Removal of small-ring polycyclic aromatic hydrocarbons using cow and goat milk iron oxide nanoparticles

ABSTRACT

This work reports an environmentally friendly method for removing small-ring polycyclic aromatic hydrocarbons (PAHs) using iron oxide nanoparticles (IONPs) synthesized from pasteurized cow and goat milk and iron salts. IONPs were prepared at room temperature using FeSO₄ (2.0 g in 500 mL) and 10.0 mL of milk at pH 12. Results show that IONPs were formed with low crystallinity. The presence of Fe was confirmed through energy-dispersive X-ray spectroscopy results. Transmission electron microscopy and X-ray diffraction data show smaller semi-spherical nanometer-size particles in an aggregated state. The PAH removal process was monitored using a fluorescence spectrometer, and the results showed that the type of iron salt used (Fe²⁺/Fe³⁺) in the preparation of IONPs had no significant effect on the removal of model PAHs. However, the ring structure of the PAHs plays a role in the time taken for the removal process. All PAHs were removed from water efficiently with naphthalene, anthracene, and fluorene removal efficiencies of 75%, >90%, and 70%, respectively. The current study shows that IONPs formed using milk have the potential to be used to remove PAHs from water.

GRAPHICAL ABSTRACT

KEYWORDS:

• Polycyclic aromatic hydrocarbons
• cow milk
• goat milk
• green nanoparticles
• iron oxide nanoparticles

Introduction

Environmental pollution is one of the major challenges the world faces today. Among the key environmental casualties, water pollution is a matter of concern (1,2). The most common hazardous pollutants present in water are heavy metals, azo dyes, polycyclic aromatic hydrocarbons (PAHs), nitrogen, and phosphorus (3–6). These contaminants are added to water from multiple pathways (7). Contaminated water affects aquatic animals, plants, and microorganisms and has a detrimental effect on human life (2). PAHs consist of over 1000 compounds and are composed of two or more aromatic rings arranged in linear, angular, or cluster form. Over 80,000 tons of PAHs are estimated to be discharged into water annually (8). They are mostly found in water bodies with proximity to crude oil exploration, gas and petrochemical production, and wood charcoal processing industries. These compounds can also be generated by both anthropogenic and natural sources (9). Produced water is one of the largest anthropogenic sources containing a considerable amount of PAHs (10). Crude oil contains 0.2–7% PAHs with configurations ranging from two to six rings (1,11). PAHs have low solubility in water, and the solubility varies depending on the structure. Of these 16 priority PAHs, some possess carcinogenic toxicities at low concentrations and exposure can occur via air, water, food, and occupation. However, the most prevalent path of human exposure is through polluted water (12).

Several studies have shown that PAHs have toxic and carcinogenic or tumorigenic effects on humans, laboratory animals, and wildlife (13–15). Early studies have shown that exposure to PAHs was linked to lung, skin, and bladder cancers, and some PAHs have caused asthma and chronic obstructive pulmonary disease (16). PAHs are also known to increase human inflammatory-mediated diseases, risks of developmental birth defects, acute diseases, human reproductive outcomes, development of atherosclerosis, and cardiovascular diseases (17–22).

Since PAHs are persistent environmental pollutants having a wide range of biological toxicity, the removal of PAHs from the environment has been a global concern for many decades. Reducing the levels of PAHs in water can be achieved by physical, biological, and chemical processes, such as environmental clean-up, adsorption, reverse osmosis, combined coagulation and flocculation, chlorination, ozonation, electrochemical oxidation, biodegradation, and photocatalysis (22–25). Waste treatment plants are not designed for PAH removal, which shows a gap in PAH remediation (26). Recently, biological methods have gained considerable attention as they are environmentally friendly and do not have the same disadvantages associated with chemical and physical methods. In microbial remediation methods, bacteria, fungi, algae, and other co-cultures have been tested with a variety of PAHs (1,25). PAH bioremediation depends on many factors, such as PAH structure, temperature, and the pH of the solution. Of the priority PAHs, many studies are focused on the higher ring PAHs. Not much attention is given to lower ring PAHs (2–3 ring priority PAHs) as they are thought to be volatile and less harmful to many living species. However, recent research suggests that small-ring naphthalene is toxic in animal studies (27–29), and many small-ring PAHs can accumulate and present in water bodies affecting organisms that may be consumed by humans (30–32).

