Synergistic Approaches for Navigating and Mitigating Agricultural Pollutants

Abstract

The alarming increase in the use of chemically driven pesticides for enhanced crop productivity has severely affected soil fertility, ecosystem balance, and consumer health. Inadequate handling protocols and ineffective remediation strategies have led to elevated pesticide concentrations, contributing to human respiratory and metabolic disorders in humans. In the current context, where agricultural activities and pesticide applications are intertwined, strong and sustainable remediation strategies are essential for environmental protection without sacrificing crop productivity. Various bio-inspired methods have been reported, such as phytoremediation, bioremediation, and in situ remediation; however, limited success has been observed with either single or combined approaches. Consequently, biopolymer biomanufacturing, nanoparticle-based bioengineering, and computational biology for improved understanding of mechanisms have been revisited to incorporate updated methodologies that detail the fate and action of harmful chemical pesticides in agriculture. An in silico mechanistic approach has been emphasized to understand the molecular mechanisms involved in agricultural pesticides’ degradation using nanomaterials. A roadmap has been created by integrating cutting-edge machine learning techniques to develop nature-inspired sustainable agricultural practices and contaminant disposal methods. This review represents a pioneering effort to explore the roles of wet-lab chemistry and in silico methods in mitigating the effects of agricultural pesticides, providing a comprehensive strategy for balancing environmental sustainability and agricultural practices.

1. Introduction

Economical and sustainable agricultural practices are the key to meeting the upsurge in food demand and, consequently, global development goals. Such practices are prerequisites to strengthening economic prosperity and feeding a projected 9.7 billion people by 2050 [1]. Moreover, adopting resilient and environmentally sustainable agricultural practices is essential for enhancing the gross domestic product of both developing and developed nations. Regrettably, the overuse of chemical stimulants, such as pesticides (i.e., neonicotinoids), in recent decades has created significant threats to agricultural practices and the planet’s ecosystem due to bio-magnification [2,3].

Pesticides encompass a range of chemical compounds, including insecticides, fungicides, herbicides, rodenticides, molluscicides, nematicides, and plant growth regulators [4]. Numerous pesticides have been introduced in agricultural practices and banned lately due to their detrimental environmental impacts. For instance, organochlorines, carbamates, and pyrethroids were introduced in the mid- and late 1900s, followed by a ban on their application, largely in Western countries [5]. Due to the adverse impacts of chemical-induced agricultural practices, several sustainable techniques have been developed simultaneously to eliminate or mitigate their effects on different crop types. A few noteworthy practices applied for agricultural pollutant remediation in both soil and water environments include (i) adopting nutrient management techniques, (ii) employing conservation drainage practices, (iii) ensuring year-round ground cover, (iv) establishing planting field buffers, (v) implementing conservation tillage, (vi) managing livestock access to streams, and (vii) engaging in watershed efforts, etc. [6]. Unfortunately, these documented approaches, whether used alone or in combination, have proven insufficient in combating soil contamination [6]. Moreover, overwhelming evidence has corroborated the potential threat posed by the banned compounds to humans and other life forms [7].

Technological and manufacturing advancements in the last couple of decades, resulting in biodegradable compounds [8], kill-switch mechanisms in genetically modified organisms, nanomaterials [9], etc., have provided a new perspective for agricultural research. These transformative practices assert minimal or no environmental and health implications for end users and have improved agribusinesses. Nanomaterials, in particular, exhibit a strong affinity for organic compounds and heavy metals, making them a transformative solution to archetypal challenges, e.g., pesticide leaching to water reservoirs, bioaugmentation, etc. [10]. From agricultural practices, nanomaterials represent technologically advanced materials ranging from 1 to 100 nanometers in size [11]. They possess diverse physical and chemical properties, giving them multiple biochemical activities, ranging from efficient agrochemical delivery to enhancing plant nutrient absorption and tolerance towards harmful chemicals. To better understand and predict the behavior of these nanomaterials in various environmental contexts, in silico analysis has emerged as a powerful tool, allowing researchers to model and simulate complex interactions at the nanoscale level [12,13,14,15]. In addition, rational design based on density functional theory (DFT) is an important strategy in optimizing the catalytic properties of nanomaterials. Nanozymes are a type of nanoparticle that can mimic the activity of natural enzymes, biological catalysts that speed up chemical reactions [16]. Using DFT, electronic structures, reaction pathways, and energy surfaces can be modeled at an atomic level for nanozymes. This helps understand the catalytic mechanisms that control their behavior, such as interaction with substrates and how they catalyze the reactions [16].

Biodegradable properties, controlled surface functionalization, customizable size, ease of distribution, and soil compatibility are a few noteworthy characteristics of nanomaterials that have garnered attention recently for soil reclamation and pesticide removal [9,17,18]. The simultaneous rise of computational biology, simulation techniques, and in silico algorithms has facilitated the in-depth ecotoxicity characterization of nanomaterials before their release and application in agricultural fields [19]. Concurrently, machine learning (ML) has emerged as a critical tool to investigate and predict the environmental fate of pesticides [20,21], with its applications rapidly expanding across various domains [22,23,24,25]. Likewise, the rise of biomanufacturing and bioengineering techniques has also complemented the synthesis of nature-inspired polymer-based nutrient carriers, enhancing crop yields [26]. Another material in the agricultural sector is biopolymers, which are in demand. Low-molecular-weight biopolymers such as silk, keratin, and spider silk are more prevalent and easier to synthesize, whereas ultra-high-molecular-weight (UHMW) biopolymers, including chitosan, polyhydroxyalkanoates, and titin protein, are less explored and challenging to recombinantly express. However, they offer significant potential owing to their superior mechanical and biologically relevant properties such as high mechanical strength and biodegradability [27,28]. UHMW biopolymers are characterized by their long chain lengths and high molecular weights, with exceptional physical and chemical properties, such as a high tensile strength and wear resistance [29].

An in-depth review of the literature published from 1991 to the present has been compiled on nanomaterials, state-of-the-art biomanufacturing techniques, and computational biology, owing to their transformative impact in facilitating sustainable agricultural practices. The literature reviewed was selected on the basis of comprehensive coverage of the topic, from the introduction of the issue to remediation strategies, while emphasizing recent, high-impact research relevant to agricultural practices. While much of the existing research focuses on post-harvest food safety and next-generation pesticide formulations [30,31,32], the compiled literature provides a holistic view of pesticide mitigation strategies. The primary deliverables of the literature review are (i) applications of nanomaterials in soil reclamation and pesticide removal, (ii) roles of computational biology and ML in assessing the mitigation mechanism of the residual chemicals, (iii) potential impacts of the residual chemicals on the agri-based value-added products, and (iv) health and societal impacts of agricultural pollutants.

2. Agricultural Pollutants

Pre- and post-processing activities targeting cleaning and commercial packaging of farming products contribute significantly to accumulating agricultural waste. A broad spectrum of residual items and shave-off materials are categorized as agri-waste, e.g., manures, bedding, plant stalks, hulls, leaves, etc. (

Table 1. Classification and characterization of agricultural contaminants: a detailed description of pollutants, common examples, origins, impacts on ecology as well as human health, and their environmental persistence patterns.

Pollutant	Example	Source	Ecotoxicity	Health Impact	Life Cycle Remarks
Pesticides	Organophosphate, carbamates, pyrethroids [34]	Herbicides, insecticides, fungicides, bactericides [34]	High [34]	Carcinogenic [34]	Bioaugmentation [35]; persist up to ~50 years [36]
Organic matter	Organic material, e.g., plant, livestock wastes [37,38]	Chemical or biochemical oxygen-demanding materials [37,38]	Moderate [38]	Diarrhea, trachoma, and helminth infection [39]	Bioleaching-based release and ecotoxicity [40]
Nutrients	Supplements of nitrogen, phosphorus [6]	Chemical fertilizers, organic material [6]	Low [6,41]	Alzheimer, diabetes mellitus [42]	Imbalance of nitrogen and phosphorus contents of soil [43]
Metals	Se, Pb, Cu, Hg, As, Mg [44]	Nutrient supplements, specifically for small-scale hydroponics and greenhouses [44,45]	High [46]	Kidney, liver, lung failure [46,47,48]	Compromised nutrient status of food materials [49]
Salts	N, P, K, Na, Cl, bicarbonates of Mg, Ca [50]	Nutrient supplements, fertilizers, growth stimulants [51]	High [52]	Cardiac arrhythmias, hypercalcemia [52]	Compromised nutrient status of food materials [53,54]
Emerging pollutants	2-aminobenzothiazole (ABA) [55]	Drug residues, synthetic hormones, and feed additives [55]	High [56]	Cardiovascular toxicity [56]	ABA acts as cytokines in human [57]
Pathogens	Fecal coliforms, Salmonella, Enterococci [58]	Organic manures, bio-manufactured supplements [58,59]	Moderate [58]	Carcinogenic, viral, and parasitic infection [60]	Repeating cycle of pathogen leads to multiple infection cycles and disease spread [60]

). These agri-waste items are often discarded, and their accumulation raises health, safety, environmental, and aesthetic concerns [33]. An overview of the various activities resulting in the accumulation and subsequent leaching into the ecosystem is provided in Figure 1.

2.1. Chemically Synthesized Control Agents

Pests in agriculture and veterinary contexts encompass insects, rodents, weeds, fungi, and other harmful microbes [61]. Pesticides are chemical substances used to control, prevent, or eradicate pests, safeguarding human interests concerning food security, property preservation, and disease prevention [7]. To further illustrate the impact of pesticides on soil health,

Table 2. Comprehensive analysis of agricultural pesticides related specific enzymes, their ecological functions, interacting pesticide compounds, and observed effects on soil biochemical processes and microbial communities.

