Nanofiber-Based Materials for Persistent Organic Pollutants in Water Remediation by Adsorption

Fresh water is one of the most precious resources for our society. As a cause of oxygen depletion, organic pollutants released into water streams from industrial discharges, fertilizers, pesticides, detergents or consumed medicines can raise toxicological concerns due to their long-range transportability, bio-accumulation and degradation into carcinogenic compounds. The Stockholm Convention has named 21 persistent organic pollutants (POP) so far. As opposed to other separation techniques, adsorption, typically performed with activated carbons, offers opportunities to combine low operation costs with high performance as well as fast kinetics of capture if custom-designed with the right choice of adsorbent structure and surface chemistry. Nanofibers possess a higher surface to volume ratio compared to commercial macro-adsorbents, and a higher stability in water than other adsorptive nanostructures, such as loose nanoparticles. This paper highlights the potential of nanofibers in organic pollutant adsorption and thus provides an up-to-date overview of their employment for the treatment of wastewater contaminated by disinfectants and pesticides, which is benchmarked with other reported adsorptive structures. The discussion further investigates the impact of adsorbent pore geometry and surface chemistry on the resulting adsorption performance against specific organic molecules. Finally, insight into the physicochemical properties required for an adsorbent against a targeted pollutant is provided.


Introduction
Organic water pollution from synthesized chemical pesticides has been increasing over the past 50 years [1]. As a counter-measure, 21 non-polar organic molecules have been classified as persistent organic pollutants (POPs) by the Stockholm Convention (2001,2017), to be banned or restricted globally [2]. Organochlorines (OCs) account for a large group of POPs, including aldrin or DDT pesticides, but also industrial chemicals such as the family of polychlorinated biphenyls (PCBs), as well as byproducts [1]. The list of POPs also includes several organophosphorus (OPs) pesticides, and among recently added compounds, flame retardants and polyaromatic hydrocarbons (PAHs) [1,2].
The release of POPs is associated with major toxicity concerns for both the environment and human health [3,4]. Their partial degradability affects the chemical oxygen demand (COD) of streams, causing oxygen depletion by consuming oxygen at a greater rate than its replenishment [4]. For some POPs, such as Heptachlor, it is the compound formed after reaction in water, Heptachlor epoxide, that causes threat to human life [5]. The persistency of POPs in the environment can be measured by their half-life duration (DT50). For instance, organochlorine DDT pesticide has a DT50 that, depending on the environment, ranges from 2 to 15 years [6]. The presence of POPs in the environment also affects human health, through the ability of POPs to bio-accumulate in organisms and bio-magnify

Moderately Hydrophobic Pesticides
In this section, the adsorptive performance of nanofiber-based adsorbents is reviewed and benchmarked against other reported adsorbents for Alachlor, Atrazine, Chlorpyrifos and Diazinon pesticides. After presenting the adsorbate molecules' characteristics, an analysis of the adsorbent structure-performance relationships will be given. First, the influence of adsorbent specific and contact surface area as well as pore volume will be reviewed. Then, the range order between adsorbent dose, initial concentration and obtained adsorption capacity will be highlighted. Table 1 presents the molecular characteristics of the four pesticides considered in this section, namely Alachlor, Atrazine, Chlorpyrifos and Diazinon. Both Alachlor and Atrazine are organochlorine pesticides (OCPs). Alachlor is composed of an aniline aromatic structure, bound to an oxoacid and a chloroacetamide group via a central nitrogen group. Atrazine possesses a chlorotriazine aromatic structure, bound to two dimethylamine groups on the aromatic carbon atoms. Although Chlorpyrifos also possesses chlorine terminal groups, it is commonly classified with Diazinon as organophosphorus pesticides (OPPs). Chlorpyrifos aromatic structure is made of a ring of trichloropyridine, with attachment of the three chlorine groups on three of the ring's carbon atoms. To form Chlorpyrifos, the chloropyridine is bound to an organophosphate structure, composed of a modified phosphate group, which is double-bonded to a sulphur atom, also called phosphorothioate. Diazinon shares the same sulphur-phosphate structure, which is here bound to a pyrimidine or 1,3-diazine aromatic ring [26]. organochlorine pesticides (OCPs). Alachlor is composed of an aniline aromatic structure, bound to an oxoacid and a chloroacetamide group via a central nitrogen group. Atrazine possesses a chlorotriazine aromatic structure, bound to two dimethylamine groups on the aromatic carbon atoms. Although Chlorpyrifos also possesses chlorine terminal groups, it is commonly classified with Diazinon as organophosphorus pesticides (OPPs). Chlorpyrifos aromatic structure is made of a ring of trichloropyridine, with attachment of the three chlorine groups on three of the ring's carbon atoms.

Adsorbate Properties
To form Chlorpyrifos, the chloropyridine is bound to an organophosphate structure, composed of a modified phosphate group, which is double-bonded to a sulphur atom, also called phosphorothioate. Diazinon shares the same sulphur-phosphate structure, which is here bound to a pyrimidine or 1,3diazine aromatic ring [26]. The reported molecular weights range from 216 g/mol for Atrazine up to 351 g/mol for Chlorpyrifos, with each of the pesticides carrying a different molecular configuration. In terms of molecular size, Atrazine followed by Alachlor are smaller than organophosphates Chlorpyrifos and Diazinon, which is of interest for the study of the molecule diffusion in an adsorbent material with reported pore features.
Hydrogen bond acceptor and donor counts are reported here to discuss the influence of chemisorption by H-bonding on the pesticide uptake by nanofiber adsorbents. Both Alachlor and Atrazine have an H-bond acceptor count of 2 due to their longer carbon chains. Atrazine, Chlorpyrifos and Diazinon have an H-bond donor count of 5, 5 and 6, respectively. H-bond donor groups are formed by amide and nitrogen groups in the case of Atrazine, and by nitrogen groups, as well as sulphur and oxygen atoms from the organophosphate backbone, as in the case of Chlorpyrifos and Diazinon. As each of the adsorbate molecules possesses an aromatic structure, adsorption by electron interaction through π-π stacking between the adsorbent surface unsaturated bonds and the molecule aromatic ring is believed to play a significant role in the pesticide uptake here [26,27].
Finally, the octanol-water partition coefficients are reported to point out the less hydrophobic character of the selected pesticides compared to persistent-classified pesticides, such as the here reviewed aliphatic organochlorines. Solubility in water at ambient temperature was also reported to give an insight into the maximum levels of water contamination achievable by these pesticides.

Nanofibers for Pesticide Adsorption
2.2.1. Impact of Adsorbent Specific and Contact Surface Area Figure 1 shows the impact of the adsorbent specific surface area (BET) on the adsorption removal. The removal percentage was overall found to decrease with increasing adsorbent specific surface area for both Alachlor and Diazinon, yet was found to increase in the case of Atrazine removal. The reported molecular weights range from 216 g/mol for Atrazine up to 351 g/mol for Chlorpyrifos, with each of the pesticides carrying a different molecular configuration. In terms of molecular size, Atrazine followed by Alachlor are smaller than organophosphates Chlorpyrifos and Diazinon, which is of interest for the study of the molecule diffusion in an adsorbent material with reported pore features.
Hydrogen bond acceptor and donor counts are reported here to discuss the influence of chemisorption by H-bonding on the pesticide uptake by nanofiber adsorbents. Both Alachlor and Atrazine have an H-bond acceptor count of 2 due to their longer carbon chains. Atrazine, Chlorpyrifos and Diazinon have an H-bond donor count of 5, 5 and 6, respectively. H-bond donor groups are formed by amide and nitrogen groups in the case of Atrazine, and by nitrogen groups, as well as sulphur and oxygen atoms from the organophosphate backbone, as in the case of Chlorpyrifos and Diazinon. As each of the adsorbate molecules possesses an aromatic structure, adsorption by electron interaction through π-π stacking between the adsorbent surface unsaturated bonds and the molecule aromatic ring is believed to play a significant role in the pesticide uptake here [26,27].
Finally, the octanol-water partition coefficients are reported to point out the less hydrophobic character of the selected pesticides compared to persistent-classified pesticides, such as the here reviewed aliphatic organochlorines. Solubility in water at ambient temperature was also reported to give an insight into the maximum levels of water contamination achievable by these pesticides.

Nanofibers for Pesticide Adsorption
2.2.1. Impact of Adsorbent Specific and Contact Surface Area Figure 1 shows the impact of the adsorbent specific surface area (BET) on the adsorption removal. The removal percentage was overall found to decrease with increasing adsorbent specific surface area for both Alachlor and Diazinon, yet was found to increase in the case of Atrazine removal. The reported molecular weights range from 216 g/mol for Atrazine up to 351 g/mol for Chlorpyrifos, with each of the pesticides carrying a different molecular configuration. In terms of molecular size, Atrazine followed by Alachlor are smaller than organophosphates Chlorpyrifos and Diazinon, which is of interest for the study of the molecule diffusion in an adsorbent material with reported pore features.
Hydrogen bond acceptor and donor counts are reported here to discuss the influence of chemisorption by H-bonding on the pesticide uptake by nanofiber adsorbents. Both Alachlor and Atrazine have an H-bond acceptor count of 2 due to their longer carbon chains. Atrazine, Chlorpyrifos and Diazinon have an H-bond donor count of 5, 5 and 6, respectively. H-bond donor groups are formed by amide and nitrogen groups in the case of Atrazine, and by nitrogen groups, as well as sulphur and oxygen atoms from the organophosphate backbone, as in the case of Chlorpyrifos and Diazinon. As each of the adsorbate molecules possesses an aromatic structure, adsorption by electron interaction through π-π stacking between the adsorbent surface unsaturated bonds and the molecule aromatic ring is believed to play a significant role in the pesticide uptake here [26,27].
Finally, the octanol-water partition coefficients are reported to point out the less hydrophobic character of the selected pesticides compared to persistent-classified pesticides, such as the here reviewed aliphatic organochlorines. Solubility in water at ambient temperature was also reported to give an insight into the maximum levels of water contamination achievable by these pesticides. Figure 1 shows the impact of the adsorbent specific surface area (BET) on the adsorption removal. The removal percentage was overall found to decrease with increasing adsorbent specific surface area for both Alachlor and Diazinon, yet was found to increase in the case of Atrazine removal. 6 0

Impact of Adsorbent Specific and Contact Surface Area
The reported molecular weights range from 216 g/mol for Atrazine up to 351 g/mol for Chlorpyrifos, with each of the pesticides carrying a different molecular configuration. In terms of molecular size, Atrazine followed by Alachlor are smaller than organophosphates Chlorpyrifos and Diazinon, which is of interest for the study of the molecule diffusion in an adsorbent material with reported pore features.
Hydrogen bond acceptor and donor counts are reported here to discuss the influence of chemisorption by H-bonding on the pesticide uptake by nanofiber adsorbents. Both Alachlor and Atrazine have an H-bond acceptor count of 2 due to their longer carbon chains. Atrazine, Chlorpyrifos and Diazinon have an H-bond donor count of 5, 5 and 6, respectively. H-bond donor groups are formed by amide and nitrogen groups in the case of Atrazine, and by nitrogen groups, as well as sulphur and oxygen atoms from the organophosphate backbone, as in the case of Chlorpyrifos and Diazinon. As each of the adsorbate molecules possesses an aromatic structure, adsorption by electron interaction through π-π stacking between the adsorbent surface unsaturated bonds and the molecule aromatic ring is believed to play a significant role in the pesticide uptake here [26,27].
Finally, the octanol-water partition coefficients are reported to point out the less hydrophobic character of the selected pesticides compared to persistent-classified pesticides, such as the here reviewed aliphatic organochlorines. Solubility in water at ambient temperature was also reported to give an insight into the maximum levels of water contamination achievable by these pesticides. Figure 1 shows the impact of the adsorbent specific surface area (BET) on the adsorption removal. The removal percentage was overall found to decrease with increasing adsorbent specific surface area for both Alachlor and Diazinon, yet was found to increase in the case of Atrazine removal. Indeed, for Alachlor, as shown in Figure 1A, the maximum reported removal performance reached 80% for acid-activated beideillite clay particulates grafted with organosilanes for a BET of 221 m 2 /g ( Table 2, N5). In comparison, the largest specific surface area of 608 m 2 /g of an activated charcoal adsorbent removed 40% of Alachlor (Table 2, N4). Nanofiber adsorbents made of alumina with surface-grafted organosilanes showed the second best performance with 58% removal for a specific surface area of 252 m 2 /g ( Table 2, N1). The higher performance of organosilane-grafted clays and alumina nanofibers suggests that there is competition between adsorption via hydrophobic interactions and via electrostatic interactions as the most efficient mechanism for Alachlor uptake [28,29]. Indeed, organosilane grafting provides the adsorbent with a hydrophobic surface more amenable to pesticides [28]. Each adsorbent reported in this study for the adsorption of Alachlor differs in both structure and composition. Figure 1B shows the different compositions of Alachlor adsorbents, revealing here a lower performance for carbon-based adsorbents ( Table 2,  Indeed, for Alachlor, as shown in Figure 1A, the maximum reported removal performance reached 80% for acid-activated beideillite clay particulates grafted with organosilanes for a BET of 221 m 2 /g ( Table 2, N5). In comparison, the largest specific surface area of 608 m 2 /g of an activated charcoal adsorbent removed 40% of Alachlor (Table 2, N4). Nanofiber adsorbents made of alumina with surface-grafted organosilanes showed the second best performance with 58% removal for a specific surface area of 252 m 2 /g ( Table 2, N1). The higher performance of organosilane-grafted clays and alumina nanofibers suggests that there is competition between adsorption via hydrophobic interactions and via electrostatic interactions as the most efficient mechanism for Alachlor uptake [28,29]. Indeed, organosilane grafting provides the adsorbent with a hydrophobic surface more amenable to pesticides [28]. Each adsorbent reported in this study for the adsorption of Alachlor differs in both structure and composition. Figure 1B shows the different compositions of Alachlor adsorbents, revealing here a lower performance for carbon-based adsorbents (Table 2, N3) followed by biomass adsorbents made of kernels and leaves (Table 2, N6-N15). Several factors may explain the lower performance of activated charcoal, such as unsuitable surface pore characteristics not reported in the literature. A different structure than crystalline honeycomb carbon adsorbent may also impact the arrangement and coordination of adsorbate molecules on the charcoal surface, virtually decreasing the proportion in active sites compared to nitrogen adsorption, as well as discourage adsorption by electron interaction due to the absence of unsaturated bonds. Biomass adsorbents had a specific surface area between 275 and 394 m 2 /g, with vegetal-based adsorbents of a lower BET than kernel or stone-based adsorbents ( Table 2, N13, N7). The lower BET and fibrous aspect of leaf material, allowing for a better diffusion of the pesticide, may contribute to the increased performance compared to kernel particulates.

