Magnetic particles, generally nanostructured and magnetite-based, have been studied extensively to remove drinking water contaminants. Compositions beyond Fe3O4 could address long-standing issues of magnetic recoverability and materials integrity in drinking waters. Herein carbon-coated iron carbide (Fe3C@C) were studied for the first time for their stability, magnetic characteristics, magnetic separability, and arsenic adsorptive properties. Experimental results show that (i) Fe3C@C with a 9-nm thick graphitic shell is chemically stable in simulated drinking water; (ii) is ferromagnetic with small magnetic remanence and a magnetic saturation that is ~ 2 × greater than Fe3O4; (iii) can be separated from water magnetically under continuous-flow conditions with greater than 99% recovery; and (iv) has a surface area-normalized adsorption capacity for arsenic (6.75 µg/m2) of the same order of magnitude as that of Fe3O4 (9.62 µg/m2). Fe3C@C can be a viable alternative to Fe3O4 with further development, for the magnetic removal of arsenic and other contaminants from drinking water sources.
A comparative look at the chemical stability, adsorptive prowess, and magnetic capturability of nanostructured carbon-coated iron carbide for arsenic removal from simulated drinking water.
Nanocrystalline magnetite is one of the most commonly studied magnetic engineered nanomaterial (ENM) for water treatment  due to: (1) its ability to be easily removed from the reaction medium via magnet capture , (2) its ability to remove a host of contaminants ranging from heavy metals to organic compounds [2,3,4,5,6,7], and (3) the relatively low toxicity concern of iron [1, 8]. Nanocrystalline Fe3O4 is commercially available, easy to synthesize, and has size dependent magnetic properties such as superparamagnetism which emerges for single crystalline particle sizes < 50 nm, enabling its recovery and reuse. Such advantages have led to pilot-scale testing, for example, layered double hydroxides (LDH)/silica/magnetite and lanthanum-based magnetite ENMs for phosphorous removal from wastewater streams [9, 10] and uncoated magnetite ENMs for treating arsenic contaminated ground water wells in Mexico . Even more so, literature is replete with studies that catalog the use of iron-based nanoparticles for the removal or organic contaminants, heavy metals (i.e., Zn(II), Cu(II), Pb(II), Cr(III)), and anions (i.e., nitrates, arsenates, etc.) [12, 13]. However, there are still barriers (i.e., selectivity, regeneration, capturability, and stability) that prevent magnetic ENM treatment from being implemented in large-scale water treatment applications [9, 14]. Most studies surrounding magnetic ENMs for wastewater applications do not address leaching and lower performance issues that could occur when testing in drinking waters [1, 8, 15, 16].
A major hurdle to using iron oxides in drinking water treatment schemes is that iron-based nanoparticles have limited chemical stability in water [7, 17]. Specifically, for Fe3O4, the surface Fe2+ sites of Fe3O4 slowly oxidize, which lowers their magnetic saturation [3, 18]. Under acidic conditions, the Fe3+ readily leaches into solution [19,20,21,22]. To combat these issues, protective coatings are used to increase material stability, but the resulting core–shell structure has lower magnetic strength . A greater magnetic field strength or gradient would be needed for complete recovery, as the attractive magnetic force is dependent in part on the material's magnetization [24,25,26,27,28].
It is notable that most reported nanocrystalline Fe3O4-based absorbents have non-zero magnetic remanence and coercivity values, and are therefore not technically superparamagnetic [29,30,31]. However, their magnetic remanence and coercivity values are sufficiently small to allow for aggregation/disaggregation in the presence/absence of a magnetic field. Such materials may be suitable for water treatment scenarios in which partially magnetized particle aggregates can be tolerated.
Alternative iron compositions could improve the magnetic recovery and stability issues of iron oxide-based magnetic ENMs in water treatment. One such material is carbon-coated iron carbide ("Fe3C@C"). Containing 6.7 wt% carbon and 93.3 wt% iron , Fe3C has been studied as a nonprecious metal catalyst for oxygen reduction reactions in neutral and alkaline solutions; it dissolves under acidic conditions though [33, 34]. Fe3C has a higher magnetic strength than bulk Fe3O4 (e.g., ~ 140 emu/g  vs. 92 emu/g ) and retains its magnetic strength when crystal domain sizes are reduced to the nanoscale [11, 33, 34, 36]. Synthesized through gas-phase pyrolysis or flame spray techniques, nano-sized Fe3C@C carries a graphitic coating that protects the Fe3C surface from oxidizing. Fe3C@C has magnetic remanence and coercivity values somewhat similar with those of "superparamagnetic" Fe3O4 nanoparticles found in literature [29, 31]. A variety of magnetic carbon-based absorbents such as graphene oxide nickel ferrite, multiwalled carbon nanotubes/iron-oxide composites, magnetic biochar, and carbon encapsulated maghemite/magnetite, etc., are used for arsenic and heavy metal adsorption due to their facile magnetic recovery and the adsorptive prowess of carbon [13, 37,38,39,40,41,42,43,44,45,46,47,48,49]. However, none—to the authors knowledge—have studied Fe3C@C for arsenic removal in simulated drinking water.
There are no studies that assess the arsenic removal properties of Fe3C@C. Arsenic removal from drinking water sources remains to be an issue of concern [15, 16, 50, 51]. Large rural populations around the world that use private wells as their main source of drinking water, continue to be exposed to arsenic levels above the WHO provisional guideline for arsenic (i.e., 10 μg/L) with arsenic levels ranging between 10–50 μg/L [52, 53]. In this work, we evaluate the arsenic adsorptive properties and chemical stability of Fe3C@C in deionized water and simulated drinking water along with its magnetic recovery under flowing conditions. We subjected a commercially sourced Fe3C@C to stability batch tests under mild and aggressive conditions (i.e., 60 °C and pH 3), monitoring for dissolved iron. We quantified its adsorption isotherms for arsenic and studied its magnetic separation from water using an in-house permanent-magnet capture device. We used four common iron-based nanopowders (nanocrystalline materials in powdered form) for comparison: commercially available Fe3O4, commercially available Fe3O4 with a SiO2 coating, as-synthesized Fe3O4, and as-synthesized α-Fe2O3 nanopowder. Our analyses show that Fe3C@C (i) is chemically and magnetically stable in simulated drinking water; (ii) has surface-area-normalized adsorption capacity for arsenic comparable to commercially used iron oxide hydroxide adsorbent Bayoxide E33; and (iii) can be effectively removed from flowing water using a permanent magnet.