Enzyme (Function in Soil)	Pesticides	Remarks
Nitrogenase (fixes atmospheric N2)	Carbendazim, imazetapir, thiram, captan, carbofuran, 2,4-D, quinalphos, monocrotophos, endosulfan, γ-HCH, butachlors Terbutryn, simazine, prometryn Brominal, fenvalerate, Cuprosan Oxafun, Funaben, Baytan, pretilachlor, benthiocarb, cinmethylin, anilofos, Methabenzthiazuron, terbutryn, linuron	Reduced or no nitrogenase activity [68,69,70,71,72,73,74,75] Field doses of the fungicides did not affect nitrogenase activity of methylotrophic bacteria Higher doses suppressed activity [76]
Phosphatase (mineralizes organic phosphorus to inorganic)	2,4-D, nitrapyrin, mefenoxam, metalaxyl Monocrotophos, chlorpyrifos, mancozeb, and carbendazim Quinalphos Diazinon, imidacloprid, lindane Glufosinate ammonium	Inhibited [77,78] Higher concentration or increasing incubation period had inhibitory effects [79,80] Initially inhibited but later activity was restored [81] Diazinon did not affect; imidacloprid increased while lindane decreased the enzyme when applied as seed treatment in groundnut field [82] Initial inhibition of phosphatase in sandy loam and clay loam soils [83]
Urease (hydrolyses urea into CO2 and NH3; key component of N2 cycle in soil)	Isoproturon, benomyl, captan, diazinon, profenofos	Increased activity [84,85] Reduced/inhibited activity [86,87]
DHA (removal of H2)	Azadirachtin, Acetamiprid, Quinalphos, Glyphosate Atrazine, and Northrin	Positive/stimulatory influence on DHA [88] Initially inhibited followed by activity restoration [89] Herbicides stimulated DHA of the microbial community at lower and inhibited it at higher concentrations [90]
Invertase (hydrolyze sucrose to fructose and glucose)	Atrazine, Carbaryl, Paraquat	Activity inhibition [91,92]
β-Glucosidase (hydrolyzes disaccharides in soil to form β-glucose)	Metalaxyl, Ridomil Gold Plus Copper	Enzyme activity increased and then decreased [93]; inhibited [94]
Cellulase (hydrolyzes cellulose to D-glucose)	Benlate, Captan, Brominal	Inhibited enzyme activity [95,96]
Arylsulphatase (hydrolyzes aryl sulfates)	Cinosulfuron, Prosulfuron, Thifensulfuron methyl, Triasulfuron	Activity reduction [97]

DHA, dehydrogenase; HCH, hexachlorocyclohexane.

represents a comprehensive analysis of specific enzymes related to agricultural pesticides use. Depending on the targeted pest organism, the pesticide can be an insecticide [62], molluscicide [63], fungicide [64], herbicide [65], or algaecide [66]. Most pesticides are carbon-containing organic or inorganic compounds, such as copper sulfate, ferrous sulfate, copper, lime, and sulfur [67]. Detailed descriptions of various target organisms and degradation techniques for insecticides, fungicides, and herbicides are compiled in

Table 3. A comprehensive assessment of commercial agricultural insecticides detailing the common brands, chemical compositions, target pathologies, and conventional degradation pathways.

Brand Name	Causative Ingredient	Target Organism or Disease	Degradation Technique	Reference
Keefun	Tolfenpyard	Diamond black moth	Microbial	[98]
Colt	Cypermethrin	Lepidopteron pest	Photo and microbial	[99,100]
Osheen	Dinotefuran	Brown plant hoppers	Photocatalytic	[101]
Colfos	Ethion and Cypermethrin	Bollworms on cotton	Photo and microbial	[99,100,102]
Fosmite	Ethion	Mites, scales, thrips, beetles	Microbial	[102]
Fluton	Flubendiamide	Tobacco caterpillar, American bollworm	Direct aqueous photolysis	[103]
Jumbo	Imidacloprid	Rice hoppers, thrips, turf insects	Photolysis and microbial	[104]
Snailkill	Metaldehyde	Snails and slugs	Biodegradation	[105]
Roket	Profenofos and Cypermethrin	Insect pests (both chewing and sucking)	Biodegradation	[106]
Carina	Profenofos	Insect pests (both chewing and sucking)	Biodegradation	[106]
Simbaa	Propargite	Red spider mites	Biodegradation and phytodegradation	[107,108]
Voltage	Spiromesifen	Whiteflies and mites	Biodegradation	[109]
Maxima	Thiamethoxam	Thrips and tea mosquito bug	Activated persulfate	[110]
Vibrant	Thiocyclam Hydrogen Oxalate	Stem borer
Cosko GR	Chlorantraniliprole (0.4%)	Lepidoptera	Photolysis, hydrolysis and biodegradation	[111]
Cosko SC	Chlorantraniliprole (18.5%)	Lepidoptera	Photolysis, hydrolysis and biodegradation	[111]
Rodeo	Bifenthrin	Whitefly, mites and jassids	Microbial	[112]
Distruptor	Dinotifuran and Pymetrozine	Brown planthopper, white-backed plant hopper	Photocatalytic and biodegradation	[101,113]
Aceveer	Acephate 75% SP	Pink bollworm, armyworm	Acid hydrolysis, ozone, and electrolyzed treatment	[114]
Imexo	Thiamethoxam 25% WDGs	Stem borer, gall midge, leaf folder, WBPH, BPH, GLH, Thrips, whorl maggot, jassid, aphids, whiteflies	Activated presulfated process	[110]
Virtako 0.6 GR	4 gm/kg Thiamethoxam + 2 gm/kg chlorantraniliprole	Stem borer, borers of sugarcane, and borer of corn	Acid hydrolysis, ozone, and electrolyzed treatment	[110,111]
Lambda Double	Lambda Cyhalothrin 5% EC	Bollworms, jassids, thrips, stem borer, shoot and fruit borer, mite, pod borer, pod fly, hopper, leaf hopper	Microbial	[115]
Lambda Strong	Lambda Cyhalothrin 4.9% CS	Bollworms, jassids, thrips, stem borer, shoot and fruit borer, mite, pod borer, pod fly, hopper, leaf hopper	Microbial	[115]
Onvix	Chlorantraniliprole 0.4% Grs	Chilo infuscatellus, Scirphophaga excerptalis, Scirphophaga insurtulas,	Photolysis, hydrolysis, and biodegradation	[111]
Pevota	Chlorantraniliprole 18.5% SC	Scirphophaga insurtulas, Chiloinfuscatellus medinalist	Photolysis, hydrolysis, and biodegradation	[111]
Stembo	Fipronil 0.3% Grs	rice gall midge, whorl maggot, white backed plant hopper	Photolysis, hydrolysis, and biodegradation	[116]
Veercombi 44	Profenofos 40% + Cyper 4% EC	Thrips, bollworm, aphid, jassid, mealybug	Biodegradation	[106,117]
Veercombi 505	Chlorpyrifos 50% + Cyper 5% EC	Spotted bollworm, American bollworm	Biodegradation	[118]
Veertap Power	Cartap Hydrochloride 4% Grs	Stem borer, leaf folder, whorl maggot	Fenton degradation	[119]
Uttam Flue 3935	Flubendiamide 39.35% w/w	Spotted boll worm, stem borer, leaf folder, fruit borer	Soil photolysis	[120]
Fenveer DP	Fenvalerate 0.4% dust	General insects	Photo and microbial	[121,122]
Imidaveer	Imidacloprid 17.8% SL	Thrips, hopper, aphid, jassid, termite, whitefly, leaf miner	Photolysis and microbial	[104]
Uttam Metroz	Pymetrozine 50% WDGs	Brown plant hoppers	Biodegradation	[113]
Toro-10	Bifenthrin 10% EC	Whitefly, aphids, stem borer, leaf folder, green leaf hopper	Microbial	[112]
Chloroveer	Chlorpyriphos 20% EC	Control of sucking, chewing, and boring insects	Biodegradation	[118]
Chlorveer Strong	Chlorpyriphos 50% EC	Control of sucking, chewing, and boring insects	Biodegradation	[118]
Uttam EMA	Emamectin Benzoate 5% SG	Pod borer, bollworms, tea looper, fruit and shoot borer	Foliar spray and ultraviolent radiation	[123]
Uttam Reon	Pyriproxyfen 5% and Diafenthiuron 25% SE	Whitefly, aphid, jassids, thrips	Photodegradation	[124,125]
Bruno	Buprofezin 25% SC	Whitefly, aphid, jassids	Biodegradation	[126]
Brofreya	Broflanilide	Leps and sucking pests	Photolysis	[127]

EC: emulsifiable concentrate; Grs: granules; SC: suspension concentrate; SL: soluble liquid; SP: soluble powder; SE: suspension emulsion; WDGs: water-dispersible granules.

,

Table 4. A comprehensive assessment of agricultural fungicides detailing the common brands, chemical compositions, target pathologies, and conventional degradation pathways.

Brand Name	Causative Ingredient	Target Organism or Disease	Degradation Technique	Reference
KItazin	Kitazin	Blast and sheath blight of rice, fruit rot of chili, and purple blotch of onion	Microbial	[128]
Sanipeb	Propineb	Broad spectrum of fungi	Biodegradation	[129]
Clutch	Pyraclostrobin and Metiram	Early blight disease, late blight disease and downy mildew disease	Irradiation with UV	[130]
Header	Pyraclostrobin	Rice blast	Irradiation with UV	[130]
Visma	Pyraclostrobin and Boscalid	Downy mildew, powdery mildew, anthracnose, botrytis	Irradiation with UV	[130,131]
Shield	Iprobenfos	Wide range of fungal species	Light-induced catalyst	[132]
Veersulp DP	Sulphur 85% DP	Mildew, rust, tikka leaf spot	Microbial degradation	[133,134]
Uttam Fulcot	Thifluzamide 24% SC	Sheath blight	Microbial degradation	[135]
Uttam Lexon	Azoxystrobin 11% + Tebuconazole 18.3% SC	Sheath blight, early blight, late blight, yellow rust, purple blotch	Bacterial degradation	[136,137]
Uttam Azole	Azoxystrobin 18.2% and Difenoconazole 11.4% SC	Anthracnose, powdery mildew, early blight and late blight, downy mildew	Bacterial degradation	[137]
Wagon	Thifluzamide	Sheath blight in rice	Bacterial degradation	[138]
Figo	Tricyclazole 75% WP	Blast disease	Use of hydrogen peroxide	[139]
Manzim	Manco 63 and Carben 12%	Blast disease, leaf spot, rust diseases	Microbial degradation	[140]
Sulfino	Sulphur 80% WDGs	Powdery mildew, scab	Microbial degradation	[133,134]
Veercon	Propiconazole 25% EC	Karnal bunt, leaf rust, sheath blight, rust, leaf spot	Biodegradation	[141]

and

Table 5. A comprehensive assessment of commercial agricultural herbicides detailing the common brands, chemical compositions (causative ingredients), target pathologies, and conventional degradation pathways.

