LITERATURE REVIEW
Pig farming in Vietnam
1.1.1 Pig farming development in Vietnam
According to the Food and Agriculture Organization (FAO), Asia will become the largest producer and consumer of livestock products (Vu, 2019) Therefore, Vietnam's livestock needs to maintain a high growth rate to meet the domestic consumption demand and to serve the export According to the Ministry of Agriculture and Rural Development (MARD), the annual growth rate of the livestock in the period of 2008-2019 was relatively high and stable, representing 5-6% In the draft Strategy for livestock development (Ministry of Industry and Trade, 2019) from 2020 to 2030, the average annual growth rate for the periods of 2020-2025 and 2025-2030 were predicted to be 4-5% and 3-4%, respectively.
Livestock in general and pig production contribute significantly to Vietnam's GDP growth According to the Global Environmental Strategy Institute, agriculture sector accounted for 25% of the country's GDP in 2014 In particular, the pig industry represented 71% of the GDP of the entire agriculture sector (Pham, 2017) The number of pig farming households with a scale of 100 pig heads or more was30,926, accounting for 1.04% of the total pig breeding households throughout the country [1] However, small-scale animal husbandry still dominates due to the conditions in Vietnam Pig production is concentrated mainly in areas Red RiverDelta and Northem Uplands according to the result presented in Figure 1.1.
Figure 1.1.Distribution of pig production in Vietnam by ecological regions
1.1.2 Environmental concerns of anaerobically treated swine wastewater
According to the study of Le (2014), the pig breeding industry develops at a very fast pace but it is mainly spontaneous and has not yet met the technical standards of breeding facilities and breeding techniques Pollution caused by pig breeding includes solid waste, air pollution and water pollution One pig emits 1.5 kg of manure daily and gradually increases with body weight Livestock solid waste contains large amounts of organic matter from manure, uneaten food, straw, etc It contains many pathogenic microorganisms and high nitrogen (N), phosphorus (P) content (National Environment Report 2014) On the other hand, livestock and slaughtering contribute up to 26% of greenhouse gas emissions (GHGs) in the total emissions caused by animals (Tambone, 2015) In addition, odor is also a cause of air pollution in the livestock sector However, livestock wastewater is the most significant source of pollution This is a type of wastewater generated from livestock activities including urine, rinse water, and bathing water for cattle and may contain part or all of animal manure Wastewater accounts for the majority of livestock wastes because 1 kg of livestock solid waste can be mixed with 20 to 49 kg of water (Nguyen, 2011 b ).
Nowadays, there are various technologies for treating livestock wastewater such as The Upflow Sludge Blanket filtration, USBF (Truong, 2010); Stabilization lakes (Nguyen, 2011); Upflow anaerobic sludge blanket, UASB (Rodrigues, 2010), Anaerobic reactor of expanded granular sludge bed - EGSB (Lee, 2012) However, the most common technology in livestock wastewater treatment is anaerobically digestion According to Cu (2012), biogas technology is commonly used to produce electricity and heat in the developing countries Anaerobically digestion have ability to reduce emissions GHGs from manure and generating renewable energy (Mứller, 2004; Sommer, 2004) There are millions of biogas tanks overworld, in which, roughly 3.8 million are located in India, about 60,000 tanks in Bangladesh and about 30 million tanks were built in China (Cu, 2012) In Vietnam, Biogas is considered an appropriate solution to treat wastes contain high concentrations of organic matter and solids such as pig wastewater However, the biogas systems are not the final treatment system to ensure the criteria for safe discharge into the environment (Nguyen, 2012).
According to previous studies, the parameters of wastewater after the anaerobically digestion exceeded the permitted standard many times.
Table 1.1.Treatment efficiency of piggery wastewater by anaerobically treatment in Thua Thien Hue (Nguyen, 2012)
Accordingly, despite the high treatment efficiency of over 70% for many parameters, the quality of water after the anaerobically treatment is still not reached the standard to be discharged into the environment In detail, BOD5 is 6 times higher, COD is 3 times, TKN is 13 times, and TP is 57 times higher than the standard This result has also been supported by other studies (Ho, 2016; Le, 2017).
In addition, the low treatment efficiency of nutrients (N, P) will create a burden on the receiving source Therefore, it is necessary to take additional steps to treat wastewater after anaerobically treatment before discharging into the environment.
Thus, it can be concluded that pigs breeding is an industry that can bring huge economic benefits but also certain environmental risks In particular, the negative effects caused by livestock wastewater are the most significant Therefore it is necessary to study and select appropriate and effective technologies to treat in order to prevent negative impacts from livestock wastewater.
Phosphorus pollution and remedy technologies
1.2.1 Phosphorus significance and environmental concern
On the one side, P is an important element in all known life forms Inorganic P in the form PO43- plays an important role in biological molecules such as DNA and RNA Living cells also use P to transport cellular energy through ATP Almost process in the cell that uses energy has P in its ATP form Despite being the 11 th most abundant element on earth, P in nature exists only in the form of P ore and this is an almost unrecoverable resource (when it takes 10-15 million years to recover) (Do, 2008) In agriculture, P is an important and essential nutrient for plants, P deficiency is one of the causes of crop productivity decline (Ryan, 2012).
Figure 1.3 P is an important and essential nutrient for plants (Pennsylvania's
On the other side, P is one of the causes of water pollution in the presence exceeding concentrations The recognizable manifestation of P pollution are eutrophication leading to algal blooms An excessive increase in nutrients,especially P in surface water, leads to low DO levels, killing fish and aquatic organisms (Iodache, 2014) Various species of algae (such as microsystis) can produce dangerous toxins during their life cycle, which are the causes of fishes and aquatic plants death, destroying ecological balance (Sathasivan, 2009) In addition, the increase of algae also indicates the occurrence of other pathogenic microorganisms such as Pfisteria (Sathasivan, 2008) However, chemicals used to remove algea react with organic matter in the water and create disinfection by-products (DBPs), in which, the most dangerous compound are Trihalomethanes (THMs) and Haloacetic acids because of their carcinogenic potential (Nguyen,
In addition, harmful algal blooms (HABs) are caused by many different types of algae in freshwater areas The most frequent and severe blooms typically are caused by cyanobacteria, the known freshwater algae with the potential for production of toxins potent enough to harm human health (Munn, 2018) Various species of cyanobacteria can use N2 in the air through nitrogen fixation Therefore, under the conditions of low N concentration and high P concentration, these species have normal development and can be produced toxin (Dolman, 2012) Thus, the disproportionate removal of P can lead to the development of toxic algae, which threatens humans and aquatic life.