Impact of Adsorbent Specific and Contact Surface Area
A similar decreasing trend can be observed in the case of Diazinon removal in Figure 1C, although not enough data has been reported in the literature to accurately analyse the impact of adsorbent specific surface area on removal capacity. Adsorbents with lower specific surface areas such as 208 m 2 /g for iron oxide and silica nanocrystals were able to completely remove Diazinon, for an initial contamination of 400 µg/L and adsorbent dose of 6 g/L ( Table 3, N4). The differences in levels of initial contamination and dose of adsorbents may impact the adsorptive performance comparison. Nanofiber adsorbents such as activated carbon nanofibers with manganese and alumina nanoparticles of a respective BET of 550 m 2 /g and 600 m 2 /g showed average removal around 55% (Table 3, N1, N2). The highest BET of the reported materials at 1112 m 2 /g for activated carbons led to the lowest adsorption performance amongst the series at 16.3% (Table 3, N6). Most of the adsorbent reported fitted the Langmuir isotherm model and the pseudo-second-order kinetic model well, which suggests a homogenous monolayer surface adsorption process that is rate-limited by chemisorption, and hence surface saturation [30]. Figure 1D presents different adsorbent compositions, showing the preferential adsorption of Diazinon on ceramic surfaces over ceramic-carbonaceous adsorbents.   Concerning Atrazine removal, as represented in Figure 1E, the lower BET value reported at 110 m 2 /g for carbon nanofibers reached the least adsorption performance with a removal of 10.2% (Table 4, N1). Carbon nanotube-nanofiber composites and carbon nanotubes, however, reached a removal of 90.4% and 35% for a specific surface area of 310 m 2 /g and 300 m 2 /g, respectively (Table 4, N6, N7). Higher specific surface areas of 2235 m 2 /g for activated carbons led to complete contaminant removal (Table 4, N13). As carbon-based materials are among the highest-performing adsorbents, this suggests that π-π electron interactions and electrostatic interactions are preferential adsorption mechanisms for the removal of Atrazine. Figure 1F shows the Atrazine adsorbents' composition and other carbon-based structures; biomass absorbance has been reported in this form or extracted from shells, kernels and leaves. For a similar range of specific surface area, kernel and shell adsorbents here showed better performance compared to their leaf counterparts, with over 99% and 89% highest removal percentages, respectively (Table 4, N20, N25). This difference in performance suggests that the adsorbent structure can significantly influence the Atrazine adsorption. Furthermore, all biomass adsorbents fitted the Freundlich isotherm and pseudo-first-order kinetic models well (R 2 > 0.85), suggesting a heterogeneously distributed adsorption on the biosorbent surface and physisorption as the reaction rate-limiting step [30]. This may suggest that Atrazine could not diffuse as much nitrogen into the adsorbent surface pores or between fibers' interstitial spaces.
The different trends of performance increase for Alachlor and Diazinon and decrease for Atrazine with increasing specific surface areas may be explained by the molecular geometry of the adsorbates and preferential adsorption mechanisms. Indeed, both Alachlor and Diazinon are planar molecules, while Atrazine is non-planar. Molecule planar geometry may favour the alignment of an adsorbate molecule onto an adsorptive site, while allowing other pesticide molecules to diffuse; a non-planar molecule when adsorbed might in a certain configuration form a blockage to other pesticide molecules. In such a scenario, a lower density of functional sites and consequently a lower specific surface area can result in improved adsorptive performance. Pesticide molecule polarity could also influence the localized adsorption on the adsorbent surface, limiting adsorption performance via hydrophobic interactions. Figure 2 shows the variation in adsorption capacity related to the adsorbent diameter or sheet thickness, in the adsorption of Atrazine. An exponential-type decreasing trend of the adsorption capacity for increasing material size can be observed. Indeed, carbon nanotubes of 40 nm diameter demonstrated a capacity of 105 mg/g (Table 4, N7). On the other hand, granular activated carbon of 37 µm size showed a capacity below 10 mg/g (Table 4, N12). Carbon nanofibers and polyacrylonitrile nanofibers, both with a mean diameter of 160 nm, showed a capacity of 2.2 mg/g and 1.6 mg/g, respectively (Table 4, N1, N5). All adsorbents represented here are carbonaceous, and diameters below 200 nm were reported for nanomaterials adsorbents. A higher contact surface area, and hence a lower diameter, would thus increase the proportion of electron vacancies on the nano-adsorbent surface by decreasing the volume of bulk material. As carbon-based adsorbents are prone to adsorb by electron interaction, such as π-π stacking between adsorbent and adsorbate respective unsaturated bonds, more electron vacancies and hence a lower adsorbent diameter would result in improved performance. The choice of adsorbent structure does not seem to impact on the performance, suggesting that the surface chemistry and charge distribution play a major role in the adsorptive performance. In this study, adsorbent diameters below 100 nm on average led to adsorption capacities over 20 mg/g for the removal of Atrazine. unsaturated bonds, more electron vacancies and hence a lower adsorbent diameter would result in improved performance. The choice of adsorbent structure does not seem to impact on the performance, suggesting that the surface chemistry and charge distribution play a major role in the adsorptive performance. In this study, adsorbent diameters below 100 nm on average led to adsorption capacities over 20 mg/g for the removal of Atrazine.  2.2.2. Impact of Adsorbent Pore Size and Pore Volume Figure 3 shows adsorption capacity related to the adsorbent pore volume in the removal of Atrazine. A global trend of increasing capacity for an increased pore volume up to 1.1 cm 3 /g can be noticed (Table 4, N9). This overall trend may be explained by the higher number of atrazine molecules diffusing in a higher number of surface pores across larger pore channels. However, several factors can imply lower performance for a higher pore volume. Adsorbent surface pore size is a critical factor for pesticide diffusion. For instance, surface pores twice as large as the adsorbate molecule cross section could still allow pesticide diffusion if one molecule was adsorbed along one pore wall, and could even allow two molecules to adsorb side to side. Yet, a larger pore channel would not favour the adsorption along pore walls due to a lower promiscuity. Furthermore, a higher pore volume could also translate to a higher proportion of pores with a smaller size distribution falling below the pesticide molecule size or cross section, hence hindering the adsorbate diffusion. Also, dead-end pores would affect the determination of pore volume. Nevertheless, the overall trend indicates that a pore volume around 1.0 ± 0.1 cm 3 /g is preferential for increased performance in the removal of Atrazine.  Figure 3 shows adsorption capacity related to the adsorbent pore volume in the removal of Atrazine. A global trend of increasing capacity for an increased pore volume up to 1.1 cm 3 /g can be noticed (Table 4, N9). This overall trend may be explained by the higher number of atrazine molecules diffusing in a higher number of surface pores across larger pore channels. However, several factors can imply lower performance for a higher pore volume. Adsorbent surface pore size is a critical factor for pesticide diffusion. For instance, surface pores twice as large as the adsorbate molecule cross section could still allow pesticide diffusion if one molecule was adsorbed along one pore wall, and could even allow two molecules to adsorb side to side. Yet, a larger pore channel would not favour the adsorption along pore walls due to a lower promiscuity. Furthermore, a higher pore volume could also translate to a higher proportion of pores with a smaller size distribution falling below the pesticide molecule size or cross section, hence hindering the adsorbate diffusion. Also, dead-end pores would affect the determination of pore volume. Nevertheless, the overall trend indicates that a pore volume around 1.0 ± 0.1 cm 3 /g is preferential for increased performance in the removal of Atrazine. Atrazine Particles Nanotubes Nanosheets Nanofibers Adsorption capacity (mg/g) Pore volume (cm 3 /g) Figure 3. Adsorption capacity (mg/g) related to the adsorbent pore volume against Atrazine.
Concerning the adsorbent pore size, not enough data were found in the literature to enable a study of a preferential pore diameter for a specific pollutant among the ones considered in this section. In the case of Alachlor adsorption, organosilane-grafted beideillite clay and alumina nanofibers were reported to have a pore diameter of 6.44 nm and 11.4 nm, respectively, and showed an efficiency of 80% and 58% (Table 2, N1, N5) [28,29]. These results may suggest that an 11.4 nm wide pore channel does not offer enough promiscuity for the adsorption of Alachlor molecules, combined with a possible decrease in the efficiency of silane grafting along the adsorbent pore walls, thus impeding adsorption via hydrophobic interactions. Indeed, organosilane compounds considered for alumina nanofiber grafting include CPTES (3-chloro-propyl-triethoxysilane) and OTES (octyl-triethoxysilane), both of a weight comparable to Alachlor, 241 g/mol and 277 g/mol, of single-molecule orientation and longer carbon chain.
2.2.3. Nanofiber Performance Comparison Regarding Initial Pesticide Concentration and Adsorbent Concerning the adsorbent pore size, not enough data were found in the literature to enable a study of a preferential pore diameter for a specific pollutant among the ones considered in this section. In the case of Alachlor adsorption, organosilane-grafted beideillite clay and alumina nanofibers were reported to have a pore diameter of 6.44 nm and 11.4 nm, respectively, and showed an efficiency of 80% and 58% (Table 2, N1, N5) [28,29]. These results may suggest that an 11.4 nm wide pore channel does not offer enough promiscuity for the adsorption of Alachlor molecules, combined with a possible decrease in the efficiency of silane grafting along the adsorbent pore walls, thus impeding adsorption via hydrophobic interactions. Indeed, organosilane compounds considered for alumina nanofiber grafting include CPTES (3-chloro-propyl-triethoxysilane) and OTES (octyl-triethoxysilane), both of a weight comparable to Alachlor, 241 g/mol and 277 g/mol, of single-molecule orientation and longer carbon chain.
2.2.3. Nanofiber Performance Comparison Regarding Initial Pesticide Concentration and Adsorbent Dose Figure 4 shows bar charts representing the adsorption capacity with the corresponding tests conditions of initial pesticide concentration and adsorbent dose for each pesticide, nanofiber adsorbent and selected benchmark adsorbents. It can be noticed from all the charts that nanofibers delivered an adsorption performance on average in the range of other benchmarked adsorbents. Indeed, nanofiber adsorption capacities ranged from 5.8 to 19.5 mg/g ( Table 2, N1 and Table 4, N6), for a similar range of initial dye concentration around 10 mg/L compared to benchmark materials, yet for an adsorbent dose on average below 1 g/L. Other macro-adsorbents were reported for a dose on average over 3 g/L. The lower adsorbent dose required for nano-adsorbents compared to macrostructures should be taken into consideration when calculating the material fabrication cost and regeneration volumes.
In the case of organochlorines, Organo-layered double hydroxide nanohybrids and single walled carbon nanotubes showed the highest adsorption capacities of 113 mg/g and 130 mg/g against Alachlor and Atrazine, respectively (Table 2, N2 and Table 4, N9). However, other ceramic and carbon adsorptive nanostructures showed a lower performance for the same adsorbate respectively. Indeed, organosilane/alumina nanofibers demonstrated a capacity of 5.8 mg/g against Alachlor (Table 2, N1), and carbon nanofiber/carbon nanotube composites showed a capacity of 19.5 mg/g against Atrazine (Table 4, N6). Other than the action of other adsorption mechanisms, these performance differences for similar adsorbent sizes and compositions may also be explained by a different adsorbent pore structure that is hindering the diffusion of the pesticides, as in the case of modified alumina nanofibers, for example. Nevertheless, the predominant efficiency of carbon and metal oxide adsorbent surfaces suggests that π-π electron and electrostatic interactions are the most efficient adsorption mechanisms towards selected pesticides, compared to hydrophobic interactions, which by comparison reached average performance when serving as the predominant mechanism at stake [29]. Furthermore, carbon nanofibers were reported to show adsorptive affinity for other contaminants, such as sulfamethoxazole, in a multi-contaminant flow through adsorption test, demonstrating the versatility of carbonaceous adsorbents towards organic molecules with unsaturated bonds such as aromatic rings [39]. Polystyrene molecular imprinted nanofibers showed limited performance against Atrazine and Chlorpyrifos (Table 4, N2 and Table 5, N1). Although polyester is a hydrophobic polymer and is thus supposed to show affinity to the pesticides studied here, it is suggested that adsorption may be limited by an unsuitable solution pH, conferring on both adsorbent and adsorbate a similar surface charge [40]. Nylon-Poly(pyrrole) core-shell nanofibers showed an adsorption capacity of 2.77 mg/g, corresponding to an efficiency of 27% (Table 4, N3). The aromatic ring of Poly(pyrrole) enables π-π interaction with the Atrazine triazine ring, while nylon does not have unsaturated bonds [41]. Experiments here were reported for a neutral solution pH. However, the Poly(pyrrole) isoelectric point has been reported around 10, meaning that at neutral pH both adsorbent and atrazine exhibited a positive surface charge [41]. It was concluded that a pH slightly above 10 would favour adsorption with a negative Poly(pyrrole) surface and still slightly protonated atrazine [41].
In the case of organophosphates, benchmark adsorbents tested in similar conditions to nanofiber adsorbents could rarely be found. For instance, activated carbon nanofibers modified with either magnesium or aluminium oxide nanoparticles showed a capacity of 0.1 mg/g against Diazinon (Table 3, N1, N2). Still, against Diazinon, granular activated carbon (GAC) modified with iron oxide particles reached a capacity of 65 mg/g ( Table 3, N6). Although iron oxide would be more reactive than magnesium or aluminium oxides, the large difference in applied adsorbent dose (1g/L for nanofibers and 0.1 g/L for GAC) and in initial concentration (200 µg/L for nanofibers and 40 mg/L for GAC) would significantly impact on the difference in performance. This influence stems from the equation used to determine the adsorption capacity. Adsorptive studies carried out on adsorbent dose and level of contamination in similar conditions to those reported for the polystyrene nanofiber study could not be found in the literature (Table 5, N1).
It can be noticed from Figure 4 that higher adsorbent doses, above 5 g/L, result on average in poorer adsorption capacities, falling below 1 mg/g ( Table 2, N4). Similarly, for an average adsorbent dose around 1 g/L, a lower initial dye concentration below 0.5 mg/L tends to result in a lower adsorption capacity (Table 4, N2). These tendencies can be explained by the theory used to calculate the adsorption capacity, which will be reduced for higher adsorbent doses or lower initial dye concentrations, yet have the same removal efficiency. The optimal adsorbent dose required for a given initial concentration and a target removal efficiency could be more consistently investigated among the adsorptive studies to enable a rigorous performance comparison. given initial concentration and a target removal efficiency could be more consistently investigated among the adsorptive studies to enable a rigorous performance comparison.