Brand Name	Causative Ingredient	Target Organism or Disease	Degradation Technique	Reference
Solaro	Atrazine	Trianthama monogyna, Digera arvensis, Echinochloa spp., Eleusine spp., Xantheium strumarium, Brachiara spp., Digitaria spp.	Ozonation	[142]
Wicket	Clodinafop-Propargyl	Phalaris minor	Photocatalytic	[143]
Nominee Gold	Bispyribac Sodium	Fimbristylis miliacea, Eclipta alba, Ludwigia parviflora, Monochoria vaginalis, Alternanthera philoxeroides, Sphenoclcea zeylenica	Microbial	[144,145]
Pimix	Metsulfuron methyl and Chlorimuron ethyl	Cyperus iria, Cyperus difformis, Fimbristylis miliacea, Eclipta alba, Ludwigia parviflora, Cynotis axillaris, Monochoria vaginalis	Biodegradation	[146]
Legacee	Fenoxaprop-p-ethyl	Echinochloa spp. (barnyard grass)	Hydrolysis	[147]
Melsa	Pinoxaden	Canary grass, Phalaris minor (wild oat), Avena ludoviciana	Microbial	[148]
Elite	Topramezone	Elusine indica, Digitaria sanguinalis, Dactyloctenium aegyptium, Echinocloa spp., Chloris barbata, Parthenium hysterophorus	Biodegradation	[149]
Awkira	Pyroxasulfone	Echinochloa colonum, Celosia argentia, Trianthema portulacastrum, Amaranthus viridis, Digeria arvensis	Biodegradation	[150]
Londax Powder	Bensulfuron Methyl and Pretilachlor	Echinochloa crusgalli, Echinochloa colonum, Cynodon dactylon, Sedges, Cyperus iria, Cyperus difformis, Cyperus rotundus	Microbial	[151]
Attract	Atrazine 50% WP	Trianthama monogyna, Digera arvensis, Echinochloa spp., Eleusine spp., Xantheium strumarium, Brachiara spp., Digitaria spp.	Ozonation	[142]
Butaveer	Butachlor 50% EC	Cyperusdifformis, Cyperusiria, Echinochloacrusgalli, Echinochloacolonum, Elusine Indica, Eclipta alba, Fimbristy lismiliacea	Microbial	[152]
Dhoomketu	Imazethapyr 30% SL	Echinocloa crusgalli, Digera arvensis, Commelina benghalensis, Amaranthus viridis	Photodegradation	[153]
Lido	Butachlor 50% EW	Cyperusdifformis, Cyperusiria, Monochoria vaginalis, Eclipta alba	Microbial	[152]
Metaveer	Metribuzin 75% WP	Capeweed, doublegee, wild radish	Biodegradation	[154]
Penveer	Pendimethalin 30% EC	Echinochloa, Euphorbia, wild amaranthus, Phyllanthus, Paspalum, Phalaris, Carnoplus	Biodegradation and phytodegradation	[155]
Penveer Plus	Pendimethalin 38.7% CS	Annual grasses and broad-leaved weeds like E. colonum, Digitaria sanguinalis, Dactyloctenium aegyptium, Amaranthus viridis	Biodegradation and phytodegradation	[155]
Pretilaveer	Pretilachlor 50% EC	Echinochloa crusgalli, Echinochloa colonum, Cyperusdifformis, Cyperusiria Fimbristylismilliacea, Eclipta alba	Microbial	[151,156]
Totto	Paraquat Dichloride 24% SL	Imperata cylindrica, Setaria spp., Commelina benghalensis, Boerhavia hispida, Paspalum conjugatum, Chenopodium spo., Anagallis arvensis	Microbial	[157]
Veerkill 80	2,4-D Sodium Salt 80% WP	Leucasaspera, Chenopodium album, Argemonemexlcana, Fimbrlstyllismiliacea, Anagallisarvensis, Vicia saliva	Microbial	[158]
Zeto	Fenoxaprop Ethyl 9.3% EC	Echinochloa colonum, Echinochloa crusgalli, Digitaria spp., Eleusine indica, Setaria spp., Brachiaria spp., Digitaria spp.	Microbial	[159]
Moto	Metsulfuron Methyl 20% WP	Chenopodium album, Melilotus indica, Lathyrus aphaca, Anagalis arvensis, Vicia sativa, Ludwigia parviflora, Sphenoclea zeylanica	Microbial	[160]
Prido	Pretilachlor 37% EW	Echinochloa colonum, E. crusgall, Cyperus difformis, C. iria, Digitaria sanguinalis, Frimbristylis miliacea, Eclipta alba	Microbial	[161]
Veerkill	2,4-D Ethyl Ester 38% EC	Cyperus Iria, Digera Arvensis, Convolvulus Arvensis, Trianthema Sp., Tridax Procumbens, Euphorbia Hirta, Phyllanthus Niruri, Trianthema	Microbial	[162]
Weedkil	2,4-D Amine Salt 58% SL	Cyperusiria, Striga sp. Trianthema sp., Tridaxprocumbens, Digeriaarvensis, Convolvulus arvensis, Euphorbia hirta, Phyllanthusniruri	Microbial	[158]
Weeza	Clodinofop Propargyl 15% WP	Phalaris minor	Photocatalytic	[143]
Wheto	Sulfosulfuron 75% WDGs	Phalaris minor, Chenopodium album, Melilotus alba	Hydrolytic	[163]

. An overview of pesticides is presented based on their (i) on-field deployment, (ii) spray applications, and (iii) mode of action (MoA) for protecting the plants from pests.

Broadly, insecticides can be classified into systemic (absorbed by the plant) or contact categories (direct contact required) and constitute ~55 chemical classes [164]. Their classification is primarily based on the physiological functions affecting nerve/muscle, growth, respiration, midgut, or other non-specific characteristics. Importantly, 30 modes of action (MoA) exist for these insecticides, including multiple target sites and some with unknown mechanisms. To prevent insecticide resistance in pest populations, it is necessary to rotate the MoA [164]. Likewise, fungicides operate either through contact or penetration mechanisms [165]. Effective spray coverage of fungicides is extremely important, as systemic movement throughout the plant is uncommon. The fungicide resistance action committee classified them into 14 MoA groups [165,166].

Contrarily, herbicides can be highly selective, i.e., impacting selective weeds, or non-selective in their mode of action, thus eliminating entire vegetation. Spray application of herbicides includes their deployment at pre-plant (non-selective applied to the soil before planting), pre-emergence (before weed seedlings emerge through the soil), or post-emergence (weed seedlings already emerged from the soil). There are 23 MoAs classified by the Herbicide Resistant Action Committee (HRAC). Rotation within these groups is critical to prevent herbicide resistance [167].

2.2. Fertilizers and Manure

Chemical fertilizer primarily consists of three nutrients, i.e., nitrogen (N), phosphorus (P), and potassium (K). Remarkably, half of the applied volume of the fertilizer remains unconsumed and enters the ecosystem through evaporation and direct leaching to groundwater [168]. Nearly one-third of the applied fertilizer is estimated to interact with the organic compounds in the agricultural fields and to be converted into other xenobiotic compounds [169]. Leaching of unutilized fertilizer is an enormous challenge for maintaining the ecological balance, specifically for the underground water quality [168]. Due to microcosmic microbial interactions, unused N-fertilizers form nitrates, a threat to the water quality due to their higher water solubility. Field applications of N-fertilizers have aggravated concerns since the reported concentration of nitrates in drinking water has crossed the ≥11.3 mg/L “safe” levels in many European and Asian countries [170]. Populations residing in these affected areas are prone to blood disorders, e.g., methemoglobinemia, blue baby syndrome, and more, due to elevated blood-nitrate concentrations [171].

Historically, excessive and unplanned manure utilization has impaired the ecological balance between flora and fauna [172]. Although manure is a rich source of N and P, its elemental concentration varies according to the collection and storage practices and material used in feed and beds [173]. Soluble salts, e.g., Ca, Mg, Na, S, and more, have been identified in various manure samples and preparation methods. Their leaching into underground water causes contamination and other societal challenges [174]. Additionally, numerous microbial species, including ruminant bacteria in manure, threaten biodiversity equilibrium in case of their uncontrolled release and spread into the ecosystem [175].

2.3. Post-Harvest Agriculture Waste

Post-harvest agricultural waste is a potential resource and a burden from an economic standpoint. For farmers, the price of disposing of garbage can add up, especially in areas without sufficient infrastructure for waste management. However, by being turned into useful goods, these residues also present chances for financial benefit. Agricultural waste, for instance, can serve as a feedstock for bioenergy production, such as biogas or bioethanol, offering farmers an additional source of income while reducing their dependency on fossil fuels [176]. Additionally, rural economies can become even more sustainable and diverse economically by processing these leftovers into bio-based products like biochar and bioplastics.

Apart from the consequences on health and finances, the handling of agricultural waste after harvest impacts soil health and environmental sustainability. Higher crop yields and more resilient agricultural systems can result from reintroducing these wastes as organic matter, improving soil structure, promoting water retention, and increasing nutrient availability [177]. Furthermore, it has been demonstrated that using agricultural waste to produce biochar can improve soil fertility, reduce climate change, and sequester carbon [178]. Therefore, managing and utilizing post-harvest agricultural waste effectively to support rural lives, advancing sustainable farming practice, and solving global environmental concerns is imperative. Effective management of post-harvest agricultural waste also plays an important role in reducing greenhouse gas emissions and, consequently, fighting climate change. When left to decompose in landfills or open fields, agricultural waste can release large amounts of methane, a potent greenhouse gas. However, with appropriate organic waste management strategies, for example, composting or anaerobic digestion [179], these emissions can be substantially reduced. Additionally, agricultural waste can be utilized in bioenergy production, reducing consumption and reliance on fossil fuel emissions. For example, some studies have proven that using crop residues for bioenergy produces a net reduction in greenhouse gas emissions compared to conventional fossil fuel combustion [180]. Equally importantly, the inclusion of agriculture-based waste into circular models enables better use of resources and reduces environmental impacts while boosting the economy in rural areas. Such holistic approaches in managing agricultural waste not only deal with environmental issues but are also in tune with global sustainable goals and foster the change needed toward a more sustainable, regenerative agricultural sector.

3. Ecological Impacts of Agricultural Pollutants

Unlike other scenarios of xenobiotics and pollution, agricultural pollutants are intentionally introduced into the ecosystem to achieve specific objectives such as controlling weeds, pests, and diseases, thereby protecting crop yields and quality [168,181]. However, their addition raises environmental and health concerns. The adverse impacts of pesticides, fertilizers, and manure are addressed directly or indirectly through agreements and policies to protect human and ecological health [182,183]. Global pesticide usage in 2024 amounted to approximately 4.326 million metric tons, with the highest contribution (by weight consumption) being herbicides (~50%), followed by, in descending weight, fungicides and bactericides (~21%), and insecticides (~20%). The remaining pesticides comprised 8% [184]. Notably, global pesticide application per cropland area increased to 2.7 kg/ha [185], and global pesticide and fertilizers sales are projected to reach USD 309 billion by 2025 if the trend continues [186].