Figure 1.4.HABs are triggered by nutrient enrichment (Munn, 2018)
In Vietnam, the highest discharge threshold for P in industrial wastewater is 6mg/L (QCVN 40: 2011/BTNMT, column B), this number is 1mg/L in Canada (Guidelines on wastewater quality and wastewater treatment at federal facilities, 1976), and 0.5 mg/L in China (China National Standard, 2006, Level A).
1.2.2 Technologies for phosphorus decontamination of anaerobically treated swine wastewater (ATSWW)
The concentration of P after the anaerobically treatment in pig wastewater exceeds the standard many times Therefore, various technologies have been used to treat P in wastewater after anaerobically treatment such as: stabilization lagoon, trickling filter, AAO Technology.
Stabilization lagoon after anaerobically treatment are popular in rural areas where there is a large land fund P removal in stabilization lagoon is associated with its uptake by algal biomass, precipitation and sedimentation (Kayombo, 2004).
However, wastewater contain high nutrient concentration as water after anaerobically treatment, Stabilization lagoon do not completely eliminate N and P, leading to pollution of wastewater receiving areas In addition, the low control options and more dependant on climatic conditions are disadvantages of this technology (State of Michigan Department of Natural Resources and Environment,
2010) Similarly, trickling filter used for livestock wastewater after anaerobically treatment was presented in Nguyen's research (2016) However, the nutrient removal especially for P treating was limitation Nguyen (2017 b ) reported in her research the possibility of treating contaminants in livestock wastewater after anaerobically treatment using AAO technology combined with coconut fiber media.
Accordingly, TP have significantly decreased but still have not reached the discharge standard In addition, the operation of AAO require high technical controlling, difficulting mud control, lacking of carbon sources (Zhang, 2018) leading to that using AAO to treat livestock wastewater after anaerobically treatment is unpopular.
Thus, P has important roles to humans, creatures and ecosystems but P also has certain negative effects Also, it is a non-renewable resource and will be exhausted in the future Therefore, the key requirements are removal and recovery P from wastewater Currently, there are various technologies used to treat livestock wastewater but the most common is anaerobically treatment technology Treatment efficiency of anaerobically treatment for the main parameters such as BOD, COD,
SS, and VSS is over 75% However, the concentration of these parameters are still not enough quality to discharge directly into the environment Therefore, the wastewater after anaerobically treatment needs to be further treated to reach the discharge standards It can be seen that, although various technologies have been used to treat livestock wastewater after the anaerobically treatment, these technologies show many weaknesses in treatment efficiency or technical requirements.
Constructed wetlands
CWs are engineered systems that have been designed and constructed to utilize the natural processes involving wetland vegetation, soils, and the associated microbial assemblages to assist in treating wastewaters (Vymazal, 2010).
CWs have been applied to treat wide range of wastewater types such as domestic wastewater, municipal sewage, leachate, industrial wastewater and rainwater (Vymazal, 2010).
According to Wu (2015), CWs are a technology that uses plants to treat different waste CWs is required large building area, low operating energy, and suitable for small communities or decentralized sources CWs are a natural treatment system where physical, chemical and biological processes occur when water, soil, plants and microorganisms interact together It is considered a natural ecosystem designed to take advantage of natural processes to treat wastewater (Qasaimeh, 2015).
1.3.2 Classification of constructed wetlands (CWs)
There are many ways to classify wetland systems depending on the structure, substrate, or kind of plants in the system By the flow model, CWs can be classified into two types: Free water surface constructed wetlands (FWS-CWs) and sub-surface flow constructed wetlands (SSF-CWs) (Vu, 2012) In addition, there is combining basic systems to form a hybrid treatment system.
Figure 1.6.Classification of CWs (Herath, 2015)
FWS-CWs have a low P removal ability because of limited contact of water with soil particles which adsorb and/or precipitate phosphorus (Vymazal, 2010) This idea was agreed in the White's study (2011), that the P-removal ability of FWS-CWs is low The removal of P by FWS-CWs stay in range between 30 to 50% in the long term (Economopoulou, 2004) In addition, FWS-CWs have many disadvantages such as easily causing odors due to regular anaerobic condition, good conditions for mosquitoes and insects living.
Based on the flow directions, SSF-CWs can be divided into two categories in the direction of flow: Subsurface flow vertical flow constructed wetlands (VSSF-CWs) and subsurface flow horizontal flow (HSSF-CWs).
Overall, P removal by VSSF-CWs is normally low Abdelhakeemre (2016) reported that, the P eliminating is 22% in VSSF-CWs with plant and 17% VSSF-CWs without plant Due to the water supply, the conditions in the FWS-CWs change from aerobic to anaerobic and vice versa, which affects the P removal ability of VSSF-CWs Anaerobic soils releasemore phosphate to soil solutions low in phosphate andsorbed more phosphate from soil solutions high in sol-uble phosphate than do aerobic soils (Vymazal, 2007) This phenomenon can be explained that organic structural phosphorus can become soluble phosphorus when the organic matter was oxidized Insoluble P form can re-dissolve under altered conditions (Kadlec, 2008).
Figure 1.7.The diagram of VSSF-CWs (Dotro, 2017)
HSSF has the ability to remove a variety of pollutants such as COD, SS, N, P.
According to Volker (2001), HSSF has a better ability to eliminate P than VSSF due to its ability to accumulate in P humic of P This result agrees with Arun's study
(2019) when showing that P removal ability of HSSF is higher than VSSF.
1.3.3 Phosphorus removal by different components in CWs
The P elimination pathway in CWs system consists of: P removal by filter media, plants, microorganisms and other factors In particular, removal of P by filter media and plants are two removing P sustainability processes (Wu, 2013) When hydraulic retention time is long and the soil has a fine structure, the process of P removal is mainly adsorption and precipitation in the substrate, because this condition provides a good opportunity for P adsorption and reaction (Nguyen, 2015 a ; Qin, 2016).
Table 1.2.P removal mechanisms of CWs components (Nguyen, 2019)
Chemical bonding of P to iron, aluminum, and calcium on soil particle exchange sites
Precipitation P binds to dissolved iron, aluminum, and calcium to form a solid or semi-solid
Sedimentation Particulate phosphorus settles out of the water column
OrthoP and some organic P taken up by plants and algae
Plant available and some organic forms of P are consumed by microbial communities and stored in their tissues a The wetland substrates
Metabolism of P between materials and wastewater in CWs is very diverse including: adsorption - desorption, precipitation and dissolution, fragment and leach, mineralize, and settle (Vymazal, 2008) Filter materials in CWs play a important role in P eliminating through two pathways: adsorption, precipitation (Wu, 2015;
Regarding P sorption mechanisms in CWs involves the movement of dissolved inorganic P from soil, wastewater to the surface of materials, where it only accumulates on the surface without penetrating the interior The balance between adsorption and desorption maintains the equilibrium between the solid phase and P in water in the pores of the material This phenomenon is defined as the phosphate buffering capacity similar to the pH buffering capacity of the soil (Dunne, 2006).
