Oil Spill Removal from Water Surfaces using Zinc Ferrite Magnetic Nanoparticles as A Sorbent Material

In this study, zinc ferrite magnetic nanoparticles (ZnFe2O4, ZFO MNPs) were employed as a sorbent for the removal of oil spill from water surfaces. ZFO MNPs were synthesized via a sol-gel process and characterized by Fourier transform infrared spectroscopy (FTIR) and X-ray powder diffraction (XRD). Both the apparent density and magnetic force were determined. ZFO MNPs presented a considerable magnetic force (40.22 mN) and an adequate density (0.5287 g/cm 3 ), which are important for the magnetic separation and flotation. Four oil samples (gasoline engine oil, crude oil, used motor oil and diesel engine oil) were used to investigate the gravimetric oil removal capability of ZFO MNPs. The oil sorption capacities were found to be 23.00-6.13, 27.65-7.42, 22.62-7.01 and 30.54-9.93 g/g for crude, diesel engine, gasoline engine, and used motor oil, respectively. The current findings demonstrate that ZFO MNPs exhibit good properties (e.g., magnetic and density) and can be used as a sorbent for oil spill cleaning-up from water surfaces.


Introduction
Oil has become an essential element to our modern industrial society and our daily life. In fact, the worldwide production and consumption of oil and petroleum products (e.g., plastics, chemical feedstocks, and fertilizers) are increasing, and oil pollution risk is increasing consequently [1]. In oil industry, spills or leakages can occur during various steps, including oil exploration, production activities, transportation, and storage [1,2]. It was estimated that about 400.000 tons of oil are spilt annually [3]. For instance, one of the biggest oil spills in human history occurred in 1991 during Kuwait oil fire, at which one billion tons of oil was burnt and poisonous gases were released into the environment [4]. Another oil spill accident occurred in 2010 in the Deep-water of Gulf of Mexico and released approximately 206 million gallons of petroleum products into the sea [4,5]. Oil spills in water bodies (seas, oceans, etc.) and land can threat the environment, as they have detrimental effects on marine life as well as human and animal health [1,2,4]. In disaster management, oil spillage cleanup is one of the most expensive processes. For example, in the USA, from 16 to 20 dollars per gallon is the estimated average cost for spilled oil cleanup [4]. Therefore, there is an urgent need to develop economic, effective, and feasible technologies for oil spills cleaning.

Experimental Part 2.1 Materials and Chemicals
The Libyan crude oil sample was received from Amal oil field, Harouge oil operations, Libya. A used motor oil sample was collected from car oil change shops. Gasoline and diesel engine oils were purchased from local shops. Zinc acetate dihydrate ((CH 3 COO) 2Zn ·2H 2 O, 98-102%) was purchased from VWR Chemicals. Iron (III) nitrate nonahydrate (Fe(NO 3 ) 3 ·9H 2 O, > 98%) was received from Berck and Scientific Supplies. Citric acid (C 6 H 8 O 7 , > 99.5%) was obtained from Labkem. Ammonia solution (35%) was purchased from Scharlau. Ethylenediaminetetraacetic acid (EDTA, C 10 H 18 N 2 O 8 , 99%) was supplied by Serva. All chemicals were used as received, with no purification.

Synthesis of ZnFe 2 O 4 Magnetic Nanoparticles
ZFO MNPs were synthesized using a sol-gel process, as described elsewhere [30]. In brief, calculated amounts of (CH 3 COO) 2 Zn·2H 2 O and Fe(NO 3 ) 3 ·9H 2 O were dissolved in deionized water. Then, EDTA and citric acid were added to the metal nitrates solution as complexing agents with a ratio of citric acid : EDTA : metal cations of 1.5 : 1 : 1. Afterward, a diluted ammonia solution was added to adjust the pH of the mixed solution to around 6. At this pH value, EDTA dissolves completely. The mixture was magnetically stirred and evaporated on a hotplate before being converted into a solid product (dried gel). Finally, the dried gel was calcined in air at 600 °C for 2 h to obtain ZFO MNPs. The sol-gel process used for ZFO MNPs synthesis is represented in Figure-1.