Substantial nutrients are lost into the environment in forms such as nitrate, ammonia, and particulate phosphorus, negatively impacting human and environmental health [187]. Persistent chemicals residues can bioaccumulate within the food chain and have been detected in products such as meat, poultry, fish, vegetable oils, and nuts [188]. These residues or pollutants enter ecosystems through direct and indirect routes. Direct routes are associated with agricultural activities driven by commercial and societal demands, whereas indirect routes often result from ambiguous storage and usage guidelines, resulting in the leaching of pollutants into ecosystems [189]. An overview of the worldwide consumption of the primary nutrients in chemical fertilizers is given in Figure 2.

Biomagnification of agricultural pollutants, including heavy metals and pesticide residue [191], consequently impacts human health and can contribute to the emergence of superbugs and incurable diseases. Indirect exposure to pollutants occurs through pesticide ingestion via contaminated food and water, inhalation of droplets, fumigation practices, and storage activities of agricultural products. Specifically, the organophosphate pesticides used in vegetables gradually accumulate in the human body and have been linked to cancer [192].

3.1. Environmental Pollution

Microbial communities, their interactions, and the secreted by-products (e.g., enzymes, exopolysaccharides, biofilm matrix components, etc.) are crucial for maintaining soil health [193]. These bio-inspired activities convert complex organic material into a simple form that crops/plants can assimilate efficiently. An environmental sample’s pesticide content mainly affects the soil’s enzyme activity, microbial diversity, and soil organic matter degradation.

Additionally, the surface topology and soil types are the rate-limiting factors in the absorption of pesticides. Upon absorption, pesticides in the soil may undergo certain structural and functional changes, resulting in reduced crop productivity [194]. Ammonia containing runoff water from a fresh manure site is toxic to aquatic life. Fish are sensitive to high ammonia levels and are killed by high ammonia levels [195]. For example, concentrations as low as 0.35 mg/L (ammonia, 96-h LC₅₀) can be lethal to some freshwater fish species [195].

3.2. Societal Challenges

The societal and end-user impacts of these chemical pesticides are enormous. The detrimental impacts of agricultural pollutants are well documented in terms of their short- and long-term consequences. Pesticide exposure has been linked to an elevated incidence of Alzheimers, Parkinsons, amyotrophic lateral sclerosis, asthma, and system failures [52,56,196,197]. Limited exposure for short durations can result in symptoms such as headache, nausea, tremors, and slurred speech. Long-term exposure to pesticides has resulted in numerous immune, metabolic, and endocrine disorders and carcinogenic conditions [198,199]. Moreover, continuous exposure over long durations may cause irreversible damage to the liver, kidneys, and central nervous system. Specifically, long-term exposure to Dichlorodiphenyldichloroethylene (DDE) and Dichlorodiphenyltrichloroethane (DDT) has been linked to neurodevelopmental disorders in children [200]. Certain classes of pollutants, such as organophosphates and carbamates, are associated with the denaturation or malfunctioning of cholinesterase enzymes [201], which are required for the hydrolysis of acetylcholine into choline and acetic acid, responsible for returning cholinergic neurons to their normal resting state after activation [202].

Significant disruptions in protein, carbohydrate, and fat metabolism and genotoxic effects on mitochondrial function have been reported, causing cellular oxidative stress and problems in the nervous and endocrine systems [203] (Figure 1). A major class of pollutants, namely, organophosphate pesticides, has been reported for their endocrine-disrupting activities [204], and leading to reproductive disorders [205,206] and metabolic disorders [203]. Their cytotoxic and genotoxic impacts cause necrosis in human T-cells and natural killer cells [207], and apoptosis in T lymphocytes [208] has been extensively investigated. These pollutants can also increase the risk of dementia [209], neurobehavioral effects [210], and non-Hodgkin’s lymphoma [211].

In addition to pesticides, other agrochemicals, such as herbicides, fungicides, and rodenticides, have severe effects on human health. Herbicides like glyphosate and atrazine have been linked to endocrine disruption, oxidative stress, and genetic harm [212]. Glyphosate, for example, has been shown to disrupt hormonal processes and may lead to reproductive and developmental issues [212]. Fungicides such as mancozeb, chlorothalonil, and azole derivatives have been shown to be cytotoxic, genotoxic, and toxic to reproductive in animal models [213] and human cell lines. Mancozeb, a commonly used dithiocarbamate fungicide, has been shown to impact thyroid hormone regulation and mitochondrial respiration [213,214], while specific azole fungicides disturb steroidogenesis and spermatogenesis [215].

4. Nano-Remediation: An Emerging Paradigm

Nano-remediation has gained status as an innovative and sustainable approach for mitigating agricultural pollutants, capitalizing on the distinct physicochemical attributes of nanomaterials [216]. These materials possess exceptionally high surface area-to-volume ratios and modifiable surface characteristics [217], allowing them to interact with contaminants at a molecular level with high precision [12,218]. Through mechanisms such as adsorption, catalytic degradation, redox transformations, and ion exchange, nanomaterials can effectively neutralize, sequester, or degrade harmful compounds, thereby reducing their environmental persistence and bioavailability. The effectiveness of these nanomaterials is further improved through surface functionalization with stabilizing agents or chemical modifications, which enhance their selectivity, dispersibility, and stability in heterogeneous soil systems. The controlled design and synthesis of these nanoparticles enable their targeted deployment for pollutant remediation, offering a strategic advantage over conventional decontamination techniques.

Recent advancements in nanotechnology have also highlighted the potential of integrating nanomaterials with biological components, such as plant-derived extracts, microbial metabolites, and organic ligands, to amplify their remediation efficiency while mitigating adverse effects on soil microbial communities [219,220]. Studies have demonstrated that nanocomposites comprising metallic nanoparticles, carbon-based structures, and polymeric nanomaterials can efficiently degrade or immobilize a wide spectrum of agricultural pollutants, including heavy metals, pesticide residues, and persistent organic pollutants (POPs) (Figure 3) [221,222,223,224]. However, despite the promising potential of nano-remediation, several concerns remain regarding its long-term stability, environmental persistence, and potential ecotoxicological effects on soil ecosystems. Understanding the interactions between nanomaterials and soil constituents is crucial for optimizing their application while minimizing unintended environmental risks.

4.1. Metal- and Metal Oxide-Based Nanoparticles for Pollutant Remediation

Metal- and metal oxide-based nanoparticles (NPs) have emerged as effective materials for the remediation of environmental pollutants due to their unique physicochemical properties, including high surface area, reactivity, and the ability to undergo various redox reactions. These NPs can be broadly categorized based on their composition and structure.

4.1.1. Iron-Based Nanomaterials

Iron-based nanomaterials, such as zero-valent iron (nZVI) and iron oxide nanoparticles, are particularly attractive for environmental remediation due to iron’s abundance and its central role in various natural processes. In soil and water systems, nano-sized iron oxides form through sedimentation mechanisms and are effective in removing heavy metals (e.g., As (III), Cr (VI), Pb (II)) via redox reactions, adsorption, and precipitation [225]. Moreover, nZVI has been applied to dehalogenate chlorinated hydrocarbons, including pentachlorophenol (PCP) and DDT-related compounds, making it a versatile option for treating a wide array of pollutants [226,227]. The nano-remediation of PCP, DDT, and other DDT-related compounds like DDE (2,2-bi’s(p-chlorophenyl)-1, 1-dichloroethylene), DDD (1,1-dichloro-2,2-bi’s(p-chlorophenyl) ethane), is achieved by using zero-valent iron NPs in a buffered aqueous solution due to the de-chlorination reaction or sorption over the surface of nZVI NPs [228].

The magnetic properties of iron oxide NPs allow for efficient dispersion in aqueous environments [229]. Iron oxide NPs, including magnetite (Fe₃O₄), hematite (α-Fe₂O₃), and maghemite (γ-Fe₂O₃), are widely used as nano-adsorbents due to their effectiveness in removing various pollutants from water sources as well as soils [230]. Additionally, studies have shown that Fe₃O₄ nanoparticles combined with indigenous soil microbes significantly enhance degradation of herbicides (2,4-dichlorophenoxyacetic acid) and improve enzymatic activities, such as amylase, acid phosphatase, catalase, and urease [231]. Fe₃O₄ composites have shown effectiveness in degrading phenanthrene through visible-light-driven photocatalysis, thereby offering a promising approach for sustainable soil remediation [232]. Notably, Fe₃O₄ and γ-Fe₂O₃ exhibit magnetic properties that facilitate their separation and recovery, making them advantageous for such applications [233]. Both have been successfully utilized as sorbent materials for removing heavy metals [234]. Additionally, metal oxide nanoparticles have proven effective in treating water pollutants, especially for adsorbing heavy metals and organic pollutants, showing promising results in water treatment [230]. However, challenges such as aggregation, oxidation, and potential magnetic-phase leakage in acidic environments can limit their effectiveness. To address these issues, surface functionalization with inorganic and organic materials has been employed to improve stability and prevent oxidation. Functional groups like carboxylic (-COOH) and amine (-NH₂) groups have been shown to enhance efficiency and surface area [235,236,237,238].

4.1.2. Silver Nanoparticles (AgNPs)

Silver nanoparticles have garnered significant attention in soil environment remediation due to their exceptional photocatalytic properties, which facilitate the degradation of various organic pollutants, acting as antimicrobial agents [239]. Synthetically, AgNPs can be produced through green chemistry methods, utilizing bacteria or other natural reducing agents [240]. These nanoparticles can be immobilized on substrates such as cellulose acetate membranes, enhancing their utility in removing pesticides like malathion and chlorpyrifos from contaminated environments [241]. The remediation mechanism of AgNPs predominantly involves photocatalysis, wherein plasmonic excitation generates hot electrons, and semiconductor behavior leads to the formation of electron–hole pairs [242]. Empirical studies have demonstrated the successful degradation of pesticides, such as malathion [243], paraquat [244], and paraoxon [245] utilizing surface-functionalized AgNPs.

4.1.3. Other Metal-Based Nanoparticles

Metal and metal oxide nanoparticles, including gold (AuNPs), titanium dioxide (TiO₂), and other metal oxides, such as MgO, Al₂O₃, MnO, CuO, have gained significant attention for their potential in environmental remediation due to their unique properties and effectiveness in pollutant removal [9,246,247]. For instance, nano-TiO₂ has been shown to reduce the half-lives of phenanthrene and atrazine, whereas AuNP-anchored zirconium dioxide nanocomposites have shown photocatalytic degradation of pesticides diuron and methyl parathion [248,249]. The interplay between negatively charged oxygen surfaces and positively charged metal surfaces forms electronic dipoles, promoting voltage and temperature polarization along specific directions and planes, thereby enhancing the efficacy of metal oxide nanoparticles in contaminant removal [230,250].