Phenolic Compounds
This section reviews the performance of nanofibers reported in the adsorption of phenol, 2-and penta-chlorophenol. Other than adsorptive performance relationships with adsorbent specific surface area and pore characteristics, the influence of the adsorbate chlorine number on adsorption will be investigated. Finally, test conditions such as pH, initial contamination and adsorbent dose will be taken into consideration. Table 6 presents the molecular characteristics of Phenol, 2-Chlorophenol and Pentachlorophenol. Phenol, or carbolic acid, is a base molecule for phenolic compounds, such as explosive nitrophenols or persistent pentachlorophenol. All three compounds considered in this study have been used as disinfectants. The phenol molecule is composed of an aromatic ring derived from benzene and bound to one hydroxyl group. Phenol's smallest cross section has been calculated at 0.43 nm. The 2-chlorophenol molecule is derived from phenol, with one chlorine group attachment on the aromatic ring, at position 2. Pentachlorophenol possesses five chlorine attachments to the aromatic ring, the sixth being the hydroxyl group, a component of the phenol molecule [26]. All three molecules possess three unsaturated bonds, conferred by the benzene-like aromatic structure, thus favouring adsorption via electron π-π stacking, provided on the adsorbent surface chemistry.

Phenolic Compounds
This section reviews the performance of nanofibers reported in the adsorption of phenol, 2-and penta-chlorophenol. Other than adsorptive performance relationships with adsorbent specific surface area and pore characteristics, the influence of the adsorbate chlorine number on adsorption will be investigated. Finally, test conditions such as pH, initial contamination and adsorbent dose will be taken into consideration. Table 6 presents the molecular characteristics of Phenol, 2-Chlorophenol and Pentachlorophenol. Phenol, or carbolic acid, is a base molecule for phenolic compounds, such as explosive nitrophenols or persistent pentachlorophenol. All three compounds considered in this study have been used as disinfectants. The phenol molecule is composed of an aromatic ring derived from benzene and bound to one hydroxyl group. Phenol's smallest cross section has been calculated at 0.43 nm. The 2-chlorophenol molecule is derived from phenol, with one chlorine group attachment on the aromatic ring, at position 2. Pentachlorophenol possesses five chlorine attachments to the aromatic ring, the sixth being the hydroxyl group, a component of the phenol molecule [26]. All three molecules possess three unsaturated bonds, conferred by the benzene-like aromatic structure, thus favouring adsorption via electron π-π stacking, provided on the adsorbent surface chemistry.

Adsorbate Properties
The number of chlorine groups influences not only the molecular weight but also molecule water solubility, decreasing from 67 to 14 mg/L for phenol and pentachlorophenol, respectively (25 °C). A decrease in water solubility results in an increase in the octanol-water partition coefficient. The higher chlorine number is also believed to promote the action of hydrophobic interactions on the capture of chlorophenol molecules, providing more surface area to react with a hydrophobic adsorbent. Finally, these three molecules possess a hydroxyl termination group as part of the phenol molecule, meaning each molecule possesses one hydrogen bond donor or acceptor group [26]. As a consequence, electrostatic interactions between the adsorbent surface and the contaminant hydroxyl group could also influence the adsorbent adsorption performance.

Phenolic Compounds
This section reviews the performance of nanofibers reported in the adsorption of phenol, 2-and penta-chlorophenol. Other than adsorptive performance relationships with adsorbent specific surface area and pore characteristics, the influence of the adsorbate chlorine number on adsorption will be investigated. Finally, test conditions such as pH, initial contamination and adsorbent dose will be taken into consideration. Table 6 presents the molecular characteristics of Phenol, 2-Chlorophenol and Pentachlorophenol. Phenol, or carbolic acid, is a base molecule for phenolic compounds, such as explosive nitrophenols or persistent pentachlorophenol. All three compounds considered in this study have been used as disinfectants. The phenol molecule is composed of an aromatic ring derived from benzene and bound to one hydroxyl group. Phenol's smallest cross section has been calculated at 0.43 nm. The 2-chlorophenol molecule is derived from phenol, with one chlorine group attachment on the aromatic ring, at position 2. Pentachlorophenol possesses five chlorine attachments to the aromatic ring, the sixth being the hydroxyl group, a component of the phenol molecule [26]. All three molecules possess three unsaturated bonds, conferred by the benzene-like aromatic structure, thus favouring adsorption via electron π-π stacking, provided on the adsorbent surface chemistry.

Adsorbate Properties
The number of chlorine groups influences not only the molecular weight but also molecule water solubility, decreasing from 67 to 14 mg/L for phenol and pentachlorophenol, respectively (25 °C). A decrease in water solubility results in an increase in the octanol-water partition coefficient. The higher chlorine number is also believed to promote the action of hydrophobic interactions on the capture of chlorophenol molecules, providing more surface area to react with a hydrophobic adsorbent. Finally, these three molecules possess a hydroxyl termination group as part of the phenol molecule, meaning each molecule possesses one hydrogen bond donor or acceptor group [26]. As a consequence, electrostatic interactions between the adsorbent surface and the contaminant hydroxyl group could also influence the adsorbent adsorption performance.

Phenolic Compounds
This section reviews the performance of nanofibers reported in the adsorption of phenol, 2-and penta-chlorophenol. Other than adsorptive performance relationships with adsorbent specific surface area and pore characteristics, the influence of the adsorbate chlorine number on adsorption will be investigated. Finally, test conditions such as pH, initial contamination and adsorbent dose will be taken into consideration. Table 6 presents the molecular characteristics of Phenol, 2-Chlorophenol and Pentachlorophenol. Phenol, or carbolic acid, is a base molecule for phenolic compounds, such as explosive nitrophenols or persistent pentachlorophenol. All three compounds considered in this study have been used as disinfectants. The phenol molecule is composed of an aromatic ring derived from benzene and bound to one hydroxyl group. Phenol's smallest cross section has been calculated at 0.43 nm. The 2-chlorophenol molecule is derived from phenol, with one chlorine group attachment on the aromatic ring, at position 2. Pentachlorophenol possesses five chlorine attachments to the aromatic ring, the sixth being the hydroxyl group, a component of the phenol molecule [26]. All three molecules possess three unsaturated bonds, conferred by the benzene-like aromatic structure, thus favouring adsorption via electron π-π stacking, provided on the adsorbent surface chemistry.

Adsorbate Properties
The number of chlorine groups influences not only the molecular weight but also molecule water solubility, decreasing from 67 to 14 mg/L for phenol and pentachlorophenol, respectively (25 °C). A decrease in water solubility results in an increase in the octanol-water partition coefficient. The higher chlorine number is also believed to promote the action of hydrophobic interactions on the capture of chlorophenol molecules, providing more surface area to react with a hydrophobic adsorbent. Finally, these three molecules possess a hydroxyl termination group as part of the phenol molecule, meaning each molecule possesses one hydrogen bond donor or acceptor group [26]. As a consequence, electrostatic interactions between the adsorbent surface and the contaminant hydroxyl group could also influence the adsorbent adsorption performance. The number of chlorine groups influences not only the molecular weight but also molecule water solubility, decreasing from 67 to 14 mg/L for phenol and pentachlorophenol, respectively (25 • C). A decrease in water solubility results in an increase in the octanol-water partition coefficient. The higher chlorine number is also believed to promote the action of hydrophobic interactions on the capture of chlorophenol molecules, providing more surface area to react with a hydrophobic adsorbent. Finally, these three molecules possess a hydroxyl termination group as part of the phenol molecule, meaning each molecule possesses one hydrogen bond donor or acceptor group [26]. As a consequence, electrostatic interactions between the adsorbent surface and the contaminant hydroxyl group could also influence the adsorbent adsorption performance. Figure 5 shows the adsorption performance of phenol and chlorophenols related to the adsorbent specific surface area. A near-linear trend of increase in the adsorption capacity for increasing adsorbent specific surface area can be observed in the case of phenol adsorption, as shown in Figure 5A. This near-linearity can be bound to a near-homogeneous surface saturation process, with a higher proportion in adsorptive sites, i.e., BET proportionally corresponding to a higher adsorption capacity. In over 80% of the reported kinetic and isotherm studies for phenol adsorption, adsorbents were found to follow pseudo-second order and Langmuir models, translating into a surface monolayer adsorption reaction that is limited by chemisorption, hence further suggesting homogeneous surface saturation (Table 7). Plain carbon nanofibers and iron-oxide-decorated carbon nanofibers demonstrated a capacity of 0.842 and 1.684 mg/g for a specific surface area of 40.7 and 72.4 m 2 /g, respectively (Table 7, N17, N18). However, activated carbon nanofibers showed a higher performance of 256.1 mg/g for a higher BET of 2921 m 2 /g ( Table 7, N20). Figure 5B details the adsorbent composition used for phenol capture, showing a majority of carbon or carbon-based adsorbents following the near-linear adsorption trend. An explanation of this adsorption performance trend formed by carbonaceous adsorbents relies on the donor-acceptor electron complex adsorption mechanism. Actors in the electron interaction for adsorption are the carbonyl groups on the carbon-adsorbent surface (electron donor), the proportion of which influences the specific surface area, and the phenol compound aromatic ring (electron acceptor) [4].

Nanofibers for Disinfectant Adsorption
3.2.1. Impact of Adsorbent Specific and Contact Surface Area Figure 5 shows the adsorption performance of phenol and chlorophenols related to the adsorbent specific surface area. A near-linear trend of increase in the adsorption capacity for increasing adsorbent specific surface area can be observed in the case of phenol adsorption, as shown in Figure 5A. This near-linearity can be bound to a near-homogeneous surface saturation process, with a higher proportion in adsorptive sites, i.e., BET proportionally corresponding to a higher adsorption capacity. In over 80% of the reported kinetic and isotherm studies for phenol adsorption, adsorbents were found to follow pseudo-second order and Langmuir models, translating into a surface monolayer adsorption reaction that is limited by chemisorption, hence further suggesting homogeneous surface saturation (Table 7). Plain carbon nanofibers and iron-oxide-decorated carbon nanofibers demonstrated a capacity of 0.842 and 1.684 mg/g for a specific surface area of 40.7 and 72.4 m 2 /g, respectively (Table 7, N17, N18). However, activated carbon nanofibers showed a higher performance of 256.1 mg/g for a higher BET of 2921 m 2 /g ( Table 7, N20). Figure 5B details the adsorbent composition used for phenol capture, showing a majority of carbon or carbon-based adsorbents following the near-linear adsorption trend. An explanation of this adsorption performance trend formed by carbonaceous adsorbents relies on the donor-acceptor electron complex adsorption mechanism. Actors in the electron interaction for adsorption are the carbonyl groups on the carbon-adsorbent surface (electron donor), the proportion of which influences the specific surface area, and the phenol compound aromatic ring (electron acceptor) [4].  The observed preference for the Langmuir isotherm model can further be linked to the role of π-π dispersion forces between the aromatic rings of phenol and activated carbon. Indeed, a process of localized monolayer adsorption could signify that each site is accommodating a single molecule with repulsive interactions between adsorbed molecules [4]. Ceramic and clay adsorbents, although less reported, also showed to follow the adsorption trend. Therefore, adsorbent structure rather than composition is affecting the performance for a given BET. Styrene divinylbenzene particles, here reported under polymer adsorbents and not following the trend, have a particle size in the macrorange, and offered a capacity of 8.83 mg/g for a BET of 1000 m 2 /g ( Table 7, N32). In the case of carbon nanotubes, which do not fit the global performance trend, their capacity of 166.58 mg/g for a BET of 357 m 2 /g ( Table 7, N25) may be explained by their hollow structure. Variations in test conditions may also significantly affect the adsorbent performance for a given specific surface area. For instance, temperature was shown to act as a catalyst for phenol adsorption on doped cryptomelane manganese nanofibers, with faster metal oxide surface hydroxylation reported for temperatures from 140 °C ( Table 7, N19). The created surface hydroxyl groups can bind to phenol and form phenolic complexes, thus enhancing the adsorption capacity even for a relatively low specific surface area of 40 m 2 /g, impeded by the higher density of metals compared to carbon atoms [47].
On the other hand, no clear trend can be pulled out in the case of 2-and pentachlorophenol adsorption, presented in Figure 5C-F respectively. Although fewer data are available, it can be seen that the higher the chlorine number on the adsorbate, the smaller the range of obtained adsorption capacities. Carbon nanofibers with a BET of 2326 m 2 /g showed the maximum capacity of 167 mg/g for 2-chlorophenol adsorption ( Table 8, N1). In the case of pentachlorophenol, carbon black with a specific surface area of 511 m 2 /g was reported to show a capacity of 250 mg/g ( Table 9, N6). However, over 90% of the adsorbents showed a capacity between 44.7 and 98 mg/g for a BET ranging from 3.7 to 958 m 2 /g for the capture of pentachlorophenol (Table 9, N1, N5, N12). The gradual decrease in observed performance for an increasing number of chlorine groups on the phenol molecule can be explained by the larger structure and the modified surface polarity of the molecule. Indeed, the chlorine group, being an electron-withdrawing group, lowers the aromatic ring electron density, or deactivates the ring, thus potentially hindering adsorption onto carbon or metal oxide surfaces by electron interaction and hindering hydrophobic interactions. The observed preference for the Langmuir isotherm model can further be linked to the role of π-π dispersion forces between the aromatic rings of phenol and activated carbon. Indeed, a process of localized monolayer adsorption could signify that each site is accommodating a single molecule with repulsive interactions between adsorbed molecules [4]. Ceramic and clay adsorbents, although less reported, also showed to follow the adsorption trend. Therefore, adsorbent structure rather than composition is affecting the performance for a given BET. Styrene divinylbenzene particles, here reported under polymer adsorbents and not following the trend, have a particle size in the macro-range, and offered a capacity of 8.83 mg/g for a BET of 1000 m 2 /g ( Table 7, N32). In the case of carbon nanotubes, which do not fit the global performance trend, their capacity of 166.58 mg/g for a BET of 357 m 2 /g ( Table 7, N25) may be explained by their hollow structure. Variations in test conditions may also significantly affect the adsorbent performance for a given specific surface area. For instance, temperature was shown to act as a catalyst for phenol adsorption on doped cryptomelane manganese nanofibers, with faster metal oxide surface hydroxylation reported for temperatures from 140 • C (Table 7, N19). The created surface hydroxyl groups can bind to phenol and form phenolic complexes, thus enhancing the adsorption capacity even for a relatively low specific surface area of 40 m 2 /g, impeded by the higher density of metals compared to carbon atoms [47].
On the other hand, no clear trend can be pulled out in the case of 2-and pentachlorophenol adsorption, presented in Figure 5C-F respectively. Although fewer data are available, it can be seen that the higher the chlorine number on the adsorbate, the smaller the range of obtained adsorption capacities. Carbon nanofibers with a BET of 2326 m 2 /g showed the maximum capacity of 167 mg/g for 2-chlorophenol adsorption ( Table 8, N1). In the case of pentachlorophenol, carbon black with a specific surface area of 511 m 2 /g was reported to show a capacity of 250 mg/g ( Table 9, N6). However, over 90% of the adsorbents showed a capacity between 44.7 and 98 mg/g for a BET ranging from 3.7 to 958 m 2 /g for the capture of pentachlorophenol (Table 9, N1, N5, N12). The gradual decrease in observed performance for an increasing number of chlorine groups on the phenol molecule can be explained by the larger structure and the modified surface polarity of the molecule. Indeed, the chlorine group, being an electron-withdrawing group, lowers the aromatic ring electron density, or deactivates the ring, thus potentially hindering adsorption onto carbon or metal oxide surfaces by electron interaction and hindering hydrophobic interactions.

