Recent advances in heavy metal removal by thin film nanocomposite membrane

Water is essential to all human activities. Given that by 2025, it is expected that half of the world's population will actually live in water-stressed areas, water demand has been highlighted as one of the most significant challenges of the new century (Mekonnen & Hoekstra 2016). The amount of wastewater produced in the home, industrial, and agricultural sectors grow in tandem with the global population. Freshwater supplies are not renewed to meet the demands of an ever-increasing population, which causes fierce competition and an uneven distribution of the limited freshwater resources among many different sectors (Obotey Ezugbe & Rathilal 2020). Poor water quality and water-related diseases also will lead to a serious impact on human health. Due to rapid industrialization and development, the number of contaminants entering freshwater sources is increasing (Hebbar et al. 2016). As a result, many people worldwide, particularly in developing countries, lack access to clean drinking water, and the international community is currently looking into all practical solutions to reduce the overuse of limited freshwater resources (Obotey Ezugbe & Rathilal 2020).

Heavy or toxic metals are trace metals that are detrimental to human health and have a density at least five times of water. Heavy metals normally are absorbed into the body through inhalation, uptake, and absorption by the skin after being released into the environment either via air, drinking water, food, or a variety of chemicals and man-made products. Heavy metals basically accumulate in organisms because they cannot be biodegraded, and most heavy metal ions are considered toxic. The World Health Organization (WHO) has set standards for the maximum acceptable limit of certain harmful heavy metals in drinking water and industrial wastewater, as well as the health effects of exceeding those limits (Shrestha et al. 2021).

A large number of studies have been focused on the removal of heavy metals, which are a major pollutant in water supplies. Copper (Cu), zinc (Zn), manganese (Mn), iron (Fe), and cobalt (Co) are heavy metals that play essential roles in biochemical processes in the human body. Heavy metals, in particular, arsenic (As), lead (Pb), and cadmium (Cd), have the ability to inflict ionic immobility and damage to numerous organs in organism, making them extremely dangerous to the environment, especially aquatic life (Hosseini et al. 2020). Pb²⁺ and Cu²⁺ have been reported to be the most frequently reported harmful metals in industrial effluent (Hoang et al. 2020) and excessive exposure to these metal ions can be detrimental. Under normal conditions, the human body can handle trace levels of metals without causing serious health concerns. However, long-term exposure to heavy metals can result in a large accumulation of toxins in the body, eventually leading to system failure and death (Sardar et al. 2013).

Heavy metals had been eliminated from industrial wastes for decades by using various techniques like chemical precipitation, ion exchange, and biological treatment to address this issue. Toxic sludge production, insufficient heavy metal removal, and high energy needs are just a few of the significant drawbacks and shortcomings of these low-cost advantage-oriented methods (He et al. 2019), followed by low metal removal efficiency, considerable reagent loss, uneconomical energy use, and post-treatment operations imposed by secondary waste production, requiring further improvement of the current methods (You et al. 2017). Membrane-based technologies such as reverse osmosis (RO), nanofiltration (NF), and electrodialysis (ED) have sparked growing interest in heavy metal removal due to high rejection and low secondary pollutant formation (You et al. 2017) also due to their benefits provided in water and wastewater treatment. This technology is frequently implemented in wastewater treatment processes due to their ease of manufacture, better properties, and enhanced separation efficiency (Kim et al. 2018).

Numerous advantages for improving wastewater treatment are provided by membrane technology, including decreased energy and equipment needs (Quist-Jensen et al. 2015). There are also some additional benefits including energy savings without phase change, high separation efficiency, ease of upgrade, and environmentally friendly. Despite the benefits and prospects for these actual applications, some inherent limitations linger, such as short membrane lifespan, low productivity, restricted selectivity, and economic impracticality membrane fouling and high energy costs continue to be stumbling challenges (Zhang et al. 2016; Roy & Ragunath 2018). Additionally, membrane-based technology has demonstrated a tremendous potential to close the gap between economics and sustainability owing to its minimal or non-existent chemical usage, environmental friendliness, and accessibility to a wide range of people. To put it another way, in recent years, membrane technology has emerged as a more favourable choice for the processes used to treat wastewater (Singh & Hankins 2016).

Over the years, several initiatives have been made to introduce different wastewater treatment technologies, and existing methods have also been modified to meet the most recent emission or reuse regulations (Obotey Ezugbe & Rathilal 2020). Among the methods under consideration, the membrane-based approach holds the most promise for wastewater treatment and heavy metal ion removal. Membrane-based materials are widely used in wastewater treatment because of their ease of preparation, superior properties, and improved separation efficiency (Kim et al. 2018). Since it has made significant advancements in recent years and is well known for the beneficial results it produces in the treatment of both water and wastewater, membrane technology has become one of the wastewater treatment technologies that has experienced tremendous advancement in recent years. Due to significant reductions in equipment size, energy requirements, and capital costs, membrane technology provides a broad range of opportunities in wastewater treatment (Quist-Jensen et al. 2015). Additionally, the minimization of membrane impurities, which is a significant challenge for the membrane process, is constantly being improved through the modification of membrane modules and membrane elements (Obotey Ezugbe & Rathilal 2020).

Membrane is basically a two-phase barrier that selectively separates and limits the transport of many chemicals and is classified into four types based on their structure: homogeneous, heterogeneous, symmetric, and asymmetric. Membranes can separate solids and liquids, and they can be positive or negative in charge, neutral, or bipolar and this membrane serves a dual purpose in separation, and the separation is simple where it started with first the membrane acts as a semi-permeable layer between two phases, and second, it transports between two phases (Yalcinkaya et al. 2020). Membranes are classified according to their material composition and structure (see Table 1). Organic membranes, also known as polymeric membranes, are made of synthetic organic polymers such as polyethylene (PE), polytetrafluorethylene (PTFE), polypropylene, and cellulose acetate. The majority of membranes used in pressure-driven separation processes (microfiltration (MF), ultrafiltration (UF), NF, and RO) are synthetic organic polymers (Obotey Ezugbe & Rathilal 2020). Meanwhile inorganic membranes are made of chemically and thermally stable materials such as ceramics, metals, zeolites, or silica and are widely used in industrial applications such as hydrogen separation, UF, and MF. Membranes with mixed matrix structures are produced by combining an inorganic or inorganic–organic hybrid material in the form of micro- or nanoparticles with a polymer matrix. TFC membrane, a new generation of membrane structure that overcomes the permeability-selectivity trade-off in conventional membrane, combines an ultrathin selective layer with an extremely porous substrate. Lastly, TFN membranes are a newer type of composite membrane created by the interfacial polymerization (IP) process, in which nanoparticles are incorporated within the selective layer of the TFC membrane in order to enhance the interfacial polymerized layer's characteristics.

