with a rust ball from which the filter membrane had been removed as the filtering material, so that they could undergo iron removal tests under the same conditions. Chemical composition of the iron-based active filter membrane Table 2 Sample name Chemical composition Fresh filter membrane Fe2O3·5H2O or Fe(OH)3·H2O Substance inside the rust balls Fe2O3·H2O or FeOOH Fresh filter membrane Fe2O3·6H2O or Fe(OH)3·2H2O Figure 4 shows the test results. As can be seen from the figure, there are rust balls with fresh filters, showing good iron reduction efficiency. The rust balls removed from the filter membrane have a poor iron-removal effect and exhibit the same properties as the new filter media, which indicates that only the loose material on the surface of the rust balls possesses catalytic activity, while the dense material inside the rust balls does not have such activity. This catalytically active, porous iron filter membrane on the surface of the filter media is known as an iron active filter membrane. 　　The iron concentration in groundwater is 14 mg/l ； Dissolved oxygen concentration: 7–8 mg/l ； Filter velocity: 10 m/h. 　　Experiments show that fresh iron-based active filters have the highest catalytic activity; as time passes, the iron-based filters gradually age, and their catalytic activity declines accordingly. The experiment was conducted using mature filter media, and the results are shown in Figure 5. As can be seen from the graph, after a few days of being out of use, the iron removal efficiency of the mature filter media has **decreased, indicating that the iron-containing filter membrane loses its catalytic activity over time as it ages. The dense material inside the rust ball is formed by the long-term accumulation of an aged iron filter membrane. Therefore, the catalytic effect of the iron-active filter membrane on the surface of the filter media can only be achieved through a continuous iron removal process. The iron-based active filter membrane on the surface of the filter media is continuously replenished during the iron removal process, with new layers of membrane being formed on top of the existing ones; this ensures that the filter membrane remains fresh and possesses high catalytic activity. The old filter membranes gradually age and lose their catalytic activity, eventually becoming dense deposits on the surface of the filter media. The continuous renewal of the iron-based active filter membrane on the surface of the filter media is a necessary condition for the proper progress of the rust sand contact oxidation iron removal process. 　　It has been understood that in the process of removing iron through contact oxidation using an iron-based active filter membrane, the membrane first carries out ion exchange to adsorb ferrous ions from water, as can be expressed as follows: Fe(OH)3·2H2O + Fe2+ → Fe(OH)2(Ofe)·2H2O + H+. When there is dissolved oxygen in the water, the adsorbed ferrous ions are rapidly hydrolyzed and oxidized under the catalysis of the active filter membrane, thereby regenerating the catalyst. The products of this reaction then act as catalysts in further reactions; hence, the removal of iron through contact oxidation using an iron-based active filter membrane is an autocatalytic process. 　　Fe(OH)2(Ofe) ·2H2O + 1/4·O2 + 9/2 ·H2O = 2Fe(OH)3·2H2O + H+. Iron sludge collected from the backwash water was analyzed and found to contain virtually no ferrous compounds. It shows that ferrous ions adsorbed by the active filter membrane can be rapidly oxidized to ferric ions. 　　Based on the concept that iron removal through contact oxidation using an active iron-based filter membrane is an autocatalytic process, the iron trapped in the filter layer during the iron removal process, due to its catalytic properties, should enhance the capacity of the filter layer for contact oxidation-based iron removal. That’s indeed the case. Figure 6 shows the variation in the iron concentration in water along the depth of the filter layer during the iron removal process. Curve 1 shows the concentration distribution 1 hour after backwashing of the filter layer, while Curve 2 shows the situation 36 hours after backwashing. As can be seen from the graph, curve 2 is located higher than curve 1, indicating that as iron accumulates in the filter layer, its contact oxidation capacity for removing iron improves significantly. This confirms the conclusion that contact oxidation using an iron-active filter membrane is an autocatalytic process. 　　V. Iron removal rate by contact oxidation in a matured filter layer Ferrous ions in water are removed in a matured filter layer through the following steps: ferrous ions diffuse from the water to the surface of the filter media ； Ferrous ions are adsorbed by the active filter membrane on the surface of the filter media ； The adsorbed ferrous ions are hydrolyzed and oxidized to form ferric hydroxide – the active iron filter membrane. Among the aforementioned steps, the one with the slowest reaction rate will become the controlling step for the iron removal rate. Experiments show that the diffusion of ferrous ions to the surface of the filter media may be the controlling factor for the iron removal rate. The experiments also showed that the active filter membrane on the filter media only adsorbs ferrous ions from water on its outer surface. 　　According to Fick’s law, as ferrous ions diffuse toward the surface of the filter membrane, the diffusion rate is proportional to the concentration difference (C–C’) of ferrous ions between the water and the surface of the filter membrane, and inversely proportional to the thickness of the boundary layer at the surface of the filter membrane, σ. If the diffusion rate is taken as the iron removal rate, and assuming that C’ is very small and can be ignored, then 　　-dc/dt=DS/D(C-C’)≈DS/σ·C　　　　　　　　(1) 　　Where t is time, with t=ml/u ； 　　　　　l —— Thickness of the filter layer ； 　　　　　m——pore porosity of the filter layer ； 　　　　　u——filtration rate ； 　　　　　D——diffusion coefficient ； 　　　　　S — the external surface area of the filter membrane per unit volume of the filter layer, S=6a(1-m)/d ； 　　　　　d——filter media particle size ； 　　　　　a —— Shape coefficient of the filter media ； 　　　　　σ——boundary layer thickness ； 　　　　　C’ — the concentration of ferrous ions on the surface of the filter membrane. 　　Substituting the various parameters listed above into equation (1) yields:
 -dc/dι=βC　　　　　　　　　　　　　　　(2)
 β=6Dam(1-m)/ σdu　　　　　　　　　　　(3)
In these equations, β is referred to as the contact catalytic activity coefficient of the filter layer. 　　When water flows in a laminar flow state through the filter layer, the thickness of the boundary layer can be considered constant (σ=const). As shown in equation (3), the catalytic activity coefficient of the filter layer is inversely proportional to the first power of the filtration rate at this time. 　　When water flows in a turbulent state through the filter layer, it can be approximated that the thickness of the boundary layer is inversely proportional to the filtration velocity: σ=a/u　　　　　　　　　　　　　　　　　(4) Where a is the proportionality constant. By substituting equation (4) into equation (3), we obtain β=6Dam(1-m)/ad　　　　　　　　　　　　　(5). In other words, in turbulent flow, the iron removal efficiency is independent of the filtration rate; this can be considered to be inversely proportional to the zeroth power of the filtration rate. 　　When the water level in the filter layer is below the transition zone between laminar and turbulent flow, it can be assumed that the catalytic activity coefficient of the filter layer is inversely proportional to the p-th power of the filtration rate: β=6Dam(1-m)/bdup  (6) Where b is the proportionality constant ； And 0