Materials and methods
All iron-based materials studied were nanopowders, i.e., in powdered form and containing XRD-detectible nanocrystalline domains. Carbon coated iron carbide nanopowder ("Fe3C@C"; prepared through a high-temperature plasma chemical vapor deposition process) from Nanostructured & Amorphous Materials Inc. and commercial Fe3O4 nanopowder ("c-Fe3O4") from Sigma Aldrich were used for the stability, magnetic recovery, and adsorption experiments (Table S1). As-received iron oxide hydroxide goethite granules (Bayoxide E33; CAS #51274-00-1; from Lanxess) were crushed and used for the adsorption studies. The silica coating for "c-Fe3O4@SiO2" was performed using HCl (35 vol%), and tetraethyl orthosilicate (TEOS) obtained from Sigma Aldrich and used as-received. For the synthesis of s-Fe3O4 and α-Fe2O3, ACS grade FeCl2·4H2O, FeCl3·6H2O, Fe(NO3)3 and FeCl3 purchased from Sigma Aldrich were used.
Simulated drinking water was prepared with the following ACS-grade chemicals (Sigma Aldrich): NaHCO3, CaCl2, MgSO4·7H2O, Na2SiO2·9H2O, NaNO3, NaH2PO·H2O, and NaF (fluoride standard solution from Ricca Chemical). A single sodium arsenate standard (Na3AsO4 in the form of an Atomic Adsorption Spectroscopy Arsenic Standard from Sigma Aldrich; 1000 mg/L As in 2% nitric acid prepared with high purity As2O3, HNO3, NaOH, and H2O) was used.
For the NEMI 3500 Fe-B iron detection method, HCl (35 vol%), hydroxylamine, ammonium acetate, acetic acid, and 1,10-phenanthroline monohydrate were obtained from Sigma Aldrich and used as-received.
Synthesis of c-Fe3O4@SiO2 nanopowders
The c-Fe3O4@SiO2 nanopowder was prepared as follows: 1 g of c-Fe3O4 nanopowder was placed in 0.1 M HCl aqueous solution (50 mL) and ultrasonicated for 10 min, magnetically separated and washed three times with DI water, and then dispersed in a solution of ethanol (80 mL), DI water (20 mL) and concentrated ammonia (1.0 mL, 28 wt% aqueous solution). The 0.1 M HCl is to clean off the surface of the c-Fe3O4 nanoparticles from any pre-existing impurities from the manufacture in order to prepare them for functionalization . TEOS (0.9 g, 4.32 mmol) was added immediately afterwards. The resulting mixture was stirred at room temperature for 12 h, and the resulting solid product was magnetically collected, washed with ethanol and water three times, and dispersed in ethanol.
Synthesis of s-Fe3O4 and α-Fe2O3 nanopowders
Fe3O4 nanopowder was synthesized by dissolving 1.4 g of FeCl2·4H2O and 2.7 g of FeCl3·6H2O in 100 mL of DI water. The resulting solution was heated to 80 °C. Ammonium hydroxide (25 wt%) was added dropwise until the nanopowder precipitated out of solution. After being separated from the solution with a neodymium magnet, the nanopowder was washed with DI water and ethanol and dried overnight under vacuum at 80 °C (synthesized Fe3O4 or "s-Fe3O4"). The α-Fe2O3 nanopowder was synthesized from Fe(NO3)3 and FeCl3 through forced hydrolysis, as described in an earlier report .
Each nanopowder was analyzed by powder x-ray diffraction (XRD) using a Rigaku diffractometer with Cu Kα radiation (1.5418 Å). The data was collected from 3° to 90°. The average crystallite size of each nanopowder was calculated using Scherrer's formula .
X-ray photoelectron spectroscopy
The atomic ratio of the Fe3C@C nanopowder was obtained from x-ray photoelectron spectroscopy (XPS) performed using a PHI Quantera SXM with an Al source (focused beam of 1.5 kV, 25 W). The pass energy of the survey spectra was 140 eV with step size of 0.5 eV, while the one of atomic spectra was 26 eV with step size of 0.1 eV. Each sample was dried and loaded to Al foil. XPS spectra was analyzed with MultiPak software. All peak positions were corrected based on C 1s at 284.8 eV.
Nitrogen physisorption analysis
The specific surface area of the nanopowders and the Bayoxide E33 powder was determined by first degassing each sample at 250 °C under vacuum overnight. Once degassed, nitrogen adsorption–desorption isotherms were collected at 77 K using a QuantaChrome (Model #AS3B) instrument. The specific surface area for each nanopowder was calculated with a standard five-point BET analysis method (P/P0 = 0.10, 0.15, 0.20, 0.25, and 0.30). The pore size distribution and pore volume of the Fe3C@C nanopowder were determined from the adsorption branch of the nitrogen isotherms using the non-localized density functional theory (NLDFT) model .
Magnetic characterization was conducted with the Superconducting Quantum Interference Device (SQUID) fitted with a MPMS XL (Quantum Design Inc.). For each hysteresis curve generated, the nanopowder was weighed, wrapped in Teflon tape, and analyzed at 27 °C from − 10 to 10 kOe.
Magnetic stability measurements were performed using 15 mg of c-Fe3O4 and 15 mg of Fe3C@C placed into 15 mL of a solution of simulated drinking water. The resulting dispersions were then sonicated for 10 min and placed on a shake table for 4 weeks at 25 °C. Periodically, liquid aliquots were taken from the suspension and dried on Teflon tape under vacuum at 40 °C. The resulting dried nanopowders were weighed, wrapped in Teflon, and measured from − 10 to 10 kOe at 27 °C.
Simulated drinking water preparation
Simulated drinking water was prepared in accordance with the NSF 53  challenge water by using the following salt concentrations: NaHCO3 (252 mg/L), CaCl2 (147 mg/L), MgSO4·7H2O (124 mg/L), Na2SiO3·9H2O (95 mg/L), NaNO3 (12 mg/L), NaF (2.2 mg/L), and NaH2PO4·H2O (0.18 mg/L). The solution was adjusted to pH 7.5 using HCl (1.0 M). The resulting ionic strength and total dissolved solids (TDS) values were 8.5 mM and 478 mg/L, respectively.