Impact of Adsorbent Pore Size
The impact of adsorbent pore size on the adsorption capacity against phenol is illustrated on Figure 6. There is a noticeable trend, characterised by an exponential-type decrease in the adsorption capacity with the increase of adsorbent pore size. The impact of adsorbent pore size on the adsorption capacity against phenol is illustrated on Figure 6. There is a noticeable trend, characterised by an exponential-type decrease in the adsorption capacity with the increase of adsorbent pore size. For phenol capture, adsorbent pore size was reported from 0.43 nm (activated carbon nanofibers, Table 7, N20) to 15 nm (diatomite nanocomposites, Table 7, N23), corresponding to an adsorption capacity of 256.1 and 2 mg/g, respectively, being the maximal and minimal capacities here reported. Maximum adsorption capacities were observed for a pore size in a range between 0.43 and 2.2 nm ( Table 7, N20, N4). The optimum pore size range of the adsorbent can be linked to the kinetics of pore blocking (i.e., phenol diffusion rate), adsorbent pore geometry and the single orientation of the phenol molecule ( Table 6). Considering that the smallest cross section of the phenol molecule is 0.43 nm, mesopores (from 2 to 50 nm) may insufficiently promote adsorbent/adsorbate contact, thus resulting in a potentially lower performance depending on the adsorption mechanisms at stake ( Table 7, N7, N32). On the other hand, activated carbon nanofibers were shown to possess an open pore structure and a dual pore size repartition of 0.43 nm and 0.8 nm, allowing the penetration of one or two phenol molecules, respectively (Table 7, N20). Dual pore size distribution on adsorbent surfaces can be further advocated by the fitting of the adsorption reaction to a Freundlich isotherm.
Furthermore, average adsorbent pore size has been calculated from a Barrett-Joyner-Halenda (BJH) method, and may not accurately reflect the adsorbent pore size distribution, depending on the actual pore tortuosity. However, a pore size calculated by BJH method and falling in the optimal range defined above may suggest an increased volume of micropores enhancing phenol adsorption. Using a BET/BJH method, the presence of micropores leads to an isotherm of type I, with increased nitrogen adsorption at low pressures [53]. Similarly to a Langmuir isotherm, the height of the plateau on the isotherm would reflect the volume of micropores, also called micropore capacity at the adsorbent surface. The extension of the isotherm hysteresis to relatively low pressures would give an indication of the micropores' diameter [53].
Beside the adsorbent morphology, voids between the nanofibers in the case of fibrous mats could also impact the diffusion of the phenol molecules; however, over 95% of the studies reported a nanofiber web fabrication by a solution blowing or electrospinning process, hence ensuring webs of high porosity and of suitably large pore sizes to enable phenol diffusion. This last aspect is critical as the hydrodynamics of the solution diffusion, and therefore the impact of the webs' wettability, is critical to the design of nanofibrous adsorbents. For phenol capture, adsorbent pore size was reported from 0.43 nm (activated carbon nanofibers, Table 7, N20) to 15 nm (diatomite nanocomposites, Table 7, N23), corresponding to an adsorption capacity of 256.1 and 2 mg/g, respectively, being the maximal and minimal capacities here reported. Maximum adsorption capacities were observed for a pore size in a range between 0.43 and 2.2 nm ( Table 7, N20, N4). The optimum pore size range of the adsorbent can be linked to the kinetics of pore blocking (i.e., phenol diffusion rate), adsorbent pore geometry and the single orientation of the phenol molecule ( Table 6). Considering that the smallest cross section of the phenol molecule is 0.43 nm, mesopores (from 2 to 50 nm) may insufficiently promote adsorbent/adsorbate contact, thus resulting in a potentially lower performance depending on the adsorption mechanisms at stake (Table 7, N7, N32). On the other hand, activated carbon nanofibers were shown to possess an open pore structure and a dual pore size repartition of 0.43 nm and 0.8 nm, allowing the penetration of one or two phenol molecules, respectively (Table 7, N20). Dual pore size distribution on adsorbent surfaces can be further advocated by the fitting of the adsorption reaction to a Freundlich isotherm.
Furthermore, average adsorbent pore size has been calculated from a Barrett-Joyner-Halenda (BJH) method, and may not accurately reflect the adsorbent pore size distribution, depending on the actual pore tortuosity. However, a pore size calculated by BJH method and falling in the optimal range defined above may suggest an increased volume of micropores enhancing phenol adsorption. Using a BET/BJH method, the presence of micropores leads to an isotherm of type I, with increased nitrogen adsorption at low pressures [53]. Similarly to a Langmuir isotherm, the height of the plateau on the isotherm would reflect the volume of micropores, also called micropore capacity at the adsorbent surface. The extension of the isotherm hysteresis to relatively low pressures would give an indication of the micropores' diameter [53].
Beside the adsorbent morphology, voids between the nanofibers in the case of fibrous mats could also impact the diffusion of the phenol molecules; however, over 95% of the studies reported a nanofiber web fabrication by a solution blowing or electrospinning process, hence ensuring webs of high porosity and of suitably large pore sizes to enable phenol diffusion. This last aspect is critical as the hydrodynamics of the solution diffusion, and therefore the impact of the webs' wettability, is critical to the design of nanofibrous adsorbents.

Impact of Adsorbent Surface Charge and Solution pH
Adsorbent isoelectric point (IEP) corresponds to the pH at which the adsorbent carries no net electrical charge and is used to determine adsorbent global surface charge in an aqueous medium in function of the solution pH. As phenol and phenolic compounds are polar molecules, the adsorbent surface charge can significantly enhance the phenolic compound adsorptive performance with the attraction of two opposite charges between adsorbent and adsorbate. Phenol adsorption up to pH 7 has been experimented on with carbon nanofibers, resulting in a 31% removal capacity [89]. Nevertheless, the adsorbent IEP was not reported. Considering that phenol is a weak acid with a pKa of 9.95 (Table 6), an iso-electric point between pH 4 and pH 6.5 would be ideal to get the phenol molecule to bind to a carbon adsorbent surface for a reaction pH slightly below the adsorbent theoretical IEP and the phenol pKa. Adsorption studies at a reaction pH superior to the adsorbate pKa have been carried out for meta-aramid/polyacrylic acid nanofibers on another phenol compound, namely 2,4-dichlorophenol [90]. It was reported that the phenol compound dissociated into chlorophenolate anions, thus impeding the adsorption performance, given the electrostatic repulsion between identical charged target molecules and fiber active sites [90]. Solution pH therefore plays a determinant role in the efficiency of adsorption by electrostatic interaction. Ionic strength also has an impact on the adsorption process. Indeed, electrical double layers can form on the adsorbent surface [41]. In a case of a neutral pH, when falling below the adsorbent isoelectric point the adsorbent surface would carry a positive charge, and would therefore attract the anions present in the solution to the interface to form what is called a stern layer. The protonated contaminant would also carry a positive surface charge and would be found in the diffuse layer, attracted by anions in the stern layer. The thickness of the electrical double layers can prevent the adsorption reaction. However, when the ionic strength increases, the electrolyte will compress the electrical double layer and thus influence electrostatic interactions between adsorbate and adsorbent [41]. Figure 7 shows the variation in adsorption performance in function of the solution pH. It can be seen from all graphs that, starting from a solution pH of 6, the performance decreases with increasing pH. A solution pH between 4.5 and 6 seems to be the optimal window for metal oxide-and carbon-based adsorbents, confirming the theory explained above. The microfiber/nanofiber carbon composite adsorbent was surface-functionalized with ammonia [57]. It was reported that the inclusion of nitrogen groups on the adsorbent surface increased the adsorbent basicity, thus facilitating phenol adsorption at a close to neutral pH [57].

Impact of Adsorbent Surface Charge and Solution pH
Adsorbent isoelectric point (IEP) corresponds to the pH at which the adsorbent carries no net electrical charge and is used to determine adsorbent global surface charge in an aqueous medium in function of the solution pH. As phenol and phenolic compounds are polar molecules, the adsorbent surface charge can significantly enhance the phenolic compound adsorptive performance with the attraction of two opposite charges between adsorbent and adsorbate. Phenol adsorption up to pH 7 has been experimented on with carbon nanofibers, resulting in a 31% removal capacity [89]. Nevertheless, the adsorbent IEP was not reported. Considering that phenol is a weak acid with a pKa of 9.95 (Table 6), an iso-electric point between pH 4 and pH 6.5 would be ideal to get the phenol molecule to bind to a carbon adsorbent surface for a reaction pH slightly below the adsorbent theoretical IEP and the phenol pKa. Adsorption studies at a reaction pH superior to the adsorbate pKa have been carried out for meta-aramid/polyacrylic acid nanofibers on another phenol compound, namely 2,4-dichlorophenol [90]. It was reported that the phenol compound dissociated into chlorophenolate anions, thus impeding the adsorption performance, given the electrostatic repulsion between identical charged target molecules and fiber active sites [90]. Solution pH therefore plays a determinant role in the efficiency of adsorption by electrostatic interaction. Ionic strength also has an impact on the adsorption process. Indeed, electrical double layers can form on the adsorbent surface [41]. In a case of a neutral pH, when falling below the adsorbent isoelectric point the adsorbent surface would carry a positive charge, and would therefore attract the anions present in the solution to the interface to form what is called a stern layer. The protonated contaminant would also carry a positive surface charge and would be found in the diffuse layer, attracted by anions in the stern layer. The thickness of the electrical double layers can prevent the adsorption reaction. However, when the ionic strength increases, the electrolyte will compress the electrical double layer and thus influence electrostatic interactions between adsorbate and adsorbent [41]. Figure 7 shows the variation in adsorption performance in function of the solution pH. It can be seen from all graphs that, starting from a solution pH of 6, the performance decreases with increasing pH. A solution pH between 4.5 and 6 seems to be the optimal window for metal oxide-and carbonbased adsorbents, confirming the theory explained above. The microfiber/nanofiber carbon composite adsorbent was surface-functionalized with ammonia [57]. It was reported that the inclusion of nitrogen groups on the adsorbent surface increased the adsorbent basicity, thus facilitating phenol adsorption at a close to neutral pH [57]. Horseradish peroxidase immobilized on polymer nanofibers was tested in the adsorption of Pentachlorophenol in several solution pH conditions (Table 9, N1). Sorption capacities were reported of 15.27 mg/L, 14.33 mg/L and 11.85 mg/L for an increasing pH of 2, 3, and 4, with no sorption occurring at pH 5 [78]. This direct proportion to solution acidity can be related to the protonation degree of Pentachlorophenol [78]. Since the hydrophobicity of Pentachlorophenol was found to influence the sorption rate, hydrophobic interactions can be identified as a significant capture mechanism. Other mechanisms reported π-π electron interactions between the Pentachlorophenol benzene ring and π electrons of the ketone group from the polymer molecular skeleton, as well as hydrogen bonding between the polymer ketone groups and Pentachlorophenol hydroxyl group [78]. Figure 8 shows the relationship between adsorption capacity and reaction equilibrium time for different adsorptive structures against phenol. These graphs aim to show the average shorter equilibrium times (under 5 h) reported for nanostructures compared to macro-adsorbents, as can be noticed in Figure 8A, as well as the yield of higher adsorption capacities for nanofibers compared to other nano-structures, with, for instance, activated carbon nanofibers reaching a capacity of 256.1 mg/g after 1.3 h (Table 7, N20). The difference in adsorptive performance and kinetics can be explained by the stability of nanofibers in water, while nanoparticle adsorbents tend to agglomerate under the effect of surface interactions such as Van der Waals, hence hindering the adsorption reaction. The difference in kinetics between nanofibers and macro-structures can be the result of the higher surface to volume ratio of nanofibers. Figure 8B details the composition of the reported adsorbents, revealing a combination of faster kinetics and higher performance for carbonaceous adsorbents, hence suggesting electron interaction as the most efficient predominant adsorption mechanism in the capture of phenol [91]. Horseradish peroxidase immobilized on polymer nanofibers was tested in the adsorption of Pentachlorophenol in several solution pH conditions (Table 9, N1). Sorption capacities were reported of 15.27 mg/L, 14.33 mg/L and 11.85 mg/L for an increasing pH of 2, 3, and 4, with no sorption occurring at pH 5 [78]. This direct proportion to solution acidity can be related to the protonation degree of Pentachlorophenol [78]. Since the hydrophobicity of Pentachlorophenol was found to influence the sorption rate, hydrophobic interactions can be identified as a significant capture mechanism. Other mechanisms reported π-π electron interactions between the Pentachlorophenol benzene ring and π electrons of the ketone group from the polymer molecular skeleton, as well as hydrogen bonding between the polymer ketone groups and Pentachlorophenol hydroxyl group [78]. Figure 8 shows the relationship between adsorption capacity and reaction equilibrium time for different adsorptive structures against phenol. These graphs aim to show the average shorter equilibrium times (under 5 h) reported for nanostructures compared to macro-adsorbents, as can be noticed in Figure 8A, as well as the yield of higher adsorption capacities for nanofibers compared to other nano-structures, with, for instance, activated carbon nanofibers reaching a capacity of 256.1 mg/g after 1.3 h (Table 7, N20). The difference in adsorptive performance and kinetics can be explained by the stability of nanofibers in water, while nanoparticle adsorbents tend to agglomerate under the effect of surface interactions such as Van der Waals, hence hindering the adsorption reaction. The difference in kinetics between nanofibers and macro-structures can be the result of the higher surface to volume ratio of nanofibers. Figure 8B details the composition of the reported adsorbents, revealing a combination of faster kinetics and higher performance for carbonaceous adsorbents, hence suggesting electron interaction as the most efficient predominant adsorption mechanism in the capture of phenol [91].  Table 7, N6). For the same initial contamination, activated carbon microfiber-nanofiber composites showed a capacity of 150 mg/g for a dose of 0.6 g/L ( Table 7, N16). It can be seen from all charts that carbon-based adsorbents are amongst the best against phenol and chlorophenols.   Figure 9 shows bar charts representing the adsorption capacity with the corresponding test conditions of initial disinfectant concentration and adsorbent dose for each nanofiber adsorbent and selected benchmark adsorbents. Comparatively to relatively hydrophilic pesticides, nanofiber adsorbents demonstrated an adsorption capacity in the range of the one performed by activated carbon adsorbents, yet for a lower adsorbent dose. For instance, in the case of phenol capture, activated carbon treated with potassium permanganate demonstrated a capacity of 142.15 mg/g for a contamination of 100 mg/L and a dose of 1 g/L ( Table 7, N6). For the same initial contamination, activated carbon microfiber-nanofiber composites showed a capacity of 150 mg/g for a dose of 0.6 g/L ( Table 7, N16). It can be seen from all charts that carbon-based adsorbents are amongst the best against phenol and chlorophenols.  Figure 9 shows bar charts representing the adsorption capacity with the corresponding test conditions of initial disinfectant concentration and adsorbent dose for each nanofiber adsorbent and selected benchmark adsorbents. Comparatively to relatively hydrophilic pesticides, nanofiber adsorbents demonstrated an adsorption capacity in the range of the one performed by activated carbon adsorbents, yet for a lower adsorbent dose. For instance, in the case of phenol capture, activated carbon treated with potassium permanganate demonstrated a capacity of 142.15 mg/g for a contamination of 100 mg/L and a dose of 1 g/L ( Table 7, N6). For the same initial contamination, activated carbon microfiber-nanofiber composites showed a capacity of 150 mg/g for a dose of 0.6 g/L ( Table 7, N16). It can be seen from all charts that carbon-based adsorbents are amongst the best against phenol and chlorophenols.

