Development of a Hybrid Nanosensor Using Eruca sativa Leaf Extract Reinforced with Graphene Oxide (GO) for Solid-Phase Microextraction (SPME) of Mercury and Arsenic in Industrial Wastewater from Baghdad

INTRODUCTION
The problem of heavy metal pollution of aquatic ecosystems, especially those caused by industrial effluents, is a severe issue on the health and environment of the urban centers of Iraq and Baghdad among them [1]. The most dangerous ones are mercury and arsenic as they are persistent, bioaccumulative, and even toxic at traces [2]. The WHO provides maximum allowed concentrations of Hg and As in drinking water 6 µg/L and 10 µg/L, respectively [3]. Nevertheless, the industrial sectors in Baghdad (e.g., Al-Rustamiya (chemical plants, pharmaceutical plants), Al-Dora (oil refining, petrochemicals plants)) periodically emit untreated or partially treated wastewater with the concentration of Hg and As that are above these values [4,5].
Traditional methods of analysis such as the inductively coupled plasma mass spectrometry (ICP-MS) and graphite furnace atomic absorption spectrometry (GFAAS) have high sensitivity but demand expensive tools, trained experts, and laboratory facilities - all of which are inaccessible in low-resource areas [6]. Solid-phase microextraction (SPME) has become a promising sampling preparation method that combines extraction, pre-concentration and cleanup into a single procedure and has become sensitive to detection and reduced solvent usage [7].
Graphene oxide (GO), with large surface area, oxygen functional groups, and controllable surface chemistry is commonly employed in SPME [8]. Nonetheless, its predilection to aggregation and selectivity restrict its usefulness to a surface modification. An environmentally friendly method to stabilize and functionalize GO with phytochemicals (e.g., flavonoids, phenolics) as capping/reducing agents involves a green synthesis through plant extracts [9]. Eruca sativa (rocket or arugula), with its sulfur- and nitrogen-containing compounds that are known to chelate heavy metals [10] is yet to be utilized in nanosensor creation to detect Hg/As.
This paper fills this gap by designing a new GO -E. sativa nanohybrid to SPME Hg(II) and As(III) in industrial wastewater in Baghdad. The article contains full material description, optimization of methods, analytical validation, application to real samples, selectivity test, and reusability test- which complies with the analytical rigor. 

MATERIALS AND METHODS
Materials and Reagents
Graphite (99.99%), potassium permanganate (KMnO4), sulfuric acid (H2SO4, 98% ), hydrogen peroxide (H2O2, 30%), mercury(ii) nitrate [Hg(NO3)2] and arsenic(iii) oxide (As2O3) were bought at Sigma-Aldrich (Germany). The rest of the chemicals were of analytical grade. Ultrapure water (18.2 M O cm) was obtained as per Milli-Q system (Merck, Germany). The leaves of Eruca sativa were obtained at the local farms at the Wasit Governorate (April 2025), botanically authenticated, washed, dried, and stored at 4 °C.

Instrumentation
The this is atomic absorption spectrometer (AAS, Shimadzu AA-7000) with hydride generation to absorb As, and cold vapor to absorb Hg.
SEM (scanning electron microscopy) (FEI Quanta 250): TEM (transmission electron microscopy) (JEOL JEM-2100).
Fourier-transform infrared spectrometer (FTIR, Thermo Nicolet iS10).
• X-ray diffractometer (XRD, Bruker D8 Advance).
• Bet surface area analyzer (Micromeritics ASAP 2020).
pH meter (Mettler Toledo SevenCompact).

GO -E. sativa Nanohybrid synthesis
GO was ready through the procedure of Hummers with modifications [11]. In short, 2 g of graphite and 1 g of NaNO3 were measured and 46 mL of H2SO4 under ice cooling was added. KMnO4 (6 g) was gradually incorporated and 2 h of stirring was done at 35 °C. A 140 mL of water and 10 mL H2O2 (30 percent) were added to the mixture to form brown GO dispersion.
Preparation of Fresh E. sativa leaf extract Fresh leaf extract was prepared by boiling 10 g dried leaves in 100 mL water using boiled water, and it was then filtered. 20 mL extract was dropwise added to 100 mL GO dispersion (2 mg/mL) stirred at 60 °C, 4h. The black precipitate that was obtained was centrifuged, rinsed with water/ethanol and dried at 60 °C [12].

Characterization
Analysis was performed in terms of morphology (SEM/TEM), functional groups (FTIR: 4000-400 cm-1), crystallinity, and surface area (BET, N2 adsorption).