Stability batch tests
The stability of the magnetic materials was assessed in simulated drinking water at 25 °C and neutral pH (~ 7.5), and also at an elevated temperature (60 °C) and neutral and low pH values (~ 7.5 and ~ 3) to accelerate any potential iron leaching. We chose the elevated temperature of 60 ºC, which was the temperature used by Sidhu et al. in their study of dissolution of iron oxides at low pH values . The amount of dissolved iron for the stability tests was quantified with a Perkins Elmer Optima 8300 Inductively Coupled Plasma-Optical Emission Spectrometer (ICP-OES) and, separately, a colorimetric procedure adapted from the National Environmental Methods Index (NEMI). For ease of sample preparation and UV–vis photometric measurements, we primarily used the NEMI 3500 Fe-B (phenanthroline) method . The pH of the bulk solution was adjusted to a desired pH (~ 7.5 or ~ 3) with dropwise additions of HCl solution (1.0 M), and the resulting solution was used immediately. An appropriate amount of water was combined with the desired nanopowder such that the final weight concentration of each nanopowder suspension was 500 mg/L.
After sonication for 10 min, 2-mL aliquots were taken from each sample, filtered with a 0.2-µm PTFE syringe filter, and stored in 2-mL microcentrifuge tubes. The sonicated solutions were then placed inside a temperature-controlled shake table (60 °C, shaking speed of 150 rpm) for 24 h. The suspension pH did not change during the 24-h period.
Liquid aliquots (2 mL) were extracted from each sample and filtered with a 0.2-µm PTFE syringe filter. All samples were then analyzed for dissolved iron concentrations by using adapted procedures from the National Environmental Methods Index (NEMI). Specifically, the NEMI 3500 Fe-B (phenanthroline) method, which is well suited for field tests, was used . For each sample, 1 mL was removed and combined with a solution containing HCl (40 μL; 38 wt% conc. aqueous solution), hydroxylamine (20 µL, 22.14 mM), ammonium acetate buffer (200 µL, 587 mM), and phenanthroline solution (80 μL, 0.31 mM). The resulting solution (final pH of ~ 3.3) was vigorously agitated and then allowed to sit for 30 min before being analyzed with UV–vis spectroscopy at 510 nm (Shimazdu UV-2450 UV-Spectrophotometer).
Arsenic adsorption experiments
Batch arsenic equilibrium adsorption experiments were conducted in simulated drinking water at 25 °C. The initial pH value of the solution was set to ~ 7.5 and the initial concentration of arsenate was set to ~ 50 μg-As(V)/L. Fe3C@C, c-Fe3O4, and Bayoxide E33 were used as the adsorbent media. The adsorbent dosages ranged from 4 to 400 mg/L. Experiments were conducted in LDPE wide-mouth bottles, and the samples were continuously agitated for ~ 3 days to ensure equilibrium. When collecting samples (2 mL) for analysis, the adsorbent was removed from the suspension via a 0.2-μm Nylon syringe filter. The equilibrium water-phase concentration of arsenic was quantified using a tuned Thermo Scientific X-Series II Inductively Coupled Plasma Mass Spectrometer (ICP-MS) .
where qe is the adsorption capacity (units of μg-adsorbate/g-adsorbent), Kf is the Freundlich adsorption capacity parameter (units of μg-adsorbate/g-adsorbent)(mg-adsorbate/L)1/n), Ce is the equilibrium concentration of the contaminant (units of μg-adsorbate/L), n is the Freundlich affinity of adsorption parameter, b is the adsorption equilibrium constant (units of L/μg-adsorbate), and qm is the maximum adsorption capacity (units of μg-adsorbate/g-adsorbent) as determined by the Langmuir isotherm fitting . The “Freundlich derived” maximum adsorption capacity, qe-max, was determined using the largest measured adsorption capacity value taken from the Freundlich isotherm fitting.
Magnetic column separation experiments
Our benchtop magnetic capture unit was set up like a conventional high gradient magnetic separator (HGMS), permanent magnets were used instead of electromagnets [26, 62, 63]. A column packed with magnetically susceptible wires (e.g., stainless steel wool, SSW) causes an applied magnetic field to dehomogenize, producing large field gradients (regions where the magnetic field is no longer uniform) around the wires to attract magnetic particles to their surface. The strong magnetic forces produced by the large field gradients at the SSW wire surface are extremely effective at capturing fine particles (< 100 microns) of weakly magnetic substances .
Our bench-top magnetic capture unit consisted of a cylindrical quartz column with an internal radius of 0.330 cm and a length of 12 cm (a total volume of 4.11 cm3) (Fig. 1). The 5-cm bed length was filled with SSW (wire diameter of 50 µm, 550 mg, grade 434 stainless steel), such that the bed density was 321.6 mg-SSW/cm3 (void fraction = 0.96). The column was placed in a 2-cm gap between two permanent magnets housed within a rectangular steel frame. The total magnetic field strength produced was estimated to be within the range of 1.5–2.0 T (15–20 kOe).
A nanopowder suspension (50 mL) was pumped through the column at a desired flow rate (1, 3, 5, or 7 mL/min) using a KD Scientific Syringe Pump System (Model number: 100, Series number: 4377). The empty bed contact time (volumeEmptyBed ÷ flow rate) for each flow rate was 1.71 min, 0.57 min, 0.34 min, and 0.24 min, respectively). The respective residence times (volumeEmptyBed × (void fraction) ÷ flow rate) were 1.64 min, 0.55 min, 0.33 min, and 0.23 min. A high inlet suspension concentration of 500 mg/L (362, 467, and 350 mg/L-Fe for c-Fe3O4, Fe3C@C, and α-Fe2O3, respectively) was chosen to test the robustness of the magnetic capture unit. Collected periodically, the dissolved iron concentration in the effluent samples were analyzed with the NEMI Fe 3500-B method (Fig. S1).