P is adsorbed on the substrate based on presence of Al, Fe, Ca, and Mg ions, where ability to eliminate P of these ions depends on the pH of the system and the amount of ions exist in the system In acidic condition, inorganic P is adsorbed on the oxides of iron and aluminum while compounds of calcium and magnesium are usually produced at higher pH.
Figure 1.9.P adsorption mechanism on material surface (Ramesh, 2008)
The adsorption capacity of materials is usually determined through the Langmuir model The other commonly used models are Freundlich and Tempkin (Hongling,
2017) The adsorption isotherms illustrate the balance between amount of adsorption and solubility under certain conditions.
P removal by chemical precipitation mechanism include the reaction between phosphate ions with metal cations such as calcium, aluminum, iron, magnesium to form amorphous crystalline salts The following chemical reaction shows immobilization P with silicate clay:
Al 3+ + H2PO4-+2H2O = 2H + + Al(OH)2H2PO4(Variscite)
Al2SiO5(OH)4 + 2H2PO4- = 2Al(OH)2H2PO4+Si2O52-
Similar precipitation reactions occur with iron ions under acid conditions, or with calcium and magnesium under neutral conditions (Ramesh, 2008).
Fe 3+ + PO4 3- → FePO4↓ 5Ca 2+ + 3 PO43-+ OH - → Ca5(PO4)3(OH)↓
In case wastewater have acid conditions, Fe and Al content may be more important because precipitation reactions with these ions are preferred at lower pH levels(Arias, 2001).
The P dominant removal mechanism depends on the physical and chemical properties of the filter material In addition, P removal capacity also depends on the content of calcium, iron, aluminum and their oxides and hydroxides (Yang, 2018).
There are many ways to classify filter materials based on usage history, sources of substrates or removal mechanism Regarding the sources of substrates, they can be classified into three types: natural natural materials, industrial by-products and man-made artificial products (Cucarella, 2009). b The wetland plants
Plants is a necessary part of CWs that plays an important role in pollutant removal.
The contribution of wetland vegetation to pollutant removal through filtration and sedimentation, stabilization of the wetland surface, light attenuation, and additional surface area for the attachment of microorganisms (Shan, 2011) The active reaction zone of constructed wetlands is the root zone (or rhizosphere) This is where physicochemical and biological processes take place that are induced by the interaction of plants, microorganisms, the soil and pollutants (Stottmeister, 2003).
This is where the most intensive interaction between the plant and microorganisms is to be expected The roots releases organic acids (oftenly citrate and oxalate) which rising the availability of P availability Amount of excreted organic acid, mycorrhizal fungi, root-zone microorganisms allow a plant to uptake P more from soil (Nguyen, 2019 b ).
Figure 1.10.Phytoremediation Using Aquatic Plants (Fletcher, 2020)
Figure 1.11.Rhizosphere in CWs plants (Ghimire, 2019)
Study subjects
Although CWs show good removal abilities for main pollutant parameters such as BOD, COD, SS, TSS, their ability to retain P is very low For P elimination, the selection of filter media is extremely important In addition, the P removal capacity of HSSF system is higher than that of VSSF system (Arun, 2019) Since this study aims at enhancing the P removal from pig wastewater after anaerobic treatment, the HSSF-CWs system were selected for investigation. a Coal slag
According to Vietnam Energy Association's report (2015), 35% of electricity was generated from thermal power plants In Vietnam, most coal-thermal power plants are located in the North, where the coal sources are closely located (Nguyen,
Figure 1.13.Source structure of the national electricity system by primary energy
The the quality of raw coal used for thermal power plant is usually low Depending on production technologies of the thermal power plants, the amounts and characteristics of coal slag can vary significantly According to a survey by theJapan Bank for International Cooperation (2003), the amount of slag discharged from five thermal power plants of in the north of Vietnam was 673,600 tons annually In the future, when the demand for electricity increases and the thermal power plants keep working, the amount of coal slag is expected to be increased substantially.
In Vietnam, coal slag is currently considered as an industrial waste Normally, coal slag is not disposed but utilized to produce construction materials.
Many studies have shown that coal slag is capable of removing dyes such as azo, Tryphenylmethane and Brilliant Blue FCP (Gupta, 2006), Vertigo Blue 49 (Nguyen,
2015 a ) and reducing COD in the wastewater from the paper mill up to 40% (Pierre, 2002).
Nguyen (2020) reported that the coal slag under optimal conditions could remove P up to 21.63 mgP/g in the P concentration range of 0-30 mg/L A great potential of the coal sal for removing P was also reported by Safaa (2013) Accordingly, with the increase of the initial concentration of phosphate from 0.1 to 25 mg/L, the percentage removal for the slag increases from 76% to 99% The effective P removal of coal slags could be explained by the presence of SiO2, Al2O3, and Fe2O3 in their compositions (Kieu, 2011).
On the other hand, Phan (2008) pointed in his study that coal slag of the thermal power plants hardly reacted directly with water Nguyen (2015 a ) presented the good water conductivity of coal slag This characteristic is extremely important in the application of materials in CWs because it helps to minimize the possibility of clogging.
However, a controversial issue regarding to the application of coal slag is the possible release of heavy metals into the aquatic environment (Li, 2016) Therefore, further studies are needed to identify effects of coal slag before applying them as CWs substrates on a large scale.
Thus, in order to make use of industrial waste, coal slag is a potential material for use in the CWs. b Ferralsols
According to the Soil Map of the World (FAO, 2005), Ferralsols can be subdivided into 6 groups: Plinthic Ferralsols (Fp), Humic Ferralsols (Fh), Acric Ferralsols (Fa), Rhodic Ferralsols (Fr), Xanthic Ferralsols (Fx), Orthic Ferralsols (Fo).
Ferralsols consist mainly of quartz, kaolinite, oxides, and organic matter (Eswaran,
2005) Ferralsols composed primarily of base cation-poor minerals such as 1:1 phyllosilicates (e.g., kaolinite) and sesquioxides (Tabor, 2017) Ferralsols are one of the major soil groups in Siaya County in western Kenya with low available P This is one of the factors limiting crop production in western Kenya (Owino, 2015).