Application of nanomaterials in water purification has been attempted using synthetic nanoparticles (NPs) as well as NPs synthesized using green sources (33–36). Metal ions, silver, gold, and, in some cases, iron (Fe) salts are used in combination with plant extracts and other natural sources. The main advantage of Fe over other metals is the low toxicity, biodegradability, low cost, abundance, and high chemical reactivity and affinity for the substrates (37). Compared to the synthetic methods, the yield of green method nanomaterials varies depending on the starting source. Research articles published thus far show some promising results of green metal/metal oxide NPs in the removal of PAHs (6,35–37). Hence, the identification of a green source that generates a significant yield and can remove all the priority PAHs will take the field of green NP-based water purification to the next level.

To investigate the efficiency of Green-Iron Oxide Nanoparticles (GIONPs) in the removal of PAHs from contaminated water, NPs were prepared using cow (CW) and goat (GT) milk as the green source. Milk has been used in the preparation of silver, iron, and platinum NPs and has the potential to reduce metal ions and stabilize NPs (38–45). In our previous work, we showed the application of milk to prepare NPs using FeCl₃ and the application of these green nanoparticles (GNPs) in the removal of azo dye methyl orange (44,45). The selection of iron to prepare the GNPs to remove PAHs was based on our work and the work of others (6,35,36,45). Herein, we report the synthesis of GNPs using ferrous sulfate (FeSO₄) and CW and GT milk. The GIONPs of Fe(II) and Fe(III) salts were used in the removal of small-ring PAHs, and the results are presented here.

Materials and methods

Materials and instrumentation

Anhydrous FeSO₄, and high-purity naphthalene, anthracene, and fluorene were purchased from Sigma Aldrich (St. Louis, MO) chemical company. Pasteurized CW and GT milk were purchased from local stores in TX. A BT Lab system shaker, a Jasco V-770 spectrophotometer, a Thermo Fisher Scientific Nicolet iS5 Fourier Transform infrared (FTIR) spectrometer, a JEOL JSM-6010LA, a scanning electron microscope (SEM), a Bruker Endeavor x-ray diffraction (XRD) spectrometer, and a Kratos Ultra DLD spectrometer were used in the characterization studies.

Synthesis of IONPs using FeSO₄

The iron oxide nanoparticle (IONP) using Fe(II) salt and milk was prepared following the method described by Hassan et al. with minor modifications (35). To 2.0 g of FeSO₄ in a 500.0 mL beaker, distilled water was added, and the mixture stirred continuously until dissolved. A 10.0 mL aliquot of milk was then added, and the mixture stirred. The pH of the mixture was adjusted to 12.0 by adding 2M NaOH, and the mixture stirred until a dark brown precipitate was formed. The precipitate was then centrifuged at 6000 rpm for 15 min. The precipitate was separated and washed several times with distilled water. After decanting the supernatant, the precipitate was dried at 70°C overnight and used in the characterization and PAH removal experiments.

SEM, EDS, TEM, and XRD measurements

Scanning electron microscopy and energy-dispersive X-ray spectroscopy (EDS) analyses were performed using a scanning electron microscope and In TouchScope software (44,45). The samples were prepared on carbon-coated adhesive tape. Backscattered electron images were collected using an accelerating voltage of 10 kV and a load current of approximately 90 µA with a working distance of 9 mm. EDS spectra were gathered at a magnification of 2000×, and the analyzed area was 0.15 mm² (110 µm × 135 µm). A silicon-drift detector performed qualitative analysis and quantification of the elemental composition with the characteristic X-rays. Semi-quantification was based on theoretical correction of the ZAF (Z-atomic number, A-absorption, and F-fluorescent excitation) effect. The transmission electron microscopy (TEM) images were obtained using a JEOL 1200 EX microscope at an accelerating voltage of 100 kV. IONPs were sonicated in distilled water and filtered through a 0.2-micron filter, and the sample solution was pipetted onto a size 300 carbon-coated Cu grid. The particle size distribution of the IONPs was obtained from TEM images and analyzed using ImageJ version 1.53t freeware (http://imagej.nih.gov/ij). Images were spatially calibrated, and the largest diameter (length) of the particles was measured manually using the line selection tool in ImageJ. XRD experiments were performed using an X-ray source of 1 kW Cu X-ray tube and maintained at an operating current of 40 kV and 25 mA. The wavelength and the scan type were 1.540598 Å and coupled Ɵ/2Ɵ, respectively (44,45).