Results and discussion
Figure 2 shows the x-ray diffraction (XRD) patterns for the nanopowders. The c-Fe3O4, c-Fe3O4@SiO2, and s-Fe3O4 materials were verified to have the magnetite Fe3O4 crystalline phase , with the s-Fe3O4 having a smaller grain size (Table 1). The silica content of c-Fe3O4@SiO2 estimated to be ~ 21 wt%, based on molar ratios, indicating a thickness of ~ 2 nm as detected through TEM (Fig. S2). The α-Fe2O3 sample was verified to be the α phase of Fe2O3 (hematite) .
The Fe3C@C nanopowder has the cementite crystalline phase . From x-ray photoelectron spectroscopy (XPS) analysis, the C:Fe atomic ratio of Fe3C@C is 2.08:1, which is higher than the C:Fe atomic ratio for Fe3C (1:3) due to the carbon coating. The coating accounts for ~ 14 wt% of the Fe3C@C material, suggesting ~ 26 monolayers of graphene (a single sheet is 0.34 nm thick) . This closely matches the carbon coating thickness (~ 9 nm) as detected through TEM (Fig. S3). As measured, Fe3C@C has a specific surface area (SSA) of 24.9 m2/g and (Table 2). In comparison, c-Fe3O4 displayed superparamagnetic-like behavior, with its low magnetic remanence (Mr) and coercivity (Hc) values (Fig. 3, Table 1) [67, 68]. The magnetic saturation (Ms) of c-Fe3O4 and c-Fe3O4@SiO2 were 77.1 emu/g and 54.4 emu/g, respectively.
Even after normalizing to only the Fe3O4 content (77.1 emu/g-Fe3O4 and 68.9 emu/g-Fe3O4), the magnetic saturation was lower after coating, which is possibly due to surface spin disorder caused by the silica shell and the interactions between the non-magnetic SiO2 atoms and the magnetic c-Fe3O4 atoms [65, 69, 70]. In addition, the c-Fe3O4@SiO2 had greater magnetic remanence and coercivity values than c-Fe3O4, due to the presence of its SiO2 shell . The s-Fe3O4, had a similar magnetic saturation to that of the c-Fe3O4, but had lower magnetic remanence and coercivity values. Fe3C@C had the highest magnetic saturation of all the iron-based nanopowders analyzed, with a measured magnetic saturation value of 138.2 emu/g, slightly lower than the value for bulk Fe3C (Ms ~ 140 emu/g)  and 80% higher than that for c-Fe3O4. Fe3C@C has a higher Ms than c-Fe3O4 because it has more iron and less oxygen; a materials magnetic saturation capacity is dependent upon the individual atoms that comprise the material and the arraignment of those atoms. Fe3C contains mostly iron (93.3 wt% iron) whereas ~ 72 wt% of Fe3O4 is iron. Oxygen has a lower magnetic moment (i.e., a single oxygen molecule has a magnetic moment of 2.85 bohr magnetons) than iron (i.e., 5.92 bohr magnetons).
The ratio of magnetic remanence to magnetic saturation (Mr/Ms) indicates the extent of ferromagnetism. s-Fe3O4 can be considered less ferromagnetic (or more superparamagnetic, Mr/Ms = 0.015) than c-Fe3O4 (Mr/Ms = 0.105), consistent with its smaller grain size . To contrast, α-Fe2O3 is antiferromagnetic  and did not reach magnetic saturation within the magnetic field range of -50 kOe to 50 kOe (Fig. 3, inset). Being technically a ferromagnetic material, Fe3C@C can be considered to have superparamagnetic-like property (Mr/Ms = 0.161) as c-Fe3O4@SiO2 (Mr/Ms = 0.165).
Within error (± 9.59 emu/g), the c-Fe3O4 and the Fe3C@C nanopowders were magnetically unchanged after one month of being immersed in simulated drinking water (pH ~ 8.5) and DI water (pH ~ 6.5) at room temperature, indicating structural stability and a lack of iron leaching (Tables S1-S2).
Iron leaching behavior in simulated drinking water
For all nanopowders tested at room temperature in simulated drinking water, the amount of dissolved iron found after the 24-h testing window was below the ICP-OES detection limit (< 0.1 mg/L) (Table 2). Dissolved iron was also not detected when the samples were subjected to testing at 60 °C (Table 2).
Under the more aggressive acidic water condition for accelerated testing, c-Fe3O4 and s-Fe3O4 both leached significantly due to low-pH dissolution (FeOOH + 3HCl → FeCl2+ + 2 Cl− + 2H2O). Dissolved iron levels reached 0.69 mg/L and 0.82 mg/L, respectively (Table 2) . α-Fe2O3 also leached, but the extent of iron leaching was considerably less; of all iron oxides, hematite is the most stable . Dissolved iron concentrations were monitored at the beginning of the leaching experiments (i.e., immediately after the nanopowders were added to the room temperature solution) also. Roughly 30% of the total amount of leached iron from s-Fe3O4 was released initially. Coating c-Fe3O4 with SiO2 is well noted in literature to reduce and even prevent leaching . Over the 24-h period, no leaching was detected for c-Fe3O4@SiO2 (Table 2).
It was hypothesized that Fe3C@C would behave similarly to c-Fe3O4@SiO2 since both have a protective coating. However, Fe3C@C leached significantly at pH 3 after 24 h (Table 2). It leached significantly at the beginning of the leaching experiment also, releasing > 80% of the total dissolved iron. While the protective carbon coating of the Fe3C@C fully covers the Fe3C core, the leaching might be attributable to its detected porosity (PSD of ~ 40 nm and PV ~ 0.04 cm3/g from N2 adsorption measurements), which would allow the acid to reach the Fe3C core.