Table 1

Strengths and weaknesses of membranes

Types of membrane	Polymeric membrane Alqaheem & Alomair (2020)	Ceramic membrane Asif & Zhang (2021)	Mixed matrix membrane Maghami & Abdelrasoul (2018)	Thin film composite (TFC) membrane Ji et al. (2021)	Thin film nanocomposite (TFN) membrane Ji et al. (2021)
Strengths	High temperature stability, fouling resistance, low lifecycle costs, simple pore-forming mechanism, and high flexibility (Warsinger et al. 2018; Kamali et al. 2019)	Stronger mechanical, thermal, and chemical stability, higher fouling resistance and more extended membrane longevity, high removal efficiency (Abdullayev et al. 2019; Kamali et al. 2019; Asif & Zhang 2021)	Possibility of combining synergistically the easy processability of polymers, ability to overcome the trade-off between selectivity and permeability which is characteristic for pure polymer membranes, both polymer and filler qualities affect separation performance (Dechnik et al. 2017; Qadir et al. 2017; Cheng et al. 2018)	Able to achieve the optimal combination of flux and salt rejection, inexpensive, have good chemical and mechanical stability and can be formed into membranes using a simple phase inversion technique (Ji et al. 2021; Khoo et al. 2021; Peng et al. 2022)	While maintaining the TFC membrane's separation effectiveness, the interfacial polymerized layer's characteristics can be improved to increase hydrophilicity and/or surface charge density (Yang et al. 2020; Zhao et al. 2020; Liao et al. 2021)
Weaknesses	High selectivity for low solutes and species, membrane fouling, and low mechanical strength in polymeric membranes (Warsinger et al. 2018; Kamali et al. 2019)	Requires a large membrane area and energy, high production costs, uneconomical for most water treatment applications (Abdullayev et al. 2019; Kamali et al. 2019; Asif & Zhang 2021)	Poor filler – polymer compatibilities and filler segregation or blocking of its porosity by the polymer, selection of inorganic filler in order to fabricate MMM is quite challenging owing to their textural characteristics as well as their interaction with polymers (Dechnik et al. 2017; Qadir et al. 2017; Cheng et al. 2018)	Agglomeration of NP due to high surface energy, in the PA layer, which reduces the surface area of nanoparticles as well as creates defects in the active layer, lack of chemical interaction between the PA layer and NP (Ji et al. 2021; Khoo et al. 2021; Peng et al. 2022)	Aggregation or agglomeration of nanoparticles in the PA layer, since nanomaterials and PA matrix do not interact chemically, they are likely to easily leach out during the filtration and/or IP processes (Yang et al. 2020; Zhao et al. 2020; Liao et al. 2021)

Types of membrane	Polymeric membrane Alqaheem & Alomair (2020)	Ceramic membrane Asif & Zhang (2021)	Mixed matrix membrane Maghami & Abdelrasoul (2018)	Thin film composite (TFC) membrane Ji et al. (2021)	Thin film nanocomposite (TFN) membrane Ji et al. (2021)
Strengths	High temperature stability, fouling resistance, low lifecycle costs, simple pore-forming mechanism, and high flexibility (Warsinger et al. 2018; Kamali et al. 2019)	Stronger mechanical, thermal, and chemical stability, higher fouling resistance and more extended membrane longevity, high removal efficiency (Abdullayev et al. 2019; Kamali et al. 2019; Asif & Zhang 2021)	Possibility of combining synergistically the easy processability of polymers, ability to overcome the trade-off between selectivity and permeability which is characteristic for pure polymer membranes, both polymer and filler qualities affect separation performance (Dechnik et al. 2017; Qadir et al. 2017; Cheng et al. 2018)	Able to achieve the optimal combination of flux and salt rejection, inexpensive, have good chemical and mechanical stability and can be formed into membranes using a simple phase inversion technique (Ji et al. 2021; Khoo et al. 2021; Peng et al. 2022)	While maintaining the TFC membrane's separation effectiveness, the interfacial polymerized layer's characteristics can be improved to increase hydrophilicity and/or surface charge density (Yang et al. 2020; Zhao et al. 2020; Liao et al. 2021)
Weaknesses	High selectivity for low solutes and species, membrane fouling, and low mechanical strength in polymeric membranes (Warsinger et al. 2018; Kamali et al. 2019)	Requires a large membrane area and energy, high production costs, uneconomical for most water treatment applications (Abdullayev et al. 2019; Kamali et al. 2019; Asif & Zhang 2021)	Poor filler – polymer compatibilities and filler segregation or blocking of its porosity by the polymer, selection of inorganic filler in order to fabricate MMM is quite challenging owing to their textural characteristics as well as their interaction with polymers (Dechnik et al. 2017; Qadir et al. 2017; Cheng et al. 2018)	Agglomeration of NP due to high surface energy, in the PA layer, which reduces the surface area of nanoparticles as well as creates defects in the active layer, lack of chemical interaction between the PA layer and NP (Ji et al. 2021; Khoo et al. 2021; Peng et al. 2022)	Aggregation or agglomeration of nanoparticles in the PA layer, since nanomaterials and PA matrix do not interact chemically, they are likely to easily leach out during the filtration and/or IP processes (Yang et al. 2020; Zhao et al. 2020; Liao et al. 2021)

View Large

In wastewater treatment, there are many different types of membrane filtration systems that had been applied and they basically rely on pore size to separate contaminants from the feed stream. There are equilibrium-based membrane processes, non-equilibrium-based membrane processes, pressure-driven, and non-pressure-driven processes (Obotey Ezugbe & Rathilal 2020). From pre-treatment to post-treatment, pressure-driven membrane processes appear to be the ones that are most frequently employed in the treatment of wastewater. These methods use the application of hydraulic pressure to achieve separation (Obotey Ezugbe & Rathilal 2020). The most often used methods for treating wastewater, from pre-treatment until post-treatment, are pressure-driven ones where separation is achieved by applying hydraulic pressure. The four different mechanisms are RO, MF, UF, and NF (Obotey Ezugbe & Rathilal 2020). The processes require more energy as the pores get smaller. The technologies are thus commonly divided into low pressure (MF, UF) and high pressure (NF, RO) categories (Tchobanoglous 2014). Both MF and UF operate under low pressure (2 bar). Separation by MF is primarily achieved through sieving and is limited to the removal of due to the relatively large pore size, separation by MF is primarily accomplished through sieving and is restricted to the removal of suspended solids or particles, bacteria, and, to a lesser extent, organic colloids (approximately 0.1–1.0 m) (Warsinger et al. 2018). UF membranes work similarly to MF membranes, but have a wider separation range and are able to remove particles, bacteria, viruses, and colloids depending on the pore size (usually between 0.01 and 0.1 m) (Warsinger et al. 2018). Both NF and RO are very similar processes that are designed to remove dissolved chemical constituents such as salts, where both require high hydraulic pressures and use similar membrane materials (Warsinger et al. 2018). NF removes many of the same solutes as RO, but to a lesser extent. The UF and MF processes are suggested over the NF and RO processes, which would additionally lower operating costs, in order to maintain a low transmembrane pressure and achieve a higher flux (Goswami et al. 2019).

The International Union of Pure and Applied Chemistry classifies porous materials into three grades based on the diameter of their pores (D), microporous (D < 2 nm), mesoporous (2 nm ≥ D < 50 nm), and macroporous (D > 50 nm). Ceramic membranes are classified as industrial porous materials (Ali et al. 2017). In water and wastewater treatment, ceramic membranes provide various advantages over polymer-based filtration systems, including high temperature stability, fouling resistance, and low maintenance requirements, all of which lead to lower lifecycle costs (Abdullayev et al. 2019).

When the pore size is small, such as in RO, UF, and NF, the fluid permeability is also influenced by the solute and solvent's affinity with the porous materials (Ohji & Fukushima 2012). Although it offers a high removal efficiency, using membrane separation techniques to treat contaminated water uses a lot of energy and membrane area. Therefore, it is crucial to create new membranes based on inexpensive components like clay and apatite that will be an effective treatment, especially when they are burned to create ceramic membranes that are regarded as industrial porous materials (Masmoudi et al. 2005; Khemakhem et al. 2007). However, the high production costs of most commercially available ceramic membranes, owing to raw material and processing costs, make such systems uneconomical for most water treatment applications (Abdullayev et al. 2019).

Polymers are probably the most widely used membrane material in wastewater treatment because of their advantages such as a simple pore-forming mechanism, low cost, and high flexibility. This includes ED, UF, NF, and RO membrane-based technologies (Ng et al. 2013). Experimental evidence has shown that polymeric membranes are effective at removing a wide range of contaminants, including suspended pollutants, organic matter, and organic and inorganic compounds (such as heavy metal ions) (Wieszczycka & Staszak 2017). When polymeric membranes (typically made from naturally occurring or chemically synthesized polymers) are used to separate two adjacent phases (feed and permeate), they create a selective interface barrier that regulates the movement of species between the two phases (Castro-Muñoz et al. 2021).