Persistent Organic Pollutants
In this section, the adsorptive performance of nanofiber-based adsorbents is benchmarked and reviewed for the capture of selected aliphatic organochlorines, and selected aromatic compounds from the PCBs and PAHs families. Adsorbate molecules considered in this section are classified POPs (persistent pesticides) and industrial products. Characteristics of the adsorbate shall first be discussed, followed by reported adsorbent structure-performance relationships. Table 10 shows the physico-chemical characteristics of selected aliphatic organochlorines in this review, namely Aldrin, Dieldrin, Endrin, Chlordane, Heptachlor, Heptachlor epoxide, and Toxaphene 8. Drin pesticides possess an alicyclic structure with 6 chlorine groups. The Aldrin molecule is formed by reaction of norbornadiene, a bicyclic hydrocarbon, and hexachlorocyclopentadiene, also called C-56. In specific conditions, such as in living organisms, Aldrin can be oxidised to form Dieldrin. Endrin is a stereoisomer of Dieldrin, differing in the spatial orientation of the oxygen group, as well as one hydrogen and one chlorine bond. Cyclodiene compounds selected in this study also comprise Chlordane and Heptachlor. Also derived from hexachlorocyclopentadiene, Chlordane and Heptachlor have eight and seven chlorine atoms, respectively, surrounding the polycyclic organic structure. It has been observed that when Heptachlor penetrates the soil, it can be turned into Heptachlor epoxide by soil microorganisms in an epoxy reaction. Toxaphene is another polycyclic organochlorine compound, formed by the chlorination of camphene, usually by chlorine gas. Also called camphechlor or polychlorocamphene, Toxaphene 8 possesses eight chlorine groups [26].

Adsorbate Properties
Except for Toxaphene, all molecules present one unsaturated bond, enabling adsorption via π-π stacking of electrons with suitable adsorbents. Dieldrin and Endrin present three chlorine groups at proximity, while other molecules show two chlorine groups, except for Aldrin, which does not show any hydrophobic group near the unsaturated bond.
Selected organochlorine POPs are of a similar weight, ranging from 365 g/mol for Aldrin to 414 g/mol for Toxaphene; however, they differ in their spatial orientation. Apolar molecules have a solubility in water below 0.6 mg/L, with octanol-water partition coefficients over 5. Dieldrin, Endrin and Heptachlor epoxide are only slightly polar molecules due to the presence of one oxygen group (an electron acceptor group), which significantly improves their water affinity (0.019 mg/L for Aldrin and 0.19 mg/L for Dieldrin) and is believed to improve binding with adsorbents by electrostatic interaction.

Persistent Organic Pollutants
In this section, the adsorptive performance of nanofiber-based adsorbents is benchmarked and reviewed for the capture of selected aliphatic organochlorines, and selected aromatic compounds from the PCBs and PAHs families. Adsorbate molecules considered in this section are classified POPs (persistent pesticides) and industrial products. Characteristics of the adsorbate shall first be discussed, followed by reported adsorbent structure-performance relationships. Table 10 shows the physico-chemical characteristics of selected aliphatic organochlorines in this review, namely Aldrin, Dieldrin, Endrin, Chlordane, Heptachlor, Heptachlor epoxide, and Toxaphene 8. Drin pesticides possess an alicyclic structure with 6 chlorine groups. The Aldrin molecule is formed by reaction of norbornadiene, a bicyclic hydrocarbon, and hexachlorocyclopentadiene, also called C-56. In specific conditions, such as in living organisms, Aldrin can be oxidised to form Dieldrin. Endrin is a stereoisomer of Dieldrin, differing in the spatial orientation of the oxygen group, as well as one hydrogen and one chlorine bond. Cyclodiene compounds selected in this study also comprise Chlordane and Heptachlor. Also derived from hexachlorocyclopentadiene, Chlordane and Heptachlor have eight and seven chlorine atoms, respectively, surrounding the polycyclic organic structure. It has been observed that when Heptachlor penetrates the soil, it can be turned into Heptachlor epoxide by soil microorganisms in an epoxy reaction. Toxaphene is another polycyclic organochlorine compound, formed by the chlorination of camphene, usually by chlorine gas. Also called camphechlor or polychlorocamphene, Toxaphene 8 possesses eight chlorine groups [26].

Adsorbate Properties
Except for Toxaphene, all molecules present one unsaturated bond, enabling adsorption via ππ stacking of electrons with suitable adsorbents. Dieldrin and Endrin present three chlorine groups at proximity, while other molecules show two chlorine groups, except for Aldrin, which does not show any hydrophobic group near the unsaturated bond.
Selected organochlorine POPs are of a similar weight, ranging from 365 g/mol for Aldrin to 414 g/mol for Toxaphene; however, they differ in their spatial orientation. Apolar molecules have a solubility in water below 0.6 mg/L, with octanol-water partition coefficients over 5. Dieldrin, Endrin and Heptachlor epoxide are only slightly polar molecules due to the presence of one oxygen group (an electron acceptor group), which significantly improves their water affinity (0.019 mg/L for Aldrin and 0.19 mg/L for Dieldrin) and is believed to improve binding with adsorbents by electrostatic interaction.

Persistent Organic Pollutants
In this section, the adsorptive performance of nanofiber-based adsorbents is benchmarked and reviewed for the capture of selected aliphatic organochlorines, and selected aromatic compounds from the PCBs and PAHs families. Adsorbate molecules considered in this section are classified POPs (persistent pesticides) and industrial products. Characteristics of the adsorbate shall first be discussed, followed by reported adsorbent structure-performance relationships. Table 10 shows the physico-chemical characteristics of selected aliphatic organochlorines in this review, namely Aldrin, Dieldrin, Endrin, Chlordane, Heptachlor, Heptachlor epoxide, and Toxaphene 8. Drin pesticides possess an alicyclic structure with 6 chlorine groups. The Aldrin molecule is formed by reaction of norbornadiene, a bicyclic hydrocarbon, and hexachlorocyclopentadiene, also called C-56. In specific conditions, such as in living organisms, Aldrin can be oxidised to form Dieldrin. Endrin is a stereoisomer of Dieldrin, differing in the spatial orientation of the oxygen group, as well as one hydrogen and one chlorine bond. Cyclodiene compounds selected in this study also comprise Chlordane and Heptachlor. Also derived from hexachlorocyclopentadiene, Chlordane and Heptachlor have eight and seven chlorine atoms, respectively, surrounding the polycyclic organic structure. It has been observed that when Heptachlor penetrates the soil, it can be turned into Heptachlor epoxide by soil microorganisms in an epoxy reaction. Toxaphene is another polycyclic organochlorine compound, formed by the chlorination of camphene, usually by chlorine gas. Also called camphechlor or polychlorocamphene, Toxaphene 8 possesses eight chlorine groups [26].

Adsorbate Properties
Except for Toxaphene, all molecules present one unsaturated bond, enabling adsorption via ππ stacking of electrons with suitable adsorbents. Dieldrin and Endrin present three chlorine groups at proximity, while other molecules show two chlorine groups, except for Aldrin, which does not show any hydrophobic group near the unsaturated bond.
Selected organochlorine POPs are of a similar weight, ranging from 365 g/mol for Aldrin to 414 g/mol for Toxaphene; however, they differ in their spatial orientation. Apolar molecules have a solubility in water below 0.6 mg/L, with octanol-water partition coefficients over 5. Dieldrin, Endrin and Heptachlor epoxide are only slightly polar molecules due to the presence of one oxygen group (an electron acceptor group), which significantly improves their water affinity (0.019 mg/L for Aldrin and 0.19 mg/L for Dieldrin) and is believed to improve binding with adsorbents by electrostatic interaction.

Persistent Organic Pollutants
In this section, the adsorptive performance of nanofiber-based adsorbents is benchmarked and reviewed for the capture of selected aliphatic organochlorines, and selected aromatic compounds from the PCBs and PAHs families. Adsorbate molecules considered in this section are classified POPs (persistent pesticides) and industrial products. Characteristics of the adsorbate shall first be discussed, followed by reported adsorbent structure-performance relationships. Table 10 shows the physico-chemical characteristics of selected aliphatic organochlorines in this review, namely Aldrin, Dieldrin, Endrin, Chlordane, Heptachlor, Heptachlor epoxide, and Toxaphene 8. Drin pesticides possess an alicyclic structure with 6 chlorine groups. The Aldrin molecule is formed by reaction of norbornadiene, a bicyclic hydrocarbon, and hexachlorocyclopentadiene, also called C-56. In specific conditions, such as in living organisms, Aldrin can be oxidised to form Dieldrin. Endrin is a stereoisomer of Dieldrin, differing in the spatial orientation of the oxygen group, as well as one hydrogen and one chlorine bond. Cyclodiene compounds selected in this study also comprise Chlordane and Heptachlor. Also derived from hexachlorocyclopentadiene, Chlordane and Heptachlor have eight and seven chlorine atoms, respectively, surrounding the polycyclic organic structure. It has been observed that when Heptachlor penetrates the soil, it can be turned into Heptachlor epoxide by soil microorganisms in an epoxy reaction. Toxaphene is another polycyclic organochlorine compound, formed by the chlorination of camphene, usually by chlorine gas. Also called camphechlor or polychlorocamphene, Toxaphene 8 possesses eight chlorine groups [26].

