Introduction

Worldwide, water pollution is rising, endangering the economic potential and development objectives of severely polluted areas because of the detrimental effects on human health and aquatic ecosystems. The improper disposal of industrial and agricultural pollutants (such as organic dyes, pesticides, and pharmaceutical residues) in water systems is becoming more and more of a global health threat. Over two billion people live in water-stressed countries, according to the World Health Organization (WHO, 2020), and it is anticipated that this situation will get worse in some areas because of the increased industrial discharge of contaminated water, population growth, and climate change [1]. According to current projections, 57% of the world's population will experience water shortages by 2050 if sustained and coordinated efforts are not made [2,3]. The estimate provided in [2,3] might have been too low. The projections for water consumption, availability, and quality are affected by a variety of unreliable geopolitical factors. Nevertheless, there is a growing need for the efficient removal of environmental pollutants and the proper treatment of industrial wastes to allowable discharge limits, which are crucial for preserving human life and protecting the environment.

Numerous techniques have been employed to treat contaminated water and wastewater, including adsorption, bioremediation, precipitation, electrocoagulation, filtration, membrane separation, flocculation, centrifugation, advanced oxidation processes based on photocatalysis, and chemical coagulation [4-11]. Each of these techniques has demonstrated varying levels of effectiveness and drawbacks that restrict their widespread use. For instance, due to deficiencies such as the formation of harmful by-products and incomplete removal of organic pollutants, traditional water treatment methods such as sedimentation, filtration, and precipitation, in particular, are believed to be ineffective [4,11]. As a result of the non-biodegradable and persistent nature of the majority of organic contaminants, some physicochemical treatment techniques, such as adsorption, are ineffective in removing them from water resources [11]. Because of their flexible design and low cost, biological approaches have been used for the treatment of various contaminated effluents. However, the process is time-consuming, can be ineffective when toxic recalcitrant pollutants are present, and may even be irreparably harmful to the environment.

Among the water treatment technologies, advanced oxidation processes (AOPs) are regarded as a practical, efficient, and fiercely competitive technology for water treatment for the removal of a variety of toxic and bio-recalcitrant organic pollutants and for the inactivation of pathogen microorganisms that cannot be treated by conventional methods [11-14]. For the oxidation of organic molecules, AOPs rely on the in situ generation of potent oxidants (reactive oxygen species, ROS) such as hydroxyl or sulfate radicals. AOPs have been broadly categorised in terms of how ROS are produced, including non-photochemical techniques, such as chemical, radiation-induced, cavitation, electrochemical techniques, and photochemical processes [11,15-17].

One of the AOPs, photocatalysis, uses natural light – a resource that is both clean and recyclable – to completely degrade a variety of organic pollutants and inactivate pathogens. The term “photocatalysis” refers to chemical reactions that use light and a photocatalyst (basically a semiconductor). A few of the requirements that an effective photocatalyst system should satisfy include high sunlight absorption, an appropriate gap (1.5–2.8 eV), long-term charge carrier separation, high photo-transporter mobility, appropriate physical and chemical properties, sufficient band alignment to meet the kinetic requirements of the target reaction, and anti-corrosion stability in reactive environments [18-20].

Figure 1 depicts the mechanism of the photocatalyst. In a nutshell, when exposed to light of the desired wavelength (enough energy), an electron (e−) in the photocatalyst's valence band absorbs photon energy and is excited to the conduction band on a femtosecond scale. This results in the formation of a hole (h+) in the valence band and a charge carrier pair (e− and h+) on the surface of the photocatalyst. Three possibilities exist at this point: (a) The generated charge carriers recombine and generate heat, (b) the generated interfacial charge carriers simultaneously reduce and oxidise contaminants, or (c) the generated charge carrier and an electron donor or acceptor on the surface of the photocatalyst may continue to interact. Nothing happens in the first scenario. In the second scenario, an electron or hole interacts with dissolved oxygen or water to produce ROS (e.g., •OH, O2•−). These ROS play a significant role in the photo-oxidation/reduction reaction, along with other species such as oxygen, hydrogen peroxide, and persulfate. This excited electron reduces an acceptor, and the acceptor's hole oxidises donor molecules. What happens to the excited electron and hole depends on the relative positions of conduction band and valence band of the semiconductor as well as the redox levels of the substrate [11,21].