The hydrophobic polymer polyvinylidene fluoride (PVDF), which is frequently used in a range of water treatment applications, with high chemical tolerance, as well as favourable mechanical and thermal properties, are just a few of its many benefits (Xia & Ni 2015). The performance of membrane separation is typically influenced by the characteristics of the transported species (such as molecule size, shape, and chemical nature) as well as the physicochemical characteristics (hydrophilicity/hydrophobicity, surface charge, roughness), particularly its porous structure, of the polymer membrane. Unfortunately, polymeric membranes still do not have all of the desired characteristics for membrane separation processes, such as high selectivity towards low solutes and species. Additionally, the main obstacle to their application is membrane fouling and polymeric membranes' weak mechanical strength (Castro-Muñoz et al. 2021).

The development of membrane technology currently focuses on the addition of inorganic particulates to create mixed matrix membranes (MMMs), which are filled hybrid polymer composite membranes (Muthukumaraswamy Rangaraj et al. 2020). By combining an inorganic or inorganic–organic hybrid material in the form of micro- or nanoparticles (the discrete or dispersed phase; additive or filler) with a polymer matrix, composite membranes known as MMMs are developed (Dechnik et al. 2017). Specifically, the complicated process of incorporating various nanomaterials into membrane matrices can give the membrane's structure unique properties and affect the behaviour of gas transport. MMMs are a viable alternative to traditional polymeric membranes because they have the potential to eliminate the ‘trade-off’ effect found in commercial membranes and one of the most important parameters for designing efficient MMMs is the choice of filler material (Muthukumaraswamy Rangaraj et al. 2020).

Several approaches have been suggested over the last few decades to improve the performance of polymeric membranes where MMMs have been created by modifying various polymers with inorganic fillers such as zeolites, mesoporous silicas, activated carbons, carbon nanotubes (CNTs), and even non-porous solids (Zornoza et al. 2013). MMM is an intimate and homogeneous dispersion of filler particles in a polymeric matrix, where the properties of both the polymer and the filler influence MMM morphology and separation performance (Zornoza et al. 2013). The MMM has a bright future in gas separation because of its outstanding physical and chemical properties when compared to organic and inorganic membranes, but on the other hand, it encounters some challenges because of interfacial defects or adhesion between the organic and inorganic phases, possibly requiring countless modifications to address these problems (Nasir et al. 2013).

One of the most crucial elements in determining separation properties and transport mechanisms across the membrane is membrane structure. Since their production typically only requires one step of flat sheet casting or hollow fibre spinning, integrally asymmetric polymeric membranes have been utilized in numerous membrane processes for decades (Nguyen et al. 2015). Nevertheless, this new kind of membrane has drawn a lot of interest in the research and industrial sectors ever since the first polyamide TFC membrane was introduced because it can achieve the ideal balance between flux as well as salt rejection (Zhou et al. 2014; Kim et al. 2016). TFC membranes are broadly used in NF, RO, and FO, mostly for seawater and brackish water desalination. The versatility in selecting and optimizing the materials used to form a microporous substrate layer and a selective thin film layer is another appealing quality of TFC membrane (Xu et al. 2017a, 2017b).

In MF and UF processes, polymers with outstanding permeability, selectivity, and chemical, mechanical, and thermal stability are frequently used as membrane materials. Examples of these polymers include PSF, PES, polyacrylonitrile (PAN), polypropylene (PP), and polytetrafluoroethylene (PTFE) (Ursino et al. 2018). For UF membranes, PSF and PES membranes have seemed to be among the most frequently used materials. These standard base polymers are also used in the creation of NF and RO composite membranes, whereas PP and PVDF are more frequently used for MF membranes (Pendergast & Hoek 2011). However, these polymeric membranes' separation performance must be optimized and improved (Alzahrani & Mohammad 2014), along with their stability, hydrophilicity profile, and fouling resistance (Abu Bakar & Tan 2016).

A more recent type of composite membrane produced by the IP procedure are TFN membranes. In order to improve the properties of the interfacial polymerized layer, such as continuing to increase the hydrophilicity and/or surface charge density, nanoparticles are added to the thin polyamide (PA) dense layer of the TFC membrane without reducing the separation efficiency of the TFC membrane (Lau et al. 2015). Before membrane casting, nanoparticles can be either coated onto the membrane surface or dispersed in the polymer solution (Al Aani et al. 2017). The most typical way for membrane preparation is to blend nanoparticles into the polymeric matrix, however membrane surface coating with nanoparticles is becoming commonly employed to adjust membrane functionality (Lau et al. 2012). Incorporating NPs into the TFC membrane during the IP process has the potential to be one of a promising method. Researchers in the field of membrane science have used a variety of inorganic nanofillers with various properties, such as titanium dioxide (TiO₂), graphene oxide (GO), CNTs, titania nanotubes (TNTs), zinc oxide (ZnO), and silicon dioxide (SiO₂), to enhance the performance of TFN membranes for both pressure-driven and osmotically driven filtration processes (Ji et al. 2017; Lee et al. 2019).

Nanoparticles can be divided into many categories based on their size, shape, physical, and chemical properties. Another way to categorize nanoparticles is based on their morphology, size, and chemical characteristics. There are basically two types of nanoparticles: organic and inorganic. The first category includes micelles, dendrimers, liposomes, hybrid, and compact polymeric nanoparticles. The second category includes fullerenes, quantum dots, silica, and metal nanoparticles. Numerous studies have used various nanoparticles, such as carbon-based nanoparticles – carbon quantum dots (CQDs) and GO, ceramic-based nanoparticles – silica/silicon dioxide (⁠⁠) and titania/titanium dioxide (⁠⁠), metal oxide nanoparticles – copper oxide (CuO), semiconductor nanoparticles – zinc oxide (ZnO), polymeric nanoparticles – cellulose and polylactic acid (PLA), and lipid-based nanoparticles. Table 2 summarizes some examples from previous works that had used nanoparticles in TFN membrane for heavy metal removal.

Table 2

TFN membrane for heavy metal removal

Substrate layer .	Nanoparticle .	Nanoparticle loading .	Contact angle .	Pure water flux .	Removal % .	Reference .
PSF	GO	0.000 wt%	64 ± 0.50	12.5	Pb (79.3), Cd (81.8), Cr (81.3)	Saeedi-Jurkuyeh et al. (2020)
		0.004 wt%	54 ± 0.37	16.4	Pb (97.6), Cd (97.4), Cr (97.2)
		0.008 wt%	49 ± 0.40	34.3	Pb (99.9), Cd (99.7), Cr (98.3)
		0.012 wt%	53 ± 0.70	19.4	Pb (98.7), Cd (98.3), Cr (97.9)
PSF		0.00 wt%	61.4	13–15	Cu 80–85	Karki & Ingole (2022)
					Hg 70–80
					Pb 80–85
		0.01 wt%	49.4	13–15	Cu 86.89 (Highest)
					Hg 70–75
					Pb 80–85
		0.02 wt%	39.3	26.75 (Highest)	Cu 80–85
					Hg 77.01 (Highest)
					Pb 83.42 (Highest)
		0.05 wt%	50.3	15–18	Cu 70–75
					Hg 50–55
					Pb 60–65
		1.00 wt%	54.0	15–18	Cu 70–75
					Hg 45–50
					Pb 60–65
PSF	UiO-66-	0.00 wt%	36.63	31.15 (NF) 9.4527 (FO)	Pb (II) 50 ppm (90)	Abdullah et al. (2021)
					Pb (II) 250 ppm (87.5)
					Pb (II) 500 ppm (84.95)
		0.005 wt%	38.18	41.51 (NF) 12.94 (FO)	Pb (II) 50 ppm (90–93)
					Pb (II) 250 ppm (90–93)
					Pb (II) 500 ppm (80–90)
		0.01 wt% (Optimum)	27.79	60.41 (NF) 18.51 (FO)	Pb (II) 50 ppm (93)
					Pb (II) 250 ppm (90.2)
					Pb (II) 500 ppm (88.93)
		0.025 wt%	37.09	47.60 (NF) 12.43 (FO)	Pb (II) 50 ppm (70–80)
					Pb (II) 250 ppm (80–90)
					Pb (II) 500 ppm (80–90)
		0.05 wt%	23.03	47.95 (NF) 10.29 (FO)	Pb (II) 50 ppm (60–70)
					Pb (II) 250 ppm (70–80)
					Pb (II) 500 ppm (70–80)
		0.1 wt%	20.92	71.47 (NF) 12.63 (FO)	Pb (II) 50 ppm (50–60)
					Pb (II) 250 ppm (70–80)
					Pb (II) 500 ppm (60–70)
PES	CQDs	0.00 wt%	54.6 ± 2.2	5.25–6.0	82.40	He et al. (2018)
					91.30
		0.02 wt%	34.9 ± 2.0	8.25–9.0	97.20
					98.10
		0.05 wt%	29.1 ± 1.8	9.75–10.0	97.60
					98.30
		0.08 wt%	26.5 ± 1.9	7.5–8.25	98.00
					98.95
		0.12 wt%	23.5 ± 1.4	6.0–6.75	98.60
					99.50