FTIR spectroscopy measurements

Interactions of biomolecules with NPs of pasteurized CW and GT were observed by recording FTIR spectra of the synthesized NPs using an FTIR spectrometer. FTIR spectra were recorded in the range of 400–4000 cm⁻¹. The samples were directly applied to the diamond crystal, and the spectra were recorded using the attenuated total reflection (ATR) method, while the data were analyzed using the Omnic FTIR software (44,45).

XPS measurements

X-ray photoelectron spectroscopy (XPS) measurements were performed on a Kratos Ultra DLD spectrometer using a monochromatic Al-Kα source operating at 150 W (1486.6 eV). The operating pressure was 2 × 10⁻⁹ Torr. Charge compensation was achieved by using low-energy electrons. All spectra were charge-referenced by adjusting the C 1s region to 284.6 eV. The survey scans were acquired at a pass energy of 120 eV. The high-resolution spectra were acquired at a pass energy of 20 eV. The XPS data were processed using Casa XPS software.

PAH removal

1 mg/L solutions of naphthalene, anthracene, and fluorene were prepared separately as described by Hassan et al. (35) in the PAH removal process. Herein, naphthalene, anthracene, and fluorene were prepared by weighing 1 mg of the PAH, dissolving it in 1 mL of acetonitrile, and then transferring the PAH in acetonitrile to water in a 1L volumetric flask. To 10.0 mL of 1 mg/L naphthalene solution, 5.0 mg of the NPs were added, and the mixture was allowed to mix for 2–3 h. The mixture was centrifuged at 1000 rpm ,and the scanning started after 30 min followed by scanning at 15 min intervals using the fluorescence spectrometer. The excitation wavelength of naphthalene was 270 nm, and the emission spectrum scanned from 300 nm to 450 nm. In a similar manner to 10.0 mL of 1 mg/L of anthracene solution, added 10.0 mg of the GIONP in a falcon tube and then centrifuged. The centrifugation was done for 90 min at 1000 rpm. The excitation wavelength of anthracene was 350 nm, and the emission spectrum scanned from 350 to 550 nm. In a falcon tube, 10.0 mg of each GIONP was added to 10.0 mL of prepared fluorene (1 mg/L) solution and centrifuged at 1000 rpm for 30 min first, and the resultant mixture was scanned using the fluorescence spectrometer. The centrifugation was performed, and the changes were observed using the fluorometer every 15 min for a total of 120–150 min for the different GIONPs. The excitation wavelength of fluorene was 279 nm, and the emission spectrum scanned from 275 to 375 nm.

Results and discussion

SEM-EDS and TEM characterization of CW and GT IONPs

The NP formation using CW and GT milk and the iron salt was visually observed with the color of the solution. This observation was consistent with the data reported previously by others (35,46). The yield generated using the Fe(II) salt was much higher than the yield obtained in our previous work, where a Fe(III) salt was used. Figure 1 shows the SEM-EDS profiles of CW and GT IONPs. The SEM data provide information on the morphology of the NPs, while the EDS spectrum provides quantitative information on the synthesized NPs. SEM data in Figure 1(upper panel) show aggregated spherical NPs similar to our previous work (44,45). The quantitative data of both types of NP samples exhibit absorption peaks for Fe between 6.00 and 8.00 keV with characteristic Fe peaks seen at 6.5 and 7.1 keV. In both preparations, the predominant elements present were C and Fe followed by O. The amount of Fe present by atom was 17.36% and 12.36% for CW and GT IONPs, respectively. Trace amounts of S, Ca, and P were detected in both preparations. The size and the shape of the NPs were observed through TEM experiments. Figure 2 shows TEM data and the average size of the NPs were 35 ± 8 and 46 ± 11 nm for CW and GT IONPs, respectively.