Titanium dioxide nanoparticles are widely studied for their photocatalytic properties in environmental remediation. Degradation processes involving photocatalysis typically rely on the activation of wide-band-gap semiconductor catalyst materials using a light source in the presence of polluted water, which generates electron–hole pairs [251]. When exposed to light of an appropriate wavelength (with photon energy equal to or greater than the band-gap energy), these materials undergo a series of oxidative and reductive reactions on the photocatalyst surface [252,253]. The generated electrons then diffuse across the photocatalyst’s interface and interact with the surrounding environment, driving oxidation and reduction reactions [254,255]. In oxidation reactions, the catalyst, activated by energy, facilitates the production of highly reactive radical species, such as OH• and •O₂⁻. These reactive radicals are generated through interactions between photogenerated electrons and molecular oxygen, as well as between photogenerated holes and water, effectively degrading organic contaminants in a non-selective manner. However, the requirement for UV light limits their practical use in natural environments. To improve their efficiency under visible light, TiO₂ has been doped with nanocomposites to expand its applicability and enhance its photocatalytic capabilities [256].

In addition to iron and titanium oxide nanoparticles, other metal oxide nanoparticles, such as magnesium oxide (MgO), aluminum oxide (Al₂O₃), and manganese oxide (MnO), have also been investigated for their ability to remove pollutants [257]. MgO nanoparticles, along with nanoscale Mg(OH)₂ particles, have shown significant effectiveness in removing contaminants such as phosphate from aqueous solutions, with Mg-enriched biochar removing up to 88.5% of phosphate, demonstrating its potential [258]. Al₂O₃ nanoparticles and aluminosilicate-based nanomaterials have been shown to have strong sorption capabilities for metal ions such as Cd (II), Pb (II), Mn (II), and Cr (III), demonstrating good adsorption efficiencies [259]. Similarly, green synthesis methods, utilizing Clitoria ternatea, have been successfully employed to produce AuNPs capable of degrading persistent organic pollutants [260].

4.2. Carbon-Based Nanomaterials

Carbon-based NPs are another promising remediation material due to their unique structural and chemical properties; for example, the versatility of elemental carbon’s hybridization state (sp, sp2, sp3) allows formation of various nanostructures, including fullerene C60, carbon nanotubes (CNTs), and graphene, which have distinct characteristics suitable for remediation applications [261,262,263]. These carbon-based nanostructures, such as CNTs and fullerene C60, can yield different structural configurations such as fullerene C60, fullerene C540, single-walled nanotubes, multi-walled nanotubes, and graphene [264]. Even though these technologies are highly efficient for remediation, the adaptation of carbon-based technologies is lower than for graphene due to the cost effectiveness of graphene [265]. Carbon-based NPs excel in contaminant removal due to their high surface area, strong adsorption capabilities, and ability to interact with pollutants through mechanisms such as π-π interactions, van der Waals forces, and electrostatic attraction [9]. Various investigations have determined the suitability of graphene, carbon nanotubes, and other carbon-based nanomaterials for remediation applications [9,266,267,268]. Carbon-based nanoparticles have been extensively studied for contaminant remediation, employing primary mechanisms such as adsorption, photocatalysis, and reduction. For instance, hydroxyl-functionalized multi-walled carbon nanotubes (MWCNTs) have demonstrated significant adsorption capacities for pesticides such as atrazine and 2,4-dichlorophenoxyacetic acid (2,4-D), reaching 47.7 mg/g and 51.4 mg/g, respectively. Furthermore, hybrid nanocomposites comprising carbon nanotubes or graphene-based materials integrated with TiO₂ have shown enhanced photocatalytic degradation of organic contaminants, notably polycyclic aromatic hydrocarbons (PAHs), in soil matrices. In terms of redox activity, carbon nanomaterials have also been shown to facilitate the reductive transformation of persistent organic pollutants, including PAHs, contributing to their detoxification and removal from environmental systems [9,269,270].

Multi-walled and single-walled carbon nanotubes (MWCNTs and SWCNTs) have been the subjects of many pollutant remediation studies [261,271,272]. MWCNTs are modified with hydroxyl or carboxyl groups and doped with magnetite for high adsorption capabilities. For example, hydroxyl functionalized MWCNTs were observed to achieve adsorption capacities of 51.4 mg/g for 2,4-D and 47.7 mg/g for atrazine [269]. Whereas SWCNTs have a reported adsorption capacity of 19.4 mg/g for bisphenol A (BPA) [273].

Through adsorption, these nanoparticles effectively remove organic and inorganic pollutants; for instance, magnetic TiO₂-graphene nanocomposites have shown high removal efficiency for herbicides (2,4-D) from water, while graphene-TiO₂ removes inorganic nitrogen oxides [274,275]. In photocatalysis, carbon-based nanoparticles combined with metal oxides like TiO₂ degrade pollutants under UV radiation [276]. Under UV irradiation, photons of energy greater than or equal to the band gap of the nanotubes promote the generation of valence-band holes and conduction-band electrons. The holes are responsible for forming hydroxyl radicals that take part in oxidizing chlorinated organic compounds [276,277]. Additionally, they facilitate reduction processes by converting toxic metal ions into less harmful forms, further enhancing their applicability in environmental remediation [9,278]. Several studies have been reported that describe the use of graphene to fabricate photocatalytic nanocomposites [279,280,281]. Graphene composites containing TiO₂ NPs show increased photocatalytic activity compared to bare TiO₂ NPs due to increased conductivity [282]. A few examples of commonly employed nanocomposites include magnetic TiO₂-graphene composites for removal of the herbicide 2,4-D from water [274]; GO/TiO₂ nanocomposites for removal of aromatic dyes, heavy metals, and crude oil from water [283]; and the plasmonic gold TiO₂-graphene nanocomposite for degradation of organic pollutants [284].

Carbon quantum dots (CQDs) and graphene-based nanomaterials (GBNMs) are advanced carbon nanomaterials with significant potential in environmental remediation [222]. These nanomaterials remove heavy metals and organic pollutants through adsorption and interaction with the functional groups (hydroxyl, carboxyl, and carbonyl) on the surface. Oxygen-containing groups are present in abundance in CQDs, aiding in binding to metal ions or organic molecules via electrostatic attraction, hydrogen bonding, and coordination interaction, enabling them to immobilize or capture the pollutant [285,286]. CQDs, including carbon nanodots (CNDs) and graphene quantum dots (GQDs), exhibit remarkable photoluminescence, large surface area, and biocompatibility, making them effective in detecting and removing heavy metals, organic pollutants, and antibiotics from water [287,288,289]. Their tunable optical properties and fluorescence-quenching capabilities enable sensitive environmental sensing; i.e., when these nanomaterials interact with specific pollutants, the intensity of their fluorescence decreases, thereby allowing easy detection. For example, when CQDs are exposed to mercury ions in water, fluorescence quenching occurs. The degree of quenching is directly proportional to the amount of mercury sensed, making an accurate, as well as sensitive, detector [290]. GBNMs, such as graphene oxide (GO) and reduced graphene oxide (rGO), possess high mechanical strength, electrical conductivity, and large surface area, facilitating efficient adsorption of contaminants [270,291,292]. Functionalization further enhances their adsorption efficiency, but concerns regarding their environmental impact necessitate improved synthesis methods and safe disposal strategies. Despite their immense potential, further research is required to optimize their large-scale application and assess their long-term ecological effects.

4.3. Rate-Limiting Variables for Nanomaterial Remediation Approaches

Several factors contribute to the degradation of pesticides in the soil environment, and the overall nanomaterial-mediated remediation of pesticides relies on the interplay of these factors. These factors primarily encompass the physiochemical characteristics of pesticides, soil characteristics, and other environmental conditions and management practices. Figure 4 represents various soil types’ pesticide degradation and absorption.

4.3.1. Physical and Chemical Characteristics of Target Pesticide

The molecular structure of the target pesticide predominantly determines its physical and chemical properties. Any alteration in the molecular structure of the pesticide, whether through addition or substitution reactions, can modify its physiochemical properties, including degradability [293]. Even minor structural changes can significantly alter the pesticide’s degradability. However, modifications involving adding polar groups, such as -COOH, -NH₂, and -OH, may facilitate nanomaterial-mediated remediation. The introduction of polar functional groups increases the pesticide’s affinity for adsorption via nanomaterials. For instance, modification of biochar with hydroxyl, carboxyl, or carbonyl enhances adsorption of polar pesticides via hydrogen bonds and electrostatic interactions. Such modifications significantly improve the removal of pesticides such as glyphosate and atrazine [294].

Introducing polar groups like -OH, -COOH, and -NH₂ at the terminal end or along a side chain of the carbon chain or aromatic ring of a pesticide as a functional group can provide sites for pesticide remediation [295]. Conversely, adding halogen or alkyl substitutes can increase resistance to degradation owing to their chemical and physical property-altering capabilities [296]. While halogen atoms form stronger and less reactive bonds that stabilize the compound against microbial and chemical breakdown, alkyl groups increase hydrophobicity and reduce water solubility, thereby enhancing resistance to degradation [297,298]. Chlorinated hydrocarbons such as DDT, pentalene, and dieldrin are insoluble in water and tightly adsorb to the soil, thus remaining relatively unavailable for biodegradation [299]. In contrast, the insecticide carbofuran and herbicide 2,4-D, despite their differing molecular structures, can be degraded within a few days in field soils [300]. Even minor differences in the position or nature of substituents in pesticides of the same class can influence the degradation rate [301]. These structural modifications govern the solubility characteristics of the pesticides and thus influence their degradability [296]. For instance, tebuconazole and chloropyridine, though both classified as triazole fungicides, differ significantly in their chemical structures and that has a significant impact. In the same environmental conditions, a study found that tebuconazole degraded at a much faster rate, i.e., 72.47–80.27% compared to chloropyridine, which was 47.76–64.82%. This difference highlights how variations in structure can influence a compound’s solubility and chemical reactivity, ultimately affecting how easily it breaks down in the environment [302].

Additionally, the degradation rates generally decrease proportionately to the residual pesticide content, with the kinetics of many pesticides following a first-order pattern (Topp et al., 1997) [301]. Gupta and Gajbhiye [303] reported that flufenacet’s half-life in inseptisols, vertisols, and ultisols ranged from 10.1 to 31.0 days at low application rates (1.0 µg/g) and from 13.0 to 29.2 days at high application rates (10.0 µg/g). Yu et al. (2003) [304] found that butachlor’s half-life in non-rhizosphere soils ranged from 6.3 to 18.0 days at 1.0 mg/kg, from 2.9 to 19.9 days at 10.0 mg/kg, and from 10.8 to 23.2 days at 100.0 mg/kg, indicating that butachlor degradation is influenced by both application rate and soil type [304].