Substrate layer .	Nanoparticle .	Nanoparticle loading .	Contact angle .	Pure water flux .	Removal % .	Reference .
PSF	GO	0.000 wt%	64 ± 0.50	12.5	Pb (79.3), Cd (81.8), Cr (81.3)	Saeedi-Jurkuyeh et al. (2020)
		0.004 wt%	54 ± 0.37	16.4	Pb (97.6), Cd (97.4), Cr (97.2)
		0.008 wt%	49 ± 0.40	34.3	Pb (99.9), Cd (99.7), Cr (98.3)
		0.012 wt%	53 ± 0.70	19.4	Pb (98.7), Cd (98.3), Cr (97.9)
PSF		0.00 wt%	61.4	13–15	Cu 80–85	Karki & Ingole (2022)
					Hg 70–80
					Pb 80–85
		0.01 wt%	49.4	13–15	Cu 86.89 (Highest)
					Hg 70–75
					Pb 80–85
		0.02 wt%	39.3	26.75 (Highest)	Cu 80–85
					Hg 77.01 (Highest)
					Pb 83.42 (Highest)
		0.05 wt%	50.3	15–18	Cu 70–75
					Hg 50–55
					Pb 60–65
		1.00 wt%	54.0	15–18	Cu 70–75
					Hg 45–50
					Pb 60–65
PSF	UiO-66-	0.00 wt%	36.63	31.15 (NF) 9.4527 (FO)	Pb (II) 50 ppm (90)	Abdullah et al. (2021)
					Pb (II) 250 ppm (87.5)
					Pb (II) 500 ppm (84.95)
		0.005 wt%	38.18	41.51 (NF) 12.94 (FO)	Pb (II) 50 ppm (90–93)
					Pb (II) 250 ppm (90–93)
					Pb (II) 500 ppm (80–90)
		0.01 wt% (Optimum)	27.79	60.41 (NF) 18.51 (FO)	Pb (II) 50 ppm (93)
					Pb (II) 250 ppm (90.2)
					Pb (II) 500 ppm (88.93)
		0.025 wt%	37.09	47.60 (NF) 12.43 (FO)	Pb (II) 50 ppm (70–80)
					Pb (II) 250 ppm (80–90)
					Pb (II) 500 ppm (80–90)
		0.05 wt%	23.03	47.95 (NF) 10.29 (FO)	Pb (II) 50 ppm (60–70)
					Pb (II) 250 ppm (70–80)
					Pb (II) 500 ppm (70–80)
		0.1 wt%	20.92	71.47 (NF) 12.63 (FO)	Pb (II) 50 ppm (50–60)
					Pb (II) 250 ppm (70–80)
					Pb (II) 500 ppm (60–70)
PES	CQDs	0.00 wt%	54.6 ± 2.2	5.25–6.0	82.40	He et al. (2018)
					91.30
		0.02 wt%	34.9 ± 2.0	8.25–9.0	97.20
					98.10
		0.05 wt%	29.1 ± 1.8	9.75–10.0	97.60
					98.30
		0.08 wt%	26.5 ± 1.9	7.5–8.25	98.00
					98.95
		0.12 wt%	23.5 ± 1.4	6.0–6.75	98.60
					99.50

View Large

From previous research, Saeedi-Jurkuyeh et al. (2020) discovered that at neutral pH, the rejection rate of a single heavy metal was way greater in the TFN 0.008 membrane than the TFC membrane. According to Table 2, water flux increased as GO concentration in membranes continued to increase. For instance, the water flux in the TFN 0.008 membrane (34.3 LMH) is essentially higher than in the conventional TFC membrane (12.5 LMH), which can be accounted for by the following factors: The presence of functional groups on the surface of GO improved the roughness and reduced the resistance to water diffusion through the active layer of TFN membranes. TFN 0.008 FO membranes rejected 99.9% of Pb, 99.7% of Cd, and 98.3% of Cr, respectively. In terms of membrane structure, FO performances, and separation properties, those modified FO membranes showed excellent improvement.

According to Karki & Ingole (2022), the TFN membrane with 0.02 wt% of TiO₂ has the highest in pure water flux of all, with 26.75 L m⁻²h⁻¹. Table 2 clearly shows that all TFN membranes have greater permeability than TFC membranes. The data showing that TFC has the lowest permeability of any membrane are supported by the combination of a higher contact angle value and a lower surface roughness (Ra). The membranes permeated pure water at a rate of 13–27 LMH. The TFN-2 membrane had the highest permeability of all, with 26.75 LMH of pure water. The permeability of the membranes increases as the concentration of nanomaterial increases. However, as shown in Table 2, regardless of having higher nanomaterial concentrations than TFN-2, the TFN-3 and TFN-4 membranes failed to show higher permeability than the TFN-2 membrane. As previously reported, this is because the functionalized TiO₂ nanoparticles self-aggregated in the solution during IP. In comparison to the TFN-3 and TFN-4 membranes, the TFC, TFN-1, and TFN-2 membranes demonstrated excellent rejection of Cu, Hg, and Pb ions from their respective feed solutions during NF performance. The membranes displayed the % rejection 70–87% for Cu ion, 45–70% for Hg ion, and 60–85% for Pb ion. TFN-1 membrane had the maximum Cu ion rejection (86.89%), whereas TFN-2 membrane seemed to have the maximum rejection of the remaining ions (Hg – 77.01% and Pb – 83.42%).

All TFN/UiO-66- membranes have higher water fluxes than the TFC membrane for both NF and FO membranes, as shown in Table 2 (Abdullah et al. 2021). According to the table, the flux obtained by the FO process has a lower value for TFN/UiO-66- 0.01 compared to the flux acquired by the NF process, which is 18.51 and 60.41 LMH, respectively. Flux steadily increased at low loading of TFN/UiO-66-NH₂ from 0.01 to 0.025 wt% and significantly decreased at 0.05 and 0.1 wt% loading. The hydrophilic properties of UiO-66-NH₂ and the additional water transport pathways provided by the UiO-66-NH₂ particles are responsible for the increased water flux. Table 2 shows the removal of Pb (II) at various starting values ranging from 50 to 500 mg/L. In each concentration, the removal trend is comparable, with TFN rejecting more than TFC at loadings of 0.005 and 0.01 wt% before declining at loadings of 0.025, 0.05, and 0.1 wt% for TFN. TFC and TFN/UiO-66-NH₂ rejection rates were 90 and 93%, respectively, at 50 ppm, 87.5 and 90.2% at 250 ppm, and 84.59 and 88.93% at 500 ppm, respectively. As can be seen, increasing the beginning concentrations resulted in a modest drop in Pb (II) rejections. The results showed that when the concentration is less than 100 ppm, the metal ion rejection is near to that predicted by the Donnan exclusion. As a result, at lower concentrations, Donnan exclusion is the primary influencing factor, whereas the solvation energy barrier effect plays a secondary role in ion rejection.