Figure 1: Mechanism of the photocatalytic process used to treat water contaminated with organic pollutants.

One of the main barriers preventing photocatalysis from being used in practical applications is the lack of suitable semiconductor photocatalysts. The commonly used nanometre-sized photocatalysts are metal oxides or sulfides (binary compounds: TiO2, CuO, CdS, MoO3; ternary compounds: Bi2Mo3O12, ZnFe2O4; quaternary compounds: Ni0.5Zn0.5Fe2O4, Bi4NbxTa1−xO8I) [19-26]. Because of its distinct features, TiO2 is the most extensively investigated photocatalytic semiconductor. However, it barely absorbs 4–5% of the ultraviolet light in the solar spectrum due to its broad bandgap of 3.2 eV, which limits the use of visible light. Because of this, the potential photocatalytic use of TiO2 is constrained and the photocatalytic effectiveness is reduced [19,20,25]. Table 1 compares some of the salient characteristics of some of the bismuth-based photocatalysts with some of the typical metal oxide-based photocatalysts. Some of these important variables and values have been extracted from articles that have been published [27-38].

Table 1: Comparison of nanometre-sized metal oxide-based and bismuth-based photocatalysts.

Features Metal oxides   TiO2 ZnO SnO2 bandgap (eV) 3.0–3.4 3.10–3.37 3.76–4.24 performance based on the light source very active in UV light very active in UV light very active in UV light semiconductor type n-type n-type n-type crystal structure anatase (tetragonal), brookite (orthorhombic), rutile (tetragonal) hexagonal wurtzite (most stable at ambient conditions) and cubic zincblende tetragonal stability photostable in solution and resistant to corrosion readily dissolves in water, photocorrosion under UV good stability toxicity nontoxic low-toxicity relatively non-toxic photon absorption efficiency and quantum yield high higher than TiO2 moderate cost low low low electron–hole pairs recombination rate high fast high magnetic properties no no no   Bismuth-based   BiFeO3 Bi2WO6 Bi2S3 bandgap (eV) 2.0–2.5 2.6–2.9 1.4–1.6 performance based on the light source both visible and UV light both visible and UV light both visible and UV light semiconductor type n-type n-type n-type crystal structure rhombohedral distorted perovskite structure orthorhombic orthorhombic stability sufficiently stable superior stability highly stable toxicity low toxicity nontoxicity low toxicity photon absorption efficiency and quantum yield very high moderately high high cost low low low electron–hole pairs recombination rate high fast moderate magnetic properties ferromagnetic at low temperatures and superparamagnetic at room temperature. (multiferroic behaviour) no no   Bismuth-based   BiOBr Bi2O3 Bi3O4Cl bandgap (eV) 2.69–2.99 1.5–2.8 2.6–2.8 performance based on the light source both visible and UV light both visible and UV light both visible and UV light semiconductor type p-type p-type n-type crystal structure tetragonal (PbFCl-type structure) monoclinic (room temperature), tetragonal β-phase or body-centred γ-phase (intermediate temperature), cubic (very high temperature) cubic (Silleń structure) stability good chemical stability highly chemically stable and photostable in solution good stability toxicity nontoxic low toxicity nontoxic photon absorption efficiency and quantum yield moderately high very high moderate cost low low low electron–hole pairs recombination rate moderate low moderate magnetic properties no paramagnetic behaviour no