Arsenic adsorption experiments
Arsenic equilibrium adsorption data for Fe3C@C and c-Fe3O4 were fit to the Freundlich isotherm model (Fig. 4, Table 3) and the Langmuir isotherm model (Fig. S5, Table S3). In literature, activated carbon and graphene-based materials have been shown to remove arsenic, although the process and optimal conditions for arsenic adsorption remain unclear [74,75,76]. In our studies, Fe3C@C shows slight adsorption aversion for the negatively charged arsenate species (i.e., H2AsO4− and HAsO42−) common in groundwaters – as indicated by its Freundlich adsorption affinity parameter (1/n = 1.13) being greater than 1. Surface chemistry is critical in arsenic removal by metal oxide species with electrostatic attraction between the anionic arsenic species being favored on positive surfaces. The pH of zero point of charge or the isoelectric point can be an indicator of the removal of anionic arsenic species . Given the negatively charged carbon surface of Fe3C@C at pH ~ 7.5, the anionic arsenate species would be averse to adsorbing onto its surface (Fig. S4; isoelectric point (IEP) of Fe3C@C ~ 4.2). Despite this aversion between the negatively charged carbon surface and the anionic arsenate species, an appreciable arsenic adsorption capacity emerges for Fe3C@C. Much like in other porous materials (i.e., iron, aluminum, titanium, zirconia, etc.), the anionic arsenate species diffuses through the porous carbon coating of Fe3C@C (porosity: PSD of ~ 40 nm and PV ~ 0.04 cm3/g) and adsorbs onto the positively charged (i.e., IEP ~ 10) iron-carbide core .
Fe3C@C was benchmarked against c-Fe3O4. c-Fe3O4 adsorbed 1.4 × more arsenic by mass than Fe3C@C at an equilibrium concentration Ce of 20 µg/L, and ~ 2 × more arsenic at Ce ~ 40 µg/L. When normalized to surface area, the maximum arsenic adsorption capacity of Fe3C@C was ~ 70% of c-Fe3O4  (qe-max-SSA = 9.62 vs. 6.75 µg/m2, respectively), which was less due to differences in arsenic affinity. As a nonmagnetic comparison material and as a commonly used As adsorbent, Bayoxide E33 adsorbed ~ 9 × more arsenic on a mass basis than Fe3C@C at an equilibrium arsenic concentration Ce of 20 µg/L and ~ 5 × more arsenic at Ce ~ 40 µg/L, due to greater arsenic affinity and a 4.8 × higher surface area. When normalized to surface area, the arsenic adsorption capacity of Fe3C@C was closer (88%) to that of Bayoxide E33 (qe-max-SSA = 6.75 vs. 7.67 mg/m2, respectively).
We gained limited insights by fitting the equilibrium adsorption data to a Langmuir model. Fitted values for the Langmuir constant and the maximum sorption capacity were negative for Fe3C@C, indicating that the monolayer adsorption assumption of the model did not apply for arsenic adsorbing onto this nanopowder (Table S3). Multilayer adsorption of arsenic seems to play more of a role in arsenic adsorption onto the surface of Fe3C@C because of its porous nature . The fitted values for the Langmuir constant and the maximum sorption capacity for c-Fe3O4 were negative also. On the other hand, the Langmuir model applied to Bayoxide E33 material; its maximum As adsorption capacity was 5 mg/g (= 41.7 µg/m2) (Table S3, Table 3).
Magnetic separation flow results
The removability of the Fe3C@C nanopowder suspended in simulated drinking was tested at different flow conditions in the bench-top magnetic capture unit. Effluent samples were collected and acidified, and any non-captured nanopowder was quantified using the NEMI 3500-B method. For the Fe3C@C nanopowder, the total dissolved iron levels in the effluent were below the US EPA secondary maximum contaminant level (SMCL ≤ 0.3 mg/L) at all flow rates tested, and below or near the lower limit of detection (0.1 mg/L) (Table 4, Table S4). The low Fe3C@C concentrations in the effluent should not raise much toxicity concerns, as the Fe3C@C composition is being studied for biomedical applications . For the comparison material, c-Fe3O4 nanopowder, the total dissolved iron levels were below the SMCL and below detection limit (< 0.1 mg/L). Both magnetic nanopowders were successfully magnetically captured from simulated drinking water regardless of contact time for the tested flow rates below. To explore the applicability of Fe3C@C for larger water treatment schemes, a magnetic capturing unit that can accommodate larger volumes and higher flow rates of water is being developed.
The aim of this work was to quantify the stability, magnetic separation, and arsenic adsorptive properties of nanostructured carbon-coated iron carbide ("Fe3C@C") in simulated drinking water for the first time to address long-standing issues of magnetic recoverability and martial integrity of iron-based adsorbents in drinking water sources. Fe3C@C is highly magnetic, with a magnetic saturation value that is at least 80% higher than that of Fe3O4 nanopowders. It retains its chemical integrity and superparamagnetic-like property in simulated drinking water, and it is completely removable using a permanent magnet flow device (i.e., we were able to achieve a percent removal of > 99% using an in-house magnetic separation column). While it adsorbs arsenic with a surface-area-normalized capacity that is modestly lower compared to Bayoxide E33 goethite (by ~ 12%) and Fe3O4 (by ~ 30%), Fe3C@C is magnetically recoverable (whereas Bayoxide E33 is not). This paper demonstrates the ability of Fe3C@C to provide a functionalizable carbon surface (that Fe3O4 does not have), to remove arsenic and plausibly other contaminants from drinking water sources.