Water flux is reportedly decreased when more Na-CQDs are added to the PA layer, according to another study by He et al. (2018). This might be a result of a thicker PA layer, as hydrophilic Na-CQDs increase phosphatidylinositol phosphate (PIP) diffusivity and hasten its transit to the interface. Additionally, the PA layer's pores could become clogged as a result of higher nanoparticle loading. When using single ion solutions as feeds, the Na-CQD loading has an impact on how the membrane separates to Se and As ions. The optimal loading of Na-CQDs is 0.12 wt% (TFN-0.12). Rejection to rises from 82.4 to 98.6% when compared to the TFC membrane, while rejection to rises from 91.3 to 99.5%. As shown in Table 2, Na-CQDs create a selective layer with small pore sizes and a narrow pore size distribution, which is largely responsible for these advantages. However, the pure water permeance (PWP) of the TFN-0.12 membrane is not the highest.

Recently, several new varieties of nanocomposite membranes have entered the market for use in various filtration applications. The characteristics of nanocomposite membranes that were used for industrial water treatment will be based on the production of tailored membranes with high selectivity, competitive flux, and self-cleaning properties while also meeting sustainability criteria such as environmental impact, flexibility, adaptability, and simplicity of use (Le & Nunes 2016). The hydrophilicity of polymeric membranes is generally increased by adding different classes of NPs, which reduces fouling in water treatment and enhances some of the polymers' other properties (mechanical, thermal, and chemical) (Ursino et al. 2018).

Another issue in TFN membranes is the incompatibility of nanomaterials with the PA matrix. Nanomaterials are likely to easily leak out during the IP and/or filtration process due to the lack of chemical interaction between them and the PA matrix, which reduces the effectiveness of the prepared TFN membrane and the effectiveness of using nanoparticles in production (Lau et al. 2015). The initiative for reducing NP release during the membrane process is to control the effective adhesion on the polymeric matrix and evaluate the stability over time and as a result, producing straightforward, inexpensive techniques to immobilize NPs on membranes without compromising their performance is one of the challenges (Qu et al. 2013). Numerous research studies are concentrated on the release of NPs, which presents a notable technical challenge for risk assessment, as well as the detection techniques, which are actually limited in number, typically sophisticated, expensive, and fraught with numerous limitations (Ursino et al. 2018). Few incompatibilities between inorganic nanomaterials and organic PA layers have been found in prior studies, but reports state that no thorough research has been done to assess nanoparticle leaching during prolonged filtration (Lau et al. 2015).

Membrane fouling, the most important limiting factor in industrial membrane applications, is yet another key problem with TFN membrane. When different solutes gather on the membrane surface or internal structure, membrane fouling occurs, generating an extra barrier at the membrane surface or clogging the internal pores (Al Aani et al. 2017). This prevents the solvent from passing through the membrane, lowering the permeability and increasing the transmembrane pressure necessary to sustain productivity and as a result, the membrane's lifespan is reduced. Additionally, clogged membranes would require a lot of cleaning agents, which could harm the membrane surface and, in extreme cases, necessitate replacing the membrane. As a result of all of these issues, the cost of operating and maintaining the water treatment unit would increase (Contreras 2011). Fouling phenomena could be somewhat reduced by selecting the right membrane materials, operating designs, pre-treatment techniques, and environmental conditions, but membrane sustainability at the industrial scale still poses a challenge because of its complexity and variety (Al Aani et al. 2017). In an effort to lessen their effects, membrane fouling phenomena have received extensive attention for a long period of time from a variety of angles, such as acknowledging fouling mechanisms, incidence, types, and factors influencing fouling growth.

Membrane surface modification is also a fundamental strategy in the fouling reduction approach. Choosing an acceptable technique for modifying the membrane surface is just as crucial as selecting specific nanoparticles. Recently, the use of engineered nanomaterials (ENMs) in polymeric membrane matrices for the treatment of water and wastewater has drawn a lot of interest. The ultimate objective of membrane development is the creation of nanocomposite membranes that preserve the benefits of polymeric membranes while addressing their drawbacks through the integration of ENMs (Kango et al. 2013; Qu et al. 2013). In contrast to conventional membranes, nanocomposite membranes are a novel type of membrane made of both organic polymers and inorganic nanoscale components. Various different ENMs have been used in countless studies to create novel composite polymer membranes for water treatment applications. Over the last decade, researchers have incorporated a wide range of solid nanofillers into polyamide layers, including titanium (⁠⁠), (Ag) cerium oxide, silver, GO, CNTs, aluminium (⁠⁠), silicon (⁠⁠), iron (⁠⁠), zirconium (⁠⁠), zinc (ZnO), clay nanoparticles, and zeolite (NaX) (Yang et al. 2017, 2019; Pang & Zhang 2018; Zhao & Chung 2018; Wang et al. 2019).

The surfaces have been altered with organic functionalities to prevent agglomeration and inadequate dispersion of nanoparticles in polymer matrices. Unmodified nanoparticles will cause polymer nanocomposites' properties to deteriorate without this surface modification. Surface-modified nanomaterials (SMNs) have received much attention in comparison to unmodified nanomaterials due to the outstanding interfacial interaction between the surface of the nanoparticles and polymers. Surface modification of NPs provides four primary purposes: (1) to enhance or alter magnetic nanoparticles (MNP) dispersion; (2) to boost MNP surface activity; (3) to improve MNP physicochemical and mechanical qualities; and (4) to enhance MNP biocompatibility (Zhu et al. 2018).

The surface-functionalization of nanomaterials involves the use of a variety of organic functional groups, such as alcohols, thiols, sulphonic, carboxylic acids, and amines. To alter the surfaces of nanomaterials, a variety of techniques have been employed, including copolymerization of functional organ silanes, macromonomers, and metal alkoxides, (ii) functionalization of organic components within sol-gel-derived silica or metallic oxides, (iii) organic functionalization of nanotubes, nano clays, or other compounds with lamellar structures, and (iv) other methods (Vengatesan & Mittal 2015). The outstanding hybrid nanocomposites produced by SMNs reinforced into polymer matrices have low weight, high strength, and improved thermal conductivity.

Adsorbents had served as vital components in TFN membranes, which are used in a variety of separation procedures. There are various advantages of using adsorbents in TFN membranes for heavy metal removal. For example, previous research (Qiu & He 2019) had applied zeolite to a PA layer in order to improve membrane water permeability. As stated in those reports, the water permeability of all TFN with different wt% of zeolite is higher compared to the TFC and the values increase with the increasing of loading (Qiu & He 2019). It is vital to note that having high water permeability while also effectively removing heavy metals may require a trade-off. Increasing permeability may reduce adsorption capacity, while decreasing permeability may increase adsorption capacity. As a result, balancing these factors based on the application's specific requirements is critical. Combining adsorbent and metal oxide to produce advanced nanocomposite materials is another approach to reduce agglomeration issues of nanoparticles. In a different study (Zhu et al. 2017), they created a TFN membrane by combining a PA layer with a hydrophilic nanoparticle called ZIF-8 that has been modified with poly(4-styrene sulphonate). The hydrophilicity of the surface of the PA layer increased, thereby enhancing water permeability. From those previous works, it can be said that adsorbents can help to increase the hydrophilicity of TFN membranes, allowing more water to pass through. Hydrophilic adsorbent particles can lessen the contact angle of the membrane surface, permitting water molecules to penetrate and flow more freely. This results in increased water permeability and better filtering or separation performance.

Various processing processes are used to create the nanocomposites, which normally are made up of organic polymers and inorganic nanoparticles. Metal, metal oxide, and carbon-based nanomaterials are some of the examples of nanoparticles that have been extensively used in the fabrication of hybrid polymer nanocomposites (Vengatesan & Mittal 2015). A wide variety of chemical forms and various geometric surface structures are accessible for solid materials that had been used as adsorbents. The primary characteristics of the materials that are extensively employed as conventional adsorbents for heavy metal removal from wastewater are high specific surface area, high ion exchange capacity, low cost, and their abundance in nature (Atkovska et al. 2018). Clay minerals (with a special emphasis on bentonite), zeolites, activated carbon, and metal oxides are some of the most often used adsorbents. The table of classification for the adsorbents was introduced by the previous researcher (Crini et al. 2018), as shown in Table 3.