The worldwide extent of Ferralsols is estimated at some 750 million hectares, almost exclusively distributed in the humid tropics on the continental shields of South America and Africa (Rattan, 2016) In Southeast Asia, about 4 percent of the land area is covered by Ferralsols, located mainly in Vietnam, Thailand and Cambodia (USAID, 1980).
Feralsols group dominates in the land distribution of Vietnam (with about 65.2% of the whole territory), in which the most common is red and brown feralite soil (Ton,
2000) According to the Vietnam Association of Land, ferralsols is diversely distributed in many provinces and cities of Vietnam such as the Northeast region (Phu Tho, Ha Giang, Tuyen Quang, Yen Bai and Bac Kan), Northwestern region (Lang Son, Hoa Binh and Cao Bang) and the Central Highlands region In general, the soil is less porous (about 40% porosity); acid condition (pHKCl 4.0 - 4.5); P content ranges from 1 to 5 mg P2O5/100 grams of soil The P content in ferralsols is very low and difficult to use for cultivation unless there are solutions to improve the soil or apply additional fertilizers (Owino, 2015).
The Central Highlands is a region with a high area of ferralsols in Vietnam The boundary of the Central Highlands region almost coincides with the administrative boundaries of 5 provinces of Kon Tum, Gia Lai, Dak Lak, Dak Nong and LamDong province (Nguyen, 2015 c ) The natural area of the Central Highlands is
5,646,127 ha including 10 soil groups Of which, Ferralsols group has the largest area (3,556,336 ha), accounting for 69.2 percent of the total regional area.
In the Central Highlands, ferralsols group is mainly used for agroforestry production In the field of environmental treatment, ferralsols has been used as a potential adsorbent to remove many heavy metals such as arsenic (Kassenga, 2008), chromium, copper, zinc, lead (Ntambi, 2020) or organic toxins such as Bisphenol S (BPs) (Shiqiu, 2019) Muindi (2015) showed a great potential of natural ferralsols in Kenya in removing P.
Explaining the P-removal ability of ferralsols, Tabor (2017) has shown that the main components of ferralsols are basic cation-poor minerals such as phyllosilicates and sesquioxides This was also presented in USDA Soil Taxonomy that yellow-red soil contained some amorphous, aluminum-iron mineral components that were easily converted into crystal minerals such as kaolinite, goethite, hematite or gibbsite These components are capable of removing P by adsorption, precipitation, and ion exchange mechanisms.
Thus, ferralsols is a natural material, with a huge reserve worldwide and widely distributed in Vietnam In addition, it has a great potential in removing P Therefore, this study focuses on the use of ferralsols as the wetland substrates for eliminating P from wastewater. c Sand
According to the General Department of Geology and Minerals of Vietnam (2017), sand is widely distributed in 09 coastal provinces in the North and Central Vietnam.
The total reserves of 13 mines, which have been explored, is 123 million tons and the forecast resource is about 3 billion tons Sand is used extensively in human activities, especially construction In environmental treatment, sand has been studied as an adsorbent for removing heavy metals (Pb, Cr, Cu, Zn) and organic matter (Awan, 2003) Many studies also showed the ability to remove P of sand.
However, P removal capacity of sand compared to other natural materials such as limestone and seashells is lower.
1.4.2 Plants a Water spinach (Aquatica ipomoea)
MATERIALS AND METHODS
Materials
In this study, CS was collected from Pha Lai Thermal Power Joint Stock Company, which is located in Chi Linh district, Hai Duong province, 65 km northeast of Hanoi, close to the northern corner of National Road 18 and left bank of Thai Binh river.
The total area of the company is around 322 hectares.
Figure 2.1 Pha Lai Thermal Power Joint Stock Company
Pha Lai Thermal Power Joint Stock Company has 02 power generation plants The construction of plant 1 was started in 1980 and completed in 1986 Plant 1 has a maximum capacity of 440MW including 4 power generation sets Plant 2 was built in 1998 and completed in 2002 It have a maximum capacity of 600MW including two power generation sets Currently, the maximum power output of Pha LaiThermal Power Joint Stock Company is 6.54 billion kWh/year Of which, the power outputs of plant 1 and plant 2 are 2.86 billion kWh/year and 3.68 billion kWh/year, respectively.
Slag and ash collecting system
Figure 2.2.Principle diagram of electricity production technology
CS of Pha Lai Thermal Power Joint Stock Company has the following chemical composition: SiO246.2%; Fe2O3 12.3%, Al2O3 24.2%, CaO 1.75%; and some other oxides such as MgO, TiO, Na2O (Nguyen, 2015 a ). b Ferralsols
In this study, the natural ferralsols (NF) was collected in Thuan Hanh T-junction,Dak Song district, Dak Nong province, which belongs to Vietnam CentralHighlands, one of regions with largest NF area in Vietnam.
Figure 2.3.Sampling location of Ferralsols in Dak Nong Province
According to Co (2012), the criteria to select NF were dark red color, homogeneous, and without agricultural activities on the sampling location First, NF was collected from the soil layer of 0.5 m below the soil surface to avoid the effects of agricultural activities Then, NF was transported to MEE laboratory by car for temporary storage Finally, NF was conveyed to Bat Trang pottery and ceramics village, which is located in Kim Lan Commune, Gia Lam District, Hanoi, for calcination NF was thermally treated in a commercial furnace (LxWxH = 3m x 2m x 3m) with the temperature sensor at 500 0 C for 2 h. c Other filter materials
In the HSSF-CWs, there were also other filter materials, such as stone and sand.
These filter materials were utilized for preventing the water clogging and supporting the vegetation growth The stone had the size of 10-25 mm on its longest edge It was washed with tap water before being transferred in CWs units with the total amount of 10 kg/unit The sand was purchased from a construction material store It was placed on the top of CWs units with the quantity of 15 kg/unit.
Figure 2.4.Stone placed at the bottom of CW units Figure 2.5.Sand added on the top of
This work investigated two kinds of plants, including water spinach (Aquatica ipomoea) and lemongrass (Cymbopogon citratus) The water spinach was planted from seeds on the soil When water spinach plants were 5 days of age with the height of 20-23 cm (including root length) and 2-3 leaves on each plant, they were taken out of the seeding place, washed with tap water to remove soils attached on the roots surface, gently dried with paper towels The plants with the same height were transferred to the CWs units with the density of 128 plants/HSSF-CW1, 128 plants/HSSF-CW2 The lemongrass was prepared from stems, which were purchased from the local market The stems, without leaves and roots, of the similar height (25 cm), were put in beakers containing tap water The beakers were placed in the shady place After 7 days, the new roots and leaves were well developed The newly developed lemongrass plants were transferred into CW units with the density of 40 plants/HSSF-CW4.