As an alternative to TiO2 for photocatalysis, nanometre-sized photocatalysts based on bismuth have recently been investigated and evaluated, because the majority of bismuth-based photocatalysts have a bandgap below 3.0 eV, making them usable in visible light. Additionally, their electrical structure produces a valence band with hybrid O 2p and Bi 6s orbitals, as opposed to the valence band of TiO2, which is made up entirely of O 2p orbitals. The mobility of the photogenerated charge carriers is increased by the well-dispersed Bi 6s orbital. Due to their distinctive structure, Bi-based photocatalysts exhibit a steeper absorption edge in the visible-light spectrum. Additionally, the reverse bond between the cation and anion is more favourable for the production and transportation of holes, which facilitates photocatalytic activity. Because of this, significant efforts have been made to synthesise bismuth-based nanomaterials (BiVO4, Bi5O7I-MoO3, Bi2O3, BiFeO3, Bi2WO6, Bi2Mo3O12, Bi2MoO6, and BiOI [24,25,39-45]) using a variety of techniques to tailor their size, morphology, and optoelectrical properties to improve their photocatalytic performance and to better understand the factors influencing their performance. Different materials based on bismuth have been developed and used for a range of environmental remediation applications. For instance, Mu et al. [46] synthesised a Bi2S3/Bi4O7 heterostructure via an in situ sulfidation approach and utilised it for the degradation of rhodamine B dye under visible-light exposure. Since the oxidation rate is still up to 96.3% after four cycles, the photocatalyst showed great performance and stability in the photocatalytic oxidation of the dye.

This review provides an overview of the recent nanostructured photocatalytic materials based on bismuth that are employed in the photocatalytic degradation of organic dyes and antibiotics in water. The general synthesis of nanometre-sized photocatalytic materials based on bismuth employing energy-efficient techniques is examined. A critical review is also given of ways to improve the photocatalytic activity of the photocatalysts. An extensive critical evaluation is given of recent findings on the photocatalysis of nanostructured materials based on bismuth and doped bismuth for the remediation of textile and pharmaceutical wastewater.

Antibiotics and organic dyes in the environment and their toxicological consequences

Antibiotics are administered therapeutically to cure/prevent pathogen infections in people, animals, or both, as well as to increase livestock yields. However, since 50–80% of the antibiotic compounds that are taken are typically eliminated through urine and faeces, there are growing concerns regarding their excessive consumption and how they affect the environment. The widespread use of pharmaceuticals, especially antibiotics, has made them prevalent in the environment, and nearly everyone in the world now acknowledges their existence in both artificial and natural systems. Particularly, it has been claimed that antibiotic residues or metabolites have contaminated groundwater, soil, sediment, tap water, sludge, wastewater, and surface water.

Chemical manufacturing facilities, effluents from wastewater treatment facilities, and animal husbandry and aquaculture are the three main entry points for antibiotics into fresh waters [47-49]. According to the paper of Wise in the year 2002 [50], nearly 200,000 tons of antibiotics are consumed globally each year, with roughly 50% being utilized for veterinary medication and growth stimulants. Notably, between the years 2000 and 2010, the amount of antibiotics consumed by humans alone increased by 36% globally, demonstrating the ongoing problem of antibiotics pollution [51].

According to a recent study by Browne et al. [52], which covered 204 nations from 2000 to 2018, the rate of antibiotic consumption worldwide grew by 46% during the last 20 years. The report offers a comparative analysis of global human consumption rates of all antibiotics, expressed in defined daily doses (DDD) per 1000 population per day, a WHO metric. In contrast to the very low rates of consumption in sub-Saharan Africa and several regions of Southeast Asia, high rates of antibiotic usage were seen in the Middle East, Europe, and North America. The regions of South Asia (116% rise) and North Africa and the Middle East (111% rise) experienced the biggest increases in antibiotic usage rates. Specifically, in South Asia, third-generation cephalosporin consumption rates surged 37-fold and fluoroquinolone consumption rates increased 1.8-fold over the course of the study.

Different geographical areas have different levels of antibiotics in the environment. For instance, aus der Beek et al. [53] reported ofloxacin and sulfamethoxazole at 17.7 μg/L and 14.3 μg/L, respectively, and sulfamethazine has been reported with a concentration of 19 ng/L in Vietnam [54]. Sulfamethoxazole was lastly detected in Africa, where it was found at 53.8 ng/L in Mozambique [55] and 38.9 ng/L in Kenya [56]. Nalidixic acid and ciprofloxacin quantities of 23 μg/L and 14 μg/L, respectively, were found in South African streams and rivers [57]. While it is critical to understand the presence and levels of antibiotics in freshwater environments, it is maybe even more crucial to understand whether the residues or metabolites of the antibiotics have any impact on the various species that live there. The concentration necessary to produce a 50% effect after a given exposure time is known as the EC50. Chemicals having an EC50 between 10 and 100 mg/L are classified as hazardous, those from 1 to 10 mg/L as toxic, and those below 1 mg/L are classified as extremely toxic to aquatic life by the Commission of the European Communities [58]. The Wikipharma statistics [59] show that EC50 values were less than 1 mg/L in 25% of all research assessing the effects of antibiotics on eukaryotic, single-celled algae and that EC50 was even less than 100 μg/L in twelve investigations.