Due to their strong chemical stability, low density, structural diversity, and applicability for field-scale applications, activated carbon-based adsorbents are extensively studied by researchers for adsorption (Crini et al. 2019). The distinctive properties are typically a result of their internal pore morphology, surface properties, porosity, pore volume, chemical structure, and presence of functional groups from their source material, including their activation (Crini et al. 2019; Sonal et al. 2020). Because of its high specific surface area and enormous volume of micropores and mesopores, activated carbon is frequently employed to remove heavy metal ions (Atkovska et al. 2018). Carbons are actually one of the oldest and most commonly used adsorbents in the industry. Furthermore, commercial activated carbons have also been considered as the most effective adsorbents because of their high capacity to adsorb contaminants, they produce an excellent output if the adsorption system is developed appropriately (Crini et al. 2019). This ability is mainly due to their structural qualities, porous texture, which gives them a large surface area, and chemical composition, which can be easily changed by chemical treatment to vary their properties. The processes that use these widely known adsorbents are typically carried out in batch mode, either by continuously passing the contaminated solution over a packed bed of carbon or by adding activated carbon to a vessel that already contains the contaminated solution (Crini et al. 2019).

Another significant advantage of using commercial activated carbons is that no by-products are formed, as compared to chemical oxidation treatment. Activated carbon powder is used in conjunction with a UF membrane or other processes, such as oxidation, to provide cost-effective technology (particularly in the field of water recycling) (Crini et al. 2019). Although activated carbons have been used for a long time, research and development are continuously ongoing, particularly due to the growing demand for very clean water. Despite all advantages that had been mentioned, activated carbon technology also has some drawbacks. Commercially available activated carbon is quite pricey and non-selective (the higher the quality, the higher the cost). Even though the great absorbing power of active carbons no longer needs to be demonstrated, there is still the issue of disposal of spent activated carbons, as well as the disadvantage of their rapid saturation and thus their regeneration. Because of these factors, their general use is limited; in particular, small- and medium-sized businesses cannot afford such therapy (Crini et al. 2019).

Considering conventional treatment methods do not completely remove all contaminants, the researchers focused on developing technologies that will successfully remove these new contaminants (Atkovska et al. 2018). Metal oxide nanoparticles are gaining popularity among researchers as a viable alternative for wastewater treatment (Liu et al. 2019; Gautam et al. 2020; Li et al. 2020; Shaheen et al. 2020) due to the fact that metal oxide is extremely effective at removing numerous types of air, water, and soil contaminants via adsorption and photodegradation. Therefore, researchers started to focus on surface modification. Coating, stabilization, and functionalization can all be used to modify the surface of metal oxide nanoparticles. However, metal oxides are prone to agglomeration due to van der Waals forces or other interactions (Sousa & Teixeira 2020), causing a reduction in capacity and selectivity. To improve the application of nanomaterial oxides in wastewater treatment, they had been combined with an AC porous matrix. This is because magnetic adsorbents have recently received a lot of attention for water treatment because of their ease of separation and collection by a magnet. Furthermore, combining with magnetite particles can reduce particle agglomeration (Alorabi et al. 2020).

The addition of transition metal oxide nanoparticles to activated carbon as a result of growing research efforts has also produced synergistic benefits for the adsorptive removal of dyes from wastewater (Anjum et al. 2019; Nayeri & Mousavi 2020; Nasrollahzadeh et al. 2021). It has been reported that metal oxides, in addition to functional groups of activated carbon, provide numerous hydroxyl and oxygen linkage sites for mutual interaction, considerably improving the adsorption capacity of activated carbon-based metal oxide nanocomposite (Farooq et al. 2022). For example, iron oxide (⁠⁠) (Alorabi et al. 2020), copper oxide (CuO) (Peternela et al. 2017; Vidovix et al. 2019), and zinc oxide (ZnO) (Saini et al. 2017; Muntean et al. 2018) loaded activated carbon have shown a surprisingly high cationic dye adsorption capacity due to effective dye inter-diffusion. Because of its stable and varied pore structure, activated carbon (AC) has proven to be an excellent support. A previous study (Liu et al. 2011) has reported on /GO composites that had greater adsorption than ⁠. Another previous researcher who used stated that it can attract water molecules in surface coatings (hydrophilicity) but does not form clusters due to particle dispersion (Jain et al. 2018). In that previous report a statement was also given that surface coating of iron oxide by activated carbon enhances their sorption capabilities because surface coating phenomenon aids in turning the closely packed cubic geometry of magnetic nanoparticles into compact and strong nanoparticles that helps to reduce aggregation (Ojemaye et al. 2017; Table 4).

Table 4

Summarized previous research on metal oxides-adsorbent nanocomposites

Metal oxides .	Adsorbents .	Surface area (⁠⁠) .	Adsorption capacity (mg/g) .	Reference .
	AC	– 41.0	Removal of Cr (VI)	Kaur et al. (2021)
		AC/ 1:2 – 441.93	– 83.33
		AC/ 1:1 – 544.69	AC/ 1:2 – 250.00
		AC/2:1 – 549.30	AC/ 1:1 – 333.33
		AC – 1045.49	AC/2:1 – 500.00
			AC – 500.00
	AC	Sunflower head waste	(Cr (VI)) – 5.50	Jain et al. (2018)
		(SHW) – 0.99	(Cu (II))– 2.70
		AC – 25.07	(Cd (II))– 0.09
		– 47.87	AC/ (Cr (VI)) – 8.06
		AC/ – 51.10	AC/ (Cu (II)) – 3.20
			AC/(Cd (II)) – 2.15
	Bentonite	Pure Bentonite – 78	Removal of Cd (II)	Abbasi et al. (2020)
ZnO		Bentonite/ – 314.7	Bentonite/ – 13.14
		Bentonite/ – 321.9	Bentonite/ – 12.03
GO	ZeO	ZeO – 72.739	Removal of Cd (II)	Liu et al. (2020)
		GO/ZeO – 64.74	ZeO – 149
			GO/ZeO – 196
GO		GO – 292	Removal of Pb (II)	Barik et al. (2020)
		– 522	GO- – 527
		GO- (water) – 252.5	Removal of As (III)
		GO- (IL) – 858	GO- – 30

Metal oxides .	Adsorbents .	Surface area (⁠⁠) .	Adsorption capacity (mg/g) .	Reference .
	AC	– 41.0	Removal of Cr (VI)	Kaur et al. (2021)
		AC/ 1:2 – 441.93	– 83.33
		AC/ 1:1 – 544.69	AC/ 1:2 – 250.00
		AC/2:1 – 549.30	AC/ 1:1 – 333.33
		AC – 1045.49	AC/2:1 – 500.00
			AC – 500.00
	AC	Sunflower head waste	(Cr (VI)) – 5.50	Jain et al. (2018)
		(SHW) – 0.99	(Cu (II))– 2.70
		AC – 25.07	(Cd (II))– 0.09
		– 47.87	AC/ (Cr (VI)) – 8.06
		AC/ – 51.10	AC/ (Cu (II)) – 3.20
			AC/(Cd (II)) – 2.15
	Bentonite	Pure Bentonite – 78	Removal of Cd (II)	Abbasi et al. (2020)
ZnO		Bentonite/ – 314.7	Bentonite/ – 13.14
		Bentonite/ – 321.9	Bentonite/ – 12.03
GO	ZeO	ZeO – 72.739	Removal of Cd (II)	Liu et al. (2020)
		GO/ZeO – 64.74	ZeO – 149
			GO/ZeO – 196
GO		GO – 292	Removal of Pb (II)	Barik et al. (2020)
		– 522	GO- – 527
		GO- (water) – 252.5	Removal of As (III)
		GO- (IL) – 858	GO- – 30

View Large

Based on these findings, efforts have been made to incorporate nanocomposite metal oxide-adsorbents into membranes for heavy metal removal applications. In order to improve the stability and performance of heavy metal removal, He et al. (2020) developed a sulphonated graphene oxide @ metal organic framework-modified forward osmosis nanocomposite (SGO@UiO-66-TFN) membrane. The stability of the nanomaterial in the membrane was improved by the cross-linking of SGO@UiO-66 and m-phenylenediamine. Moreover, the new water channels created by SGO@UiO-66 improved the water permeability. The diffusion path of the solute was made more circuitous by the lamellar structure of SGO and the frame structure of UiO-66, which enhanced salt rejection and heavy metal removal. The hydrophilic layer of the SGO@UiO-66-TFN membrane was also capable of blocking impurities and loosen the structure of the pollution layer, helping to ensure that the membrane remained at a high removal rate. In comparison to the thin-film composite membrane, the SGO@UiO-66-TFN membrane's water flux and reverse solute flux were 14.77 LMH and 2.95 gMH, respectively, and managed to improve by 41 and 64%. Removing heavy metal ions from the membrane was another area of success. More than 99.4% of the heavy metal ions were removed in just 2 h, and more than 97.5% in just 10 h.