SPME Procedure
GO -E. sativa (10 mg) was put into 50 mL wastewater sample (pH adjusted). The adsorbent was then separated by centrifugation or with a magnetic field (after doping with Fe3O4, not a compulsory step). The 2 ml of 0.1 M HNO3 was used to desorb the metals (10 min), and the eluent was measured with AAS.

Real Sample Collection
The samples of wastewater (n = 12) were taken in the April-May of 2025 through the outfalls of Al-Rustamiya and Al-Dora (Baghdad). The samples were filtered (in 0.45 µm) and acidified (with HNO3, pH < 2), and kept at 4 °C [13].

RESULTS AND DISCUSSION
Physicochemical Characterization of GO Eruca sativa Nanohybrid.
The scanning electron microscopy (SEM) was initially used to determine the morphological characteristics of the synthesized nanohybrid. The pure graphene oxide (GO) presented a smooth, layered, and lightly wrinkled sheet-like structure (Fig. 1A), which is in line with the morphology that is usually reported by Marcano et al. [11]. After functionalization of the surface with Eruca sativa leaf extract, the surface appeared significantly coarser and covered with nano-sized clusters (Fig. 1B), meaning the successful phytochemical anchoring. The observation can be compared with recent studies by Singh et al. that proved that plant polyphenols are both reducing and capping agents and consequently heterogeneous nucleation on carbonaceous supports [12]. The exfoliation character of the GO sheets was further confirmed by transmission electron microscopy (TEM), which provided uniform deposition of nanoparticles with a diameter of 515 nm (Fig. 2), which is considered to maximize the surface-volume ratio, as well as binding capacity of the metals [13].
Molecular evidence of functionalization was given by Fourier-transform infrared (FTIR) spectroscopy. Typical peaks in the spectrum of pure GO were 1720 cm -1 (carboxyl C = O stretch) and 1220 cm -1 (epoxy C -O stretch) (Fig. 3). These peaks were less intense after E. sativa treatment, but new bands appeared at 1600 cm -1 (N -H bending of amines) and 1050 cm -1 (C -S stretch), which is indicative of the presence of sulfur- and nitrogen-containing phytoconstituents - such as glucosinolates and erucin - in the reduction and stabilization of GO [10,14]. This is in line with the results of Al-Jumaili et al. that described the presence of sulfur-rich metabolites in E. sativa as major chelators of soft metal ions such as Hg2+ [10]. Conversely, Zhang et al. indicated that extracts of polyphenols (e.g. green tea) mostly give C-O and C=O bonds, with no significant C-S bonds formed [14], which is a characteristic chemical signature that E. sativa is able to provide.
The X-ray diffraction (XRD) analysis (Fig. 4) revealed that GO had a high sharpness at 2 theta =10.5° which was a (001) plane with the interlayer spacing of approximately 0.84 nm. Following phyto-functionalization, this peak has disappeared and a hump with a center at 24.2° appeared and this is evidence of a partly reconstituted graphitic structure with decreased interlayer spacing (~0.37 nm) [11,15]. The change proves the loss of oxygen functionalities and the decrease of GO, which is a necessary step to increase the electronic conductivity and adsorption kinetics.
Notably, adsorption-desorption isotherms of nitrogen indicated that the surface area of GO changed drastically as compared to GO to 32 m2/g to 148 m2/g, respectively, of the GO-E. sativa nanohybrid (Table 1). This almost five-fold improvement is explained by the fact that restacking is inhibited with steric hindrance through the presence of bound phytochemicals which has been previously seen with tannin-functionalized graphene systems [16]. Mean pore size also reduced to 11.0 nm, in comparison with 15.2 nm, indicating the creation of mesoporous networks that are good in diffusion and binding of metal ions [17].

Maximization of SPME Conditions.
The pH-dependence of the extraction efficiency of Hg(II) and As(III) was very high. The two analytes adsorbed (maximum of 95 percent) at pH 5.5 (Fig. 5). High proton concentration caused the indicators to compete with metal cations in the binding sites at lower pH (<3), whereas, in high pH (>7), hydroxides started hydrolyzing or precipitating Hg 2 and As(III) to yield available species to adsorb it [18]. Optimal pH is consistent with speciation diagrams in which Hg2+ and neutral H3AsO3 are reactants in this range [15]. This observation is consistent with that of Chen et al., who observed the maximum uptake of As(III) by thiol-modified adsorbents at pH 5-6 [18], however, it is different to those of Wang et al., who found that optimal uptake of Hg occurred at pH 7 because of the addition of amine-rich ligands that prefer deprotonated species [19].
The equilibrium was achieved in 30 min (Fig. 6), much quicker than most reported SPME systems (>60 min) [20], presumably because the hierarchical porosity and high diffusivity of the nanohybrid translate into high diffusivity. It was observed that the best dosage of adsorbent was 10 mg (Table 2); beyond this dosage, agglomeration and lower accessibility to active sites occur, also observed by Li et al. in MOF-based extractions [21].