Figure 2.6.Aquatica ipomoeaplanted from seeds on the soil before being transferred into CW units
Figure 2.7 Cymbopogon citratus kept alive in the tap water before being transferred into CW units
Figure 2.8 Aquatica ipomoeaat the time being tranferred into CWs
Synthetic wastewater, which was prepared from chemicals (KH2PO4and NH4Cl) of analytical grades, simulated the composition of the real 4-fold diluted swine wastewater after anaerobic treatment The composition of the real wastewater was adapted from a previous study by Nguyen (2019) and is presented in Table 2.1 The phosphate aqueous solutions were utilized for adsorption tests The simulated wastewater containing only ortho phosphate (PO43-) and amonium (NH4+) was used for operating CW units.
Table 2.1.Parameters of real post-anarobically swine wastewater in Chuong My,
Experimental set-up
Ferralsols calcination was conducted at both lab-scale and large-scale In the MEE lab, NF was calcined at three categories of temperature (300, 500, and 700 o C) using the Carbolite furnace (CWF 12/13, England) in the laboratory of Master’s Program in Environmental Engineering (MEE).
Figure 2.9.Carbolite furnace (CWF 12/13, England)
Then, the best calcination temperature was selected by comparing P adsorption capacity and the pH of the post-adsorption solutions In Bat Trang village, NF was loaded on 5 different trays (LxWxH = 180 cm x 45cm x 10cm) inside a commerical furnace.
Figure 2.10.Preparing NF for calcination with the commcerical furnace
It took around 6 hours to set up the stable temperature of 500 o C The NF was calcined at the constant temperature of 500 o C for 2 h Then the furnace was turned off and cooled down to the ambient temperature The calcined Ferrralsols (CF500) was packed and transported by car to MEE lab for further investigation.
There are many factors that may affect the P adsorption of CF500 This study investigated three influential factors, namely pH, adsorbent dose, and temperature.
Effects of pH: So as to examine the effect of solution pH on P adsorption of filter materials, different solution pH values (3, 5, 7, 9, and 11) were employed The pH values were adjusted using NaOH and H2SO4 solutions of different concentrations to minimize the change in the solution volume Other adsorption conditions were kept the same, such as the room temperature of 25 o C, Ci = 50 mg/L shaking speed of 120 rpm, shaking time of 24 h, and adsorbent dose of 1g/75ml.
Effects of adsorbent weight:In order to investigate the effect of adsorbent dose on
P adsorption of filter materials, different adsorbent doses were used including 1, 3, and 5g/75 mL of solutions with Ci = 50 mg/L Other adsorption conditions were kept the same, such as the room temperature of 25 o C, shaking speed of 120 rpm, shaking time of 24 h were the same, pH of the solution was adjusted to the optimal pH value determined in the above test.
Effects of temperature: Different temperatures including 25, 30, 40, and 50 o C were used to evaluate the effect of temperature on P adsorption Other adsorption conditions were kept the same, such as shaking speed of 120 rpm, shaking time of
24 h were the same, solution pH and adsorbent dose were obtained from the above tests. b Isotherm study
Isotherm studies at room temperature (25 o C) were conducted to determine the maximum adsorption capacity of the filter materials The initial P concentrations of the aqueous solutions were varied in the range of 10-1000 mg/L Other adsorption conditions were determined from the experiments on the influential factors mentioned above Two most commonly used isotherm models, including Langmuir and Freundlich, were utilized to evaluate their suitability in describing the experimental data of the present study.
The Langmuir adsorption isotherm model is expressed in the linear form as follows: q e
1 (2.5) where KLis a constant related to the binding strength of phosphate (L/mg), qmis the maximum sorption capacity (mg/g) These parameters were calculated through the slope and the intercept of the plot of Ce/qeversus Ce.
The Freundlich model is expressed as following equation: qe= KFCe1/n (2.6) lnqe= lnKF+ ( n
1)lnCe (2.7) in which KF and n were constants related to adsorption capacity and energy of adsorption, respectively. c Kinetic study
In order to examine the effect of contact time on P adsorption of the filter materials, the adsorption kinetic tests were conducted at the same adsorption conditions, such as initial P concentration of 50 mg/L, adsorbent dose of 1 g/75 mL, room temperature of 25 0 C, and shaking speed of 120 rpm in a series of flasks After different time intervals, each flask was removed from the shaking machine, filted with the filter paper ỉ15 cm (GB/T 1914-2007, China), and analyzed for determining P concentration using the UV/Vis Diode Array Spectrophotometer (S2100 UV, Unico, USA) In order to simulate the experimental kinetic data, two widely used kinetic models were tested.
Pseudo-first-order kinetic model:
The Pseudo-First-Order kinetic model were given as follows: ln(qe- qt) = lnqe- k1t (2.8) where, qt: the amount of P/mass of sorbent (mg/g) at any time t qe: the amount of P/mass of sorbent (mg/g) at equilibrium k1: the rate constant of first order sorption (1/min)
Pseudo-second-order kinetic model:
Pseudo-Second-Order kinetic model was shown in following expression: q t t = 2 e
The initial sorption rate is defined by the equation below: h = k2qe2 in which, k2: the rate constant qt: the P uptake capacity at any time t The kinetic parameters of Pseudo-second-order model were calculated using Solver sofware according to non-linear method.
2.2.3 Constructed wetlands design and operation
CWs units were made of stainless steel (inox 340) with the dimensions as following LxWxH = 68.5x33x42 There were 4 CW units in parallel in system.
The CW units were first tested in term of water leakage Then, they were filled with filter materials All 4 CW units contained a layer of stone at the bottom and a layer of sand on top In the middle layer, CW unit 1 was packed with coal slag, whereas three remaing CWs units were filled with a mixture of coal slag and calcined ferrasols (CF500) with the mixing ratio of 1.25:1 by volume for comparision purpose The water spinach (Aquatica ipomoea) was planted in the CW units of 1&2 while the lemongrass (Cymbopogon citratus) was vegetated in CW unit 4 The
CW unit 3 was unplanted and used as the control.
Figure 2.11.Layers structure of tanks in CWs
Nutrient solutions were stored in a tank of 200 L It was supplied to CW units using
04 peristaltic pumps Four CW units were fed in the HSSF direction with the same hydraulic retention time (HRT) of 5.4 days The effluents of CW units were sampled every three days for determining pH, conductivity and P concentration.