Adsorbate Properties
Except for Toxaphene, all molecules present one unsaturated bond, enabling adsorption via ππ stacking of electrons with suitable adsorbents. Dieldrin and Endrin present three chlorine groups at proximity, while other molecules show two chlorine groups, except for Aldrin, which does not show any hydrophobic group near the unsaturated bond.
Selected organochlorine POPs are of a similar weight, ranging from 365 g/mol for Aldrin to 414 g/mol for Toxaphene; however, they differ in their spatial orientation. Apolar molecules have a solubility in water below 0.6 mg/L, with octanol-water partition coefficients over 5. Dieldrin, Endrin and Heptachlor epoxide are only slightly polar molecules due to the presence of one oxygen group (an electron acceptor group), which significantly improves their water affinity (0.019 mg/L for Aldrin and 0.19 mg/L for Dieldrin) and is believed to improve binding with adsorbents by electrostatic interaction.   Figure 10 shows the relationship between adsorption capacity and test conditions of initial contaminant concentration and adsorbent dose for Aldrin, Dieldrin, Endrin pesticides plus Heptachlor and its epoxide. Polystyrene (PES) nanofibers decorated with titania nanoparticles showed limited adsorptive performance for these pesticides, as well as against Chlordane, showing a capacity of 0.02 mg/g (Table 11, N1) [92]. Adsorbent surface properties were not reported, yet the poor affinity of hydrophobic polyester for hydrophobic organochlorines could have impacted the performance, with adsorption reaction localized on the titania surface. In the case of Dieldrin removal, bromopropyl functionalised silica nanofibers revealed an adsorption capacity of 0.19 mg/g for contamination and adsorbent dose of 20 µg/L and 0.1 g/L, respectively (Table 12, N1). In comparison, untreated silica nanofibers achieved a capacity of 0.014 mg/g (Table 12, N2). The addition of a dense bromopropyl layer on silica nanofibers produced a hydrophobic adsorbent surface, with chemical affinity with pesticide chlorine groups.
Concerning Aldrin removal, vegetal activated carbon showed a relatively high performance with a capacity of 0.38 mg/g for a contamination of 2 mg/L, yet for a relatively large dose of 5 g/L (Table 13, N2). Vegetal activated carbon performed a similar role in the case of Dieldrin adsorption, in the same conditions (Table 12, N8). Against Heptachlor epoxide, the vegetal activated carbon was reported to show a capacity of 0.42 mg/g for a contamination of 2.2 mg/L for a similar dose of 5 g/L (Table 14, N7). Finally, in the case of Endrin, powdered activated carbon was reported to show a capacity of 0.085 mg/g for a contamination of 0.8 µg/L, and a relatively large dose of 0.2 g/L (Table  15, N6). These results show that although adsorbent specific surface area and pore channels could be adapted to improve the adsorptive performance, activated carbons yield removal efficiencies over 95%. However, compared to nanomaterial adsorbents, a large adsorbent dose is required.
Acid-treated date and olive stones showed a significantly improved performance over nontreated biomass adsorbents. For instance, acid-treated date stones showed a capacity of 0.45 mg/g in the removal of Aldrin, while untreated date stones yielded 0.05 µg/g for conditions of contamination and adsorbent dose of respectively 500 µg/L, 1 g/L and 0.5 µ g/L and 10 g/L (Table 13, N9,N12). Date stones were thermally treated with hydrochloric acid, thus creating a hydrophobic layer on the adsorbent surface, and facilitating the anchorage of organochlorines [93]. In the adsorption of Dieldrin and Endrin, acid-treated date stones showed a capacity of 0.44 mg/g and 0.40 mg/g, in the same conditions as Aldrin capture (  Figure 10 shows the relationship between adsorption capacity and test conditions of initial contaminant concentration and adsorbent dose for Aldrin, Dieldrin, Endrin pesticides plus Heptachlor and its epoxide. Polystyrene (PES) nanofibers decorated with titania nanoparticles showed limited adsorptive performance for these pesticides, as well as against Chlordane, showing a capacity of 0.02 mg/g (Table 11, N1) [92]. Adsorbent surface properties were not reported, yet the poor affinity of hydrophobic polyester for hydrophobic organochlorines could have impacted the performance, with adsorption reaction localized on the titania surface. In the case of Dieldrin removal, bromopropyl functionalised silica nanofibers revealed an adsorption capacity of 0.19 mg/g for contamination and adsorbent dose of 20 µg/L and 0.1 g/L, respectively (Table 12, N1). In comparison, untreated silica nanofibers achieved a capacity of 0.014 mg/g (Table 12, N2). The addition of a dense bromopropyl layer on silica nanofibers produced a hydrophobic adsorbent surface, with chemical affinity with pesticide chlorine groups.
Concerning Aldrin removal, vegetal activated carbon showed a relatively high performance with a capacity of 0.38 mg/g for a contamination of 2 mg/L, yet for a relatively large dose of 5 g/L (Table 13, N2). Vegetal activated carbon performed a similar role in the case of Dieldrin adsorption, in the same conditions (Table 12, N8). Against Heptachlor epoxide, the vegetal activated carbon was reported to show a capacity of 0.42 mg/g for a contamination of 2.2 mg/L for a similar dose of 5 g/L (Table 14, N7). Finally, in the case of Endrin, powdered activated carbon was reported to show a capacity of 0.085 mg/g for a contamination of 0.8 µg/L, and a relatively large dose of 0.2 g/L (Table  15, N6). These results show that although adsorbent specific surface area and pore channels could be adapted to improve the adsorptive performance, activated carbons yield removal efficiencies over 95%. However, compared to nanomaterial adsorbents, a large adsorbent dose is required.
Acid-treated date and olive stones showed a significantly improved performance over nontreated biomass adsorbents. For instance, acid-treated date stones showed a capacity of 0.45 mg/g in the removal of Aldrin, while untreated date stones yielded 0.05 µg/g for conditions of contamination and adsorbent dose of respectively 500 µg/L, 1 g/L and 0.5 µ g/L and 10 g/L (Table 13, N9,N12). Date stones were thermally treated with hydrochloric acid, thus creating a hydrophobic layer on the adsorbent surface, and facilitating the anchorage of organochlorines [93]. In the adsorption of Dieldrin and Endrin, acid-treated date stones showed a capacity of 0.44 mg/g and 0.40 mg/g, in the same conditions as Aldrin capture (  Figure 10 shows the relationship between adsorption capacity and test conditions of initial contaminant concentration and adsorbent dose for Aldrin, Dieldrin, Endrin pesticides plus Heptachlor and its epoxide. Polystyrene (PES) nanofibers decorated with titania nanoparticles showed limited adsorptive performance for these pesticides, as well as against Chlordane, showing a capacity of 0.02 mg/g (Table 11, N1) [92]. Adsorbent surface properties were not reported, yet the poor affinity of hydrophobic polyester for hydrophobic organochlorines could have impacted the performance, with adsorption reaction localized on the titania surface. In the case of Dieldrin removal, bromopropyl functionalised silica nanofibers revealed an adsorption capacity of 0.19 mg/g for contamination and adsorbent dose of 20 µg/L and 0.1 g/L, respectively (Table 12, N1). In comparison, untreated silica nanofibers achieved a capacity of 0.014 mg/g (Table 12, N2). The addition of a dense bromopropyl layer on silica nanofibers produced a hydrophobic adsorbent surface, with chemical affinity with pesticide chlorine groups.
Concerning Aldrin removal, vegetal activated carbon showed a relatively high performance with a capacity of 0.38 mg/g for a contamination of 2 mg/L, yet for a relatively large dose of 5 g/L (Table 13, N2). Vegetal activated carbon performed a similar role in the case of Dieldrin adsorption, in the same conditions (Table 12, N8). Against Heptachlor epoxide, the vegetal activated carbon was reported to show a capacity of 0.42 mg/g for a contamination of 2.2 mg/L for a similar dose of 5 g/L (Table 14, N7). Finally, in the case of Endrin, powdered activated carbon was reported to show a capacity of 0.085 mg/g for a contamination of 0.8 µg/L, and a relatively large dose of 0.2 g/L (Table  15, N6). These results show that although adsorbent specific surface area and pore channels could be adapted to improve the adsorptive performance, activated carbons yield removal efficiencies over 95%. However, compared to nanomaterial adsorbents, a large adsorbent dose is required.
Acid-treated date and olive stones showed a significantly improved performance over nontreated biomass adsorbents. For instance, acid-treated date stones showed a capacity of 0.45 mg/g in the removal of Aldrin, while untreated date stones yielded 0.05 µg/g for conditions of contamination and adsorbent dose of respectively 500 µg/L, 1 g/L and 0.5 µ g/L and 10 g/L (Table 13, N9,N12). Date stones were thermally treated with hydrochloric acid, thus creating a hydrophobic layer on the adsorbent surface, and facilitating the anchorage of organochlorines [93]. In the adsorption of Dieldrin and Endrin, acid-treated date stones showed a capacity of 0.44 mg/g and 0.40 mg/g, in the same conditions as Aldrin capture (  Figure 10 shows the relationship between adsorption capacity and test conditions of initial contaminant concentration and adsorbent dose for Aldrin, Dieldrin, Endrin pesticides plus Heptachlor and its epoxide. Polystyrene (PES) nanofibers decorated with titania nanoparticles showed limited adsorptive performance for these pesticides, as well as against Chlordane, showing a capacity of 0.02 mg/g (Table 11, N1) [92]. Adsorbent surface properties were not reported, yet the poor affinity of hydrophobic polyester for hydrophobic organochlorines could have impacted the performance, with adsorption reaction localized on the titania surface. In the case of Dieldrin removal, bromopropyl functionalised silica nanofibers revealed an adsorption capacity of 0.19 mg/g for contamination and adsorbent dose of 20 µg/L and 0.1 g/L, respectively (Table 12, N1). In comparison, untreated silica nanofibers achieved a capacity of 0.014 mg/g (Table 12, N2). The addition of a dense bromopropyl layer on silica nanofibers produced a hydrophobic adsorbent surface, with chemical affinity with pesticide chlorine groups.
Concerning Aldrin removal, vegetal activated carbon showed a relatively high performance with a capacity of 0.38 mg/g for a contamination of 2 mg/L, yet for a relatively large dose of 5 g/L (Table 13, N2). Vegetal activated carbon performed a similar role in the case of Dieldrin adsorption, in the same conditions (Table 12, N8). Against Heptachlor epoxide, the vegetal activated carbon was reported to show a capacity of 0.42 mg/g for a contamination of 2.2 mg/L for a similar dose of 5 g/L (Table 14, N7). Finally, in the case of Endrin, powdered activated carbon was reported to show a capacity of 0.085 mg/g for a contamination of 0.8 µg/L, and a relatively large dose of 0.2 g/L (Table  15, N6). These results show that although adsorbent specific surface area and pore channels could be adapted to improve the adsorptive performance, activated carbons yield removal efficiencies over 95%. However, compared to nanomaterial adsorbents, a large adsorbent dose is required.
Acid-treated date and olive stones showed a significantly improved performance over nontreated biomass adsorbents. For instance, acid-treated date stones showed a capacity of 0.45 mg/g in the removal of Aldrin, while untreated date stones yielded 0.05 µg/g for conditions of contamination and adsorbent dose of respectively 500 µg/L, 1 g/L and 0.5 µ g/L and 10 g/L (Table 13, N9,N12). Date stones were thermally treated with hydrochloric acid, thus creating a hydrophobic layer on the adsorbent surface, and facilitating the anchorage of organochlorines [93]. In the adsorption of Dieldrin and Endrin, acid-treated date stones showed a capacity of 0.44 mg/g and 0.40 mg/g, in the same conditions as Aldrin capture (Table 12, N18 and Table 15, N10). The lower performance in the adsorption of Endrin compared to Dieldrin can be explained by the spatial organization of the Endrin molecule, with several orientations. The similar adsorption capacities of Aldrin and Dieldrin indicate that the pesticide oxide groups of Dieldrin and Endrin do not significantly influence the adsorption 0 0 4.1.2. Nanofibers for Pesticide Adsorption Figure 10 shows the relationship between adsorption capacity and test conditions of initial contaminant concentration and adsorbent dose for Aldrin, Dieldrin, Endrin pesticides plus Heptachlor and its epoxide. Polystyrene (PES) nanofibers decorated with titania nanoparticles showed limited adsorptive performance for these pesticides, as well as against Chlordane, showing a capacity of 0.02 mg/g (Table 11, N1) [92]. Adsorbent surface properties were not reported, yet the poor affinity of hydrophobic polyester for hydrophobic organochlorines could have impacted the performance, with adsorption reaction localized on the titania surface. In the case of Dieldrin removal, bromopropyl functionalised silica nanofibers revealed an adsorption capacity of 0.19 mg/g for contamination and adsorbent dose of 20 µg/L and 0.1 g/L, respectively (Table 12, N1). In comparison, untreated silica nanofibers achieved a capacity of 0.014 mg/g (Table 12, N2). The addition of a dense bromopropyl layer on silica nanofibers produced a hydrophobic adsorbent surface, with chemical affinity with pesticide chlorine groups.
Concerning Aldrin removal, vegetal activated carbon showed a relatively high performance with a capacity of 0.38 mg/g for a contamination of 2 mg/L, yet for a relatively large dose of 5 g/L (Table 13, N2). Vegetal activated carbon performed a similar role in the case of Dieldrin adsorption, in the same conditions (Table 12, N8). Against Heptachlor epoxide, the vegetal activated carbon was reported to show a capacity of 0.42 mg/g for a contamination of 2.2 mg/L for a similar dose of 5 g/L (Table 14,  N7). Finally, in the case of Endrin, powdered activated carbon was reported to show a capacity of 0.085 mg/g for a contamination of 0.8 µg/L, and a relatively large dose of 0.2 g/L (Table 15, N6). These results show that although adsorbent specific surface area and pore channels could be adapted to improve the adsorptive performance, activated carbons yield removal efficiencies over 95%. However, compared to nanomaterial adsorbents, a large adsorbent dose is required.
Acid-treated date and olive stones showed a significantly improved performance over non-treated biomass adsorbents. For instance, acid-treated date stones showed a capacity of 0.45 mg/g in the removal of Aldrin, while untreated date stones yielded 0.05 µg/g for conditions of contamination and adsorbent dose of respectively 500 µg/L, 1 g/L and 0.5 µ g/L and 10 g/L (Table 13, N9, N12). Date stones were thermally treated with hydrochloric acid, thus creating a hydrophobic layer on the adsorbent surface, and facilitating the anchorage of organochlorines [93]. In the adsorption of Dieldrin and Endrin, acid-treated date stones showed a capacity of 0.44 mg/g and 0.40 mg/g, in the same conditions as Aldrin capture (Table 12, N18 and Table 15, N10). The lower performance in the adsorption of Endrin compared to Dieldrin can be explained by the spatial organization of the Endrin molecule, with several orientations. The similar adsorption capacities of Aldrin and Dieldrin indicate that the pesticide oxide groups of Dieldrin and Endrin do not significantly influence the adsorption reaction, meaning electrostatic interactions played a minor role in the capture of these pesticides. Finally, cellulose nanofibers modified with manganese oxide have been reported to yield 96.5% removal against Toxaphene 8, highlighting the potential of metal oxide-based nanofiber adsorbents ( reaction, meaning electrostatic interactions played a minor role in the capture of these pesticides. Finally, cellulose nanofibers modified with manganese oxide have been reported to yield 96.5 % removal against Toxaphene 8, highlighting the potential of metal oxide-based nanofiber adsorbents (Table 16, N1).        Table 17 presents the different molecular structures for the selected polychlorinated biphenyls PCBs in this study, which are namely 2 to 6 -PCB, PCB 28 and PCB 77. PCBs are polycyclic aromatic organochlorines of chemical formula C 12 H 10−x Cl x . The PCB structure is composed of two benzene rings, thus sharing the aromatic structure of phenols molecules, and decorated by a number of chlorine groups. PCBs 2 and 3, also called 2-and 3-chlorobiphenyls, possess one chlorine group attached on position 2 and 3, respectively, on one of the benzene rings. Sharing a molecular weight of 188.7 g/mol, water solubility was determined at 1.45 mg/L (25 • C), corresponding to a Log K ow of 4.54. PCBs 4-6 possess two chlorine groups, for a weight of 223.1 g/mol. Their chemical names are 2,2 -2,3-and 2,3 -dichlorobiphenyls, respectively, corresponding to the chlorine groups' respective positions on one or both benzene rings. PCB 28, called 2,4,4 -trichlorobiphenyl, possesses three chlorine groups for a weight of 257.5 g/mol. PCB 77, called 3,3 ,4,4 -tetrachlorobiphenyl, possesses four chlorine group for a weight of 292.0 g/mol. Compared to monochloro-biphenyls, PCB 77 water solubility decreased to 0.18 mg/L (25 • C, log K ow 6.72). All molecules are apolar.  Table 17 presents the different molecular structures for the selected polychlorinated biphenyls PCBs in this study, which are namely 2 to 6 -PCB, PCB 28 and PCB 77. PCBs are polycyclic aromatic organochlorines of chemical formula C12H10−xClx. The PCB structure is composed of two benzene rings, thus sharing the aromatic structure of phenols molecules, and decorated by a number of chlorine groups. PCBs 2 and 3, also called 2-and 3-chlorobiphenyls, possess one chlorine group attached on position 2 and 3, respectively, on one of the benzene rings. Sharing a molecular weight of 188.7 g/mol, water solubility was determined at 1.45 mg/L (25 °C), corresponding to a Log Kow of 4.54. PCBs 4-6 possess two chlorine groups, for a weight of 223.1 g/mol. Their chemical names are 2,2′-2,3-and 2,3′-dichlorobiphenyls, respectively, corresponding to the chlorine groups' respective positions on one or both benzene rings. PCB 28, called 2,4,4′-trichlorobiphenyl, possesses three chlorine groups for a weight of 257.5 g/mol. PCB 77, called 3,3′,4,4′-tetrachlorobiphenyl, possesses four chlorine group for a weight of 292.0 g/mol. Compared to monochloro-biphenyls, PCB 77 water solubility decreased to 0.18 mg/L (25 °C, log Kow 6.72). All molecules are apolar.  Table 18 presents selected characteristics of the Phenantrene molecule, non-POP-classified, from the family of polycyclic aromatic hydrocarbon compounds. Of a molecular weight in the same order as monochlorinated biphenyls, its water solubility was determined to be 1.15 mg/L (25 °C). An apolar molecule, Phenantrene is composed of three fused benzene rings.  Figure 11 shows the adsorption capacity related to the adsorbent dose and initial contamination for adsorbents of PCBs and Phenantrene. Oriented porous magnetite nanofibers were tested in the adsorption of PCB 77 and showed a capacity of 0.18 mg/g in hexane for a contamination of 30 µg/L and a dose of 0.15 g/L (Table 19, N8). However, activated carbon showed the largest adsorption capacity of 649 mg/g for a contamination of 4 mg/L and a relatively small dose of 0.015 g/L (Table 19, N1). To study the influence of the adsorbate aromatic structure on the adsorption performance, Phenantrene, with three aromatic cycles, has been considered here. Cellulose acetate unmodified nanofibers showed limited performance in the capture of Phenantrene, which could be due to the hydrophilic nature of cellulose acetate; they are thus poorly adapted for the adsorption of apolar compounds [112]. Cyclodextrin-grafted cellulose acetate nanofibers were tested for the capture of  Table 17 presents the different molecular structures for the selected polychlorinated biphenyls PCBs in this study, which are namely 2 to 6 -PCB, PCB 28 and PCB 77. PCBs are polycyclic aromatic organochlorines of chemical formula C12H10−xClx. The PCB structure is composed of two benzene rings, thus sharing the aromatic structure of phenols molecules, and decorated by a number of chlorine groups. PCBs 2 and 3, also called 2-and 3-chlorobiphenyls, possess one chlorine group attached on position 2 and 3, respectively, on one of the benzene rings. Sharing a molecular weight of 188.7 g/mol, water solubility was determined at 1.45 mg/L (25 °C), corresponding to a Log Kow of 4.54. PCBs 4-6 possess two chlorine groups, for a weight of 223.1 g/mol. Their chemical names are 2,2′-2,3-and 2,3′-dichlorobiphenyls, respectively, corresponding to the chlorine groups' respective positions on one or both benzene rings. PCB 28, called 2,4,4′-trichlorobiphenyl, possesses three chlorine groups for a weight of 257.5 g/mol. PCB 77, called 3,3′,4,4′-tetrachlorobiphenyl, possesses four chlorine group for a weight of 292.0 g/mol. Compared to monochloro-biphenyls, PCB 77 water solubility decreased to 0.18 mg/L (25 °C, log Kow 6.72). All molecules are apolar.  Table 18 presents selected characteristics of the Phenantrene molecule, non-POP-classified, from the family of polycyclic aromatic hydrocarbon compounds. Of a molecular weight in the same order as monochlorinated biphenyls, its water solubility was determined to be 1.15 mg/L (25 °C). An apolar molecule, Phenantrene is composed of three fused benzene rings.  Figure 11 shows the adsorption capacity related to the adsorbent dose and initial contamination for adsorbents of PCBs and Phenantrene. Oriented porous magnetite nanofibers were tested in the adsorption of PCB 77 and showed a capacity of 0.18 mg/g in hexane for a contamination of 30 µg/L and a dose of 0.15 g/L (Table 19, N8). However, activated carbon showed the largest adsorption capacity of 649 mg/g for a contamination of 4 mg/L and a relatively small dose of 0.015 g/L (Table 19, N1). To study the influence of the adsorbate aromatic structure on the adsorption performance, Phenantrene, with three aromatic cycles, has been considered here. Cellulose acetate unmodified nanofibers showed limited performance in the capture of Phenantrene, which could be due to the hydrophilic nature of cellulose acetate; they are thus poorly adapted for the adsorption of apolar compounds [112]. Cyclodextrin-grafted cellulose acetate nanofibers were tested for the capture of  Table 17 presents the different molecular structures for the selected polychlorinated biphenyls PCBs in this study, which are namely 2 to 6 -PCB, PCB 28 and PCB 77. PCBs are polycyclic aromatic organochlorines of chemical formula C12H10−xClx. The PCB structure is composed of two benzene rings, thus sharing the aromatic structure of phenols molecules, and decorated by a number of chlorine groups. PCBs 2 and 3, also called 2-and 3-chlorobiphenyls, possess one chlorine group attached on position 2 and 3, respectively, on one of the benzene rings. Sharing a molecular weight of 188.7 g/mol, water solubility was determined at 1.45 mg/L (25 °C), corresponding to a Log Kow of 4.54. PCBs 4-6 possess two chlorine groups, for a weight of 223.1 g/mol. Their chemical names are 2,2′-2,3-and 2,3′-dichlorobiphenyls, respectively, corresponding to the chlorine groups' respective positions on one or both benzene rings. PCB 28, called 2,4,4′-trichlorobiphenyl, possesses three chlorine groups for a weight of 257.5 g/mol. PCB 77, called 3,3′,4,4′-tetrachlorobiphenyl, possesses four chlorine group for a weight of 292.0 g/mol. Compared to monochloro-biphenyls, PCB 77 water solubility decreased to 0.18 mg/L (25 °C, log Kow 6.72). All molecules are apolar.  Table 18 presents selected characteristics of the Phenantrene molecule, non-POP-classified, from the family of polycyclic aromatic hydrocarbon compounds. Of a molecular weight in the same order as monochlorinated biphenyls, its water solubility was determined to be 1.15 mg/L (25 °C). An apolar molecule, Phenantrene is composed of three fused benzene rings.  Figure 11 shows the adsorption capacity related to the adsorbent dose and initial contamination for adsorbents of PCBs and Phenantrene. Oriented porous magnetite nanofibers were tested in the adsorption of PCB 77 and showed a capacity of 0.18 mg/g in hexane for a contamination of 30 µg/L and a dose of 0.15 g/L (Table 19, N8). However, activated carbon showed the largest adsorption capacity of 649 mg/g for a contamination of 4 mg/L and a relatively small dose of 0.015 g/L (Table 19, N1). To study the influence of the adsorbate aromatic structure on the adsorption performance, Phenantrene, with three aromatic cycles, has been considered here. Cellulose acetate unmodified nanofibers showed limited performance in the capture of Phenantrene, which could be due to the hydrophilic nature of cellulose acetate; they are thus poorly adapted for the adsorption of apolar compounds [112]. Cyclodextrin-grafted cellulose acetate nanofibers were tested for the capture of  Table 17 presents the different molecular structures for the selected polychlorinated biphenyls PCBs in this study, which are namely 2 to 6 -PCB, PCB 28 and PCB 77. PCBs are polycyclic aromatic organochlorines of chemical formula C12H10−xClx. The PCB structure is composed of two benzene rings, thus sharing the aromatic structure of phenols molecules, and decorated by a number of chlorine groups. PCBs 2 and 3, also called 2-and 3-chlorobiphenyls, possess one chlorine group attached on position 2 and 3, respectively, on one of the benzene rings. Sharing a molecular weight of 188.7 g/mol, water solubility was determined at 1.45 mg/L (25 °C), corresponding to a Log Kow of 4.54. PCBs 4-6 possess two chlorine groups, for a weight of 223.1 g/mol. Their chemical names are 2,2′-2,3-and 2,3′-dichlorobiphenyls, respectively, corresponding to the chlorine groups' respective positions on one or both benzene rings. PCB 28, called 2,4,4′-trichlorobiphenyl, possesses three chlorine groups for a weight of 257.5 g/mol. PCB 77, called 3,3′,4,4′-tetrachlorobiphenyl, possesses four chlorine group for a weight of 292.0 g/mol. Compared to monochloro-biphenyls, PCB 77 water solubility decreased to 0.18 mg/L (25 °C, log Kow 6.72). All molecules are apolar.  Table 18 presents selected characteristics of the Phenantrene molecule, non-POP-classified, from the family of polycyclic aromatic hydrocarbon compounds. Of a molecular weight in the same order as monochlorinated biphenyls, its water solubility was determined to be 1.15 mg/L (25 °C). An apolar molecule, Phenantrene is composed of three fused benzene rings.  Figure 11 shows the adsorption capacity related to the adsorbent dose and initial contamination for adsorbents of PCBs and Phenantrene. Oriented porous magnetite nanofibers were tested in the adsorption of PCB 77 and showed a capacity of 0.18 mg/g in hexane for a contamination of 30 µg/L and a dose of 0.15 g/L (Table 19, N8). However, activated carbon showed the largest adsorption capacity of 649 mg/g for a contamination of 4 mg/L and a relatively small dose of 0.015 g/L (Table 19, N1). To study the influence of the adsorbate aromatic structure on the adsorption performance, Phenantrene, with three aromatic cycles, has been considered here. Cellulose acetate unmodified nanofibers showed limited performance in the capture of Phenantrene, which could be due to the hydrophilic nature of cellulose acetate; they are thus poorly adapted for the adsorption of apolar compounds [112]. Cyclodextrin-grafted cellulose acetate nanofibers were tested for the capture of  Table 17 presents the different molecular structures for the selected polychlorinated biphenyls PCBs in this study, which are namely 2 to 6 -PCB, PCB 28 and PCB 77. PCBs are polycyclic aromatic organochlorines of chemical formula C12H10−xClx. The PCB structure is composed of two benzene rings, thus sharing the aromatic structure of phenols molecules, and decorated by a number of chlorine groups. PCBs 2 and 3, also called 2-and 3-chlorobiphenyls, possess one chlorine group attached on position 2 and 3, respectively, on one of the benzene rings. Sharing a molecular weight of 188.7 g/mol, water solubility was determined at 1.45 mg/L (25 °C), corresponding to a Log Kow of 4.54. PCBs 4-6 possess two chlorine groups, for a weight of 223.1 g/mol. Their chemical names are 2,2′-2,3-and 2,3′-dichlorobiphenyls, respectively, corresponding to the chlorine groups' respective positions on one or both benzene rings. PCB 28, called 2,4,4′-trichlorobiphenyl, possesses three chlorine groups for a weight of 257.5 g/mol. PCB 77, called 3,3′,4,4′-tetrachlorobiphenyl, possesses four chlorine group for a weight of 292.0 g/mol. Compared to monochloro-biphenyls, PCB 77 water solubility decreased to 0.18 mg/L (25 °C, log Kow 6.72). All molecules are apolar.  Table 18 presents selected characteristics of the Phenantrene molecule, non-POP-classified, from the family of polycyclic aromatic hydrocarbon compounds. Of a molecular weight in the same order as monochlorinated biphenyls, its water solubility was determined to be 1.15 mg/L (25 °C). An apolar molecule, Phenantrene is composed of three fused benzene rings.  Figure 11 shows the adsorption capacity related to the adsorbent dose and initial contamination for adsorbents of PCBs and Phenantrene. Oriented porous magnetite nanofibers were tested in the adsorption of PCB 77 and showed a capacity of 0.18 mg/g in hexane for a contamination of 30 µg/L and a dose of 0.15 g/L (Table 19, N8). However, activated carbon showed the largest adsorption capacity of 649 mg/g for a contamination of 4 mg/L and a relatively small dose of 0.015 g/L (Table 19, N1). To study the influence of the adsorbate aromatic structure on the adsorption performance, Phenantrene, with three aromatic cycles, has been considered here. Cellulose acetate unmodified nanofibers showed limited performance in the capture of Phenantrene, which could be due to the hydrophilic nature of cellulose acetate; they are thus poorly adapted for the adsorption of apolar compounds [112]. Cyclodextrin-grafted cellulose acetate nanofibers were tested for the capture of  Table 17 presents the different molecular structures for the selected polychlorinated biphenyls PCBs in this study, which are namely 2 to 6 -PCB, PCB 28 and PCB 77. PCBs are polycyclic aromatic organochlorines of chemical formula C12H10−xClx. The PCB structure is composed of two benzene rings, thus sharing the aromatic structure of phenols molecules, and decorated by a number of chlorine groups. PCBs 2 and 3, also called 2-and 3-chlorobiphenyls, possess one chlorine group attached on position 2 and 3, respectively, on one of the benzene rings. Sharing a molecular weight of 188.7 g/mol, water solubility was determined at 1.45 mg/L (25 °C), corresponding to a Log Kow of 4.54. PCBs 4-6 possess two chlorine groups, for a weight of 223.1 g/mol. Their chemical names are 2,2′-2,3-and 2,3′-dichlorobiphenyls, respectively, corresponding to the chlorine groups' respective positions on one or both benzene rings. PCB 28, called 2,4,4′-trichlorobiphenyl, possesses three chlorine groups for a weight of 257.5 g/mol. PCB 77, called 3,3′,4,4′-tetrachlorobiphenyl, possesses four chlorine group for a weight of 292.0 g/mol. Compared to monochloro-biphenyls, PCB 77 water solubility decreased to 0.18 mg/L (25 °C, log Kow 6.72). All molecules are apolar.  Table 18 presents selected characteristics of the Phenantrene molecule, non-POP-classified, from the family of polycyclic aromatic hydrocarbon compounds. Of a molecular weight in the same order as monochlorinated biphenyls, its water solubility was determined to be 1.15 mg/L (25 °C). An apolar molecule, Phenantrene is composed of three fused benzene rings.  Figure 11 shows the adsorption capacity related to the adsorbent dose and initial contamination for adsorbents of PCBs and Phenantrene. Oriented porous magnetite nanofibers were tested in the adsorption of PCB 77 and showed a capacity of 0.18 mg/g in hexane for a contamination of 30 µg/L and a dose of 0.15 g/L (Table 19, N8). However, activated carbon showed the largest adsorption capacity of 649 mg/g for a contamination of 4 mg/L and a relatively small dose of 0.015 g/L (Table 19, N1). To study the influence of the adsorbate aromatic structure on the adsorption performance, Phenantrene, with three aromatic cycles, has been considered here. Cellulose acetate unmodified nanofibers showed limited performance in the capture of Phenantrene, which could be due to the hydrophilic nature of cellulose acetate; they are thus poorly adapted for the adsorption of apolar compounds [112]. Cyclodextrin-grafted cellulose acetate nanofibers were tested for the capture of  Table 17 presents the different molecular structures for the selected polychlorinated biphenyls PCBs in this study, which are namely 2 to 6 -PCB, PCB 28 and PCB 77. PCBs are polycyclic aromatic organochlorines of chemical formula C12H10−xClx. The PCB structure is composed of two benzene rings, thus sharing the aromatic structure of phenols molecules, and decorated by a number of chlorine groups. PCBs 2 and 3, also called 2-and 3-chlorobiphenyls, possess one chlorine group attached on position 2 and 3, respectively, on one of the benzene rings. Sharing a molecular weight of 188.7 g/mol, water solubility was determined at 1.45 mg/L (25 °C), corresponding to a Log Kow of 4.54. PCBs 4-6 possess two chlorine groups, for a weight of 223.1 g/mol. Their chemical names are 2,2′-2,3-and 2,3′-dichlorobiphenyls, respectively, corresponding to the chlorine groups' respective positions on one or both benzene rings. PCB 28, called 2,4,4′-trichlorobiphenyl, possesses three chlorine groups for a weight of 257.5 g/mol. PCB 77, called 3,3′,4,4′-tetrachlorobiphenyl, possesses four chlorine group for a weight of 292.0 g/mol. Compared to monochloro-biphenyls, PCB 77 water solubility decreased to 0.18 mg/L (25 °C, log Kow 6.72). All molecules are apolar.  Table 18 presents selected characteristics of the Phenantrene molecule, non-POP-classified, from the family of polycyclic aromatic hydrocarbon compounds. Of a molecular weight in the same order as monochlorinated biphenyls, its water solubility was determined to be 1.15 mg/L (25 °C). An apolar molecule, Phenantrene is composed of three fused benzene rings.  Figure 11 shows the adsorption capacity related to the adsorbent dose and initial contamination for adsorbents of PCBs and Phenantrene. Oriented porous magnetite nanofibers were tested in the adsorption of PCB 77 and showed a capacity of 0.18 mg/g in hexane for a contamination of 30 µg/L and a dose of 0.15 g/L (Table 19, N8). However, activated carbon showed the largest adsorption capacity of 649 mg/g for a contamination of 4 mg/L and a relatively small dose of 0.015 g/L (Table 19, N1). To study the influence of the adsorbate aromatic structure on the adsorption performance, Phenantrene, with three aromatic cycles, has been considered here. Cellulose acetate unmodified nanofibers showed limited performance in the capture of Phenantrene, which could be due to the hydrophilic nature of cellulose acetate; they are thus poorly adapted for the adsorption of apolar compounds [112]. Cyclodextrin-grafted cellulose acetate nanofibers were tested for the capture of Table 18 presents selected characteristics of the Phenantrene molecule, non-POP-classified, from the family of polycyclic aromatic hydrocarbon compounds. Of a molecular weight in the same order as monochlorinated biphenyls, its water solubility was determined to be 1.15 mg/L (25 • C). An apolar molecule, Phenantrene is composed of three fused benzene rings.  Table 17 presents the different molecular structures for the selected polychlorinated biphenyls PCBs in this study, which are namely 2 to 6 -PCB, PCB 28 and PCB 77. PCBs are polycyclic aromatic organochlorines of chemical formula C12H10−xClx. The PCB structure is composed of two benzene rings, thus sharing the aromatic structure of phenols molecules, and decorated by a number of chlorine groups. PCBs 2 and 3, also called 2-and 3-chlorobiphenyls, possess one chlorine group attached on position 2 and 3, respectively, on one of the benzene rings. Sharing a molecular weight of 188.7 g/mol, water solubility was determined at 1.45 mg/L (25 °C), corresponding to a Log Kow of 4.54. PCBs 4-6 possess two chlorine groups, for a weight of 223.1 g/mol. Their chemical names are 2,2′-2,3-and 2,3′-dichlorobiphenyls, respectively, corresponding to the chlorine groups' respective positions on one or both benzene rings. PCB 28, called 2,4,4′-trichlorobiphenyl, possesses three chlorine groups for a weight of 257.5 g/mol. PCB 77, called 3,3′,4,4′-tetrachlorobiphenyl, possesses four chlorine group for a weight of 292.0 g/mol. Compared to monochloro-biphenyls, PCB 77 water solubility decreased to 0.18 mg/L (25 °C, log Kow 6.72). All molecules are apolar.  Table 18 presents selected characteristics of the Phenantrene molecule, non-POP-classified, from the family of polycyclic aromatic hydrocarbon compounds. Of a molecular weight in the same order as monochlorinated biphenyls, its water solubility was determined to be 1.15 mg/L (25 °C). An apolar molecule, Phenantrene is composed of three fused benzene rings.  Figure 11 shows the adsorption capacity related to the adsorbent dose and initial contamination for adsorbents of PCBs and Phenantrene. Oriented porous magnetite nanofibers were tested in the adsorption of PCB 77 and showed a capacity of 0.18 mg/g in hexane for a contamination of 30 µg/L and a dose of 0.15 g/L (Table 19, N8). However, activated carbon showed the largest adsorption capacity of 649 mg/g for a contamination of 4 mg/L and a relatively small dose of 0.015 g/L (Table 19, N1). To study the influence of the adsorbate aromatic structure on the adsorption performance, Phenantrene, with three aromatic cycles, has been considered here. Cellulose acetate unmodified nanofibers showed limited performance in the capture of Phenantrene, which could be due to the hydrophilic nature of cellulose acetate; they are thus poorly adapted for the adsorption of apolar compounds [112]. Cyclodextrin-grafted cellulose acetate nanofibers were tested for the capture of 4.2.2. Nanofibers for Pesticide Adsorption Figure 11 shows the adsorption capacity related to the adsorbent dose and initial contamination for adsorbents of PCBs and Phenantrene. Oriented porous magnetite nanofibers were tested in the adsorption of PCB 77 and showed a capacity of 0.18 mg/g in hexane for a contamination of 30 µg/L and a dose of 0.15 g/L (Table 19, N8). However, activated carbon showed the largest adsorption capacity of 649 mg/g for a contamination of 4 mg/L and a relatively small dose of 0.015 g/L (Table 19, N1). To study the influence of the adsorbate aromatic structure on the adsorption performance, Phenantrene, with three aromatic cycles, has been considered here. Cellulose acetate unmodified nanofibers showed limited performance in the capture of Phenantrene, which could be due to the hydrophilic nature of cellulose acetate; they are thus poorly adapted for the adsorption of apolar compounds [112]. Cyclodextrin-grafted cellulose acetate nanofibers were tested for the capture of Phenantrene via molecular filtration, by inclusion-complexation of the hydrophobic contaminants in the apolar cyclodextrin cavity [112]. Hydrophobic interactions led to a contaminant removal of 97% (Table 20, N2). Cross-linking of cellulose acetate nanofibers with benzene rings by polybenzoxazine provided a hydrophobic adsorptive structure [113]. The adsorbent benzene rings further enabled π-π interactions with Phenantrene aromatic rings, thus leading to 98.5% contaminant removal ( Phenantrene via molecular filtration, by inclusion-complexation of the hydrophobic contaminants in the apolar cyclodextrin cavity [112]. Hydrophobic interactions led to a contaminant removal of 97% (Table 20, N2). Cross-linking of cellulose acetate nanofibers with benzene rings by polybenzoxazine provided a hydrophobic adsorptive structure [113]. The adsorbent benzene rings further enabled ππ interactions with Phenantrene aromatic rings, thus leading to 98.5% contaminant removal (Table  20, N1). Activated carbons and carbon nanotubes however, demonstrated an adsorption capacity of 85 mg/g for a dose of 4 mg/L (initial concentration was not reported) ( Table 20, N7). It can be suggested that π-π electron interaction between the unsaturated Phenantrene aromatics and the carbon adsorbent surface is the predominant reaction mechanism; in that sense, a higher carbon content and full crystalline structure would lead to an increased specific surface area and, for a large pore volume of adapted size, Phenantrene adsorption. Among the reported carbon-based materials, granular activated carbon possessed a BET of 706 m 2 /g and a pore volume of 0.711 cm 3 /g, yielding the highest performance. On the other hand, carbon nanotubes with a higher pore volume of 0.818 cm 3 /g but a lower BET of 537 m 2 /g showed a capacity of 75 mg/g (Table 20, N4). Furthermore, activated carbon fibers with the highest BET of 949 m 2 /g but the lowest pore volume of 0.381 cm 3 /g yielded a lower capacity of 60 mg/g (Table 20, N6), indicating a synergetic effect of pore volume and surface area (BET). Activated carbons and carbon nanotubes however, demonstrated an adsorption capacity of 85 mg/g for a dose of 4 mg/L (initial concentration was not reported) ( Table 20, N7). It can be suggested that π-π electron interaction between the unsaturated Phenantrene aromatics and the carbon adsorbent surface is the predominant reaction mechanism; in that sense, a higher carbon content and full crystalline structure would lead to an increased specific surface area and, for a large pore volume of adapted size, Phenantrene adsorption. Among the reported carbon-based materials, granular activated carbon possessed a BET of 706 m 2 /g and a pore volume of 0.711 cm 3 /g, yielding the highest performance. On the other hand, carbon nanotubes with a higher pore volume of 0.818 cm 3 /g but a lower BET of 537 m 2 /g showed a capacity of 75 mg/g (Table 20, N4). Furthermore, activated carbon fibers with the highest BET of 949 m 2 /g but the lowest pore volume of 0.381 cm 3 /g yielded a lower capacity of 60 mg/g (Table 20, N6), indicating a synergetic effect of pore volume and surface area (BET).