Once these antibiotics are released into the environment, non-target species are unavoidably exposed [47]. The development of antibiotic resistance, which has reduced the therapeutic capacity against human and animal infections, is the most significant issue associated with the release of antibiotics into the environment. It is not true that antibiotic resistance has never been observed in the natural environment; rather, it had previously only been linked to a small number of bacterial strains, but recent research has discovered antibiotic resistance genes in many other bacterial strains, raising serious health concerns. Antibiotic resistance is brought on by a high concentration of antibiotics that enter aquatic systems and interact with native species [47,60-62]. For instance, it may start to alter the genetic makeup and structure of the microbial community [47]. Antibiotic-resistant microbes (algae, fungi, and bacteria) pose a threat to both human and ecological health. The active ingredients of antibiotics and their fragments may cause kidney and liver cell damage in humans if they are exposed to antibiotic residues for an extended time [63-65]. Additionally, it has been noted that prolonged exposure to antibiotic-contaminated water might result in several allergic and respiratory conditions [62-65]. Additionally, an overabundance of antibiotics in the environment causes structural changes in the ecosystem, disruptions in ecological function, and impacts the processes of sulfate reduction, methanogenesis, and nitrogen conversion [61,63].

Antibiotics are persistent for long periods of time in natural environment. It is important to note that bacteria that develop resistance to one antibiotic also exhibit resistance to other drugs and chemicals. For example, Dickinson et al. [64] reported that the focal strain isolates from pond sediments in the northwest of the United Kingdom exhibited resistance to heavy metals and antibiotics (trimethoprim, oxacillin, and cefotaxime) where the intI1 gene was involved. A growing body of research indicates that parent antibiotics and their metabolites, which are released into the environment in low concentrations (micrograms per litre to nanograms per litre), are persistent and bioactive, potentially posing a threat to the food chain.

Macrolides, fluoroquinolones, and tetracycline also have an impact on the synthesis of mitochondrial proteins and chloroplasts in plants [48,66]. Fluoroquinolones have a detrimental impact on the morphology and photosynthesis of plants, as well as on the ability of eukaryotic cells to synthesise DNA and replicate plastids. Streptomycin prevents Hordeum vulgare from producing chlorophyll, while ciprofloxacin, enrofloxacin, and sulfadimethoxine considerably slow down plant growth. Additionally, tetracyclines have phytotoxic effects that may result in chromosomal abnormalities and the reduction of plant growth. Although β-lactams are thought to be less harmful, they also have an impact on the plastid division in lower plants [48,67].

The textile industry, in addition to the pharmaceutical sector, is another sector that supports global economic expansion. It is one of the major sources of global pollution, although its importance cannot be disputed. Due to its high water demand when producing textiles and the limitations of conventional wastewater treatment techniques, the textile industry is causing concern. The direct release of textile waste into bodies of water without proper treatment to an acceptable level has a negative impact on its aesthetic quality. The presence of organic dyes in bodies of water, even in minute amounts, raises the chemical and biochemical oxygen demand and inhibits photosynthesis. Additionally, the uptake of dye molecules or their by-products in excess may be mutagenic, teratogenic, or carcinogenic [68,69]. Myocardial depression and hypertension are reportedly exacerbated by oral exposure to methylene blue dye. Additionally, some dyes, such as xanthene and erythrosine, have been related to allergic reactions, neurotoxins, and DNA damage in both humans and animals [70]. An eco-friendly, practical, and efficient treatment method is urgently needed because of the increasing pollution and health and ecological concerns of excess antibiotics and dyes in the environment. This article discusses the use of nanomaterials based on bismuth for the remediation of persistent organic pollutants.