HSSF-CW1 HSSF-CW2 HSSF-CW3 HSSF-CW4
Figure 2.12.Nutrient solution storage tank
Figure 2.13.The HSSF-CWs system planted with water spinach and lemongrass
Analytical methods and equipment
Porosity is the ratio of void volume to the total volume of a fiter materials (Irwin,
2015) The porosity is calculated using the following equation:
Porosity volume Total voids of
Hydraulic conductivity of a material is defined as the ability of the fluid to pass through the pores and fractured rocks (Subbarayan, 2016) Based on Darcy's Law, an equation was used to determine the hydraulic conductivity experimentally In the lab conditions, hydraulic conductivity was determined for NF, CS, CF500, and mixture of CS and CF500 with different mixing ratios (1:1, 1.25:1, and 1.5:1) For each kind of experiment, the material(s) was/were placed into a cylinder Water was pumped into the first cyclinder with the water column of H2 (cm) From which, the water naturally ran into the second cylinder with the water column of H1 (cm) H2 was varied while H1 was remained a constant, thus resulting in different (H2-H1).
The effluent flow rates corresponding various (H2-H1) values were measured using a stop watch and a graduated cylinder The hydraulic conductivity was determined using the following expression:
K: hydraulic conductivity (cm/s) V: the volume of water that flows through the system during time t (mL) A: cylindrical cross sectional area (cm 2 )
L: height of the material column (cm) c Scanning electronic microscopy (SEM)
Scanning Electron Microscope (SEM) images provide information on: topography,morphology, composition (Lametschwandtner, 1992) The SEM analysis was implemented with beams of electrons to render high resolution, three-dimensional images The results were used to clarify differences in structure and morphology of ferralsols before and after calcination.
Figure 2.14.AMRAY Model 1830 Scanning Electron Microscope d X-ray powder diffraction (XRD)
X-ray powder diffraction (XRD) is used for phase identification of a crystalline material and can provide information on unit cell dimensions (Moore, 1997).
XRD measurements were implemented with Empyrean equipment (PANalytical, Netherlands) for NF and CF500 at Institute of Geophysics (IGP), Vietnam Academy of Science and Technology (VAST).
Figure 2.15.Empyrean equipment (PANalytical, Netherlands) e X-ray fluorescence (XRF)
XRF is a non-destructive analytical technique, which is to determine the elemental composition of a material (Andrew, 2018) In this study, XRF analyses were performed with NF and CF500 using X-ray fluorescence spectrometer (WD- XRF
-S4 Bruker, Germany) in the Institute of Geological Sciences (IGS), Vietnam Academy of Science and Technology (VAST).
Figure 2.16.X-ray fluorescence spectrometer (WD- XRF -S4 Bruker, Germany) f Fourier transform infrared spectroscopy (FTIR)
Fourier transform infrared spectroscopy (FTIR) analysis is performed to identify the key functional groups (Berna, 2016) that are reponsbile for adsorption reactions In this work, FTIR measurement was implemented with NF and CF500 using FTIR Spectrometer (FT/IR 6300 typeA, Jasco, Europe) in the Lab of Environmental Research (LER), VNU University of Science (HUS).
Figure 2.17.FTIR Spectrometer (FT/IR 6300 typeA, Jasco, Europe)
Though this study focused on the P concentration, some other environmental parameters were also measured, including pH, EC, heavy metal concentrations All analyses were triplicated The average was calculated The adsorption experiments were shaken by the Shaker (OS 3000, Jeiotech, Korea) at the MEE laboratory.
Figure 2.18.Shaker (OS 3000, Jeiotech, Korea)
The concentration of ortho-P was measured according to Method 365.3 (EPA) using UV/Vis Diode Array Spectrophotometer (S2100 UV, Unico, USA) at the wavelength number of 710 nm.
Figure 2.19.UV/Vis Diode Array Spectrophotometer (S2100 UV, Unico, USA)
The pH value of solution was measured by pH meter (S220-Kit, Mettler Toledo, China) at the MEE laboratory.
Figure 2.20.The pH meter (S220-Kit, Mettler Toledo, China)
Electrical conductivity (EC) values of simulated wastewater before and after treatment by CWs were measured by Quick Shipment & Quality Service sensION+
EC5 (sensION+ EC5, Hach, China) at the MEE laboratory.
Figure 2.21.The SensION + EC5, Hach, China
The contents of heavy metals in water after treatment with CWs were measured by Atomic Absorption Spectrophotometer (AAS) (ZA3000) Methods for analyzing of specific heavy metals as folows: Lead (SMEWW 3113B:2012), Arsenic (SMEWW 3114B:2012), Copper (SMEWW 3111B:2012), Cadmium (SMEWW 3113B:2012), Mercury (SMEWW 3112B:2012), Zinc (SMEWW 3111B:2012), Iron (SMEWW 3111B:2012), Aluminium (SMEWW 3111B:2012).
Figure 2.22.Atomic absorption spectrophotometer (AAS, ZA3000)
These experiments were performed at Research Center for EnvironmentalMonitoring and Modeling (CEMM), VNU University of Science (VNU-HUS).
Calculation and statistical analysis
Ci: the initial P concentration (mg/L)
Ce: the equilibrium P concentration (mg/L) b P adsorption capacity qe m
C i e x V (2.13) where: qe: adsorption capacity at equilibrium (mg/g)
Ci: the initial P concentration (mg/L)
Ce: the equilibrium P concentration (mg/L)
The statistical analysis (average, standard deviation) was implemented using Excel software.
RESULTS AND DISCUSSION
Ferrasols calcination for P removal enhancement
The substrate plays an important role in the removal of P P can be eliminated mainly through exchange, adsorption, and precipitation processes (Nguyen, 2015 a ).
Therefore, choosing suitable filter media is a crucial process.
To further enhance the P adsorption capacity of raw materials, various modification methodologies were utilized, such as thermal treatment (calcination) and chemical treatment This study investigated the 3 calcination temperatures (300, 500, and 700 oC.) to identify the optimal one to achieve higher P removal ability The optimal calcination temperature was selected based on the P sorption capacity and pH value of post-adsorption solutions.