Characterization of ZFO MNPs and oil samples
Philips -PW 1800 X-ray powder diffractometer with Cu-Kα radiation (λ = 1.54186 Å) was used to record the XRD patterns of ZFO MNPs at room temperature. The lattice parameters ( ), unit cell volume ( ), X-ray density ( ), crystallite size (D), and the specific surface area ( ) of the prepared magnetic nanoparticles were estimated using Equations (1) to (5), respectively [31,32]; where d is the interplanar distance, hkl are the Miller indices, Z is the number of molecules per formula unit (Z = 8 for spinel system), N A is the Avogadro's number, M is the molecular weight of the sample, β (in radiance) is the full width at half maximum (FWHM) of the peak, λ is the wavelength of the X-ray, and θ is the Bragg angle.
Nicolet 380 spectrometer was used to perform the FTIR analysis of ZFO MNPs in the wavenumber range of 400-4000 cm -1 using the KBr disc method. Bruker Tensor 27 spectrometer was used to record the FTIR spectra of the oil samples in the wavenumber range of 600-4000 cm -1 using the Attenuated Total Reflectance mode (ATR).
The apparent density ( ) of ZFO MNPs was measured as described elsewhere [33]. In this experiment, the weight and volume of ZFO MNPs inside a 10 mL graduated cylinder were determined. Then, the following equation was used to estimate the apparent density [34]; a (g cm ) (6) where and are the volume and the weight of ZFO MNPs inside a 10 mL graduated cylinder, respectively. The magnetic force ( ) test of ZFO MNPs was carried out using an electromagnet, triple beam balance (700/800 Series, OHAUS®), Teslameter (LEYBOLD, DIDACTIC GMBH), and power supply (Hochstrom-Netzerat, LEYBOLD). The ZFO MNPs was subjected to a magnetic field. The intensity of the magnetic field was increased by applying an electric current ranging from 0 to 0.9 A. Then, the (mN) was estimated using the following equation [35]; (7) where ∆m is the apparent mass variation of ZFO MNPs in the presence of magnetic field and g is the acceleration of gravity.

Oil Removal Experiments
The oil removal study was carried out using the gravimetric oil removal method, as described elsewhere [35,36]. In this experiment, a 25 mL beaker was filled with a 20 mL of seawater. To this beaker, a known amount of oil sample under investigation (crude, diesel, gasoline, and used oil) was splilt on the top of seawater surface. Then, the required amount of ZFO MNPs (0.01-0.05 g) was added to the top of the oil spill. After a period of 5 min, the ZFO MNPs containing the sorbed oil were collected using a permanent magnet. The oil removal experiments were triplicated and equation (8) was used to estimate the gravimetric oil removal (GOR, g/g), as follows; (8) where is the mass of ZFO MNPs and is the total mass of the beaker (containing the oil spot, ZFO MNPs, and seawater) whereas stands for the mass of oil residue (beaker mass after removing ZFO MNPs-loaded tested oil).

Characterization of ZFO MNPs and oil samples
The XRD pattern of the ZFO MNPs prepared via the sol-gel process is shown in Figure- [37]. In addition, small amounts of impurities were observed in the XRD pattern, which were identified as α-Fe 2 O 3 and ZnO [37,38]. The structural parameter values of ZFO MNPs, including lattice parameters (a), unit cell volume ( ), X-ray density ( ), crystal size (D), and specific surface area ( ), were estimated using XRD data of the high-intensity peak (311  [8] and higher than the one for maghemite/lignin-CNSL-formol nanocomposite (5.1 mN) [35]. The achieved magnetic force for ZFO MNPs was sufficient for removing the oil-loaded ZFO MNPs from the surface of water (Figure-6 c). In addition, the apparent density ( ) of the ZFO MNPs was found to be 0.5287 g/cm 3 , which is lower than the density of seawater (1.0216 g/cm 3 ). This reported value indicates that ZFO MNPs will float easily on the top of oil-contaminated surfaces [3,36].
The FTIR spectra of the dried ash and ZFO MNPs (400 to 4000 cm -1 ) are shown in Figure-3. It should be noted that both samples exhibit similar FTIR spectra. As shown (Figure-3b), the absorption bands located at 468 (ν 2 ) and 549 (ν 1 ) cm -1 could be ascribed to the bond vibrations of oxygen ionoctahedral metal ion (O-M Oct ) and oxygen ion-tetrahedral metal ion (O-M Tet ), respectively [39]. These characteristic bands are observed in FTIR spectra of most spinel ferrites and confirm their formation [40]. Within the wavenumber range (794 to 1385 cm -1 ), the observed bands are assigned to vibration [41]. In addition, the band located at 1635 cm -1 is attributed to bending vibration of O-H bonds, while that observed at 3433 cm -1 is ascribed to stretching vibration of O-H bonds [16,41]. Figure-4 displays the ATR-FTIR spectra of the selected oil samples (crude, gasoline, diesel and used motor oil) in the wavenumber ranging from 600 to 4000 cm -1 . It can be clearly seen that similar ATR-FTIR spectra were observed for all tested oil samples. As shown, an absorption band that is accredited to CH 2 rocking vibration was observed at 722 cm -1 . The band located at 1046 cm -1 is attributed to the linkage of ester which is present in the asphaltene molecule. The absorption bands at 1372 and 1456 cm -1 are ascribed to C-H bonds bending vibration in methyl and methylene, respectively. The bands located at approximately 2854 and 2919 cm -1 , respectively, are assigned to stretching vibration of C-H bonds (asymmetric and symmetric) in aliphatic groups [42,43].