Su C (2017) Environmental implications and applications of engineered nanoscale magnetite and its hybrid nanocomposites: a review of recent literature. J Hazard Mater 322:48–84. https://doi.org/10.1016/j.jhazmat.2016.06.060
Jun Y, Choi J, Cheon J (2007) Heterostructured magnetic nanoparticles: their versatility and high performance capabilities. Chem Commun. https://doi.org/10.1039/B614735F
Zhao Y, Li J, Zhao L et al (2014) Synthesis of amidoxime-functionalized Fe3O4@SiO2 core-shell magnetic microspheres for highly efficient sorption of U(VI). Chem Eng J 235:275–283. https://doi.org/10.1016/j.cej.2013.09.034
Prucek R, Hermanek M, Zbořil R (2009) An effect of iron(III) oxides crystallinity on their catalytic efficiency and applicability in phenol degradation-A competition between homogeneous and heterogeneous catalysis. Appl Catal A Gen 366:325–332. https://doi.org/10.1016/j.apcata.2009.07.019
Pastrana-Martínez LM, Pereira N, Lima R et al (2015) Degradation of diphenhydramine by photo-Fenton using magnetically recoverable iron oxide nanoparticles as catalyst. Chem Eng J 261:45–52. https://doi.org/10.1016/j.cej.2014.04.117
Kharisov BI, Rasika Dias HV, Kharissova OV et al (2012) Iron-containing nanomaterials: synthesis, properties, and environmental applications. RSC Adv 2:9325. https://doi.org/10.1039/c2ra20812a
Kim K, Qiu P, Cui M, Khim J (2016) Development and application of Fe3O4-Pd nanospheres as catalyst for electrochemical-heterogeneous Fenton process. Chem Eng J 284:1165–1173. https://doi.org/10.1016/j.cej.2015.09.035
Westerhoff P, Alvarez P, Li Q et al (2016) Overcoming implementation barriers for nanotechnology in drinking water treatment. Environ Sci Nano 3:1241–1253. https://doi.org/10.1039/C6EN00183A
Drenkova-Tuhtan A, Schneider M, Franzreb M et al (2017) Pilot-scale removal and recovery of dissolved phosphate from secondary wastewater effluents with reusable ZnFeZr adsorbent @ Fe3O4/SiO2 particles with magnetic harvesting. Water Res 109:77–87. https://doi.org/10.1016/j.watres.2016.11.039
Wu B, Fang L, Fortner JD et al (2017) Highly efficient and selective phosphate removal from wastewater by magnetically recoverable La(OH)3/Fe3O4nanocomposites. Water Res 126:179–188. https://doi.org/10.1016/j.watres.2017.09.034
Farrell JW, Fortner J, Work S et al (2014) Arsenic removal by nanoscale magnetite in Guanajuato, Mexico. Environ Eng Sci 31:1–10. https://doi.org/10.1089/ees.2013.0425
Kharisov BI, Dias HVR, Kharissova OV et al (2012) Iron-containing nanomaterials : synthesis, properties, and environmental applications. RSC Adv 2:9325–9358. https://doi.org/10.1039/c2ra20812a
Lingamdinne LP, Chang Y, Yang J et al (2017) Biogenic reductive preparation of magnetic inverse spinel iron oxide nanoparticles for the adsorption removal of heavy metals. Chem Eng J 307:74–84. https://doi.org/10.1016/j.cej.2016.08.067
Mandel K, Drenkova-Tuhtan A, Hutter F et al (2013) Layered double hydroxide ion exchangers on superparamagnetic microparticles for recovery of phosphate from waste water. J Mater Chem A 1:1840–1848. https://doi.org/10.1039/C2TA00571A
Alvarez PJJ, Chan CK, Elimelech M et al (2018) Emerging opportunities for nanotechnology to enhance water security. Nat Nanotechnol 13:634–641
Hodges BC, Cates EL, Kim J-H (2018) Challenges and prospects of advanced oxidation water treatment processes using catalytic nanomaterials. Nat Nanotechnol 13:642–650
Garrido-Ramírez EG, Theng BKG, Mora ML (2010) Clays and oxide minerals as catalysts and nanocatalysts in Fenton-like reactions—A review. Appl Clay Sci 47:182–192. https://doi.org/10.1016/j.clay.2009.11.044
Jiang W, Chen X, Niu Y, Pan B (2012) Spherical polystyrene-supported nano-Fe3O4of high capacity and low-field separation for arsenate removal from water. J Hazard Mater 243:319–325. https://doi.org/10.1016/j.jhazmat.2012.10.036
Zhao X, Shi Y, Wang T et al (2008) Preparation of silica-magnetite nanoparticle mixed hemimicelle sorbents for extraction of several typical phenolic compounds from environmental water samples. J Chromatogr A 1188:140–147. https://doi.org/10.1016/j.chroma.2008.02.069
Zhang F, Braun GB, Pallaoro A et al (2012) Mesoporous multifunctional upconversion luminescent and magnetic “nanorattle” materials for targeted chemotherapy. Nano Lett 12:61–67. https://doi.org/10.1021/nl202949y
Wu W, He Q, Jiang C (2008) Magnetic iron oxide nanoparticles: Synthesis and surface functionalization strategies. Nanoscale Res Lett 3:397–415. https://doi.org/10.1007/s11671-008-9174-9
Álvarez PM, Jaramillo J, López-Piñero F, Plucinski PK (2010) Preparation and characterization of magnetic TiO2 nanoparticles and their utilization for the degradation of emerging pollutants in water. Appl Catal B Environ 100:338–345. https://doi.org/10.1016/j.apcatb.2010.08.010
Issa B, Obaidat IM, Albiss BA, Haik Y (2013) Magnetic nanoparticles: Surface effects and properties related to biomedicine applications. Int J Mol Sci 14:21266–21305. https://doi.org/10.3390/ijms141121266
Andreu JS, Camacho J, Faraudo J et al (2011) Simple analytical model for the magnetophoretic separation of superparamagnetic dispersions in a uniform magnetic gradient. Phys Rev E Stat Nonlinear Soft Matter Phys 84:1–8. https://doi.org/10.1103/PhysRevE.84.021402
Hütten A, Sudfeld D, Ennen I et al (2004) New magnetic nanoparticles for biotechnology. J Biotechnol 112:47–63. https://doi.org/10.1016/j.jbiotec.2004.04.019