Adsorbent Regenerability
Two studies reported the adsorbent desorption and recyclability capacity for further use in the industry, and are reviewed below.
Nylon 6-Poly(pyrrole) core-shell nanofibers have been investigated for atrazine desorption and adsorbent regenerability [41]. The desorption procedure consisted of successive washing in methanol and deionized water for 5 min each to elute atrazine, followed by air-drying of nanofibers at room temperature. With this procedure, Nylon-Poly(pyrrole) nanofibers demonstrated the same capacity of 2.77 mg/g over five cycles, with a slight decrease of 0.2 mg/g in adsorption capacity observed at the 6th cycle (Table 4, N3) [41]. With an adsorption plateau reached after 5 h and a desorption procedure estimated around 3 h, such an adsorbent has a life-span of 37 h at its maximal efficiency. Electrostatic interaction and π-π stacking of electrons had been identified as the two adsorption mechanisms at stake in the uptake of Atrazine by Nylon-Poly(pyrrole) nanofibers [41]. Although absorption of Atrazine would be a recyclability limitation, the methanol wash procedure might also imply methanol adsorption.
Manganese oxide-cellulose fiber nanocomposites were investigated for the desorption of Toxaphene (Table 16, N1). Desorption rates were shown to be of 12%, 13% and 25%, respectively, at pH 3, 7, and 11 [111]. The preferential affinity for alkaline pH could suggest the reversibility of the adsorption reaction. Indeed, cellulose and manganese oxide (MnO 2 ) were reported in the literature to have an isoelectric point of 2.5 and below 5.1, respectively [119,120]. At pH 11, both components of the adsorbent material therefore carry a strong negative surface charge, while the decreasing protonation of Toxaphene with increasing pH could suggest that both adsorbent and adsorbate carry a negative surface charge, hence the desorption. However, the desorption protocol followed was not reported [111].