Bismuth and bismuth-based nanostructured photocatalysts

Bismuth (Bi) is a semimetal and a member of the p-block with a d10 configuration (6s26p3) in the sixth period of group V of the periodic table. Because of their intriguing optical, catalytic, electrical, ferroelectric, and piezoelectric properties, bismuth-based nanostructures are used in several significant fields, including optoelectronics, pollutant sensing [71], and environmental remediation via photocatalysis [25]. Bi-based semiconductors, in particular, are thought to be able to surpass the limitation of the solar light-harvesting capacity of TiO2-based photocatalytic materials because of their smaller bandgaps. Because of its highly anisotropic Fermi surface charge, low carrier density, small electron effective mass, long electron mean free path, and extremely low band overlap energy, bismuth can transition from a semimetal to a semiconductor by shrinking its crystallite size [25,71-77].

To hasten the separation of photogenerated charges and, hence, increase photocatalytic activity, metallic bismuth can function as a direct plasmonic photocatalyst (similar to Au and Ag) or a co-catalyst [77]. Also, the unique layered crystal structure of Aurivillius-type bismuth oxide-based semiconductors allows for the induction of an internal static electric field, which effectively aids in the separation and transfer of photogenerated carriers. Bulk Bi and Bi-based nanostructure morphologies can also be easily altered using a variety of synthesis techniques due to their unique electrical and optical properties, which are directly tied to the plasmonic and photocatalytic properties. The typical and most recently applied bismuth-based nanostructure photocatalysts are depicted in Figure 2.

Figure 2: Most recently studied and common bismuth-based nanostructured photocatalysts.

Structural, optoelectronic, and magnetic properties

Bismuth's peculiar optical, electronic, and more recently discovered photocatalytic and plasmonic properties have attracted the interest of a large community of scientists. With a low melting point of just above 544 K, Bi is less toxic than its neighbours in the periodic table, antimony, lead, and polonium. The structure of the bismuth crystal, which has rhombohedral symmetry, is typical of the group-V semimetals. Bi atoms form puckered bilayers of atoms perpendicular to the rhombohedral plane with three equidistant nearest neighbours and three equidistant next-nearest neighbours that are slightly farther away.

Bi is widely used in photocatalysis, in part because of its quantum confinement effect, which is important for electronic transport and semimetal-to-semiconductor transition, as well as its highly anisotropic Fermi surface (with an electron and hole Fermi energies of 27.2 and 10.8 meV, respectively), which results in an extremely low carrier density of around 3 × 1017 cm−3[78] and very little overlap between the T-point band (valence) and the L-point band (conduction) [76-78]. Note that a reduction of the crystallite size below a critical value can result in a semimetal-to-semiconductor transition [77-80]. For instance, according to Qi et al. [81], indirect bandgap semiconductors were visible in Bi nanowires with a diameter of around 1–3 nm, but as the diameter increased, they became less visible because of the intense quantum confinement effect.

In addition to the electronic properties of Bi, its outstanding optical properties have a big impact on how effective it is as a photocatalyst. Bulk Bi exhibits high interband electronic transition rates that result in a negative ultraviolet–visible permittivity and a large infrared refractive index. Numerous investigations have shown that the quantum confinement effect affects the optical properties of Bi [25,71-80]. Furthermore, nanostructured materials exhibit unique optical properties that set them apart from the corresponding bulk materials as a result of this quantum confinement. Also, note that the optical responses of Bi nanoparticles are strongly influenced by their size, morphology, bandgap structure, shape, and environment. If these parameters are adjusted, the optical responses of Bi nanoparticles can be tuned from the near-ultraviolet to the near-infrared region.

According to Figure 3, the bandgap of different bismuth-based photocatalysts has been observed to fall between 1.30 and 3.85 eV. From an optoelectronic structure standpoint, the majority of bismuth-based photocatalysts have a bandgap below 3.0 eV, which qualifies them for use in visible light. The hybridisation of the O 2p orbital and the 6s orbital in Bi is thought to be the cause of the narrow bandgap [82]. The valence band electrons are elevated by the hybridisation, which benefits the separation of photogenerated electron–hole pairs and the rate of charge carrier migration. Numerous visible-light photocatalysts based on bismuth have been used for the degradation of micropollutants because of their appropriate bandgap and non-toxic nature.