Yavuz CT, Mayo JT, Yu WW et al (2006) Low-field magnetic separation of monodisperse Fe3O4 nanocrystals. Science 80(314):964–967. https://doi.org/10.1126/science.1131475
Ambashta RD, Sillanpää M (2010) Water purification using magnetic assistance: a review. J Hazard Mater 180:38–49. https://doi.org/10.1016/j.jhazmat.2010.04.105
Gómez-Pastora J, Bringas E, Ortiz I (2014) Recent progress and future challenges on the use of high performance magnetic nano-adsorbents in environmental applications. Chem Eng J 256:187–204. https://doi.org/10.1016/j.cej.2014.06.119
Kolhatkar AG, Jamison AC, Litvinov D, Willson RC (2013) Tuning the magnetic properties of nanoparticles. ChemInform 45:51
Cullity BD, Graham CD (2009) Introduction to magnetic materials. John Wiley & Sons, New Jersey
Girginova PI, Daniel-da-silva AL, Lopes CB et al (2010) Journal of Colloid and Interface Science Silica coated magnetite particles for magnetic removal of Hg 2 + from water. J Colloid Interface Sci 345:234–240. https://doi.org/10.1016/j.jcis.2010.01.087
Jack DH, Jack KH (1973) Invited review: carbides and nitrides in steel. Mater Sci Eng 11:1–27. https://doi.org/10.1016/0025-5416(73)90055-4
Hu Y, Jensen JO, Zhang W et al (2014) Hollow spheres of iron carbide nanoparticles encased in graphitic layers as oxygen reduction catalysts. Angew Chemie Int Ed 53:3675–3679. https://doi.org/10.1002/anie.201400358
Herrmann IK, Grass RN, Mazunin D, Stark WJ (2009) Synthesis and covalent surface functionalization of nonoxidic iron core-shell nanomagnets. Chem Mater 21:3275–3281. https://doi.org/10.1021/cm900785u
Cohn EM, Hofer BLJ, Cohn EM (1959) Saturation magnetizations of iron carbides1. J Am Chem Soc 81:1576–1582. https://doi.org/10.1021/ja01516a016
Brame J, Li Q, Alvarez PJJ (2011) Nanotechnology- enabled water treatment and reuse : emerging opportunities and challenges for developing countries. Trends Food Sci Technol 22:618–624. https://doi.org/10.1016/j.tifs.2011.01.004
Lingamdinne LP, Koduru JR, Choi Y et al (2016) Studies on removal of Pb ( II ) and Cr ( III ) using graphene oxide based inverse spinel nickel ferrite nano-composite as sorbent. Hydrometallurgy 165:64–72. https://doi.org/10.1016/j.hydromet.2015.11.005
Lingamdinne LP, Choi Y, Kim I et al (2016) Porous graphene oxide based inverse spinel nickel ferrite nanocomposites for the enhanced adsorption removal of arsenic †. RSC Adv 6:73776–73789. https://doi.org/10.1039/c6ra10134h
Lingamdinne LP, Koduru JR, Karri RR (2019) A comprehensive review of applications of magnetic graphene oxide based nanocomposites for sustainable water purification. J Environ Manag 231:622–634. https://doi.org/10.1016/j.jenvman.2018.10.063
Koduru JR, Karri RR, Mubarak NM (2019) Smart materials magnetic graphene oxide-based nanocomposites for sustainable water purification. In: Mishra RK, Asiri AM (eds) Sustainable polymer composites and nanocomposites. Springer, Cham, pp 759–781
Ma J, Zhu Z, Chen B et al (2013) One-pot, large-scale synthesis of magnetic activated carbon nanotubes and their applications for arsenic removal. J Mater Chem A 1:4662–4666. https://doi.org/10.1039/c3ta10329c
Lingamdinne LP, Choi Y, Kim I et al (2017) Preparation and characterization of porous reduced graphene oxide based inverse spinel nickel ferrite nanocomposite for adsorption removal of radionuclides. J Hazard Mater 326:145–156. https://doi.org/10.1016/j.jhazmat.2016.12.035
Lingamdinne LP, Koduru JR, Chang Y, Karri RR (2018) Process optimization and adsorption modeling of Pb ( II ) on nickel ferrite-reduced graphene oxide nano-composite. J Mol Liq 250:202–211. https://doi.org/10.1016/j.molliq.2017.11.174
Lingamdinne LP, Reddy J, Chang Y et al (2019) Facile synthesis of flowered mesoporous graphene oxide-lanthanum fluoride nanocomposite for adsorptive removal of arsenic. J Mol Liq 279:32–42. https://doi.org/10.1016/j.molliq.2019.01.103
Chen B, Zhu Z, Ma J et al (2014) Journal of Colloid and Interface Science One-pot, solid-phase synthesis of magnetic multiwalled carbon nanotube/iron oxide composites and their application in arsenic removal. J Colloid Interface Sci 434:9–17. https://doi.org/10.1016/j.jcis.2014.07.046
Lingamdinne LP, Choi J, Choi Y et al (2020) Process modeling and optimization of an iron oxide immobilized graphene oxide gadolinium nanocomposite for arsenic adsorption. J Mol Liq 299:112261. https://doi.org/10.1016/j.molliq.2019.112261
Wu Z, Li W, Webley PA, Zhao D (2012) General and controllable synthesis of novel mesoporous magnetic iron oxide @ carbon encapsulates for efficient arsenic removal. Adv Mater 24:485–491. https://doi.org/10.1002/adma.201103789
Mahmoodi NM, Abdi J, Bastani D (2014) Direct dyes removal using modified magnetic ferrite nanoparticle. J Environ Heal Sci Eng 12:96. https://doi.org/10.1186/2052-336X-12-96
Zhou Z, Liu Y, Liu S et al (2017) Sorption performance and mechanisms of arsenic (V) removal by magnetic gelatin-modified biochar. Chem Eng J 314:223–231. https://doi.org/10.1016/j.cej.2016.12.113
Fawell J, Nieuwenhuijsen MJ (2003) Contaminants in drinking water. Br Med Bull 68:199–208. https://doi.org/10.1093/bmb/ldg027
Sharma SK, Sanghi R (2013) Wastewater reuse and management. Springer, Dordrecht
Westerhoff P, Benn T, Chen AS, et al (2008) Assessing arsenic removal by metal (hydr) oxide adsorptive media using rapid small scale column tests.