Zhang et al. (2018), stated in their study, in which MWCNTs-COOH were modified with poly(amidoamine) (PAMAM), and the as-produced MWCNTs-PAMAM were embedded into the inner polyamide skin-layer of the NF membranes. It is discovered that the rigid MWCNTs and dendrimer PAMAM molecules provide the NF membranes with outstanding porosity and hydrophilicity in their as-fabricated state. The NH₂ groups in PAMAM also offer a few positive sites in the polyamide layer. The most effective membrane PIP/MWCNTs-PAMAM (MPCP) membrane has a water flux of 12.2⁠, which is 1.7 times greater than that of the unaltered NF membranes, while experiencing a slight decline in rejection from 94.9 to 92.6%. In fact, the MWCNTs-PAMAM addition created positive sites in the polyamide layer, giving the MPCP membrane a potent sulphate and metal cation rejection.

Even though there have been numerous studies on TFC membrane development over the past few decades, there is still a pressing need to develop TFC membranes that are both technically and economically more feasible and have excellent separation performance and low energy requirement for industrial applications. The physicochemical properties of TFN membranes, such as hydrophilicity, mechanical stability, thermal resistance, and perm selectivity, can be improved, according to recent research (Kim et al. 2013; Zhao et al. 2014). Three different types of nanoparticles are used as nanofillers in TFN membranes: one-dimensional nanotubes, two-dimensional porous NPs, and zero-dimensional metal and metal oxides (Dong et al. 2015). Both direct NP deposition on TFC membranes as well as encapsulating NPs inside the TFC membrane during the IP stage can be used to create TFN membranes. A post-modification method called surface deposition is frequently used to improve the antifouling properties of membranes. However, the high-pressure filtration process could cause the deposited NPs to leak. Consequently, to increase NP stability in the polymer matrix, chemical functionalization techniques such as NP surface modification are needed (Zargar et al. 2017).

Incorporating NPs into the TFC membrane during the IP process has the potential to be one of a promising method. Researchers in the field of membrane science have used a variety of inorganic nanofillers with various properties, such as titanium dioxide (TiO₂), GO, CNTs, TNTs, zinc oxide (ZnO), and silicon dioxide (SiO₂), to enhance the performance of TFN membranes for both pressure-driven and osmotically driven filtration processes (Ji et al. 2017; Lee et al. 2019). Zeolite-A NPs (50–200 nm) had been incorporated into the polyamide (PA) layer in-situ by earlier researchers, who had created the first TFN membrane. These zeolite-incorporated TFN membranes showed improved separation performance and fouling resistance (Jeong et al. 2007). In the membrane matrix, NP stability and dispersion play a key role in the long-term structural and performance improvement of TFN membranes (Kang & Cao 2012). It is anticipated that these NPs added via IP will be more stable overall than those added via deposition. This is due to the fact that during the IP process, a large number of NPs are partially or entirely enclosed within the PA thin film layers (Zargar et al. 2017).

In a previous research report, FeO nanoparticles have been used to prepare surface-modified nanocomposite membranes. However, one of the main drawbacks of using FeO nanoparticles is their rapid aggregation during membrane preparation, which decreases their hydrophobic properties, obstructs surface pores, and worsens membrane fouling. In order to prevent FeO nanoparticles from aggregating, an intermediate pre-seeding step is carried out. This step involves introducing a hexane solution containing ethanol, nanoparticles, and TMC at a low concentration between the MPD and TMC layers. This helps to ensure that FeO nanoparticles are properly dispersed on the membrane surface (Lakhotia et al. 2019). As demonstrated in Figure 5, this results in an IP zone at the same layer of the membrane where the nanoparticles are present (neither above nor below the nanoparticles).

In another previous research that had been incorporating NPs into the TFC membrane (Soo et al. 2020), they used TNT- as nanoparticles to remove cadmium (Cd) and lead (Pb). The results of the work showed that all of the membranes had rejection rates of more than 98% for both Cd and Pb heavy metals. The theoretical transport model states that membrane selectivity is dependent on mechanisms such as sieving and charge repulsion (You et al. 2017). Particularly, the PA layer played a role in the rejection of heavy metals by TFNs. The dissociation and protonation of active functional groups, like amine groups, produce the charges on the surface and pore structure of the PA layer. These charges improve separation efficiency by electrostatically repelling co-ionic species from the membrane surface (Sum et al. 2019). In addition, TFNs incorporated with TNT-Fe₃O₄ hybrid nanoparticles demonstrated the maximum rejection of Cd²⁺ and Pb²⁺ heavy metals, with 99.81 and 98.84%, respectively.

Membranes successfully remove impurities from drinking water; however, fouling will lead to a rise in operating costs and limits the applicability of membrane process. According to the definition of membrane fouling, it is ‘a condition that occurs when the membrane is blocked or coated by a component present in the processed stream, causing a reduction in flux’. Membrane separation is a phenomenon that occurs on surfaces, and the top selective layer is vital (Saqib & Aljundi 2016). Consequently, it is normal to see a lot of research projects with a membrane surface modification focus. These initiatives had specific objectives, like increasing permeate flux and/or rejection (Wang et al. 2015; Ma et al. 2016). Nevertheless, reducing membrane fouling is the most important objective (Ray et al. 2015; Shafi et al. 2015; Wu et al. 2015; Matin et al. 2016). Since many foulants are naturally hydrophobic, membranes that are more hydrophilic are believed to be more resistant to fouling because stronger hydrogen bonds develop between the membrane surface and the nearby water, forming a protective water-layer from foulants (Saqib & Aljundi 2016).

‘The plane of membrane ends’ or ‘the last layer of atom before the next phase begins’ are two definitions given for a surface (Xu et al. 2009). However, the magnitude of both a short-term and long-term flux reduction is greatly influenced by membrane surface properties in the membrane transport system. As previously mentioned, interactions between surface and foulant particles can be whittled down by cleaning the membrane with chemicals like sodium hypochlorite solution. Regrettably, the use of chemicals may cause the polymeric membrane material to instantly deteriorate, shortening the lifetime of the membrane (Ayyavoo et al. 2016). By using efficient surface modification techniques, the complexity and high energy consumption of the pressure-driven membrane process (MF, UF, and NF) can be prevented. Modifying membrane characteristics that prevent membrane fouling can be made easier, including their surface hydrophilicity, charge, roughness, and functional groups. The development of completely fouling-free membranes appears to be highly complicated, but it is doable and has been accomplished through numerous modification methods (Ayyavoo et al. 2016) and many studies have shown that reducing fouling on the membrane has a positive effect (Choi et al. 2012; Kwon et al. 2012).