Figure 3: Bandgaps of some bismuth-based photocatalysts extracted from various research articles [27,35-37,83-86].

BiFeO3, one of those Bi-based photocatalysts, has been the subject of intensive research in recent years because it is the only naturally occurring magnetoelectric material with ferromagnetic and ferroelectric properties at room temperature [39,75,87-89]. Bismuth ferrite has a distorted rhombohedral perovskite structure (ABO3), where A is a corner cation, B is a body-centred middle atom, and O is an oxygen atom or anions attached to the crystal faces. BiFeO3 has strong magnetic and multiferroic, and sufficient photocatalytic properties due to this unique structure. BiFeO3 is an effective photocatalyst in the visible-light region, because in contrast to other semiconductors such as TiO2, it has a very narrow bandgap (Figure 3) and slow electron–hole recombination.

As 44% of solar radiation falls within the visible-light spectrum, BiFeO3 can be activated by direct sunlight, further lowering the cost of treatment. Aside from its magnetic and optical properties, BiFeO3 also exhibits piezoelectric characteristics, photovoltaic effects, switchable ferroelectric diode effects, and spontaneous polarisation enhancement. It is also sensitive to epitaxial strain [88]. Given its intriguing properties, a lot of researchers [90] have used bismuth ferrite to efficiently degrade organic pollutants, as shown in Table 2.

Table 2: Treatment of water containing antibiotics and dyes by bismuth ferrite nanoparticles (BiFeO3).

Particle size (nm) Target pollutant Source of light Experimental conditions Degradation (%) Ref. Remarks on the synthesis and main findings 5.5 rhodamine B dye visible light (high-power LEDs) catalyst dosage: 1.25 g/L; solution pH 2; reaction time: 50 min; initial concentration of rhodamine B: 5 mg/L; observed bandgap: 2.07 eV 100.0 [75] Monodisperse BiFeO3 nanoparticles were synthesised using a nanocasting approach, and they outperformed BiFeO3 nanoparticles prepared using other synthetic techniques in terms of photocatalytic efficiency and stability when exposed to visible light. When compared to particles of comparable size, the photocatalytic activity of the nanocast BiFeO3 particles is significantly higher. A low density of surface defects and few local strains contributed to this higher performance. 35 rhodamine B dye visible light (500 W Xe lamp) catalyst dosage: 2 g/L; solution pH 0.5; reaction time: 60 min; initial concentration of rhodamine B: 10−5 mol/L; observed bandgap: 2.06 eV 100.0 [91] By using a rapid sol–gel calcination approach, multiferroic BiFeO3 nanoparticles with rhombohedral crystal structures were synthesised, and they had stronger photocatalytic activity than the bulk. Mild room-temperature ferromagnetism was shown by the BiFeO3 nanoparticles. 150–200 methyl orange dye visible light (70 W 365 nm UV lamp) with a catalyst loading of 6.4 mmol/L, the initial concentration of the methyl orange dye was 20 mg/L; the optimum reaction time was 260 min; the bandgap of the catalyst is 2.10 eV. 92.0 [92] Chemical co-precipitation was used to synthesise the BiFeO3 nanoparticles, and analysis of the samples reveals that they have a perovskite structure that is distorted rhombohedrally and belongs to the polar R3c space group (no. 161). The nanoparticles' bandgap energy was lower than that of the bulk BiFeO3 (2.5 eV) due to the thinness of the sample. 128 methylene blue dye simulated solar light (Xe lamp 500 W) catalyst concentration: 5 ppm; initial concentration of the dye: 1 ppm; pH 1–2; optimum reaction time: 50 min; the bandgap of the 1D nanofiber is 2.38 eV. nanofiber 98.0
nanoparticulate 68.0 [93] Electrospinning and the sol–gel method were used to synthesise BiFeO3 nanofibers and nanoparticles, respectively. According to the XRD findings, the BiFeO3 phase exhibits a rhombohedral structure with average crystallite sizes of 60 and 24 nm for BiFeO3 nanoparticles and nanofibers, respectively. Due to 1-dimensional c