Figure 1. SEM (A&B) and EDS (C&D) spectrum of synthesized IONPs obtained from CW and GT milk. The upper panels represent the SEM images, and the lower panels represent the EDS data. The insets in (C) and (D) represent the atom% of elements present in each case.

Figure 2. TEM images of synthesized IONPs obtained from CW (A) and GT (B) milk.

XRD and XPS characterization of CW and GT IONPs

The XRD patterns of the synthesized NPs are shown in Figure 3, and the data provided detailed information on the chemical composition and crystallographic nature of the synthesized NPs. Reflection peaks in Figure 3 appear broader, indicating the formation of the NPs. Both NPs (CW and GT and Fe(II) salt) showed diffraction peaks at 2Ɵ = 19.21° & 20.16°, 35.9° & 36.03°, 50.17° & 50.81°, and 63.00° & 63.07° for NPs of CW and GT, respectively. These peaks correspond to crystalline peak planes of (012), (104), (116), and (214) of Fe₂O₃ and are matched with the standard Fe₂O₃ PDF card (JCPDS Number 24-0072). The broader peaks, which lack all the crystalline peaks of Fe₂O₃, indicate that the NPs are of lower crystallinity (higher amorphicity) and possibly heterogeneous. The lower crystallinity observed is consistent with previous observations (45,47,48). XRD data reported here are consistent with our previous work on silver and Fe NP preparations (44,45) with these types of milk, and this indicates that in the NPs, crystallization of the bio-organic phase may occur on the surface of the NPs. The average crystal size (D) for the prepared NPs was calculated using the Debye–Scherrer’s equation, D = 0.89λ/βcosƟ, where λ is the X-ray wavelength, β is the full width at half maximum of the diffraction line, and Ɵ is the diffraction angle of the XRD spectra. The average crystal size was calculated to be ∼ 2 and 3 nm for CW and GT IONPs.

Figure 3. XRD pattern of IONPs synthesized using CW (A) and GT (B) milk.

Figure 4 shows the XPS data obtained for CW and GT IONPs. XPS analysis gives a better understanding of the core levels. Figure 4(A,B) displays the survey scans obtained for the IONPs and shows the presence of elements C, O, and Fe in both NP preparations. Survey scans showed that in both preparations, the amount of Fe was around 1% and was less than that of C (80%) and O (18%). Changes in the oxidation state of Fe associated with IONP formation were observed using XPS scans. Similar scans were performed for the elements C and O, and the individual scans of C (1S) (4C, 4D), O (1S) (4E, 4F), and Fe (2p) (4G, 4F) are shown in Figure 4. Similar to our previous work and work from other groups, the high-resolution spectra of C1s for CW and GT IONPs can be fitted with four peaks corresponding to C–C, C–H, C–O, C=O, and O–C=O bonds, respectively (45). O1s spectra for both samples were fitted with three peaks corresponding to O–C = O, C–O, and Fe-O bonds. Scans on the Fe (2P) curve were fitted with four peaks corresponding to Fe metal, Fe–O, Fe₂O₃, and FeOOH. The amount of free Fe present was less than 1%, and the predominant forms were Fe–O and Fe₂O₃ (>85%). Both NP preparations had similar amounts of Fe₂O₃ and were about 39%. CW IONP preparation had a higher amount of FeOOH (13%), whereas GT IONP showed less than 0.5% of the same substance.

Figure 4. XPS survey spectrum (A and B), and high resolution spectra of C1s (C and D), O1s (E and F), and Fe2p (G and H) of IONP synthesized using CW (left) and GT (right) milk.