Gravimetric oil removal
Four different types of oily liquids, namely crude, gasoline engine oil, diesel engine oil, and used motor oil, were tested for gravimetric oil removal (GOR) experiments. Table-1 summarizes the main characteristics of the tested oils (i.e., density and viscosity). As shown in the table, the densities of these oils are similar, but their viscosities are different. Figure-5 displays the GOR (g/g) values of the sorbate (tested oils) as a function of the amount of ZFO MNPs (sorbent). As shown, the GOR for all tested oil samples was decreased significantly with increasing the amount of ZFO MNPs from 0.01 to 0.05 g. When the sorbent mass was 0.01 g, the GOR values were about 23.00, 27.65, 22.62, and 30.54 g/g for crude, diesel engine, gasoline engine, and used motor oil, respectively. However, by increasing the sorbent amount to 0.05 g, the GOR values were decreased to 6.13, 7.42, 7.01, and 9.93 g/g for crude, diesel engine, gasoline engine, and used motor oil, respectively. The decrease in the GOR values by increasing the amount of the sorbent was also reported by others [8,44]. For instance, the motor oil sorption capacity was reported to be decreased from 33 g/g to 23 g/g by increasing the sorbent (natural wool fibers) amount from 0.04 g to 0.18 g [44]. In another study, Amar et al. [8] reported the decrease in oil sorption capacity from 13.72 g/g to 5.5 g/g when the sorbent (spinel ferrite) amount was increased from 0.01 to 0.04 g. This decrease in the sorption capacity could be attributed to the aggregation of the sorbent materials. As a result, the surface area of the sorbent was decreased and oil spills were prevented from entering into the inner available pores for oil sorption [8]. Figure-6 presents images for the gasoline engine oil removal experiment using ZFO MNPs as a sorbent material. As shown, the magnetic sorbent loaded with gasoline oil was separated easily from the surface of seawater using a magnetic bar (Figure-6 c).   In general, the process of oil sorption is very complicated and depends on the properties of both the tested oil (e.g., density, viscosity, etc.) and the sorbent material (e.g., porosity, surface area, etc.). Figure-7 presents the relationship between the GOR or oil sorption capacity and the viscosity of the tested oil at two different sorbent amounts (0.01 and 0.05 g). As shown in the figure, oil samples with the lower viscosities exhibited the highest oil sorption capacities. When ZFO MNPs amount was 0.05 g ( Figure-7 b), the oil sorption capacities were in the following order; used oil > diesel oil > gasoline oil > crude oil. The decreased oil sorption capacities for oils with higher viscosities could be due to the fact that high viscosity prevented oils from penetrating into the inner pores of the sorbent, resulting in low oil sorption capacity [45]. This implies that low viscosity is preferred for oil diffusion in the pores of the sorbent, providing high oil sorption capacities [46]. Thus, the used motor oil has a lower viscosity (16.00 mPa s) and, among the other tested oils, it exhibits the highest oil sorption capacities of 30.54 and 9.93 g/g when ZFO MNPs amounts were 0.01 g (Figure-7 a) and 0.05 g (Figure-7 b), respectively.
A literature-based comparison of the gravimetric oil removal of the prepared material (ZFO MNPs) with those of other magnetic nanosorbents [2,28,29,[47][48][49] is given in Table-2. As shown in the table, the gravimetric oil removal capacity of ZFO MNPs is comparable to those of other oil sorbents. These results demonstrate that the prepared material (ZFO MNPs) can be used as an oil sorbent for oil spillage cleanup from water surfaces. Although the proposed sorbent material (ZFO MNPs) exhibited an acceptable performance, the oil sorption capacity might be enhanced by improving the properties of the sorbent materials (e.g., surface area and porosity). This can be achieved using alternative methods to synthesize the sorbent material, such as the hydrothermal method. Beside the viscosity of the tested oil, water salinity and sorption time should be also investigated to address their influences on oil sorption capacity. In addition, the reusability of the proposed sorbent material should be taken into consideration as it plays an economic role in the whole oil removal process.

Conclusions
In conclusion, ZFO MNPs were used as oil sorbent materials, after being synthesized via a sol-gel method and characterized using XRD and FTIR techniques. The XRD data revealed that ZFO MNPs and small amounts of impurities (α-Fe 2 O 3 and ZnO) were obtained after burning their ash at 600 °C for 2 h. The oil sorption capability of ZFO MNPs was studied using crude oil, gasoline engine oil, diesel engine oil, and used motor oil as a water pollutant model. The prepared material exhibited a good magnetic property and a density value that is lower than that of seawater, allowing its easy magnetic separation and flotation. The oil sorption capacity of ZFO MNPs is related to the viscosity of the tested oil. The sorption capacity of the used motor oil, which has the lower viscosity, was higher than those of the other oily liquid samples. The results revealed that ZFO MNPs have a good combination of the magnetic property and the removal of oil spills from water surfaces.

Conflict of Interest
We have no conflict of interest to disclose.