Analytical Performing and Method Testing.
The Hg and As calibration curves were linear at 1-100 µg/L and had correlation coefficients (R2) of 0.998 and above (Table 3). The minimum detection limits (LOD) were 0.12 µg/L and 0.18 µg/L of Hg and As respectively, which was much lower than the WHO drinking water standards (6 and 10 µg/L of Hg and As respectively) [2]. These LODs are more effective than some of the more recent nanosorbents: e.g., Fe3O4 at SiO2 thiol had an Hg LOD of 0.21 ug/L [19], and MOF at 0.48 ug/L with As [21]. The intra-day precision (RSD < 4.2%) shows a very high reproductive ability, which is acceptable (<5%) when trace metal is analysed in environmental matrices [16].
The spiked real samples provided recovery values that were 96.4 to 103.7 (Table 5), which indicates low interference and high accuracy of the matrix. Such findings are also similar to those achieved by Alwan et al. who used phyto-nanohybrids to detect Hg in the Iraqi river water in the presence of functionalized silica nanoparticles (recoveries: 95102) [5], which support the robustness of phyto-nanohybrids in wastewater.
Real Sample Analysis and Environmental Relevancy.
The Hg levels were found to be between 2.8 and 9.3 ug/L and As between 3.5-11.7 ug/L in 12 samples of effluent of the Al-Rustamiya and Al-Dora districts (Table 4). Interestingly, 8 of 12 samples were over the WHO threshold, which supports previous studies about severe heavy metal contamination in the industrial areas of Baghdad [4,22]. Their maximum was found around pharmaceutical and petrochemical outfalls, which were in line with the established Hg/As-usage in catalysts and preservatives [23]. The results provided bring about the extra necessity of the sophisticated monitoring mechanisms within the regulatory framework of Iraq.

Mechanism of Selectivity, Reusability and Adsorption
The nanohybrid was also found to be highly selective with Hg(II) and As(III) with a 10 times greater excess of normal interferents including Pb2+ and Cd2+ and Cu2+ and Zn2+ (Fig. 7). Said selectivity is explained by the soft acid-soft base principle: As(III) and Hg2+ are soft acids and prefer to bind soft bases such as thiol (-SH) and amine (-NH2) groups that are highly represented in E. sativa phytochemicals [10,24]. Hard acids, such as Ca2+ or Mg2+, did not interact, in contrast. This is in contrast to non-selective carbon nanotube systems which depend on electrostatic interactions majorly and have cation competition [25].
After adsorption-desorption activity, of 0.1 M HNO3 was conducted five times on the adsorbent, the remaining capacity was found to be more than 92% of the original capacity (Fig. 8) which implies a strong structural stability. This is better than reusability of chitosan-based composite which can undergo 3-4 cycles of degradation through scission of chains because of the acid [25].
The Langmuir model (R2 0.99, Table 7) was found to describe the adsorption isotherm data indicating a monolayer cover and homogeneous active sites. The highest adsorption capacities of 89.4mg/g Hg and 76.2mg/g As are one of the best adsorption capacities recorded in green-synthesized SPME materials [10,19]. Spontaneous adsorption (Delta G less than 0) and exothermic adsorption (Delta H less than 0) were confirmed using thermodynamic parameters (Table 7), which is in line with physisorption-dominated processes currently including van der Waals forces and electrostatic adsorption [26].

Comparative Assessment
Compared to the recent literature (Table6), the GO–E. sativa system is better in terms of sensitivity, sustainability, and cost-effectiveness. In contrast to metal-organic frameworks (MOFs) or magnetic composites, which are multi-step synthesized and need ligands which are costly to buy [21,19], we have employed a one-pot, aqueous synthesis with locally sourced plant material, an important factor in the context of low-resource settings such as Iraq.

CONCLUSION
Through green synthesis of Hg and As on industrial wastewater, a new GO-Eruca sativa nanohybrid was obtained successfully to be used in the SPME process. The sensor had a high surface area and also was highly selective and had low LODs and also it was reusable. It was verified on actual samples of Baghdad and showed that there is a lot of contamination, which necessitated more stringent control. This solution corresponds to the principles of green analytical chemistry and provides a scaled system of environmental monitoring in Iraq or other countries.

CONFLICT OF INTEREST
The authors declare that there is no conflict of interests regarding the publication of this manuscript.