Recommendations and Conclusions
Nanofiber-based materials showed their fast adsorption kinetics and required low adsorbent dose compared to benchmark adsorbents in this study. For instance, polyacrylonitrile nanofibers attained equilibrium after 0.5 h in the adsorption of Atrazine, for an adsorbent dose of 1 g/L (Table 4, N5) [41]. Regarding the adsorption performance, engineered nanofibers yielded over 95% removal efficiency with high adsorption capacities comparable to granular activated carbons and acid-treated biomass. For example, activated carbon nanofibers against Phenol showed a capacity of 256.1 mg/g (Table 7, N20), and cross-linked cellulose acetate-polybenzoxazine nanofibers demonstrated against Phenantrene 98.5% removal (Table 20, N1) [61,113]. Such a nanofiber web presented adequate structural and chemical surface characteristics for the adsorption of organic contaminants.
Essential structural characteristics for a nanofiber adsorbent were found to be surface pores and density of functional sites. Indeed, the size of pores across the fiber surface was shown to have a major impact on the saturation of the adsorbent, with pore channels engineered to take into account the size or cross section of the contaminant showing higher performance. Activated carbon nanofibers for phenol adsorption present a dual pore repartition of 0.4 nm and 0.8 nm, adapted to the phenol molecule size, with a smallest cross section of 0.43 nm [61]. Molecular orientation of the contaminant is also considered to prevent size-exclusion effects when one molecule is already adsorbed along a pore wall, as highlighted by the performance comparison of Dieldrin and stereoisomer Endrin adsorption [73]. Adsorbent contact surface area ruled by the nanofiber average diameter impacted on the contaminant capture, with diameters below 150 nm yielding higher performance in the capture of Atrazine [39]. Higher surface specific areas led in general to higher adsorption capacities, provided there was contaminant diffusion along the surface pore channels. Activated carbon nanofibers demonstrated a high BET of 2921 m 2 /g [61].
Regarding adsorbent surface chemistry, a total hydrophobic surface is preferable to show affinity for adsorption-in the case of surface grafting of hydrophobic groups, adsorption may be impeded along the pore walls [28]. Compared to hydrophobic interactions, electrostatic interactions and foremost electron interactions were shown to be more efficient adsorption mechanisms [41,113]. To this end, adsorbent isoelectric point should be systematically determined to improve the contribution of electrostatic interactions in the contaminant capture. Electron interaction in this study happens via π-π stacking of electrons between adsorbate and adsorbent unsaturated bonds, justifying the high efficiency of carbon-based adsorbents [61]. Other than pore and surface area characteristics, carbon adsorbent crystallinity and carbon content are features of interest that could be investigated in order to improve adsorption by electron interaction. The tuning of the carbon adsorbent surface chemistry to shift its isoelectric point could also be investigated to improve the contaminant adsorption by electrostatic interaction contribution. Other structures such as cellulose acetate crosslinked with polybenzoxazine lead to a highly hydrophobic surface presenting unsaturated bonds, resulting in high performance in the capture of Phenantrene [113]. Material functionalisation to reach a surface chemistry allowing the three adsorption mechanism defined above could be the focus of future research in order to yield complete contaminant removal.
Testing conditions in terms of initial contaminant concentration were found to vary between 0.5 µg/L and 700 mg/L in this study [16,81]. The initial level of contamination should be justified by the occurrence of the selected contaminant in the environment rather than chosen as a function of the available solution characterisation technique, as adsorbent performance may be affected by a significant change in contaminant concentration. Isotherm analysis was found to be in over 80% of the reported isotherm studies fitted to Langmuir and Freundlich models, with in more than 40% of cases a report of the Freundlich constant with a correlation coefficient below 0.90 for performance comparison with other adsorbents. Adsorption isotherms reflecting the adsorption mechanisms at stake should be more carefully studied to avoid inconsistent performance comparisons.
So far, engineered nanofibers showed the highest performance in the one-off removal of pesticides and disinfectants. However, these adsorbents also show versatility in the capture of contaminants. Ways to create selectivity for a target molecule, such as fiber layers of different pore channels and surface chemistry, could be further investigated. Adsorption being a conventional water treatment technique, the adsorbent material fabrication and operational costs are critical to the implementation of any new classes of materials. Industrial applications of nanofiber adsorbent material solutions should be rigorously assessed and benchmarked against commercially available fabrication routes. The adsorption kinetics and material dosage are also to be considered when evaluating adsorbent life cycle and operational costs. Industrial-scale nanofiber fabrication devices by electrospinning or multi-component fiber extrusion were reported to yield mass output varying between 0.02 kg/h and 0.2 kg/h, and up to 6.5 kg/h, respectively, with a production cost of $2-5/kg [121,122]. Other competitive techniques such as force-spinning have been developed to simultaneously increase mass output and fiber quality [123]. Furthermore, to develop a sustainable industry, adsorbent recyclability must be further investigated. As the best-performing materials in the removal of POP may not offer the best recyclability, it is critical to systematically investigate the adsorbent hysteresis of adsorption-desorption. Regeneration of the adsorbent without increasing toxicity concerns linked to adsorbed contaminants must also be investigated.