UV–visible and FTIR characterization of CW and GT IONPs

UV–visible and IR spectroscopy are widely used characterization methods of NP preparations. Previous work on IONPs has resulted in absorption maxima owing to surface plasmon resonance (SPR), where peaks in an absorption band with Fe₂O₃-NP exhibited SPR at approximately 370 nm (46,49). The peak maximum in our IONP preparations resulted at 360 nm (Supplementary figure 1) is consistent with previous observations. SPR ranges with varying energy bandgaps, and the spectra are shifted from the observed spectra for pure metallic Fe (268 nm) (50) to other wavelengths depending on the size and the shape of the metal oxide/s present (51). Bandgaps are important as they help us to understand the electronic behavior of a material. As both NPs of CW and GT resulted in similar broad peaks with a maximum at 370 nm, our results indicate that the geometrical parameters of the metal oxides remain similar in all preparations. The broad peak observed for the preparations confirmed the presence of aggregated NPs. As observed previously, aggregation could be a result of H-bonding present in organic moieties present in milk samples (39,40,45,46).

ATR-FTIR analysis is a method used to identify the functional groups in a green source responsible for reducing metal ions in NP preparations. When both types of IONP preparations were subjected to FTIR analysis, spectra were obtained from 500 to 4000 cm⁻¹. Figure 5 displays the FTIR spectra, and Table 1 lists the vibrational modes and wavenumbers for different stretching and bending modes of different functional groups. Consistent with our previous work, strong and medium bands corresponding to the O–H stretching and bending of carboxylic acids were observed: C = C stretching and bending of alkenes, N–O stretching of nitro compounds, C–H bending of alkanes and aromatic compounds, strong C–O stretching of alkyl aryl compounds, strong N–O stretching of nitro compounds, and C–N stretching of amines were also observed (45). The band at 600 cm⁻¹ indicates a possible Fe–O stretching vibration (44,45), and the FTIR data demonstrate that these groups (present in milk) surround the NPs (44,45).

Figure 5. FTIR spectrum of IONPs synthesized using CW (A) and GT (B) milk.

Table 1. FTIR spectrum data of CW and GT milk IONP.

Absorption range (cm^−1)	Group	Compound class	CW	GT
3269–3281	O–H Stretching	Carboxylic acid/ Alcohol	3269.93	3271.48
2916–2921	C–H Stretching	Alkane	2920.80	2921.23
2850–2852	C–H Stretching	Alkane	2851.30	2851.72
1451–1457	C–H Bending	Alkane (methyl)	1456.36	1456.64
1633–1645	C=C Stretching	Alkene	1636.77	1637.05
1740–1743	C–H Bending	Aromatic Compound	1742.42	1742.76
1537–1542	N–O Stretching	Nitro Compound	1540.80	1541.27
1378–1426	O–H Bending	Carboxylic Acid /Alcohol	1404.89	1425.15
1236–1239	C–N Stretching	Amine	1238.57	1238.84
1156–1173	C–N Stretching/C-O Stretching	Amine/Tertiary Alcohol	1158.59	1156.94
1093–1098	C–N Stretching	Amine	1093.62	1094.65
1000–650	C=C Bending/Fe–O stretching	Alkene/Metal-O	987.44	925.64