Table 3.1.Comparison of P adsorption capacity of investigated filter materials
P adsorption capacity, qe(mg/g) pH of post-adsorption solutions
It can be seen from Table 3.1 that the qe value rose from 3.83 to 4.68 mg/g when calcination temperature was increased form 300 to 500 o C This may be due to the increase in the BET surface area and pore volume in the CF500 However, as a result of increment in the calcination temperature from 500 to 700 o C, the qe value declined from 4.68 to 4.25 mg/g This can be attributed to the collape of pore walls at 700 o C All the pH values of the post-solutions were in the neutral range of 6.67-6.95 Of which the most neutral pH value was obtained with the calcination temperature of 500 o C Based on the values of qe and post-adsorption pH, 500 o C was considered as the optimal calcination temperature and applied for next investigations.
The optimal calcination temperature was used for modifying ferralsols at a large-scale in Bat Trang village Then, two categories of CF500 produced at lab-scale and large-scale were tested in terms of qe and post-adsorption pH for comparison purpose.
Table 3.2.Coparing CF500 produced in lab-scale and large-scale
Materials qe(mg/g) Post-adsorption pH
The values of qe and post-adsorption pH for two categories of CF500 produced at lab-scale and large-scale are presented in Table 3.2 It was found that both of these criteria were quite similar This indicates that the large-scale CF500 could be used for the next experiments including adsorption tests and CWs applications.
Adsorptive behaviours of calcined Ferrasols
Various factors can affect to P removal efficiency of materials In this research, three influential factors were investigated consisting of pH, dosage, and temperature. a Effect of pH
The pH plays a crucial role in P adsorption capacity as pH can affect the existing P species in the solution as well as functional groups on the surface of the filter materials (Nguyen, 2019 a ) In order to optimize the P sorption of the investigated materials, the effects of pH on P sorption by NF and CF500 were examined at different pH values of 3, 5, 7, 9, and 11 The result is displayed in Fig 3.1 It can be observed that for both categories of materials, the P removal efficiencies were reduced with a rise in pH values.
Figure 3.1.Effect of pH of NF and CF500 on P removal
Both CF500 and NF exhibited the highest P removal efficiencies pH 3, being 88% and 55% for CF500 and NF, respectively The P removal efficiency of CF500 was decreased from 88.18 to 61.21 % when pH increased 3 to 9 This value obtained at pH 11 (53.62 %) was much lower than that at pH 3 (88.18 %) A similar trend can be observed for NF The P removal efficiency of NF reached 52.89 % at pH 3 It declined from 52.89% to 23.62% when pH values increased from 3 to 11.
This result agreed with that reported by Nguyen (2019 a ), observing that the P removal was best in acidic medium (pH 3, 5, 7) Two mechanisms can be used to explain this result Firstly, in the acidium medium, H + cations were attached to the surface of filter media, making a positively charged layer on the surface of filter materials Consequently, the P retention on filter materials was enhanced owing to eletrostatic interactions Secondly, in the acidic medium, positive Fe 3+ and Al 3 + cations were released into the solution These cations can react with PO43- to form precipitations Contrarily, in the alkaline medium, a negatively charged OH - layer was formed on the surface of filter media This leads to a decrease in P sorption because of the repulsion force between the OH- on the materials’ surface and PO43- in the solution (Agyeia, 2002).
In conclusion, pH = 3 is the best value for P adsorption Besides, CF500 demonstrated to be effective in a wide range of pH values This can be considered as an advantage when selecting materials for CWs since this material can be suitable for the treatment of wastewater with vaious pH values. b Effect of adsorbent dosage
The percentage of P adsorption with various adsorbent doses of CF500 and NF were presented in Fig 3.4 This result shows that at equilibrium, the percentage of P removal increased while P adsorption capacity decreased with increasing adsorbent doses for both of filter media.
Figure 3.2 Effect of dosage of NF and CF500 on P removal
P removal efficiency of NF increased from 52.9 to 99.83 % and the number of CF500 is from 89.08 to 99.88 % with increasing adsorbent dose from 1 to 5 g/75 mL This is because adsorbent doses greater offered comparatively larger surface areas However, the P adsorption capacity of NF decreased from 2.30 to 0.87 mg/g and decline from 3.87 to 0.87 mg/g for CF500 respectively when rising adsorbent dose from 1 to 5 g/75mL It can be explained that the higher adsorbent dose gives a greater number of adsorption points on the surface of filter media But, the amount of ortho-P anions in the solution was limited When all ortho-P anions were kept on the surface of filter media, the increase in adsorbent dosage could not increase the P uptake anymore (Deng, 2018) In conclusion, the best dosage for P eliminating by CF500 is 1 g/75mL. c Effect of temperature
Fig 3.3 describes the effect of different temperatures on the adsorption ability ofCF500 Accordingly, a trend of increasing P removal efficiency was noticed when rising the temperature of adsorption.
Figure 3.3 Effect of temperature of CF500 on P removal it is clearly shown that the P removal efficiency and P adsorption capacity increased steadily from 25 o C (86.31%) to 50 o C (90.16%) These results show that the adsorption of CF500 is the endothermic process, therefore, as the reaction temperature increases, the adsorption equilibrium shifts in the direction that reduces the concentration of adsorbent (P) in the solution which leads to increasing of P removal efficiency (Salman, 2011) In addition, although the temperature range varies widely (from 25 o C to 50 o C), the efficiency of P removal of materials was always over 80% This is an advantage of CF500, especially when applied in Vietnam, where there is a huge temperature difference between seasons.
Sorption studies were conducted by the batch technique Batch experiments were carried out to determine the adsorption isotherms of P onto the surface of adsorbents.
Data on adsorption equilibrium by CF500 (Fig 3.4 and Fig 3.7), NF (Fig 3.5 andFig 3.8), and CS (Fig 3.6 and Fig 3.9) were presented based on two adsorption isothermal models Langmuir and Freundlich.
Figure 3.4.The fitting Langmuir and Freundlich isotherm models for P adsorption by CF500
Figure 3.5.The fitting Langmuir and Freundlich isotherm models for P adsorption by NF
Figure 3.6 The fitting Langmuir and Freundlich isotherm modelts for P adsorption by CS
Figure 3.7.Linear form of adsorption isortherms: a) Langmuir model and b)
Figure 3.8.Linear form of adsorption isortherm following a) Langmuir model and b) Freundlich model of NF
Figure 3.9.Linear forms of adsorption isortherms: a) Langmuir model and b)
Table 3.3.Langmuir and Freundlich adsorption isotherm constants
It can be concluded that the experimental data of both NF and CS can be described by Freundlich model due to the higher correlation coefficient values R 2 than Langmuir model However, CF500 better fitted the Langmuir model than the Freundlich because the R 2 values of Langmuir model is much closer to 1 than those obtained from the Freundlich isotherm Thus, the P adsorption by CF500 is monolayer adsorption isotherm while this mechanism at NF and CS are multilayer adsorption isotherm.