4.3.2. Physical and Chemical Characteristics of Contaminated Soil

The soil type plays a pivotal role in nanomaterial-mediated pesticide remediation, providing the environmental context for this process. Soils exhibit various properties such as organic matter concentration, pH, and others [303], which influence the adsorption of pesticides onto soil surfaces, limiting their availability and enhancing their persistence. The soil’s clay content, organic matter concentration, and pH also directly influence the soil microflora, which contributes to pesticide remediation [305]. Therefore, it is crucial to investigate the rate of pesticide degradation in various soil environments. Gold et al. (1996) [306] studied the impact of soil pH and clay content on the stability of several pesticides, including bifenthrin, chlorpyrifos, cypermethrin, fenvalerate, permethrin, and isofenphos, under field conditions. The half-lives of these pesticides varied significantly under different field conditions. For instance, rinsulfuron exhibited a half-life ranging from 5.6 days in sandy clay loam soil in the United States [307] to 120 days in light sandy soil in Denmark [308].

Similarly, Jones and Ananyeva (2001) [308] reported different metalaxyl and propachlor degradation rates in various soil types, such as pasture, arable, and pine forest soil. The half-life of metalaxyl was 10, 19, and 36 days in these soils, respectively, while for propachlor, the rates of degradation were 2.6, 6.1, and 8.2 days, respectively [303,308]. The rate of degradation of flufenacet varied significantly with soil type, with the half-life ranging from 10.1 to 22.3 days in weakly developed soil and from 10.5 to 24.1 days in red soil (iron-rich soil) and black soil (vertisol). Additionally, Hafez and Thiemann (2003) [309] reported that the degradation rates of imidacloprid and diazinon were higher in salty loam soil compared to sandy loam and sandy soil. The degradation rates of other pesticides, such as pencycuron, were higher in coastal saline soil than alluvial soil and soil amended with decomposed cow manure [310].

Pesticide degradation tends to be sluggish in dry soils, with degradation rates generally increasing with higher water content. For example, phorate exhibited greater persistence in flooded soil than in non-flooded soil [311]. The herbicides atrazine and trifluralin were found to disappear more rapidly under moist conditions compared to dry conditions. Insecticides like BHC can persist for several years in dry soils but may undergo partial biodegradation in moist soils; however, high organic matter content can impede degradation. DDT demonstrates relative stability in dry soils but undergoes rapid degradation to DDD in submerged or moist soils [301]. Soil temperature also alters adsorption by changing the solubility and hydrolysis of pesticides in soil [312,313]. As the adsorption process is exothermic and desorption is endothermic, adsorption is anticipated to decrease as temperature increases, leading to higher pesticide solubility. Perucci et al. (1999) [314] observed that in a clay loam soil incubated at 75% humidity, the half-life of rimsulfuron varied from 14.8 days at 10 °C to 3.5 days at 25 °C. Nanomaterials primarily degrade pesticides through adsorption or catalysis. Higher temperatures notably impact the adsorption of pesticides on the surface of nanomaterials. However, although increased temperatures may not be conducive to adsorption, they can enhance the catalytic degradation of pesticides, as catalytic activity tends to increase with temperature up to a certain threshold.

Soil pH can influence pesticide adsorption and abiotic and biotic degradation processes [312]. The impact of soil pH on the degradation of a specific pesticide largely depends on whether the compound is susceptible to alkaline- or acid-catalyzed hydrolysis [313]. Rimsulfuron, a selective herbicide for residual control of grass and weed, undergoes hydrolysis rapidly in soil under conditions of high temperature [315]. Additionally, the preliminary report suggests that saline soil may facilitate pesticide remediation. Hafez and Thiemann (2003) [309] noted that pesticide degradation occurred more swiftly in silty loam soil compared to sandy and sandy loam soil. A laboratory-based study illustrated that the half-lives of diazinon and imidacloprid in salty loam soil were 4.15 and 9.90 weeks, respectively. In contrast, in sandy loam soil, they were 7.53 and 12.16 weeks; in sandy soil, they were 10.34 and 12.60 weeks.

4.4. Mechanism of Nanomaterial-Mediated Pollutant Remediation

Nanomaterials possess unique physicochemical properties that contribute significantly to the reduction of agricultural contaminants. These nanomaterials primarily degrade agricultural pollutants through chemical reduction, catalysis, and adsorption mechanisms (Figure 3) [316]. The small size of nanoparticles confers advantages in terms of their ability to penetrate small spaces and disperse more rapidly than larger particles, enhancing their efficacy in degrading agricultural contaminants within complex soil matrices [317]. Various nanomaterials have been employed for the remediation of agricultural pollution, including zeolites, noble metals (e.g., copper, palladium, silver, gold), metal oxides (e.g., Fe₃O₄, TiO₂, CuO), carbon nanotubes, nanofibers, and enzymes [9,176,246,318,319]. These materials exhibit high surface area-to-volume ratios and unique surface chemistries, which contribute to their enhanced reactivity and adsorption capacities compared to their bulk counterparts.

Adsorption

In the adsorption process, contaminants (adsorbates) bind to the surface of solid materials known as adsorbents. Among various adsorbents reported in the literature, metal oxides have demonstrated excellent performance in pesticide remediation due to their large surface area for adsorption [320]. The interaction between nanomaterials and pesticides generally occurs through two mechanisms: physisorption and chemisorption.

Physisorption is generally controlled by van der Waals forces, dipole–dipole interactions, and London dispersion forces [321]. These types of process are reversible and has low adsorption enthalpy. For instance, a common herbicide, 2,4-dichlorophenoxyacetic acid, is removed via physisorption onto SBA-15-templated mesoporous carbon [300]. This adsorption is primarily driven by van der Waals forces at low pH (acidic) values. Another key example of physisorption is the utilization of carbon nanofibers to remove organophosphate pesticides, which is governed by interactions such as π-π stacking, hydrophobic interactions, and van der Waals forces [322]. The activation energy required for physisorption is very low compared to chemisorption [323]. Physical adsorption is generally favorable at low temperatures, as higher temperatures increase the molecular movement of the compound, which is unfavorable for physical adsorption [324]. The physical adsorption process is generally based on weak interactions between the functionalized nanomaterials and pesticides. Weak physical interactions or adsorption are inversely proportional to temperature and exhibit low selectivity for the template [325].

Chemisorption is based on irreversible chemical processes between the adsorbent and the adsorbate, which are controlled by the formation of chemical bonds such as covalent bonds, chelation, complex formation, proton displacement, and redox reactions [325]. Chemical interactions between functionalized nanoparticles and pesticides are strong due to the presence of chemical bonds between them. As chemical adsorption depends on the formation of chemical bonds between the functionalized nanomaterial and the pesticide, the requirement for bond formation increases the adsorption enthalpy of the process and also makes it irreversible in nature. For example, adsorption of deltamethrin onto modified bentonite includes valency forces due to sharing or exchange of electrons [326]. Gold nanospheres and nanorods have also shown efficient removal of organophosphate pesticide dimethoate via formation of a chemical bond between the pesticide molecule and the thiol group present on the nanoparticle surface [327]. Chemisorption generally requires high activation energy; hence, high temperatures favor the chemisorption process.

Another significant adsorption mechanism for agricultural pollutants is ion exchange, owing to the ionic nature of the pollutant. For instance, acid-modified corncob and other agricultural by-products show adsorption capacities for metal ions such as Cd²⁺ and Cu²⁺ via ion exchange, with carboxylic and other functional groups added during the modification process [328]. In addition to ion exchange, hydrogen bonding is an important adsorption process for organic pollutants, for example, pesticides containing functional groups such as -OH, -NH₂ and -COOH, which are capable of forming hydrogen bonds. Adsorption of bisphenol A onto agricultural waste-derived adsorbents showed that hydrogen bonding along with hydrophobic and π–π interaction play crucial roles in the adsorption process [329].

5. Protein-Based Biomaterials for Sustainable Agriculture Pollutant Mitigation

Biopolymers, e.g., chitosan, tannin, and cellulose, are increasingly favored across numerous disciplines due to their biodegradable nature, low toxicity, and biocompatibility, distinguishing them from environmentally harmful and non-biodegradable synthetic polymers (e.g., polyethylene, polypropylene, and polystyrene) [330]. Commercially established and extensively adopted examples include silk mimetic fibers [331], hyaluronic acid [332], polyhydroxyalkanonates [333], and keratin [334]. The remediation mechanism of synthetic polymers generally relies on adsorption, encapsulation of beneficial microorganisms, or controlled release of fertilizers and pesticides [335].

Polypeptides, or proteins, are naturally occurring polymers forming an amorphous structure [336]. Protein-based polymers are gaining prominence for their applications in agricultural industries. The functional properties of protein-based polymers are predominantly influenced by their structural, thermal, and hydrophilic characteristics [336]. In the agricultural sector, protein-based polymers find utility in various applications, including leveraging their effective moisture diffusivity to palletize entomopathogenic nematodes in producing granular biopesticides [337]. Additionally, protein polymers are employed in seed coating and encapsulating plant-beneficial microbes (PBMs) to enhance plant defense mechanisms [338]. Protein-based polymers also find application in encapsulating biocontrol agents, indirectly aiding in reducing chemical pesticide utilization, thereby decreasing agricultural pollution [339].

Furthermore, these polymers can adsorb pollutants, increasing the soil’s structure and properties [340]. Beyond these applications, protein-based polymers such as zein and soy are extensively employed in the formation of nanoparticles due to their ability to undergo phase separation and self-assembly. In the coacervation, i.e., phase separation, method, a protein polymer solution is mixed with a solvent, and the addition of salt or adjustment of pH disrupts the equilibrium, causing the proteins to aggregate and form nanoparticles [341]. Alternatively, in the self-assembly approach, protein molecules spontaneously organize into nanoscale structures under controlled conditions such as pH shifts, solvent exchange, or temperature changes [341]. Additionally, techniques like spray-drying and emulsification can be applied to dissolved proteins to produce nanoparticles suitable for agricultural applications [342]. Protein-based nanoparticles have shown significant utility in biomedical, bioimaging, and drug delivery applications, as well as in agriculture, serving as the foundation for nanocarriers and encapsulation materials for microbes and biocontrol agents [343]. This is primarily due to their ease of processing and ability to be modified to meet specific requirements [341]. The versatility of protein-based nanoparticles proves beneficial, given their properties such as biodegradability, minute size, and biocompatibility, making them applicable across various industries, including agriculture [341,344]. The nascent development of protein-based nanoparticles holds promise in addressing challenges such as nutrient-depleted or arid soil conditions and the controlled release of insecticides or plant-beneficial microbes (PMBs) into the soil [345,346]. Protein-based nanoparticles offer advantages by enabling the design of a selective release system, where factors such as pH and enzyme activity can trigger controlled release. For instance, soy protein nanoparticles encapsulating active agents such as curcumin remain stable at neutral pH. However, a slight variation in soil pH to either acidic or alkaline leads to release of the cargo in the soil [342]. Additionally, these nanoparticles can be tailored to encapsulate and neutralize a specific compound, such as curcumin, anthocyanin-rich extracts, and folic acid [347], within the protein matrix, shielded by a polysaccharide or lipid shell. For example, curcumin is first complexed with soy protein isolate to form a protein core, followed by coating with soy-soluble polysaccharides [348]. This design ultimately disrupts the matrix, releasing the content to deactivate or break down the target compound [349].