Zheng Y (2017) Lessons learned from arsenic mitigation among private well households. Curr Environ Heal reports 4:373–382
Deng Y, Qi D, Deng C et al (2008) Superparamagnetic high-magnetization microspheres with an Fe3O4@SiO2 core and perpendicularly aligned mesoporous SiO2 shell for removal of microcystins. J Am Chem Soc 130:28–29
Lounsbury AW, Yamani JS, Johnston CP et al (2016) The role of counter ions in nano-hematite synthesis: implications for surface area and selenium adsorption capacity. J Hazard Mater 310:117–124. https://doi.org/10.1016/j.jhazmat.2016.01.078
Girolami GS (2016) X-ray crystallography. University Science Books, Mill Valley, California
Ersan G, Apul OG, Perreault F, Karan T (2017) Adsorption of organic contaminants by graphene nanosheets : a review. Water Res. https://doi.org/10.1016/j.watres.2017.08.010
National Science Foundation NSF/ANSI 53 Standards for Water Treatment Systems. https://www.nsf.org/consumer-resources/water-quality/water-filters-testing-treatment/standards-water-treatment-systems. Accessed 20 Jul 2017
Sidhu PS (1981) Dissolution of iron oxides and oxyhydroxides in hydrochloric and perchloric acids. Clays Clay Miner 29:269–276. https://doi.org/10.1346/CCMN.1981.0290404
American Public Health Association, American Water Works Association, Water Environment Federation (1999) 3500-Fe B. Phenanthroline Method. In: Standard Methods for the Examination of Water and Wastewater Standard
Sandoval R, Cooper AM, Aymar K et al (2011) Removal of arsenic and methylene blue from water by granular activated carbon media impregnated with zirconium dioxide nanoparticles. J Hazard Mater 193:296–303. https://doi.org/10.1016/j.jhazmat.2011.07.061
Oberteuffer J (1973) High gradient magnetic separation. Magn IEEE Trans 9:303–306. https://doi.org/10.1109/TMAG.1973.1067673
Moeser GD, Roach KA, Green WH et al (2004) High-gradient magnetic separation of coated magnetic nanoparticles. AIChE J 50:2835–2848. https://doi.org/10.1002/aic.10270
Oberteuffer J (1974) Magnetic separation: a review of principles, devices, and applications. Magn IEEE Trans 10:223–238. https://doi.org/10.1109/TMAG.1974.1058315
Larumbe S, Gómez-Polo C, Pérez-Landazábal JI, Pastor JM (2012) Effect of a SiO2 coating on the magnetic properties of Fe3O4 nanoparticles. J Phys Condens Matter 24:266007. https://doi.org/10.1088/0953-8984/24/26/266007
El Mendili Y, Bardeau J-F, Randrianantoandro N et al (2012) Insights into the mechanism related to the phase transition from γ-Fe2O3 to α-Fe2O3 nanoparticles induced by thermal treatment and laser irradiation. J Phys Chem C 116:23785–23792. https://doi.org/10.1021/jp308418x
Mahmoudi M, Sant S, Wang B et al (2011) Superparamagnetic iron oxide nanoparticles (SPIONs): development, surface modification and applications in chemotherapy. Adv Drug Deliv Rev 63:24–46. https://doi.org/10.1016/j.addr.2010.05.006
Qu X, Alvarez PJJ, Li Q (2013) Applications of nanotechnology in water and wastewater treatment. Water Res 47:3931–3946. https://doi.org/10.1016/j.watres.2012.09.058
Tang T, Liu F, Liu Y et al (2014) Identifying the magnetic properties of graphene oxide. Appl Phys Lett 123104:27–32. https://doi.org/10.1063/1.4869827
Veligatla M, Katakam S, Das S et al (2015) Effect of iron on the enhancement of magnetic properties for cobalt-based soft magnetic metallic glasses. Metall Mater Trans A 46:1019–1023. https://doi.org/10.1007/s11661-014-2714-2
Cheng Y, Tan R, Wang W (2010) Controllable synthesis and magnetic properties of Fe3O4 and Fe3O4 @ SiO2 microspheres. J Mater Sci 45:5347–5352. https://doi.org/10.1007/s10853-010-4583-4
Cornell RM, Posner AM, Quirk JP (1976) Kinetics and mechanisms of the acid dissolution of goethite (a-FeOOH). J Inorg Nucl Chem 38:563–567. https://doi.org/10.1016/0022-1902(76)80305-3
Gawande MB, Monga Y, Zboril R, Sharma RK (2015) Silica-decorated magnetic nanocomposites for catalytic applications. Coord Chem Rev 288:118–143. https://doi.org/10.1016/j.ccr.2015.01.001
Ungureanu G, Santo S, Boaventura R, Botelho C (2015) Arsenic and antimony in water and wasterwater: overview of removal techniques with special reference to latest advances in adsorption. J Environ Manag 151:326–346
Mondal MK, Garg R (2017) A comprehensive review on removal of arsenic using activated carbon prepared from easily available waste materials. Environ Sci Pollut Res 24:13295–13306
Mohan D, Pittman CU (2007) Arsenic removal from water/wastewater using adsorbents—A critical review. J Hazard Mater 142:1–53. https://doi.org/10.1016/j.jhazmat.2007.01.006
Gutierrez-muñiz OE, García-rosales G, Ordoñez-regil E et al (2013) Synthesis, characterization and adsorptive properties of carbon with iron nanoparticles and iron carbide for the removal of As (V) from water. J Environ Manag 114:1–7. https://doi.org/10.1016/j.jenvman.2012.09.027
Clark CA, Heck KN, Powell CD, Wong MS (2019) Highly defective UiO-66 materials for the adsorptive removal of Perfluorooctanesulfonate. ACS Sustain Chem Eng 7:6619–6628. https://doi.org/10.1021/acssuschemeng.8b05572
Yu J, Chen F, Gao W et al (2017) Iron carbide nanoparticles : an innovative nanoplatform for biomedical applications. Nanoscale Horizons 2:81–88. https://doi.org/10.1039/C6NH00173D
This work was supported by the National Science Foundation (EEC-1449500) Nanosystems Engineering Research Center on Nanotechnology-Enabled Water Treatment. The authors thank Dr. K. N. Heck and Ms. M. Marcos for conducting BET measurements, Ms. Y. Xu for conducting XPS measurements, and Mr. R. Turley for conducting ICP-OES measurements. The authors wish to acknowledge the staff and facilities of the Shared Equipment Authority at Rice University, and Dr. T. Shen at Nanostructured & Amorphous Materials Inc. for helpful discussions about the Fe3C@C nanopowder.
This work was supported by the National Science Foundation (EEC-1449500) Nanosystems Engineering Research Center on Nanotechnology-Enabled Water Treatment.
Conflict of interest
There are no conflicts to declare.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Below is the link to the electronic supplementary material.
About this article
Cite this article
Powell, C.D., Guo, S., Godret-Miertschin, L.M. et al. Magnetically recoverable carbon-coated iron carbide with arsenic adsorptive removal properties. SN Appl. Sci. 2, 1690 (2020). https://doi.org/10.1007/s42452-020-03491-7
- Environmental nanotechnology