It was concluded that CF500 is the biggest potential adsorbents for P removal from wastewater (qmax = 19.38 mg/g), following by NF (qmax = 12.092 mg/g) and the smallest potential adsorbents is CS (qmax= 2.12 mg/g).
A comparison of the maximum adsorption capacity of different adsorbent materials was shown in Table 3.4.
Table 3.4.P adsortion capacity of different materials
Fiter media qmax(mg/g) Reference
Accordingly, CF500 shows a higher adsorption capacity than other materials such as bauxitie, zeolite or natural white hard clam (WHC) but lower than modified white hard clam (WHC-M800) However, the adsorption capacity of CF500 is very high and can be considered as a potential P adsorbent for CWs.
In order to determine the information about adsorption rate and elucidating adsorption mechanism, the kinetic study was investigated in this thesis.
Figure 3.10.Kinetic curve of CF500
Figure 3.11.Kinetic curve of NF
Figure 3.12.Kinetic curve of of CS
Pseudo-First-Order kinetic model ln(qe- qt) = lnqe- k1t
Pseudo-Second-Order kinetic model q t t = 2 e
1 t qe(mg/g) k1(min -1 ) R 2 qe(mg/g) k2
It is easy to see that, Pseudo second-order kinetic model has better description P sorption by CF500 than Pseudo first-order kinetic model due to the higher correlation coefficient (R 2 ) In addition, the adsorption capacity values (qe) given by these two models are quite similar.
A higher correlation coefficient (R 2 ) of the Pseudo second-order kinetic model than the Pseudo first-order model of the NF was also recorded, indicating a better description P sorption by the Pseudo second-order kinetic model.
In contrast, the Pseudo first-order kinetic model shows the ability to more accurately describe the adsorption process of CS through the higher value of the correlation coefficient.
The adsorption rate coefficient from the quadratic kinematic model was used to compare the adsorption rates of CF500, NF, and CS Accordingly, CF500 shows the fastest adsorption rate, followed by CS and the lowest is NF.
Characterization of the filter materials
3.3.1 Characterization of natural and calcined Ferralsols a SEM
Scanning electron microscopy (SEM) was used to observe the surface morphology of NF and CF500 The SEM analysis results are shown in Fig 3.13 Accordingly, the surface of NF had a dense and fine structure with very few holes, which were not typical.
Figure 3.13.SEM observation for a) NF and b) CF500
In contrast, the surface of CF500 showed a typical porous structure The pore diameter was significantly increased This change in the morphology of CF500 facilitates its P adsorption capacity. b XRD
The XRD spectra of NF and CF500 are shown in Figs 3.14 & 3.15 while their mineral composition is presented in Table 3.6 The diffraction peaks in Figs 3.14 &
3.15 indicated that NF contained 4 main minerals, namely Kaolinite, Hematite, Gibbsite, and Goethite, whereas CF500 comprised 3 minerals (Kaolinite, Hematite, and Goethite) with the absence of Gibbsite From Table 3.6, it can be seen that after calcination, the percentage of Gibbsite was reduced to zero, while those of Kaolinite and Hematite was increased by 44% and 8%, respectively The proportion of Goethite was almost unchanged under the effect of high temperature Due to the change in the mineral composition of ferralsols after calcination, the P retaining ability was boosted.
Figure 3.14.XRD spectrum of NF
Figure 3.15.XRD spectrum of CF500
Table 3.6.Mineral composition of NF and CF500 samples c XRF
To determine the elemental composition of filter materials (NF, CF500, and CS), XRF (X-ray fluorescence) analysis was implemented and the obtained results are given in Table 3.7.
Table 3.7.Main chemical compositions of NF, CF500 and CS
No Composition (%) CF500 NF CS
It is evident from Table 3.7 that CF500 possessed the highest percentages of Al2O3 and Fe2O3 compared to NF and CS In a previous study, it was reported that higher contents of metal oxides favored P sorption capacity of materials (Nguyen, 2019 a ).
Therefore, it is expected that CF500 may exhibit better P capture ability.
Fourier-transform infrared spectroscopy (FTIR) was used to determine the functional groups of NF and CF500 The spectra for NF and CF500 were shown in Fig 3.2 Accordingly, the dominant peaks were 909 cm -1 , 1100 cm -1 , 1036 cm -1 ,
3698 cm -1 , and 3622 cm -1 These peaks appeared in the spectra of both NF and CF500 However, in the spectrum of CF500, the peaks at position of 3526 cm -1 ,
3437 cm -1 , 750 cm -1 , 788 cm -1 were disappeared These peaks were attributed to functional groups of OH- of Gibbsite The decrease in the number of negatively discharged functional groups can lead to being a decrease in the repulsion between
PO43- in the aqueous solution and the OH- on the surface of CF500 As the result, the P retention on CF500 can be improved This result agrees well with the above XRD result in confirming the disappearance of the Gibbsite in the CF500.
Figure 3.16.FTIR analysis for NF and CF500
Table 3.8.Types of vibration peak in NF and CF500
Position (cm -1 ) NF CF500 Type of vibration Materials
1100, 1036 medium medium Si - O stretching Kaolinite
909 sharp sharp Si - OH Kaolinite,
750, 788 weak no Al-OH Gibbsite e Chemical composition
In this study, the chemical compositions of NF and CF500 were examined to explain the changes in P removal capacity of ferralsols after calcination The investigated parameters included pH, organic matters, Bio-available P, Feoxalate, and
Aloxalate The attained results are presented in Table 3.9.
Table 3.9.The chemical compositions in CF500 and NF
Parameters Unit Analytical method CF500 NF pH - TCVN 5979:2007 6.18 5.14
Aloxalate mg/g 2.167 0.848 pH: The use of filter materials can affect pH of the post-adsorption solutions As shown in Table 3.9, CF500 resulted in nearly neutral medium (pH = 6.18) while NF led to acidic medium (pH = 5.14) The former pH value can favor the growth of plants and microorganisms, whereas the latter can inhibit growth of some plant species (James, 2006) In other words, the CF500 applies to CWs as a substrate without harmful effects on wetland plants and microbes.
Organic matters are one of the most important factors that can influence the P removal According to Pham (2017), the iron and aluminum in the soil can react with organic matters including humic substances and some organic acids (e.g malic, oxalic, fulvic acids) This leads to less amount of iron and aluminum available for P retention It is demonstrated from Table 3.9 that the percentage of organic matter in CF500 (