In addition to protein polymers commonly utilized in agriculture, the integration of ultra-high-molecular-weight (UHMW) protein polymers, exemplified by titin, offers a robust foundation for applications in the agricultural industry. These high-performance fibers with exceptional mechanical properties exhibit high strength and toughness, showing high potential for application in agriculture for activities requiring durable and biodegradable polymer materials [28]. As a UHMW protein, titin demonstrates heightened resistance to degradation owing to its intricate structure, a feature not commonly found in smaller proteins [27,28]. The complexity of the titin protein encompasses distinct bands with diverse functionalities. For instance, the immunoglobulin region of the band (I-band) imparts elasticity, a pivotal factor in sustaining the controlled release of nutrients or PBMs [350,351]. This elasticity also contributes to improved soil structure and enhanced nutrient uptake by enabling these protein-based polymers to form a cohesive network structure within the soil, which aids in binding of soil particles together and acts as a bridge between particles [352]. Similarly, the myomesin band (M-band) within the titin protein provides structural integrity, flexibility, and stability essential for nanoparticles. The unique combination of these bands within titin’s complex structure enhances the overall performance of the protein-based nanoparticles, making them promising candidates for applications aimed at fortifying soil quality, nutrient delivery, and sustainable agricultural practices [28,350].

UHMW proteins, such as titin, exhibit a substantial size and molecular weight, characterized by an abundance of repetitive sequences. The mechanical properties of titin protein polymers, notably high tensile strength and toughness [28], significantly contribute to the enhanced physical characteristics of nanoparticles synthesized using these UHMW protein-based polymers. These attributes prove advantageous in conferring resistance to mechanical and environmental stresses and safeguarding encapsulated compounds or microbes. Titin polymer, in the first place, is a biocompatible and biodegradable compound for synthesizing nanoparticles that are not ecologically toxic and sustainable [28]. The emerging field of UHMW protein polymers, exemplified by titin, further expands the possibilities by offering robust mechanical properties that enhance the durability and performance of nanoparticles, and opening new avenues for sustainable agricultural practices. The development and application of titin protein nanoparticles hold significant potential for fortifying soil quality, facilitating nutrient delivery, and promoting sustainable farming practices.

The high-strength UHMW protein polymers are expected to be utilized primarily for encapsulating microorganisms, biocontrol agents, and nutrients. Microbes known for degrading pesticides through enzymatic or nonenzymatic reactions are encapsulated within semipermeable polymer microcapsules [353]. These capsules are designed to prevent the release of nutrients, microbes, or control agents until the desired environmental conditions are present to facilitate controlled release [340]. Upon release, these microbes facilitate enzymatic and nonenzymatic degradation of chemicals in the surrounding environment through physiological, biochemical, mineralization, or co-metabolism pathways [353]. These biomaterials utilize mechanisms such as diffusion and swelling, degradation-based release, or stimuli-responsive release systems [354]. In the diffusion and swelling mechanism, the polymer matrix swells, causing the enlargement of pores that facilitate the diffusion. The concentration gradient and the diffusivity of nutrients within the matrix govern the release rate. For instance, poly(ε-caprolactone) (PCL) is used as a polymer matrix for controlled nutrient release, where the concentration gradient of the encapsulated nutrient and diffusivity within the PCL matrix are key determinants of the release rate upon swelling of the polymer [355]. In the degradation-based release mechanism, the polymer matrix undergoes enzymatic or hydrolytic degradation, resulting in the formation of larger pores and channels that enhance the diffusivity of the encapsulated material and release. For example, whey protein is considered a natural polymer for encapsulation and proteinoid polymers show efficient self-assembly into nano-capsules for controlled release of agrochemicals [356,357]. Some protein-based polymers are engineered to respond to stimuli, such as changes in pH, temperature, or the presence of specific enzymes. In these systems, the matrix undergoes a phase transition in response to stimuli, leading to the controlled release of the encapsulate into the environment.

Like metallic and non-metallic nanoparticles, protein polymers can also remove heavy metals such as cadmium and lead since the functional group present in the protein can potentially adsorb these pollutants. Along with heavy metal contaminants, organic pollutants such as pesticides and herbicides can also be removed owing to their interaction with hydrophobic regions of proteins [9].

6. In Silico-Driven Remediation Model

With the demonstrated capabilities of nanoparticles to adsorb agropollutants, numerous papers have explored the mechanisms of their adsorption at the atomic level [12]. DFT and molecular dynamics (MD) simulations have been at the forefront, describing the atomistic interaction and molecular mechanisms of remediation by nanoparticles [12,13,14]. Rational design, aided by MD simulations, has been instrumental in designing functionalized magnetic nanoparticles for the adsorption of weakly polar pesticides from human serum, where van der Waals and electrostatic interactions play pivotal roles [14]. Nanoparticles, like fullerenes, rely solely on π–π, van der Waals, and hydrophobic interactions for adsorption, while some nanoparticles, such as fullerol-8, further interact with the help of hydrogen bonding (Figure 5a) [358].

Not limited to the surface properties, the motifs present in the pesticides can also contribute to nanoparticle adsorption. The phosphonic acid and carboxylic acid motifs present in glyphosate, a herbicide, promote efficient and strong adsorption of glyphosate on magnetite nanoparticles via covalent bonding [360]. Furthermore, indigenous bacterial genera such as Achromobacter, Pseudomonas, and Arthrobacter have demonstrated the ability to degrade glyphosate, primarily through two key enzymes: glyphosate oxidoreductase and C–P lyase. GOX breaks glyphosate’s C–N bond, forming intermediate metabolites like aminomethylphosphonic acid (AMPA) and glyoxylate [361]. C–P lyase, on the other hand, cleaves the C–P bond, producing phosphate and sarcosine, facilitating the degradation of both glyphosate and AMPA. The active-site amino acids interact with glyphosate’s C–O and C–N bonds, playing a critical role in C–N bond cleavage. Specifically, with its OH group, serine likely performs catalysis through nucleophilic attack [362].

Ag nanoparticles can also be used as pesticides, reducing the harmful impacts of traditional pesticide use. Ag nanoparticles have been found to interact with the endo-β-1,4-glucanases and xylanase enzyme in the gut of termites, impairing the digestive system and causing toxic impacts on the termites [220]. Molecular docking studies revealed that Ag nanoparticles can interact with enzymes near the catalytic and Ca²⁺-binding domains, affecting the yields and kinetics for cellulose degradation, leading to digestive dysfunction (Figure 5c) [220]. Furthermore, integrating this strategy into a pest management program can effectively decrease pesticide reliance and mitigate environmental risks.

Not confined to mere adsorption, functionalized nanoparticles offer additional advantages by catalyzing the breakdown of pollutants [363,364,365]. Enzymes play a pivotal role in transforming pesticides, inducing structural modifications to the molecules, such as altering reactive groups and removing electrons or protons, thereby mitigating the toxic effects of pesticides [366]. For instance, nanoparticles functionalized with laccase have demonstrated an ability to degrade both pesticides and pollutants such as cefixime antibiotics and safranin O dye [363,365,367]. Laccase, a multicopper oxidase, catalyzes various compounds, including phenolic compounds, aromatic amines, diamines, nonphenolic compounds, and heterocyclic compounds [368]. Laccase binds to pesticides such as glyphosate and parathion by forming hydrophobic interactions, hydrogen bonds, and van der Waals forces. The predicted active site, comprising Asn217, Val247, Glu249, Gln237, Tyr244, and Arg423, forms a stable complex with laccase-glyphosate/parathion, with predicted binding energies of −6.21 and −7.37 kcal/mol⁻¹, respectively (Figure 5b) [359]. In addition to these active sites, ionization energy and radical heat of formation serve as crucial indicators that determine the laccase mediator activity of phenolic compounds [369].

Nanozymes are nanomaterials with enzyme-like properties, and they have been considered as robust and cost-effective alternatives in environmental applications for pollutant degradation [370]. They exhibit diverse catalytic activities, mimicking natural enzymes such as peroxidase, oxidase, and hydrolase to break down pollutants efficiently [16]. The degradation mechanism involves pollutant and water adsorption onto the nanozyme surface, where catalytic sites facilitate bond cleavage. For instance DFT simulation of hydrolase-like (HYL) nanozymes showed water molecules decompose into hydroxyl radicals, which attack the pollutant’s molecular bonds, breaking them into smaller, non-toxic products [371]. Catalase-like (CAT) nanozymes catalyze the conversion of hydrogen peroxide (H₂O₂) into water and oxygen through a homolytic or heterolytic cleavage of the O-O bond. Such processes can be explained by molecular dynamics simulations that model atomic interactions as functions of time [372]. The design of nanozymes involves leveraging both experimental and computational approaches to optimize their catalytic activity and specificity. Computational tools like DFT and MD simulations are employed to study elementary reactions, energy barriers, and thermodynamics, allowing for accurate prediction of catalytic activity under various conditions [16]. These approaches facilitate high-throughput screening and rational design by incorporating activity descriptors such as adsorption energy and electron occupancy, enabling the development of highly efficient nanozymes tailored for specific applications. For instance, for development of antimicrobial peptide nanozymes, MD and DFT were employed to identify and design a minimal peptide building block that combines essential amino acids from natural antimicrobial peptides and enzymes with hydrophobic isoleucine residues to promote assembly into a functional nanozyme [373]. These identify optimal reaction pathways, analyze energy barriers, and predict how nanozymes will interact with pollutants and substrates under varying environmental conditions, such as changes in pH or temperature [374]. In addition, they show how structural features, such as surface modifications or oxygen vacancies, enhance catalytic efficiency. This approach provides deep insight into nanozyme functionality, thus allowing for their rational design for pollutant degradation [374].