In previous research work (Soo et al. 2020), a novel TFN forward osmosis (FO) membrane with a positively charged and nano functional substrate layer has been developed for efficient removal of heavy metal ions. Titanium nanotubes and magnetite oxide hybrid nanoparticles (TNT-Fe₃O₄) are incorporated to the PSF matrix in order to form the substrate layer. The addition of nanoparticles enhanced the hydrophilicity and loose structure of the substrate layer. The modified substrate layer also increased the affinities between the polymer matrix and the nanofillers, maintaining the membrane selectivity. The modified substrate hydrophilicity, which has significantly decreased the internal polarization concentration (ICP), was primarily responsible for the improvement. The membrane with the best performance increased water flux by 73% when compared to TFC and displayed a high level of Cd²⁺ and Pb²⁺ heavy metal ion rejection of >98%. This study demonstrated the viability of improving the performance of FO membrane for treating heavy metal wastewater using a quick and easy technique by designing the substrate layer (Soo et al. 2020).

In order to modify the surface properties of antifouling membranes, several modification approaches have been developed. These modifications are classified as either physical (coating, blending, and employing composite membranes) or chemical (functionalization of polymers, plasma treatment, and graft polymerization) (Ayyavoo et al. 2016).

Surface coating refers to the thin selective layer that is straightforwardly deposited on the top layer of the membrane surface. The three theoretical mechanisms by which hydrophilic functional materials are physically deposited on membrane surfaces are: (1) adhesion/adsorption, in which the binding energy is enhanced by multiple interactions between functional groups in the macromolecular layer and on the hard surface; (2) interpenetration, in which the functional material and base polymer are combined in an interphase; and (3) macroscopic entanglement of functional group and pore structure of membrane. The coating procedure typically entails submersion in a coating solution followed by the evaporation of any remaining solvent at the proper temperature to produce the coating layer. Performance of the membrane is significantly influenced by the preparation conditions (flux and rejection) (Xu et al. 2013).

Polymer blending has become a widely known, effective, and flexible method for improving membrane properties for a variety of applications. In the preparation and processing of materials, blending is a common method for distributing liquid-state powders that effectively mixes molecules at the molecular level (Ayyavoo et al. 2016). The basic transport properties of the underlying polymer must be combined with the necessary qualities in order to create a polymer blend membrane. For all pressure-driven membranes, the proper choice and combination of polymeric components in a precise ratio can produce a blend that has advantageous properties for antifouling membranes (Ayyavoo et al. 2016). According to previous studies, the presence and concentration of hydrophilic PEG 600 have a significant influence on changing membrane properties like porosity and pore size. The recently founded membranes were more effective at removing metal ions and proteins from aqueous streams (Nagendran & Mohan 2008).

The surface of membranes with different functional groups (such as hydroxyl, amine, carboxyl, and sulphone) can be altered chemically, which is a common procedure. The primary chemical treatment techniques are hydrolysis, substitution, addition, and oxidation; bulk modification also uses these techniques (used before membrane formation, not on membrane). Direct application of bulk modification techniques to the polymer solution is simple and efficient (Ayyavoo et al. 2016). Typical bulk modification techniques used to increase surface hydrophilicity include carboxylation, sulphonation, amination, and epoxidation of aromatic polymer backbones. The modified membrane surface still has the original membrane's mechanical characteristics because it is moderately stable. The different functional groups that are attached to polymer or membrane surfaces activate them to produce superior properties (Ayyavoo et al. 2016).

Based on previous research (Khoo et al. 2022) that was inspired by the benefits of surface coating and nanoparticle incorporation, this study aims to integrate both techniques to produce TFN membrane with minimal surface defects for RO application by employing a rapid surface deposition technique – plasma enhanced chemical vapour deposition (PECVD) to form a hydrophilic thin layer atop nanofiller-incorporated PA layer. This coating method is expected to prevent TNT leaking from the membrane. The PECVD approach is used for surface modification of nanomaterials and TFN membranes because it can change the material surface functionality (Hassan & Carr 2019) with its lower degree of plasma ionization without causing damage to the bulk material. In addition, unlike conventional surface grafting and coating processes, this technology has a quick reaction time and does not require any hazardous ingredients (Khoo et al. 2020).

By grafting monomers together, graft polymerization produces distinct properties while shielding the membrane surface from external agents. It is the best and most efficient method for surface modification. Grafting-to and grafting-from reactions are the two main categories used to categorise grafting reactions. A grafting-to procedure involves covalently attaching reactive functional polymer chains to the membrane surface. For grafting-from, the surface-based monomers are first polymerized by the active species on the polymer backbone (Ayyavoo et al. 2016). Multiple types of reactive group propagation occur after the grafting-from process, which is started by chemical, radiation, photochemical, and plasma-induced mechanisms. By using initiators and copolymerizing monomers on the polymer backbone, the chemical grafting technique creates free radical and ionic active sites for a variety of functional groups on the membrane surface (Ayyavoo et al. 2016). The acrylic acid (AA) monomer was successfully free radical polymerized onto polyamide (PA) by prior researchers (Makhlouf et al. 2007) and grafting was produced as a result of heat cleaving the peroxide group (Benzoyl peroxide) (Makhlouf et al. 2007).

All of the aforementioned membrane surface modifications were carried out in an attempt to avoid or lessen fouling by altering the surface's charge, hydrophilicity, functional groups, and roughness. The elimination of unwanted interactions and the creation of cutting-edge separation functions are made possible by modifications to membrane surfaces. Graft polymerization that is initiated by photochemical or redox reactions is more practical, less complicated, and much less expensive in commercial settings (Ayyavoo et al. 2016). Numerous studies on fouling have revealed that modified polymeric membranes are more impervious to foulants of all kinds than virgin membranes. Colloidal deposition can be effectively avoided by creating a fouling-resistant coating on the membrane surface using blending, photochemical grafting, and chemical grafting. Membrane washing is necessary to prevent foulants from accumulating on the membrane, and fouling resistance can be improved by incorporating membrane module design (Ayyavoo et al. 2016).

Given their high toxicity, heavy metals are known to be extremely dangerous to the environment and human health if released directly. Heavy metals in industrial wastewater should be treated carefully to meet the discharge standard and prevent any negative effects. It has been reported that heavy metal removal can be succeeded in using membrane-based processes, such as pressure-driven membrane. Despite their effectiveness, laboratory-scale tests were still conducted to examine the majority of heavy metal removals by membranes. At the pilot or commercial scale, membrane durability, scaling and fouling formation, and long-term performance must be assessed for future applications. Many industries and sciences require a low cost, dependable, and high yield technology for separation and manufacturing of high-quality materials, and TFN membrane can prepare the perfect setting for these purposes. Membranes are ideal due to their great selectivity and other unique properties, despite the fact that many laboratory size projects have yet to be commercialized. In recent years, advancement has been made in the fabrication and development of high-performance TFN membranes using functionalized fillers derived from GO, CNT, TiO₂, Ag-TiO₂, metal organic frameworks (MOFs), and organic fillers for water purification applications using membrane separation processes such as NF, UF, and RO. The nanofiller is then applied to develop TFN membranes through IP between an organic solution of the monomers in non-polar solvents and an aqueous-solution confined nanofiller. The kind of functionalized nanofiller used and how much is loaded into the PA-selective layers of the TFN membranes affect how well water purification and water treatment work.

However, agglomeration of nanofiller in the selective layer of matrix of the membranes at high loading concentrations is one of the major barriers to large-scale production of TFN membranes and how they are used in the filtration of water. In order to prevent the agglomeration of functionalized nanofiller in the selective layers at high loading quantities, new types of functionalized nanofiller that work better with the membranes' selective layers should be developed. As a result, future research should concentrate on developing high-performance TFN membranes through the use of surface engineered nanofiller or surface modification of the PA and/or substrate layer to achieve high permeate flux and/or rejection. The development of TFN membranes made from functionalized nanofiller from metal oxide-adsorbent nanocomposite can be an interesting area of scientific or industrial study for those involved in the low-cost separation and removal of heavy metals from wastewater and aqueous solutions.

We gratefully acknowledge the financial support for the work described in this paper from the Ministry of Higher Education for HICOE Research Grant (Grant No.: R.J090301.7851.4J431) and Universiti Teknologi Malaysia for Industry-International Incentive Grant (Q.J130000.3651.04M0).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.