PAH removal using IONPs

Green IONPs have been used in PAH removal and also dye degradation studies (6,35,36,46–48,52). In many studies, the IONPs were formed using Fe(II) or Fe(III) salts in conjunction with plant extracts or other sources. Our previous work demonstrated the applicability of NPs synthesized using FeCl₃ in dye degradation studies. Herein, we used NPs synthesized using both Fe salts in the PAH removal studies. Small-ring PAHs are in general volatile substances. However, some amounts may still remain in aquatic environments. PAHs are fluorophores and can be studied using fluorescence spectroscopy. The excitation wavelengths of the PAHs vary depending on the molecule, and all PAHs are excited at different wavelengths. Figure 6 represents the fluorescence emission scans obtained for the removal of naphthalene, anthracene, and fluorene using IONPs synthesized using FeSO₄ and CW and GT milk. The removal of these PAHs using IONPs synthesized using FeCl₃, and CW and GT milk are given in the supplementary material (Supplementary figure 2). In all cases, the removal process was over 70% complete in less than 2 h with a minimal amount of IONPs. Figure 7 and supplementary figure (Supplementary figure 3) represent the fluorescence intensity versus time plots for the removal process for all three PAHs. Results indicate that the removal of naphthalene and fluorene was slightly higher with IONPs made of CW and GT milk and FeCl₃, while anthracene did not show any significant effect. This shows that how the ring structure is a key factor in the removal process. Preliminary experiments carried out with another PAH phenanthrene (data not shown) showed that the removal process using all four types of IONPs was similar and comparable with anthracene. However, the time taken was much shorter than the other PAHs discussed here, possibly due to the ring arrangement of phenanthrene as compared to anthracene. When the milk sources were compared, we did not see any significant difference in the type of milk used. In our study, both IONPs had similar sizes, and adsorption may occur only on the surface of metal NPs (35). The effectiveness of NPs can be achieved by increasing its surface area (52–55). Based on previous studies, the removal is thought to occur via an adsorption process (35). Several of the metal NPs suggest low toxicity excreted by Fe/FeO NPs (56,57). Work by Hassan et al. demonstrated that IONPs can be successfully used to remove higher ring PAHs pyrene and benzo[a]pyrene. Based on our work and work by Hassan et al., IONPs synthesized that using CW and GT milk can potentially be used in the removal of all PAHs. More work needs to be carried out in optimizing the parameters as the size of the NPs plays a critical role in the removal process and the size may vary based on many factors.

Figure 6. Fluorescence emission scans obtained for naphthalene, anthracene, and fluorene removal. The upper panels represent data obtained using CW IONPs, and the lower panel represent data obtained using GT IONPs. (A) and (D) represent data obtained for naphthalene, (B) and (E) represent data obtained for anthracene, and (C) and (F) represent data obtained for fluorene. The excitation wave lengths for the three PAHs were 270, 350, and 279 nm, respectively.

Figure 7. Fluorescence intensity vs time plots for the removal of PAHs. Figures (7A, 7B, and 7C) represent data for naphthalene, anthracene, and fluorene removal using both types of milk NPs. In all cases, the blue line represents data for CW IONPs, and the orange line represents data for GT IONPs.

Our goal in this work was to examine the synthesis process and investigate the applicability of IONPs (24) made using milk in the removal of PAHs. NPs were synthesized using biological sources; however, the presence of NPs in water is a major concern (58). Hence, using less toxic metal NPs and recovering and reusing them to remove multiple pollutants would be more effective in the wastewater treatment methods.

Conclusions

IONPs have gained considerable attention in recent years because they have the potential to be used in biomedical applications and environmental remediation. The availability of green sources limits the large-scale production of green metallic NPs. Hence, it is essential to explore the synthesis of IONPs using green sources, which can be obtained easily. In this study, we demonstrate the synthesis of IONPs using commercially available milk. The milk-based synthesis of IONPs is an environmentally friendly nontoxic method; it can be used to remove PAHs without the addition of any oxidizing agents or the presence of light. Further studies are needed to optimize the synthesis parameters, understand the adsorption mechanisms, ensure the reusability of the IONPs (35), and explore the possibility of obtaining particles of different sizes for other applications.

Acknowledgements

The authors would like to thank Mr. Tony Grady for the SEM-EDS data, Dr. Richard Littleton and Dr. Vitha Stanislav of Texas A&M University (Microscopy Imaging Center) for the TEM analysis of the samples, Dr. Joseph Reibenspies of Texas A&M University (X-ray Diffraction laboratory) for his assistance with the XRD analysis of the samples, and Dr. Angelica Benavidez of the University of New Mexico for the XPS analysis of the samples. Author contributions: Conceptualization, methodology development, data analysis, original draft preparation, H.F.; experimentation, data generation, and analysis, I.G.; and writing, figures, reviewing, and editing, H.F. and I.G. All authors have read and agreed to the content in the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Financial assistance from the PVAMU Research Office, Department of Chemistry Assistantship (IG), Welch Foundation, and DOE grant 419760-00001 is greatly acknowledged.