Beyond adsorption, the catalytic potential of functionalized nanoparticles (e.g., Ag nanoparticles functionalized with laccase or glyphosate oxidoreductase) and nanozymes in pesticide breakdown present promising avenues for future environmental solutions. These findings, backed by molecular insights and machine learning predictions, not only deepen our understanding but also hold key implications for designing effective and environmentally conscious strategies in pesticide remediation.

7. Machine Learning-Based Toxicity Prediction

Predictive models, including ML models, are a subdomain of artificial intelligence (AI) that utilizes mathematical algorithms to learn from data and make predictions. Most algorithms used for developing predictive models for toxicity prediction encompass random forest (RF), support vector machines (SVMs), deep learning, and deep neural networks [375]. As most ML models are data-driven, high-quality data is essential to ensure model accuracy, reliability, and the interpretability of the crucial parameters. Most ML-based models are developed using literature, publication, or open-source data mining. However, data mining in this manner encounters inherent problems with poor data quality and missing data, as the model learns from empirical data and hidden patterns. Data used for toxicity prediction typically contains features such as structural/chemical properties and toxicity data.

Emerging pesticides (e.g., neonicotiniods, broad-spectrum insecticides) and chemicals (e.g., chlorothalonil, a broad-spectrum fungicide) are integral to various human activities. Over 144 million chemical substances and sequences have been registered in the Chemical Abstracts Service (CAS). Moreover, approximately 12,000 new substances are added daily [376]. With the increasing number of synthesized chemicals, the National Institute of Health conducted a worldwide competition in 2014 to develop toxicity prediction models, called the Tox21 Data Challenge [377]. The aim was to predict the toxicity of chemicals using structural data [377]. Compared to traditional environmental toxicology analysis, predictive models can provide accurate predictions and are less expensive than quantum chemical calculations [376]. In light of the advantages offered by predictive models in environmental toxicology analysis, there is a pressing need to enhance the interpretability of ML models for pesticide toxicity prediction. These necessitating methodologies can effectively leverage available experimental data.

A quantitative structure–activity relationship (QSAR) is a statistical approach widely used in toxicity prediction models to establish a relationship between chemical structure and biological activity such as toxicity [378]. Hence, two main methods are typically employed to convert the chemical structure into a numeric vector recognized by a machine: molecular descriptors and molecular fingerprints. Molecular descriptors, representing local and global salient characteristics of chemical structure, include physicochemical characteristics such as molecular weight, atom count, and conformational surface charge distribution [379], and can be calculated using tools such as Dragon, RDKit, Chemistry Development Kit (CDK), PaDEL-Descriptor, and Mordred. On the other hand, molecular fingerprints represent the presence or absence of specific substructures in molecules. They are commonly provided by tools such as molecular access systems (MACCS), extended-connectivity fingerprints (ECPs) and PubChem fingerprints. Nevertheless, constitutive descriptors representing counts of atoms, bonds, and functional groups, and geometrical descriptors representing 3D features, e.g., molecular shape and size, are also widely used for extracting hidden patterns within chemical structures [380].

Along with the tools for feature extraction and generation, several databases have been built to support the development of toxicity predictive models. Tox21 [381] and ChEMBL [382] are widely used toxicity databases. Furthermore, the comparable dose and toxicity potential can be extrapolated using quantitative results from chemical concentration–response curves [383].

The ML model aims to learn by mapping the features to toxicity labels, i.e., supervised learning, or by determining patterns or clusters in unlabeled data. For instance, linear regression models predict the target as a weighted sum of features with a minimum root mean square error. However, linear regression models are best suited for mapping linear relationships [384]. The SVM can be trained for linear and non-linear classification by projecting features in higher dimensions using kernel tricks such that a hyperplane separates the chemicals and their toxicity labels [385]. The SVM has been widely used for binary classification problems and for predicting inflammation, hepatocyte toxicity, carcinogenicity prediction, and more [386,387,388].

On the other hand, non-parametric-based learning, e.g., RF, combines decision trees to generate one prediction [389]. In essence, decision trees make predictions based on the decision path formed by graphs representing tests on features (root node) to denote multiple true/false questions in tree-like structures, where the final decision corresponds to the class label (leaf nodes). RF has been used to predict toxicity in various contexts, such as acute toxicity, aquatic toxicity, estrogen receptor activity, and more [390,391,392].

Nevertheless, these models can estimate feature importance and contributions in the model. RF revealed that the physical properties of pesticides, particularly dissociation constant, molecular weight, and water solubility, play a significant role in determining the ecological impacts of the pesticide [393]. Furthermore, toxic compounds tend to be hydrophobic [391].

Although ML models are widely used daily, a common limitation of ML models is their “black box” nature, indicating that users cannot easily understand the inner working mechanism. For example, projecting the original data in a hyperplane can make the model less interpretable [385] when dealing with non-linear relationships in support vector machines. Similarly, an ensemble decision tree, such as random forest, can be difficult to interpret [389]. Moreover, these black box models can lead to misinterpretation of results [394]. This opacity obstructs identifying pivotal features necessary for regulatory and safety assessments. To tackle this challenge, the present review explores methodologies aimed at improving the interpretability of ML models in predicting pesticide toxicity. A recent development in visual networks utilizes the integration of pathway hierarchy in design for interpretable models [395,396,397,398]. This network uses genes or proteins as inputs connected to specific neurons representing their associated pathway, subsequently connecting to the parent pathway and making the model interpretable. An interpretable ML model for pesticides would not only elucidate the importance of the key features but also shed light on the most vulnerable pathways, aiding in decision making for selecting the least toxic pesticides. For example, in a recent study, replacing the nitrile group in Cyantraniliprole, a compound originally toxic to honeybees, with a methyl group significantly altered its electronegativity and reactivity, while preserving the key benzamide-pyrazole scaffold; this modification was predicted by the model to reduce honeybee toxicity (changing the output from “toxic” to “non-toxic”), likely due to reduced formation of cyanide or other reactive intermediates [399]. By understanding the pathways most affected by pesticide exposure, stakeholders can make informed decisions and alter chemical structures to mitigate potential risks for humans and the environment by prioritizing the use of safe alternatives.

8. Future Prospects

This review discusses the various nanomaterial-based methods for removing agricultural pollutants from complex environments such as soil. The field of pesticide removal from soil environments is still in the development phase at the laboratory scale. There are various physical and chemical methods, e.g., the use of polymeric material such as cyclodextrins, dendrimers, hyper-cross-linked polymers, advanced oxidation processes, UV-H₂O₂, UV-ozone, zero-valent iron, Fenton reactions, photocatalysis, photo-degradation, and ultrasound-assisted remediation, that are in the development stage.

Significant efforts toward developing agricultural pollutant remedies using nanoparticles are anticipated in scaling up these techniques from lab-scale to field-scale applications, ensuring that the methods are cost-effective and seamlessly integrate with existing agricultural practices. Nanoparticle advancements are essential to improve their efficiency and selectivity for different pollutants. Precise optimization of oxidation processes in conjunction with nanoparticles can further enhance pollutant degradation. Photocatalysis and photo-degradation methods will need refinement to increase their effectiveness under diverse environmental conditions. At the same time, ultrasound-assisted remediation offers an emerging approach that requires optimization of ultrasound parameters for maximum pollutant removal with minimal energy consumption. Ensuring the environmental impact and safety of nanomaterials, alongside addressing regulatory and economic challenges, will be crucial for successful implementation. Interdisciplinary collaboration will drive innovation, emphasizing sustainable practices by developing environmentally friendly and resource-efficient methods. Consequently, there is considerable potential for advancement in nanoparticle-based pesticide removal from soil, with research paving the way for innovative and effective large-scale solutions.

The utilization of UHMW biopolymers is a promising development towards mitigating agricultural pollutants. Specific characteristics, including exceptional durability, strength, and elasticity, can be utilized to address various challenges such as chemical resistance for nanoparticles and encapsulators. Using higher-strength biopolymers as coating for the seeds enables the controlled release of nutrients, thereby preventing leaching. The emerging field of UHMW protein polymers, exemplified by titin, further expands the possibilities by offering robust mechanical properties that enhance the durability and performance of nanoparticles, and opening new avenues for sustainable agricultural practices. The development and application of titin protein nanoparticles holds significant potential for fortifying soil quality, facilitating nutrient delivery, and promoting sustainable farming practices. With the further advancement of technology, nanomaterials or smart responsive elements can be integrated, enabling more precise and efficient pollutant capture and degradation.

Incorporating ML to predict pesticide toxicity has great potential to advance sustainable agriculture. ML is improving at identifying intricate patterns in high-dimensional toxicity data due to the ongoing advancements in deep learning architectures. Enhanced explainable AI techniques will help stakeholders gain more confidence by making it easier to comprehend how these models make decisions. Incorporating multi-omics data, including genomics, transcriptomics, proteomics, and metabolomics, offers a comprehensive view of the biological impacts of pesticides, potentially leading to more accurate predictions. Improved data sharing regulations and the growth of open databases like ToxCast and PubChem will greatly increase the quality and quantity of data available for model validation and training. The application of ML to environmental toxicology prediction will hopefully bridge the knowledge gap between academia and industry by providing insightful information on environmental toxicity assessment. By addressing current challenges such as data scarcity and model interpretability, the agricultural sector can leverage ML to move toward safer and more sustainable pest management practices, ultimately reducing pesticide use’s environmental and health impacts. In the future, molecular dynamics simulation-based studies can be utilized to evaluate the potential of various enzymes to biodegrade diverse pesticides, paving the way for more eco-friendly pest control solutions. Future research should also focus on combining wet laboratory experiments with in silico studies to understand the effects of enzyme-based biodegradation better.

9. Conclusions

Addressing the issue of agricultural pollutants is essential for sustainable development and food security in a growing population. The available nanomaterials present promising avenues for tackling agricultural pollutants, such as pesticides, chlorinated hydrocarbons, and heavy metals, which pose significant threats to ecosystems and human health. Remediating pesticides in soil through nanoparticle-mediated processes involves a combination of factors, including the physiochemical characteristics of the pesticide, soil properties, environmental conditions, and management practices. Structural modifications in pesticides can greatly influence their degradability, with the addition of polar groups potentially enhancing nanoparticle-mediated remediation. Furthermore, pesticide concentration, solubility, soil type, moisture content, temperature, pH, and salinity are all crucial in determining the effectiveness of nano-remediation strategies. Additionally, integrating ultra-high-molecular-weight protein polymers, such as titin, provides promising prospects for improving nanoparticle durability and performance in agricultural applications. Insights at the molecular level and predictions from machine learning further aid in understanding and developing effective pesticide remediation strategies for sustainable farming practices. Despite ongoing development and research in this area, additional exploration and refinement of nanoparticle-mediated remediation methods are still needed to address agricultural pollution’s challenges effectively.